ie ft oo: a wes ‘ ef ; eS a as Bo | Rey Ne. i aN 19 ee SRB i Be ee HES G 2 SPOR, agra’ )\ SAA a a A, = ae 4 nh 4 hates Aine { wee Se lig “eanane” yrs WS. Qs o , nd Papere _ ~ ie peas a ot eure. ws i w yu Be wie Soe Z ~ : ’ a) . annttna TTT Bava, y bat _ : : 4 ; ag y | Alyy if’ 7 | . Tid 7) LUE AAPA : ul a eget APEN aR ITNT oe Se See aa cy “y' r ; eal at at ngs: wa as OT ieee

Maureen in ae a AAG rinse r sea nf The 9d (AA x fant fh Beye, TAIT) Ma Ju, * a Jf Ya le "Ney so Wak Lae g wd Wg 4 mig 1] Wr PORN ec cerer ones uaa Jatin aug we wee 8 mites. ty Wi. ee ) jn) we a ‘ Siebel a TOR "LoS. i: a ON i ‘ J "94 e~ 4 wl ow “5 ieee “ora dl " - Herma cctit een at a oan t}. mtr BAT tte 1 be (Seraiie “= weed veeaugt? ee Py ere j pers ] 5 yc ete aay ow nil Very os Ss TEN ‘es ea Ym vrs A ty, 2+ “dis pus, ~ & < res ee w ae oe 5 W v | ed Lid Mov © Oe * _ fe Dati if enon ryan tt | is, Ws, N. ann ace T 1 yr oe. bth rein Ud oa P pease an f ‘ee We * Oyu Bw Se a pet TS "a 3 ape - i inpy *,. Wu a Ld a vv nade OC @ J -° Ub Sel te Racca: cata eee ad e%y 4 re ‘yr? h 4 Wwe 3 wy ed Vouue Swe Pri Laer TTT eo9 al a(HAD 4 al 4aager F me eee EEL thas a8 sna ey pap we wal hy ig “1 - pe co, © » _™ PEPEeee > Miley : aé iw. Dp j pitty Pe: Yoh image iJ ; d edi piban’ i ee orl hs id past acm ~¥ kK Sk» FOURNAL OF GEOLOGY THE UNIVERSITY OF CHICAGO PRESS CHICAGO, ILLINOIS Fgents THE CAMBRIDGE UNIVERSITY PRESS LONDON AND EDINBURGH THE MARUZEN-KABUSHIKI-KAISHA TOKYO, OSAKA, KYOTO, FUKUOKA, SENDAI THE MISSION BOOK COMPANY SHANGHAI KARL W. HIERSEMANN LEIPZIG 42 ihigls BOUKNAL OF GEOLOGY A Semi-Quarterly Magazine of Geology and Related Sciences EDITORS Tuomas C. CHAMBERLIN, 2x General Charge; Roun D, Satispury, Geographic Geology, SAMUEL W. Wittiston, Vertebrate Paleontology; STuaART WELLER, /nxvertebrate Paleontology; ALBERT JOHANNSEN, Petrology; Rotiin T. CHAMBERLIN, Dynamic Geology, ASSOCIATE EDITORS SrrR ARCHIBALD GEIKIE, Great Brittain; CHARLES Barrois, Prance; ALBRECHT PENCK, Germany; Hans Reuscu, Morway,; GERARD DEGEER, Sweden; T. W. EpGEwortu Davin, Australia; Baitey Wit.I1s, Leland Stanford Juntor University; Grove K, GILBERT, Washing- ton, D.C.» CHartes D, Watcott, Smzthsontan Institution; Henry S. Witiiams, Cornell University, JosErH P. Ippincs, Washington, D.C.; JoHN C, BRANNER, Leland Stanford Juntor University; Ricuarp A. F. PENROSE, Jr., Philadelphia, Pa.; WituiaM B. Ciark, Johus Hopkins Untversity; Witit1am H. Hopss, Uxzverszty of Michigan; FRANK D, Apams, McGill University; CHARLES K, LEITH, Universzty of Wisconsin; WaLtLacE W. Atwoop, Harvard University; Witttam H. Emmons, Uzzversity of Minnesota; ArtTHUR L. Day, Carnegie Iustitution, VOLUME XXIV JANUARY-DECEMBER, 1916 e—\sonian Institure My eo uiEN THE UNIVERSITY OF CHICAGO PRESS CHICAGO, ILLINOIS Published February, March, May, June, August, September, November, December, 1916 Composed and Printed By The University of Chicago Press Chicago, Illinois, U.S.A. CONTENTS OF VOLUME XXIV NUMBER I Tue AcapiIAn Triassic. PartI. Sidney Powers . AVERAGE REGIONAL SLOPE, A CRITERION FOR THE SUBDIVISION OF OLD EROSION SURFACES. Leopold Reinecke CAIMANOIDEA VISHERI, A NEW CROCODILIAN FROM THE OLIGOCENE OF SoutH DAKOTA. Maurice G. Mehl ON THE STRUCTURE AND CLASSIFICATION OF THE STROMATOPOROIDEA. M. Heinrich THE PHYSIOGRAPHY OF Mexico. Warren N. Thayer REVIEWS NUMBER II THE AcADIAN Triassic. PartII. Sidney Powers Notes ON RipprE Marks. J. A.Udden . PYROPHYLLITIZATION, PINITIZATION, AND SILICIFICATION OF Rocks AROUND CONCEPTION BAY, NEWFOUNDLAND. A. F. Buddington . THE ORIGIN OF RED BEeps. A STUDY OF THE CONDITIONS OF ORIGIN OF THE PERMOCARBONIFEROUS AND TRIASSIC RED BEDS OF THE WESTERN UNITED STATES. PartlI. C. W. Tomlinson STUDIES IN HyDROTHERMAL ALTERATION. I, E. A. Stephenson . ZONAL WEATHERING OF A HORNBLENDE GABBRO. Albert D. Brokaw and Leon P. Smith . REVIEWS NUMBER III ORVILLE A. DERBY. John C. Branner SVANGRET Me] Reena TYPES OF PRISMATIC STRUCTURE IN IGNEOUS Rocks. Robert B. Sos- RTL ATING eae ee Rat by 8 MS Ay lees ELLipsomaL LAVAS IN THE GLACIER NaTIONAL Park, MOonrTANA. Lancaster D. Burling . . . Vv PAGE 105 123 130 153 180 200 206 209 - 215 235 vl CONTENTS OF VOLUME XXIV THE ORIGIN OF RED Beps. A STUDY OF THE CONDITIONS OF ORIGIN OF THE PERMOCARBONIFEROUS AND TRIASSIC RED BEDS OF THE WESTERN UNITED STATES. PaArTII. C. W. Tomlinson Tue ACADIAN Triassic. PartTIII. Sidney Powers Tue LoMBARD OVERTHRUST AND RELATED GEOLOGICAL FEATURES. Winthrop P. Haynes THE SKELETON OF Trimerorhachis. S.W. Williston REVIEWS : RECENT PUBLICATIONS NUMBER IV Note ON THE LINEAR ForRCE OF GROWING CrysTALs. George F. Becker and Arthur L. Day Tue CLASSIFICATION OF THE NIAGARAN FORMATIONS OF WESTERN Oxuto. Charles S. Prosser ORIGIN OF THE LyMAN ScuHIsts OF NEw Hampsuire. Frederic H. Lahee NOTES ON THE DISINTEGRATION OF GRANITE IN Ecyrpt. Donald C. Barton A RECORDING MICROMETER FOR GEOMETRICAL Rock ANALysIs. S. J. Shand Bae: REVIEWS RECENT PUBLICATIONS NUMBER V THE GEOLOGICAL SIGNIFICANCE AND GENETIC CLASSIFICATION OF ARKOSE Deposits. Donald C. Barton AN UnusvuAL Form oF Votcanic Ejecta. Wallace E. Pratt RrippLE-MARKS IN Onto Limestones. Charles S. Prosser THe RELATIONSHIPS OF THE OLENTANGY SHALE AND ASSOCIATED DEVONIAN DEposiIts OF NORTHERN On10. C. R. Stauffer EVOLUTION OF THE BASAL PLATES IN Monocyc Lic Crinoidea camerata. I. Herrick E. Wilson VARIATIONS OF GLACIERS. XX. Harry Fielding Reid . REVIEWS RECENT PUBLICATIONS PAGE 238 254 269 291 298 310 313 334 366 382 394 405 414 417 450 456 476 488 511 515 519 CONTENTS OF VOLUME XXIV NUMBER VI THE PRE-WISCONSIN Drirt oF NortH Dakota. A. G. Leonard EVOLUTION OF THE BASAL PLATES IN Monocyctic Crinoidea camerata. II. Herrick E. Wilson . DIFFERENTIATION IN INTERCRUSTAL MAGMA Hee Alfred Harker STRATIGRAPHY OF THE SKYKOMISH BASIN, WASHINGTON. Warren S. Smith “Purr” ConES ON Mount Usu. Y. Gite i ORIGIN OF FOLIATION IN THE PRE-CAMBRIAM Rocks oF NORTHERN New York. William J. Miller THE COMPOSITION OF THE AVERAGE IGNEOUS Roe Adolph Knopf . REVIEWS NUMBER VII Tue GENESIS OF LAKE Acassiz: A CONFIRMATION. W. A. Johnston THE LOWER EMBAR OF WYOMING AND Its Fauna. E. B. Branson EVOLUTION OF THE BASAL PLATES IN Monocyctic Crinoidea camerata. Ill. Herrick E. Wilson DISCOVERY OF THE GREAT LAKE TROUT, Cristivomer namaycush, IN THE PLEISTOCENE OF WISCONSIN. L. Hussakof. ASSUMPTIONS INVOLVED IN THE DocTRINE OF ISoSTATIC COMPENSATION, witH A NoTE ON HECKER’S DETERMINATION OF GRAVITY AT SEA. William Herbert Hobbs . A STAGE ATTACHMENT FOR THE METALLOGRAPHIC Microscope. Albert D. Brokaw REVIEWS NUMBER VIII Tue ROLE oF INORGANIC AGENCIES IN THE DEPOSITION OF CALCIUM CARBONATE. John Johnston and E. D. Williamson. NOTES ON THE STRUCTURAL RELATIONS BETWEEN AUSTRALASIA, NEW GuIneEA, and NEw ZEaLAnp. E. C. Andrews A BOotANICAL CRITERION OF THE ANTIQUITY OF THE ANGIOSPERMS. Edmund W. Sinnott ARE THE ‘‘ BATHOLITHS” OF THE HALIBURTON-BANCROFT mae On- TARIO, CORRECTLY NAMED? W.G. Foye . A CONTRIBUTION TO THE OOLITE PROBLEM. Francis M. Van Tuy! vil PAGE 521 533 554 559 583 587 620 623 625 639 665 685 690 718 720 729 751 777 783 792 Vili CONTENTS OF VOLUME XXIV PAGE Some EFFECTS OF CAPILLARITY ON Ort AccumuULATION. A. W. McCoy «30. 60d, SRE ie ite i ee ca ee A PEcuLtar Process oF SULPHUR Deposition. Y. Oinouye | en STUDIES FOR STUDENTS. CONTRIBUTIONS TO THE STUDY OF RIPPLE Marks. Douglas'W. Johnson’ ip "25 52.) 27.) se eee REVIEWS a) 50 4i5 Spl ee ee ce Recent PUBLICATIONS « ©.) G00 4) @ap eeaee:, | 2 4 © on INDEX TO VOLUME XCXTV | 5) (aero eee ERRATUM In the article by George T. Becker and Arthur L. Day, published in No. 4 of Vol. XXIV of this Journal, the following erratum should be noted: P. 329, line 6 from the top, the word failure is to be replaced by the word presence. " SIR ARCHIBALD GEIKIE, Great Britain se THE pane SEMI- QUARTERLY EDITED BY THOMAS GC. CHAMBERLIN AND ROLLIN D. SALISBURY ™& With the Active Collaboration of “VOLUME XXIV _ | atu oti NUMBER 1 OURNAL oF GEOLOGY, a csonian aa 35042 “ “ee 6 Nationa Muses SAMUEL W. WILLISTON ALBERT JOHANNSEN Vertebrate Paleontology _ Petrology STUART WELLER ROLLIN T. CHAMBERLIN Invertebrate Paleontology ; Dynamic Ge ASSOCIATE EDITORS JOSEPH P.IDDINGS, Washington, D.C. ~ HENRY S. WILLIAMS, Comell University ARTHUR L. DAY, Carnegie Institution JANUARY-FEBRUARY 10916 oo ee CHICAGO, ILLINOIS, U.S.A. AGENTS THE CAMBRIDGE UNIVERSITY PRESS, Lonpon anp EDINBURGH KARL W. HIERSEMANN, LErrzic THE MARUZEN-KABUSHIKI-KAISHA, Toxyo, Osaka, Kyoto CHARLES BARROIS, France JOHN C. BRANNER, Leland Stanford Junior University ALBRECHT PENCK, Germany ~ RICHARD A. F. PENROSE, Jr., Philadelphia, Pa. _ HANS REUSCH, Norway WILLIAM B. CLARK, Johns Hopkins ayers GERARD DEGEER, Sweden : WILLIAM H. HOBBS, University of Michigan _ T.W. EDGEWORTH DAVID, Australia FRANK D. ADAMS, McGill University BAILEY WILLIS, Leland Stanford Junior University CHARLES K. LEITH, University of Wisconsin ' GROVE K. GILBERT, Washington, D.C. WALLACE W. ATWOOD, Harvard University CHARLES D. WALCOTT, Smithsonian Institution WILLIAM H. EMMONS, University of Minnesota 27 47 THE ACADIAN TRIASSIC. PART I - - - - - - - SIDNEY POWERS AVERAGE REGIONAL SLOPE, A CRITERION FOR THE SUBDIVISION OF OLD — -' EROSION SURFACES 2 ORRIN S Ge AES PO iy ee Shee ehh Ee OP OED RO ENE GCE ‘ ‘CAIMANOIDEA VISHERI, A NEW CROCODILIAN FROM THE OLIGOCENE OF SOUTH DAKOTA EERE STE SSNS Le Ra a eae SRR re IVE EEE, _ ON THE STRUCTURE AND CLASSIFICATION OF THE STROMATOPOROIDEA M. HEINRICH cnn PHYSIOGRAPHY OF MEXICO ek Oe es Warren N.- THAYER ee oe OPE UNIVERSITY OF. CHICAGO: PRESS Z ian teases sxwsontan Stil, o g. P. Iddings ¥ COLLECTION y atonal mused™ Che Journal of Geology — Vol. XXIV CONTENTS FOR JANUARY-FEBRUARY 1936 No. 4 THE ACADIAN TRIASSIC. PARTI = =) =) 05 thi ie ee ee OS = = OSIDNEY “PowERs© AVERAGE REGIONAL SLOPE, A CRITERION FOR THE SUBDIVISION OF OLD EROSION SURFACES LEOPOLD REINECKE 27 CAIMANOIDEA VISHERI, A NEW CROCODILIAN FROM THE OLIGOCENE OF SOUTH DAKOTA ! AURICE G. MEHL 47 ON THE STRUCTURE AND CLASSIFICATION OF THE STROMATOPOROIDEA- - -_ - M. Hemneice 57 THE PHYSIOGRAPHY OF MEXICO) nooo = iahie! Uae ee a) ie hee oe en WAR, Nie REVIEWS.) om) cit opi dt logue tad Syelle Re sel iizln’ pilhea Blac Spdees is mst eileen Sdn 40 ket eve The Journal of Geology is published by the University of Chicago at the University Press semi- quarterly, on or about the following dates: February 1, 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 follows: For Canada, 30 cents on annua! 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 single copies, 11 cents (total 76 cents). ] Patrons are requested to make all remittances payable to the University of Chicago Press in postal or express money orders or bank drafts. The following agents have been appointed and are authorized to quote the prices indicated: For the British Empire: The Cambridge University Press, Fetter Lane, London, E.C. Yearly subscriptions, including postage, 18s. 8d. each; single copies, including postage, 3s. 2d. each. For the continent of Europe: Karl W. Hiersemann, KGnigstrasse 29, Leipzig, Germany. Yearly subscriptions, including postage, M. 22.65 each; single copies, including postage, M. 3.80 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, including postage, Yen §.50 each. Claims for missing numbers should be made within the month following the regular month of publication. The publishers expect to supply missing numbers free only when they have been lost in transit. Business correspondence should be addressed to The University of Chicago Press, Chicago, Ill. Communications for the editors and manuscripts should be addressed to the Editors of THE JOURNAL OF GEOLOGY, the University of Chicago, Chicago, Ill. Contributors are requested to write scientific and proper names with particular care and 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 leading articles will be printed as ordered, of which 50 (without covers) will be furnished grafis, the actual cost of the remainder (and covers, if desired) to be paid for by the author. The actual cost may vary, and will depend upon the amount of work in re-making the pages into forms, presswork, paper, binding, etc. Separates containing half-tones may be expected to cost somewhat more, the increase depending upon the number of cuts and the amount of work required upon them. Entered as second-class matter, March 20, 1893, at the Post-office at Chicago, III., under the Act of March 3, 1879, VOLUME XXIV NUMBER 1 THE fOURNAL OF GEOLOGY JANUARY-FEBRUARY 1916 THE ACADIAN TRIASSIC SIDNEY POWERS Troy, New York CONTENTS PART I INTRODUCTION General Relations ’ General Geography and Geology of the Region Topography Geology THe NEWARK GROUP IN THE ACADIAN AREA DESCRIPTIVE GEOLOGY Stratigraphy Grand Manan Split Rock Quaco Martin Head Waterside Advocate Harbour Cape Sharp Partridge Island Greenhill—Five Islands Five Islands Vol. XXIV, No: 1 I 2 SIDNEY POWERS PART II Gerrish Mountain Gerrish Mountain—Truro Truro—Wolfville Wolfville-Scots Bay Scots Bay—Bennetts Bay Digby Gut < Rossway—Brter Island Age PART III STRUCTURE Folds Faults Theories of Origin IGNEOUS Rocks Distribution Five Islands Volcanics North Mountain Basalt ORIGIN PART I INTRODUCTION GENERAL RELATIONS The Acadian Triassic is preserved as a rather narrow, discon- tinuous border to the Bay of Fundy and Minas Basin. The Bay of Fundy has a northeast-southwest trend, with the island of Grand Manan at the entrance. The Triassic extends from the southwest end of Grand Manan to Truro, a distance of 195 miles. The width of the Bay of Fundy, from Digby to St. John, is 45 miles. On both the New Brunswick and the Nova Scotia shore, the Triassic appears: in New Brunswick on Grand Manan Island, at Split Rock (Gard- ner’s Creek), Quaco, Martin Head, and Waterside; in Nova Scotia on the island of Isle Haute, and in a quite continuous strip from Advocate around Minas Basin and down the Bay of Fundy to Brier Island at the entrance. The field work in connection with this paper occupied a portion of the summers of 1913 and 1914. In the field work, the writer is indebted for suggestions to Professor J. W. Goldthwait, Dr. THE ACADIAN TRIASSIC 3 A. O. Hayes, and Mr. W. A. Bell of the Geological Survey of Canada, and to Professor D. S. McIntosh of Halifax, Nova Scotia. Professor Alfred C. Lane of Tufts College has kindly permitted the use of thin sections and drill cores of the Cape D’Or basalt. To Professor R. A. Daly, under whose direction this paper was pre- pared as a portion of a thesis for the degree of Doctor of Philosophy in Harvard University, special thanks are due for helpful criticism. GENERAL GEOGRAPHY AND GEOLOGY OF THE REGION Topography—tThe most important topographic feature of the Bay of Fundy region is North Mountain, extending from Cape Blomidon to Brier Island, 125 miles. South of North Mountain is the Annapolis Valley and the land of Evangeline, a broad fertile plain extending from Minas Basin to the Annapolis Basin. South of the Annapolis Valley is South Mountain, whose crest stands on a level with North Mountain, at an elevation of about Aoo feet. On the northern side of the Bay is the island of Grand Manan, ' presenting an abrupt escarpment on the west, rising 200 to 4oo feet out of the sea. The tops of these basaltic cliffs is again at the level of North Mountain. On the east side of Grand Manan is a rolling lowland fronted by many islands. The New Brunswick shore is bounded by rocky cliffs rising to a height of 50 to 200 feet, but between the Triassic exposures at Quaco and Waterside, the clifis rise to the summit level of 400 feet. Minas Basin is surrounded by lowlands, presenting a rather flat surface at elevations of 100 to 150 feet, except for the tidal marshes. On the north, the Cobequid Mountains rise to heights of 500 to 800 feet, with “peaks”’ at 1,000 feet. Geology.—The controlling factor of the topography and geology of the region is the direction of the orographic axes, from north- east to southwest (Fig. 1). The Bay of Fundy is confined between a broad belt of pre-Cambrian rocks in Nova Scotia, fronted by the Triassic; and a less broad belt of pre-Cambrian rocks in New Bruns- wick, fronted by Carboniferous and Triassic strata (Fig. 2). The peninsula of Nova Scotia is composed largely of pre-Cambrian strata intruded by Devonian granite. 4 2 SIDNEY POWERS Minas Basin is bounded by Triassic rocks, below and around which are Carboniferous and Permian strata. To the north, separated by a normal fault, are the Cobequid Mountains, which are composed of Silurian schists and quartzites with various AEA ENG ONNECTICUT_VALLEY AREA ae POMPERAIUG AAR iN x y PS ee vi x << We Some y TRIASSIG AREAS sansounshusancns™ TANG 7 EASTERN NORTH AMERICA tee ° fea UW AREA Based on Maps hy ta Cit Anca RUSSELL AND HOBBS ie AR Wn ye Veep P SCALE is $0 o 400 200 360 mites, Aeproduced from Bull. 6, Connecticur Geological and Nawwraf History Survey with modifications by Sidney Fowers Frc. 1.—The Newark group in North America associated igneous rocks. North and west of the Cobequids are ' Carboniferous and Permian strata, stretching northward over the Magdalen Islands and over Prince Edward Island. THE NEWARK GROUP IN THE ACADIAN AREA The first reference to the Triassic in the Maritime Provinces is by Alger, in 1827.7 Six years later, a description of the Triassic «F. Alger, ‘““Notes on the Mineralogy of Nova Scotia,” Am. Jour. Sci., XII (1827), THE ACADIAN TRIASSIC rocks was published by Jackson and Alger.’ At about that time Gesner began his work in connection with the Geological Survey of New Brunswick and mentions the Triassic in each of his reports.’ From 1835 to the present, a number of writers have discussed the Acadian Triassic in part or in whole, and, as Russell has given a bibliography of these papers in his corre- lation paper on “The Newark System,’ only the more important and the recent papers need be noted here. Dana gave the name “Acadian Area”’ to this mass of Triassic rocks in his Manual of Geology (2d ed., 1875). Three years later, Dawson issued the second edition of his Acadian Geology, in which volume is the only account of the Triassic in Nova Scotia and New Brunswick. From 1863 to 1880, G. F. Matthew and L. W. Bailey were employed by the Geological Survey of Canada in mapping southern New Brunswick. In their reports are descriptions of the outliers of Triassic’ strata in New Brunswick. The more recent work on the Acadian Triassic is that published by L. W. Bailey on the Digby Neck region‘ and by H. Fletcher tC. T. Jackson and F. Alger, ‘““Remarks on the Mineralogy and Geology of Nova Scotia,’ Am. Acad., Mem., N.S., 1, 217-330. 2The reports are given in the bibliography of all literature on the Newark System, in I. C. Russell’s corre- lation paper on “The Newark System,” U.S. Geol. Surv., Bull. 85, 1892. 3 [bid. 4L. W. Bailey, “Report on the Geology of South- western Nova Scotia,” Geol. Surv. Canada, Ann. Rept., IX (1898), Part M. Fic. 2.—Geologic cross-section, AA, of the entrance to the Bay of Fundy, through Quoddy Head, Maine, on the northwest, the island of Grand Manan near by, and Long Island on the southeast. Diabase (Devonian ?); 7, Triassic. P, pre-Cambrian; B, Meguma series (pre-Cambrian); S, Silurian; D7, The Triassic shales under the basalt on Grand Manan appear only on the north. The Triassic sedi- ments south of Long Island are submerged. Note the longitudinal valley on Long Island between the flows. 6 SIDNEY POWERS who mapped the Minas Basin region" for the Geological Survey of Canada. The former extent of the Acadian area is indeterminate, but it is probable that it extended several miles in all directions beyond the present boundaries. The original form appears to have been a basin with its northern limit near the Cobequid Mountains and its southern limit not far south of Grand Manan and Brier Island. There may have been a ridge of older rock extending out into the basin where the eastern side of Grand Manan now is, provided that the interpretation of the geology of that island, as given below, is correct. : The Newark group in the Acadian area has been divided into the following formation: Thickness in Feet Top, an erosion surface Scots Bay formation (calcareous white sandstone)............ 25-(2,000 ?) North Mountain basalt (a succession of lava flows) .......... 800-— 1,000 Annapolis formation (red beds, largely calcareous) Blonudantshales 3 2h s..-27 500- 1,000 Wolfville sandstone........ 2,000— 2,500 3,325- 6,500 Base, an unconformity with Paleozoic or older rocks Interbedded in the Annapolis formation, near its top, are certain basalt flows: agglomerate and tuff beds near the Five Islands, grouped under the name Five Islands volcanics. At Quaco, New Brunswick, there is a conglomerate horizon in the center of the red sandstones correlated with the Annapolis formation, and this conglomerate is called the Quaco conglomerate. DESCRIPTIVE GEOLOGY The descriptive geology of the Acadian Triassic will be taken up by localities, giving a brief description of the lithological and structural details. A number of detailed maps and sections are introduced, which may be connected with the region as a whole, by reference to the general map (Fig. 3), and the columnar sections « H. Fletcher, various papers which have been printed in the annual and summary reports of the Geological Survey of Canada from 1887 to 1907. See especially the Annual Report, V (1892), Part P. THE ACADIAN TRIASSIC 7 (Fig. 4). The principal structural features and the details concerning the igne- ous rocks are summarized in a separate section. (TM sonra mounrarm ease EZ AAAnwaronis FORMATION [oJ SCQTS BAY FORMATION LECEND ACADIAN TRIASSIC STRATIGRAPHY Grand Manan.— The island of Grand Manan is situated 4 miles southwest of Quoddy Head, the most easterly point in the Uni- ted States. It is 15 miles long and 6 miles wide in the widest part. To the east and south are numer- ous islands and reefs. Grand Manan was first visited by a _ geologist— Abraham Gesner—in 1838, then by Bailey,’ Verrill,? and Ells. Bailey visited most of the adjoining islands and gives the best description of the geology. tL. W. Bailey, “The Physiog- raphy and Geology of the Island of Grand Manan,’ Can. Nat., VI (1872), 43 ff.; also see Bailey, Matthew, and Ells, “‘ Preliminary Report on the Geology of Southern New Brunswick,” Geol. Surv. Canada, Rept. Progress (1870), pp. 216-21. 2A. E. Verrill, Appendix E to Dawson’s Acadian Geology, 3d ed. (1878), pp. 679-80. 3R. W. Ells, Geol. Surv. [Pr sacr Canada, Summary Rept., VIII (1894), 271A. ~ a > 3S = S x : 3 z Fic. 3.—Map of the Acadian Triassic SIDNEY POWERS 4204 dG o20Nf) ISSVLL, ULIPRIY dy} JO SUIS AvULUNJOO—F “Oly 4794 o001 _—SSEaE——]ase 31VIS JISSWIYL NYIGVOV 3HL IO SNOILIAS YVNNNIO9 S21U02/0A spuvjs} aaly Spuvjs| aary 40,9 2409 Uouvpy pussy Aomssoy JISSVIVL- Jud auojspuog atAFIOM > 2 2 > v SS Ps a 3 > Es > Fa 3 2 2/04 vopiwojq LIVSVE NIVINNOW H1LYON NOILUWH OF AVG SL109S hog s102S ~9/[IA$[OM THE ACADIAN TRIASSIC SVaIv OISSVIT], ‘prop UAL pur ‘oovn?y “yooy qyds oy} jo de sap Ss SSS SS SSS SSI W—'S ‘oly DI8SVI ws MIOW LITIS \ @ ~-K Ws ie) SIDNEY POWERS The physiography and geology of Grand Manan divide the island into two provinces: a western upland underlain by Triassic basalt flows, and an eastern lowland underlain by pre-Silurian’ metamorphic rocks. The upland represents the level of the Summit? peneplain, at elevations of 200-400 feet. The western coast of the island is fronted by cliffs rising abruptly in an almost straight line to a height of 100-300 feet. The Triassic rocks of the island consist of basalt flows under- lain by purple shale. The shale is exposed to a thickness of only 50 feet, and this exposure is at Dark Harbour, on the western side of the island. The basalts rest directly on older, metamorphic rocks on the eastern side of the island; at Red Head, on the south, and at the northwestern end of Flag Cove, on the north. At the former locality the contact dips 35 degrees, suggesting a fault, but the recent weathering of the rocks near the contact obscures the exposure. The basalt flows are thick at the base and thin at the top. The exact thicknesses are given below with the résumé of the igneous rocks. The dip of the flows is variable. At Dark Harbour it is practically horizontal, but north of this place, the dip is down toward the north at angles of 5 to 15 degrees. It is difficult to determine the horizontal extent of any one flow. Diabase dikes are reported by Bailey at several places, one of which is Swallow-Tail Light. The faults on Grand Manan are obscured by the massive character of the basalt. The major fault bounds the west side of the island, and the cliffs, which run in an almost straight line for 15 miles, mark the fault-line scarp. East-west faults are less prominent. One may occur at Dark Harbour, as many of the streams flowing across the Newark basalts follow fault-lines. Minor faults are seen in the shore exposures. ji Split Rock.—Eighteen miles east of St. John, between Gardner’s Creek and Tynemouth Creek, there is a strip of Triassic sediments «The age of these rocks is probably pre-Cambrian, as they are older than the Silurian rocks of the Eastport Quadrangle. See E. S. Bastin, U.S. Geol. Survey, Geol. Atlas, Eastport Folio (No. 192) (1914), p. 14. 2 The term “Summit” peneplain is used instead of “Cretaceous”’ peneplain in order to avoid any reference to the age of this topographic feature. THE ACADIAN TRIASSIC II about two miles long, and three-quarters of a mile wide at Split Rock itself (Figs. 5 and 6). This point should not be confused with one southwest of St. John by the same name. This area is bounded on the north by a fault which has brought the Triassic into contact with the Carboniferous. The dip of the Fic. 6.—The Triassic shales near Gardner’s Creek (Split Rock), dipping north- ward. The marine shelf has been cut at high tide level. Triassic red sandstones, shales, and occasional conglomerates is in general northward at angles of about 45 degrees, but the beds flatten out at split rock itself. The sediments show occasional cross-bedding. The conglomerates contain only occasional pebbles, and these pebbles are subangular, with occasionally very angular, and rarely rounded surfaces. They do not show striations or sand-blasted surfaces. Fragments of silicified wood have been found near Gardner’s Creek, and are noted by Dawson in his Acadian Geology. 12 SIDNEY POWERS Quaco.—A large area of Triassic sediments occurs at Quaco, now known as St. Martin. The Triassic extends from West Quaco to Melvin’s Beach, and includes the large area mapped as Fic. 7.—Geologic cross-section, BB, from Quaco, on the north, through North Mountain, the Annapolis Valley, and South Mountain, on the south. 4A, pre- Cambrian; D, Devonian; Gr, Devonian granite; T, Triassic. Fic. 8.—The contact of the lower red sandstone and the Quaco conglomerate above, near Macomber Brook, Quaco. The contact is a conformity. There is a very sharp boundary both in lithology and in color which is obscured by a pile of talus in the center of the picture. ° Lower Carboniferous on the map of the Geological Survey of Canada, issued in 1880. The length of the area is 73 miles and the width 3 miles, as shown in Fig. 5. The shore exposures furnish an excellent structural cross-section of the area. THE ACADIAN TRIASSIC 13 The structure of the Quaco area is synclinal, with an east-west axis (Figs. 7 and 8). The contact of the Newark rocks with older rocks is shown at West Quaco, where there is an unconformity of red Triassic sandstone on greatly pointed, pre-Triassic traps. The basal sandstone contains occasional pebbles of various kinds of rock but contains no residual soil of the trap. At the uncon- formity there is minor cross-faulting in a northeasterly direction. At Melvin’s Beach, on the northeast, the Triassic sandstones are seen, with steeply dragged dips, in fault-contact with pre-Cambrian metamorphics. The stratigraphy of the area shows two normal red sandstone members, separated by a conglomerate of pale yellow color. Inter- bedded in the conglomerate are persistent beds of sandstone, a few inches in thickness, at stratigraphic distances of 10-30 feet, as shown in Fig. 8. The conglomerate is composed of rather loosely consolidated subangular to rounded stream gravels. Many of the pebbles show impressions of one pebble into another, and other recemented fractures. In no other locality of the Acadian Triassic is such a conglomerate found. The section was estimated as follows: Wppenmedisandstomencn Wit) ee a ccisn ee 800-1,000+ @uacorconglomeraten gy sae aualsles orale 450- 700 Wowermnedssandstonem ne ah, oye cnn ie) 300—- 300 T,550—-2,000 UNCONFORMITY WITH CARBONIFEROUS Plant remains occur at several horizons in the Quaco series. Silicified wood was found by the writer within 50 feet of the top of the lower red sandstone, at Vaughan Creek, and by members of the Geological Survey of Canada at other localities. These fossils are correlated with the fragments of lignite from Split Rock and Martin Head. On account of the exposure of the basal uncon- formity of the Newark at West Quaco, it is probable that the Quaco exposure is to be correlated with the Annapolis formation, t These conglomerates are similar to those of Upper Devonian age on the north side of Scaumenac Bay, Province of Quebec, described by J. M. Clarke, Bull. Geol. Soc. Am., XXVI (1015). 14 SIDNEY POWERS and that it is below the horizon of the North Mountain basalt, as shown in Fig. 4. Martin Head—Martin Head is 20 miles northeast of Quaco. The Head itself is composed of a mass of pre-Cambrian strata, too feet in height, connected to a point of land by a shingle beach (Fig. 9). On the northern side of the Head is some red clay, apparently of Pleistocene age, which may be underlain by Triassic sediments. North of the barrier beach are low cliffs of Triassic strata, exposed only on the west side of the peninsula. The bedrock is exposed for half a mile in the form of a syncline, with the longer Fic. 9.—Cross-section of the Martin Head Triassic area. Martin Head itself, composed of pre-Cambrian strata, is on the right. Between it and the pre-Cambrian uplands on the northwest are the Triassic sediments. The unconformable contact of the Triassic with the pre-Cambrian on the southeast is hypothetical. limb on the south. On the north, the Triassic is faulted against the pre-Cambrian, as shown in Fig. 9. In the southern limb 335 feet of sediments are exposed, in the northern limb 85 feet, and between these two limbs there are no exposures. A fault probably exists between the exposures, as the strata do not match on either side of the gap. The sediments in the Martin Head area are principally yellow sandstones and shales, with occasional pale-red beds, and transition colors. The yellow is a bright-chrome shade, much brighter than that of the Quaco conglomerate, which is merely that of the common stream gravels. Conglomeratic beds occasionally appear. The sediments are characterized by a notable amount of muscovite and of calcite. The former has evidently been derived from the pre-Cambrian mica schists on the north. Lignite occurs at several horizons in the yellow beds, as car- bonized twigs, limbs, and bits of wood, often 2—3 inches in diameter. This lignite has been described by Miss Ruth Holden,’ and the * Ruth Holden, “Fossil Plants from Eastern Canada,” Annals of Botany, XXVII (1913), 243-55. THE ACADIAN TRIASSIC 15 paper is summarized below in treating of the age of the Acadian Triassic. As the plant remains appear to be similar to those found at Quaco, the sediments in the two areas may be correlated with each other. Waterside —On the north side of Chignecto Bay is a strip of Triassic at Waterside, 20 miles northeast of Martin Head. The length of the strip is about 4 miles, ex- tending from Den- nis’ break-water on the west to the east- ern end of the marsh at Little Rocher on the east (Fig. 10), but the length of the actual exposures of Triassic sediments is about 14 miles. The structure of this area is anticlinal on the west and synclinal] on the east, with the axis of the syncline at the Waterside wharf. The dip of the folds is gentle, exposing only 320 feet on the eastern limit of the anticline. On the west, Fic. 1o.—Map of the Waterside area of Triassic sediments. T c sD C T 18 D =B \ Cn Fic. 11.—Geologic cross-section, CC, from Waterside, on the north, through Cape d’Or, across Minas Channel, and through North Mountain. A, pre-Cambrian, B, Meguma series (pre-Cambrian); SD, Cobequid group (Silurian, cut by Devonian ~ igneous rocks); D, Devonian; C, Carboniferous; 7, Triassic. The major fault of the region is shown in the center of the section, where the Carboniferous and Triassic are dropped down against the Cobequid group. the Triassic is faulted against a sheared Carboniferous conglomer- ate, but the contact is concealed by a Pleistocene delta deposit. This same fault probably bounds the Triassic area on the north and east, but it is not again exposed. SIDNEY POWERS 16 The Waterside Newark strata consist of pale-red sandstones with occasional conglomerates. At the top of the series there are voIv 10.P adej-o}eo0Apy oy} Jo depy—*z1 “OL See syjney Bjoy-a00g x NYIYNTS suosepuog [ ] NWINYATASNNIA “WMAoy syodouuy Res ANOGAD, sie) AIPSOD YW YL4°A/ ae ap,OIOAPY e IISSVIUL SN NN NN ee SS a \ S\ a4 b20004 \ SS SO QNIDI7I PHOT p2yig EN NEN to peculiar several calcitized sandstone beds, which weather out in Persistent thin green shale or sandstone beds are present. forms. THE ACADIAN TRIASSIC 17) Contemporaneous erosion channels are seen west of the Waterside wharf. The Waterside section is to be correlated with the Annapolis formation, but, as no plant remains have been found in it, a more definite correlation is impossible. Advocate Harbour—At the southeastern end of Minas Basin is Cape d’Or and north of it is Advocate Harbour. The shore from Cape d’Or to Cape Spencer (Figs. 11, 12) is fronted by basalt FARTRIOGE Is. LEGEND North Mt. Basalt Annapolis Form. TRIASSI | Fic. 13.—Map of the Cape Sharp-Partridge Island region cliffs too feet or more in height. The lowlands on the north are underlain by Triassic sandstone, and, farther north, by Carbonif- erous sediments. . North of the latter are the Cobequid Mountains, fronted by a fault-line scarp rising abruptly from the lowlands. The upland from Cape d’Or to Cape Spencer, and the islands, Isle Haute and Spencer Island, are composed of basalt flows dipping southward at a low angle. Five flows are exposed at Cape d’Or in drill-cores. The base of the flows rests on Newark shale and sandstone of a white or red color. Both the basal amygdaloid and the underlying sediments are penetrated by gypsum veins. North of Cape Spencer, very coarse conglomerates, which are probably of Newark age, are exposed. The bowlders in this con- glomerate are a foot or two in diameter. On the shore near West Advocate is the only other exposure of Newark sediments in this area. Red sandstone, with some red 18 SIDNEY POWERS shale, and thin greenish or white bands of sandstone, are dipping eastward. Adjoining these sandstones, on the north, is Silurian carbonaceous schist of the Cobequid group, separated by a fault which may be traced down the beach in a S. 71° W. direction. This is the main Cobequid fault, shown in Fig. 10. Other minor faults are shown in the same map. Ep Fic. 14.—Partridge Island from the west, showing basalt flows overlying red sandstones. The sandstones appear along the gentle slope at the left-hand side of the cliff. Cape Sharp.—To the east of Cape Spencer, Cape Sharp is the first promontory. It consists of basalt, as does Black Rock on the west (Fig. 13). On the north of Cape Sharp is a lowland underlain by red Triassic sandstone and shale with occasional green bands, and north of this is a rolling country underlain by Carbonif- erous sediments. The basalt on Cape Sharp consists of one or more flows— probably two flows—which dip to the south at an angle of about ug) THE ACADIAN TRIASSIC UOISII Spurs] OMT, 9y} Jo def{—'s1 “OI SapliWs ; ITVIS HOUOUIIOS syodouuy SS DISSYIUL Q SIIUOIY spug/sy aAly POAT SONV7SI OML nN SING UHOSSDM 20 SIDNEY POWERS ~\2) 5-10. The shale and sandstone underlying the basalt are exposed on the northeast side, for a short distance only. The sandstone Frc. 16.—Structure section, G, through Clarke’s Head, on the south, and Swan Creek, on the north. The structure of the pre-Triassic rocks isnot shown. Overlying the Triassic red sandstone is a bed of tuff, overlain, in turn, by agglomerate. At Swan Creek only agglomerate appears, but it 1s probably a part of the same flow. Fic. 17.—A detailed view of the agglomerate which comprises a large portion of the Five Islands volcanics. The photograph shows the center of a 1oo-foot bed of the agglomerate east of Blue Sack. The angular blocks are composed of basalt, and the matrix is tuffaceous. is seen to be in fault contact with the Carboniferous shales and sandstone to the north. Partridge Island——East of Cape Sharp is a peninsula called Partridge Island, formed of a mass of basalt connected with the THE ACADIAN TRIASSIC shore by a low, swampy area which is covered by the sea at very high tides. On the northwest is Gilbert’s Cliff, rising to a height of 60 feet, and on the northeast Parrsboro Pier (Fig. 13). The basalt on Partridge Island (Fig. 14) is partly columnar and partly vesicular. It probably consists of two flows. Stilbite is very abundant in geodes in the amygdaloid. At the base of the lower flow the highly weathered amygdaloid is 15 feet thick. The Triassic shales and sandstones are seen underlying the basalt on the west side of the island, dipping southward at an angle of.10 to 35 degrees. . Near Gilbert’s Cliff are red clays of Recent age, overlying the beveled, upturned edges of the ripple- marked Pennsylvania shales. The Triassic is not exposed to show whether this surface was the one on which the Triassic was deposited or whether it was the one made by the Pleistocene ice-sheet. Greenhill_-Five Islands—From a point 3 miles east of Parrsboro, near Greenhill, to Five Islands, there is an almost continuous strip of Triassic, faulted down against the Riversdale—Union series of Pennsylvanian age (Fig.15). The entire region has suffered extensive faulting and, with these move- ments, gypsum veins have been introduced. The Newark comprises red sandstones and some shales, with occasional beds of green sandstone or shale; tuff and ag- glomerate beds; and basalt flows. The dis- tribution of these rocks is very irregular, owing to the faulting, and is disturbed by extensive landslides in the volcanics. ( u E at r 243 PRWIZ s Hi? ire ve Aap Ra ae IS Fic, 18.—Structure section from Moose Creek toward Blue Sack, a distance of about 2 miles, showing the character of the On the extreme The height of the cliffs is 150-200 feet. right, a bed of agglomerate of the Five Islands volcanics is shown, overlying sandstone. folding and faulting in the sediments as seen in the shore exposures. 22 SIDNEY POWERS West of Clarke Head is an excellent exposure showing sandstones overlain by black ash, and this by agglomerate, the whole being cut off by a fault (Fig. 16). North of Clarke Head, agglomerate is exposed, and copper has been sought at this locality. East of Swan Creek, sandstone appears, overlain by a volcanic conglomerate—a mass of fragments 6 inches to 2 feet in diameter, of various kinds of basalt and agglomerate, imbedded in a red sand- stone matrix. Above this, with a gradual transition, comes a true agglomerate of angular blocks imbedded in a tuffaceous matrix (Fig. 17). Sandstone appears above the agglomerate, and the contact is locally cross-cutting. At Wasson’s Bluff, agglomerate appears above red sandstone, with a conformable contact. East of Wasson’s Bluff the agglom- erate rests unconformably on the sandstone. Another local uncon- formity is found at McKay Head, where the sediments are overlain by agglomerate with columnar basalt above. East of McKay Head are two remnants of the same basalt, with faulted contacts. The cross-cutting nature of some of the contacts may indicate volcanic vents. Two Islands, known also as The Brothers, consist of basalt flows dipping gently northwest. The islands are probably sepa- rated from each other and from the mainland by faults. East of Moose River the sediments reappear and extend from this point around Minas Basin, continuously as far as the Shube- nacadie River. The structure of the sandstones between Moose River and Five Islands may be seen in Fig. 18. Near Blue Sack (see Fig. 19) they are greatly slickensided, as shown in detail in Fig. 20. On the top of the cliffs east of Moose River, tuff, overlain by agglomerate, forms a capping for the sandstone. ‘The thickness of the volcanics varies, but is only too feet at a maximum. Near Blue Sack are two beds of agglomerate interbedded with sandstone, the lower being too feet or more, the upper 20 feet, with to feet of intervening sandstone. One of the contacts is cross-cutting, but there can be little doubt that the volcanics were formed contem- poraneously with the sandstone, as blocks of basalt occur in the latter. THE ACADIAN TRIASSIC swsof syodouuy [S| CH 8S09 WWW ‘gacd a1 sabwoong soom/y 97/1) Vit UOIZAI SPURS] OAT 94} Jo dvyy— Or “oy Saw S 24 SIDNEY POWERS This strip of Triassic is bounded on the north by a fault which probably continues east almost as far as Truro. The fault may be seen near the town of Two Islands, and between Moose Creek and Blue Sack the older rocks may be seen in one place at the top of the cliffs, in contact with the Triassic volcanics. Fic. 20.—A detailed view of the slickensides north of Long Island of the Five Islands. The polished surfaces strike at right angles to the beach and the movement has been horizontal. Little or no vertical displacement is shown, but the slicken- sided surfaces strike toward the Five Islands between each of which there is a fault. It is impossible to determine which are the major fault-planes. Five Islands.—The Five Islands are situated west of Gerrish Mountain. Their names, from east to west, are: Moose, Diamond, Long, Egg, and Pinnacle islands, and Pinnacle Peak (Fig. 21). Moose Island is nearly a mile in length and half a mile in width. The highest point on it is 350 feet above sea-level. Diamond Island is a small, round island, Long Island is one-quarter of a mile long and 180 feet in height at the center. Egg Island is smaller THE ACADIAN TRIASSIC and round. Pinnacle Island is a quarter of a mile long and 130 feet in height in the center. Pinnacle Peak is merely an ero- sion pinnacle. East of Egg Island at ebb tide is Egg Rock. At low tide, with a high run of tides, the sea-bottom between the islands and the mainland is left dry except for numerous deep river channels through the soft red clay. Moose Island consists of basalt flows on the north, under- lain by red sandstones on the south. On the west end there is a fault between red amygda- loid and the sandstone, with a number of gypsum veins near the contact. On the east side the amygdaloid and basalt above dip north at angles of 45° above the sandstone. At the base of the amygdaloid is greenish- white ash, 2 feet thick, similar to that at Gerrish Mountain. Fletcher's mapping of Moose Island and of the other Islands is largely incorrect. Diamond Island consists of a portion of the same basalt flows as on Moose Island and the other islands. The dip of the basalt is about 40° north- east. ; Long Island consists of basalt on the north and sandstone on On the left is Gerrish Mountain. Egg, and Pinnacle islands, and Pinnacle Peak. Complex block-faulting separates these islands. The Five Islands are, from left to right: Moose, Diamond, Long, Fic. 21.—The Five Islands. iS) ial 26 SIDNEY POWERS the south, with the contact striking nearly east and west. The dip of the sandstone and basalt is variable owing to minor folds, but is in general in a northerly direction about 20°. Egg Island and Egg Rock on the east consist wholly of red sandstone, dipping northwestward about 15°. Pinnacle Island consists of red sandstone on the south and basalt on the north, dipping northwest at angles of 20°-40°._ Pinnacle Peak consists of basalt. Each of the islands is separated from the others by a fault, and they are probably bounded on the north by a continuation of the Gerrish Mountain fault. On each island the flows or sand- stones dip northward, but at different angles. The basalt flows of the Five Islands were undoubtedly originally connected with the flow on Gerrish Mountain and with those on the Two Islands. They do not, however, appear to be directly con- nected with the agglomerate and tuff which are exposed along the shore. Probably the dike in Gerrish Mountain was the source of most of the igneous material, part of which flowed out, and part of which was blown out.. The relative age of the pyroclastic material and the flows could not be determined. [To be continued] AVERAGE REGIONAL SLOPE, A CRITERION FOR THE SUBDIVISION OF OLD EROSION SURFACES" LEOPOLD REINECKE Geological Survey of Canada CONTENTS INTRODUCTION MetuHop oF MEASURING REGIONAL SLOPES VALUE OF THE MEASUREMENT OF REGIONAL SLOPES a) In Determining the Agencies Which Have Formed the Surface 6) In Separating Forms Due to Different Erosion Cycles c) In Furnishing Accurate Data for the Quantitative Measurement of Earth Movements Value of Certain Criteria for Peneplanation PROPOSED SUBDIVISION GENETIC SIGNIFICANCE OF REGIONAL SLOPES OBJECTIONS TO THE SUBDIVISION a) The Subdivision Is Arbitrary b) An Accurate Average of the Regional Slopes Is Not Easily Obtained c) It Entails Added Field and Office Work SUMMARY INTRODUCTION During the four field seasons from 1908 to 1911 the writer was engaged in topographic and geologic work in the southern part of the Interior Plateaus of British Columbia. Certain questions which arose in the study of the physiography of that region are discussed in this paper. Information regarding the physiography was acquired from a study of the Tulameen and Beaverdell map areas at the Southern end of the Plateaus, of the Kamloops and Shuswap? map sheets covering 9,000 square miles to the north of them, and from the t Published by the permission of the Director of the Geological Survey of Canada. 2 The geological work upon the Tulameen map area was done by C. Camsell, and upon the Kamloops and Shuswap areas by G. M. Dawson of the Geological Sur- vey of Canada. Explorations in the country between these areas have been made by the same men. 28 LEOPOLD REINECKE descriptions of the region lying between. The locations of the areas in question and of the Interior Plateaus are shown on the index map, Fig. t. No detailed work has been done to the north of the Kamloops area, but explorations have indicated that the 000 7, \Hudsone Hope <2 A Kamloops map area B Shuswap 4 ” C Tulameen « D Sesverde// . 48 Boundary of the | Sr eS, Plateau 128 25 Scale yoo Miles Fic. 1.—Index map character of the surface resembles that farther south although the general elevation is said to be less. The southern part of the Plateaus consists of an old erosion surface or upland, dissected by younger deep valleys (Figs. 2, 3, and 4). The region appears to have been overridden by a conti- Weird Creek East- west pra Ferroux Me Arlington Mt N°2 EFast-west profile | Wallace Mountain N°o3 North-south profile Nipple Mt N2S fast-west 2 2 = S 8 o ~ 5S Tulameen River N°6 East-west S 22 POS 2 es ss00' (55 frexetairexetenes 800° Beier ey 200° ——| 3600" 3000'. North-east profile or VD ei Triassic Complex JSurass, (mainly andesite Javes and wt¥s) and & Fic. 2.—Character CHINA noon - Re ron Raeree® Net East. west orofive scross mnyves in the Beaverdel! me arca, Uhistentings regional and local slopes By JOM nook UTTMM re ”) > Mall Cr Arhegtan we East. west erotife across muyos 02 the Beavenyel! mup ares, lustratag regional slopes towards large stream valleys Matiace Mewrtarn Corry me ea enathiition ats i“ Sey DP er Nortihseet) profile ao the Seavervie¥ map eared, iiustnating slopes along ridges and, on Wallace Mountain, accordance between topography and structure King Solomon Mauntarn Curry Mountain Crystal Mountain West form River 10 the Beaverdell map ares iHustratng variations in the elevations of interstream blocks “ i > 2 Fatlx Crock 3 DISS ri Stock: focene Batholith (pa) Po f Oligocene and River Allawium Post haves Fic. 2.—Character of the surface in the Interior Plateaus of British Columbia. Horizontal scale: 1 in.=about 2? miles AVERAGE REGIONAL SLOPE 20 ‘“nental ice sheet which removed the soil covering from the upland, carved a few shallow rock basins, and left a thin irregular blanket of drift upon its retreat, but which does not appear to have modified the upland slopes in any essential manner. This old upland sur- face has all the essential characters which are commonly used as criteria to distinguish peneplains, but the average slopes on it, measured from the higher areas or ridges toward main drainage lines, vary from 150 to over 300 feet to the mile, and these slopes Fic. 3.—View of a portion of the Interior Plateaus near the Tulameen Valley are found on the surface as a whole, not merely on isolated portions Giiestite If the slopes on this land form had not been measured, the dominant discordance of topography and structure and the general evenness of sky line would have caused it to be classed as a peneplain. The degree of slope present, however, led to a search for indications that would point to the fact that in this instance the stage of old age had just begun. It was found that the drainage system upon the upland was apparently related to a system of shear zones in the underlying rocks, and that in some of the areas underlain by certain Tertiary sediments and lavas, topographic form was governed by 30 LEOPOLD REINECKE geological structure. The accordance of topography and structure might in this case be described as a dimmed and accessory char- acter, while the discordance was sharply defined and extensively developed, an essential character. It was found that on an old surface, with slopes of 3 to 6 per cent, the characters developed in maturity were just disappearing and those related to peneplanation were strongly developed but had not yet entirely supplanted the others. According to the hypothesis of the geographic cycle this land form, if it had been left undisturbed, would have been gradually worn down; that is, its average slopes would have been gradually diminished, the characters of maturity have gradually disappeared, and the characters of peneplanation have prevailed everywhere. It seemed possible then, by using the criterion of regional slopes, to subdivide old land forms on a quantitative basis. The primary object of this paper is to point out the importance of the measurement of average regional slopes upon “‘old erosion surfaces,” and to show that such data assist materially in the more accurate study of the physiographic development of the region in which these surfaces occur, and of the diastrophic movements which have taken place there. The writer believes that it will be possible eventually to subdivide old land surfaces on the basis of their average slopes, and has attempted to do so here. The sub- division proposed is necessarily imperfect, partly because of the lack of accurate data regarding the slopes of old erosion surfaces, and largely because of the writer’s imperfect knowledge of the literature describing such surfaces. As more data upon the slopes of old surfaces become available, however, the imperfections of a subdivision of this kind can be remedied. METHOD OF MEASURING REGIONAL SLOPES By regional slope is meant the general slope of the land toward main drainage lines. Slope is stated here as the percentage of vertical to horizontal distance rather than as an angle, because the measurement of the angles of slope on a land form of moderate relief is generally impracticable in the field, and for that reason the degree of slope stated as an angle, especially if the angle is small, does not 31 AVERAGE REGIONAL SLOPE JOATY UsIUEIN,T, oy} JO AoypeA oy7 Ssor9e qsvoy}IOU SuIxOo] snvoye]q JOU] 94} JO Mor A—? ‘Ol 32 "LEOPOLD REINECKE carry a suggestion of the actual land form to the mind of the reader. For the measurement of such slopes either topographic maps or a number of traverses across the region to be examined are essential. An appreciation of the significance of the slopes can, however, be attained only by traveling over them. The following is an outline of the methods followed in obtaining data from topographic maps of portions of the southern section of the Interior Plateaus of British Columbia. These methods were applied, in part, to the measuring of slopes upon topographic maps of certain sections of the United States, with satisfactory results. The first step to be taken is the drawing of a number ot profiles, some in the direction of the main drainage and others at right angles to it. The profiles should include as many of the pertinent topographic features of the region as possible. If they are plotted with a vertical scale somewhat larger than the horizontal, they will assist both in determining whether the land form under considera- tion is the result of one or more cycles of erosion, in discovering whether processes other than subaérial erosion have been respon- sible for existent forms, and in detecting tilting or warping of the crust subsequent to the formation of the surface. The usefulness of profiles is discussed more fully farther on. If forms due to more than one cycle are present, slopes should of course be measured on each of those forms separately. In the Interior Plateaus two cycles are represented, an older upland and younger valleys intrenched in it. There is a distinct break or topographic unconformity between the upland and the valley forms (Fig. 2, profiles 1 to 7). In this instance the measurements of slopes on the older land form, the upland, were made by taking a large number of horizontal measurements on the topographic map from a dominant ridge line to the bottom of a large upland valley. If a deep valley of the younger cycle occupied the site of the bottom of the old valley, the measurements were made to the point where the break in slope occurred between the old and the young forms (profiles 1 and 2). Horizontal measurements were made as long as possible, and never less than one mile. The vertical difference was read directly from the topographic map. Measurements were taken in both directions AVERAGE REGIONAL SLOPE 33 - at right angles to the trend of the ridge, and also along its crest. Variations in the slopes along ridge crests in the Beaverdell map area of southern British Columbia are illustrated in Fig. 2, profiles 3and 4. The average slope measured on four or five ridges in this area lay between 100 and 300 feet to the mile, and averaged over 200 feet except in places where certain Tertiary formations occurred over which the slopes ranged between 500 and goo feet to the mile, and averaged 600. The Tertiary areas occupied less than one- eighth of the area of the whole upland; the average slope along ridges therefore averaged between 200 and 300 feet to the mile. Slopes across ridges were in this area of very nearly the same magnitude and did not average over 300 feet to the mile. Slopes as high as goo to 1,000 feet to the mile were found in a few places only, and could have been omitted from the general average without changing the result to any great extent. Such local irregularities of slope are more likely to occur in land forms with fairly high slopes than in those which are of a plainlike char- acter. Of twelve measurements on the Caldwell, Kansas, map sheet for instance, six lay between 14 and 21 feet to the mile, four between 32 and 35 feet, one was 47 feet, and one ro feet to the mile. VALUE OF THE MEASUREMENT OF REGIONAL SLOPES The study and determination of the regional slopes upon old erosion surfaces is both useful and necessary. It is useful: (@) in helping to determine the agencies which have carved and molded the topography to its present form, and (0) in separating forms due to different erosion cycles. It becomes necessary (c) when an old erosion surface is to be used as a datum for measuring diastrophic movements. : a) The study of regional slopes often will indicate the agencies which have carved or assisted in carving a land form. This is illustrated by Barrell’st work along the New England coast. He found that certain flat-topped ridges in the interior sloped gently toward the coast, and the plainlike surfaces, of which the ridge tops were residuals, occurred in terraces of successively higher tJoseph Barrell, ‘‘Piedmont Terraces of the Northern Appalachians,” Bull. Geol. Soc. Am., XXIV, No. 4 (December, 1913), 688-90. 34 LEOPOLD REINECKE elevations, each being separated from the next by a shorter and steeper slope. Further study proved that a number of flat hill- tops in this region, which have for a long time been regarded as residuals of a tilted peneplain formed by subaérial erosion, were in reality parts of a series of wave-cut marine terraces. b) The measurement of slopes and’ the study of profiles were found very useful in determining the number of cycles of erosion through which the uplands of the Interior Plateaus of British Columbia had passed. The older uplands and younger valley cycles were separated with comparative ease, but detailed study of the slopes was needed to show that no remnants of an older plain- like surface existed within the upland itself. c) The measurement and recording of regional slopes will be of the greatest value, however, in cases where old surfaces and their residuals are used to determine the manner in which earth move- ments have taken place or the amount through which a section of the crust has moved. The manner in which movements of the crust have taken place sometimes can be brought out by profiles (see Fig. 2, profile 6), but it is necessary to determine the original internal relief and average slope of an elevated or warped old erosion surface before such a surface can be used for quantitative measurements of movements of the earth’s crust. In the uplands of the Interior Plateaus, for instance, the relative relief of points within 10 miles of each other quite commonly is from 1,500 to 2,000 feet. If such a surface be uplifted and dis- sected until only remnants of it remain, the difference of elevation between them could be 1,500 feet without the section of the crust within which they occur having been either warped or tilted. Cal- culations of earth movements based on the assumption that such a surface was plainlike before uplift would be liable to errors of 1,500 feet or more. If the slopes are not measured, however, old surfaces of marked relief are likely to be thought nearly flat or of much lower relief than actually is the case. For instance in describing an old erosion surface in the Colorado Front Range, Davis' says: ‘In the highland west of Palmer Lake, between Denver and Colorado Springs the sky line seems to be essentially tW. M. Davis, ‘‘The Colorado Front Range,” Ann. Assoc. Am. Geog., I, 42. AVERAGE REGIONAL SLOPE 35 ~ level; much more so in the actual view than would be inferred from the crowded contours of the Platte Canyon map sheet.” Value of certain criteria of peneplanation—The reasons for mis- takes of the kind referred to are, first, that an uplifted old erosion surface of moderate relief is often seen in juxtaposition to younger topographic forms upon which the slopes are much steeper, so that by contrast the relief on the older surface appears much less than it really is. A second and less obvious reason is that certain of the more important characteristics of plainlike erosion surfaces with average slopes of less than 10 feet to the mile are found also on old erosion surfaces with slopes as high as 300 feet to the mile. The criteria referred to are a general flatness of sky line and the planing of a rather flat topographic surface across rocks of different hardness and texture without any apparent change in the character of the topography. Flat sky lines: In a rolling hill country ridge lines sloping from 100 to 300 feet to the mile may appear quite flat and the complete- ness of the illusion will depend partly on the position of the observer and partly on the distance of the ridge line or lines from him. Flat sky lines often are caused by the blending of more than one ridge line of entirely different elevation in the observer’s line of sight, the irregularities of one being neutralized by the other (Figs. 3 and 4). The writer knows of at least one locality in the Beaverdell map area of British Columbia where an observer, climbing up one side of the West Fork River valley and looking across to the opposite side, would see first a flat sky line on a ridge with an elevation of 4,000 feet, and as he climbed higher another flat ridge line would come into view with an elevation approximately 700 feet higher and lying 3 miles farther away. The two ridges are shown in cross- section in Fig. 2, profile 2, the flat top of the higher, the St. John ridge, in profile 3. Between the two positions there is doubtlessly one where both ridge lines would blend and appear as one. The flat sky line in this instance evidently does not mean that the ridge tops represented in the sky line are remnants of one nearly flat plain, for the lower one is next to a large river, the higher 3 miles from it, and the slope between them over 200 feet to the mile. 36 LEOPOLD REINECKE Nor can the lower and flatter of the two ridges be considered a peneplain remnant. Both ridges are, in fact, part of one surface in the stage of early old age, a surface with average slopes of about 23 per cent. Their nearly flat surface is doubtlessly due to their lying between nearly parallel drainage lines. Measurements made along apparently flat ridges, moreover, often show that they slope at a quite appreciable rate. The slopes upon St. John ridge, one of the ridges referred to in the preceding paragraph (Fig. 2, profile 3), vary from 100 to 300 feet to the mile. In Fig. 4, an apparently flat sky line is shown between the points a and 6 which are about 73 and 6 miles respectively, from the camera. From the photographic work done at this place it is known that a vertical shift of ;},5 of an inch in the sky line of the picture represents an actual fall of 90 feet in the topography, and that between a and 0 there is a broad upland draw which is 250 feet deep, and whose sides slope at the rate of 100 feet to the mile. If the sky line in Fig. 4 were farther away, it would, without doubt, appear much flatter. In the clear western air, ridges 20 miles ~ away often are plainly visible. Discordance of structure and topography: Discordance of topography and structure must also not be considered a final proof that the land form being examined is at all plainlike. Relatively flat surfaces planing across the contacts of rocks of different hard- ness are quite common in the Interior Plateaus, but sloping surfaces which plane across the structure are much more common. The flat areas are local developments on the rolling-hill type of Interior Plateau topography. In one instance a flat surface was seen planing across a centroclinal basin of relatively soft rocks which were pro- tected on the outside by hard layers. The flat surface is shown in Fig. 2, profile 6, just east of the point marked Hamilton Hill, and a part of the same surface in the foreground of Fig. 4. This is undoubtedly a case of local base-leveling and not a proof of uni- versal peneplanation. In another locality a flat ridge top lying next to a large river at an elevation of 3,800 feet was found planing across the structure. The ridge, a part of King Solomon Mountain in the Beaverdell area, is shown in cross-section in Fig. 2, profile 4, but the change of structure is not shown in the profile. This ridge AVERAGE REGIONAL SLOPE 37 top represents the lowest part of the upland within an area of several hundred square miles and within 10 miles of it there are numerous ridges from 1,000 to 2,000 feet higher. The flat surface is a small but integral part, not of a plainlike, but a decidedly hilly, land form. Discordance between topography and structure is, moreover, as well developed on the sloping hillsides of that land form as on the few flat surfaces that are present within it. Neither approximately even sky lines, nor flat or nearly flat areas planing across the structure, are therefore in themselves a proof that the land form within which they occur is of more moder- ate relief than the upland of Interior Plateaus with average slopes as high as 6 per cent. The measurement of the slopes on old erosion surfaces must therefore be made before one can venture to judge of its actual relief or use it in quantitative measurements of earth warping. PROPOSED SUBDIVISION The following subdivision is concerned only with the stage of old age in the normal cycle of erosion as outlined by Davis. An old erosion surface is for the purposes of this discussion defined as a geographic unit worn down by subaérial processes alone to a state of moderate relief. By geographic unit is meant | a portion of the earth’s surface over which topographic conditions and the underlying rock structure were essentially similar at the beginning of the erosion cycle, and over which conditions of erosion remained essentially the same while the cycle was in progress. It is proposed to treat all surfaces in this stage as varying from two types, those of plainlike forms of peneplains and forms corre- sponding in general features to the uplands of the Interior Plateaus of British Columbia which may be referred to as “‘beveled hills.” Following Smith? and Davis’ peneplains are defined as geographic units worn down by subaérial processes alone to a condition of very moderate relief. The theory of the formation of such plainlike land forms does not necessarily imply that all parts of them lay tW. M. Davis, ‘The Geographic Cycle,”’ Geog. Jour., XIV (18099), 481. 2W.S. Tangier Smith, “Some Aspects of Erosion in Relation to the Theory of the Peneplain,”’ Univ. of California Bull. Dept. of Geol., II (1899), 155-77. 3 W. M. Davis, “The Geographic Cycle,” Geog. Jour., XIV (1899), 486. 38 LEOPOLD REINECKE near the ocean at the time of their formation. If the geographic unit be large, parts of it must lie far from the ocean and may be at a considerable elevation above it.‘ The ‘Almost plains”’ are char- acterized as presenting absolute discordance between topography and structure, graded streams and hill slopes, that is, practically a lack of cliff and local flat surfaces, and by deep soil covering. These are the commonly accepted criteria for determining pene- planation. In addition, we suggest that the term be restricted to surfaces with average slopes of less than 2 per cent, 105 feet to the mile. For examples of peneplains one may cite the Laurentian peneplain? of Canada, and a peneplain in the Mississippi Valley illustrated by the Caldwell, Kansas, topographic map sheet. The Laurentian peneplain has an area of about two million square miles, with average slopes of about one-tenth of 1 per cent. It differs from an ideal type in that it has been modified by the accident of glaciation in removing the residual soil, in substituting an irregular drift mantle, and in slightly altering the form of the original surface. Upon the Caldwell area, average slopes vary from 10 to 50 feet to the mile, that is from one-fifth of 1 per cent to Ze per Cent: “Beveled hills’ are characterized as geographic units worn down to moderate relief by subaérial processes alone. Their ‘“‘essential’”’ characters are discordance of topography and structure, graded slopes, smooth sky lines and contours, and a deep soil cover- ing. ‘Their ‘‘accessory”’ characters are local accordance between topography and structure, and the local occurrence of cliff faces and flat areas, that is, of ungraded slopes. In this instance the terms ‘‘essential’’ and “accessory”? are used in the same way as they are in petrography, essential characters being those which predominate within the land form, accessory those of which but few examples can be found and which may be entirely absent. “‘Peneplains” and ‘‘beveled hills” are distinguished therefore by their degree of slope and also by the occasional finding in the «W. M. Davis, “The Colorado Front Range,” Aun. Assoc. Am. Geog., I, 42. 2 A.W. G. Wilson, “The Laurentian Peneplain,” Jour. Geol., XI (1903), 628-29. 3Henry Gannett, “Topographic Atlas of the United States. Physiographic Types,” U.S.G.S., Folio 1, Caldwell, Kansas, sheet, 1898. AVERAGE REGIONAL SLOPE 39 “beveled hills” type of characters which are characteristic of the stage of maturity. It is suggested that the upland portion of the Interior Plateaus of British Columbia be taken as a type of the “beveled hills” form, and that the term therefore be restricted to old erosion surfaces upon which the average regional slopes are from 3 to 6 per cent. The upland of the Plateaus differs from an ideal type in that gla- ciation has removed the soil covering and substituted an irregular mantle of drift. The term “‘beveling’’ was introduced into physiographic litera- ture by Tarr. He applied the term to the process of the cutting down of certain of the peaks and ridges on a land form, by differ- ential erosion, to approximately uniform elevations. According to Murray’s New Dictionary one of the meanings of “‘to bevel”’ is “to reduce (a square edge) to a more obtuse angle.”’ As used in this paper the adjective ‘‘beveled” is meant to suggest that the land form so designated has been reduced to one on which nearly uniform sky lines are a common characteristic, and one upon which ridge tops have broadened or become rounded in cross-section; that is, the angles which ridge sides make at their crests have been increased to obtuse angles. ‘‘Hills” are meant to suggest that the land form is composed of numerous eminences of moderate relief and of smooth and rounded contours. Except in so far as it sug- gests reduction from a higher and more rugged form, the term “beveled hills” is intended to be descriptive, and is not meant to suggest the agencies by which reduction was effected. GENETIC SIGNIFICANCE OF REGIONAL SLOPES The desirability of a subdivision of this kind is that it will stimulate the gathering of data on the slopes of old erosion sur- faces, and that it places definite limits on the term “peneplain.”’ The value of such data and of such a restriction have been referred to before. An added argument in its favor is that the subdivision is based upon a factor which is of genetic significance in the develop- ment of land forms. For the slopes of a topographic form are not only one of the results of its development, but the amount of slope tR.S. Tarr, “The Peneplain,” Am. Geol., XXI (January—June, 1898), 351-70. 40 LEOPOLD REINECKE present is also a factor in the rate of its further development. Moreover, the rate of development decreases so rapidly with decrease of slope that ‘‘beveled hills” are probably chronologically closer to forms in early maturity than to peneplains. It is proposed in the following section to give proofs for the hypothesis that the rate at which a land surface progresses through the geographic cycle is dependent on its average regional slope, and that its progress becomes slower as the slopes become less. ‘This hypothesis has of course been accepted by physiographers* for a long time, and is discussed only because of the emphasis placed in this paper on ‘“‘average regional slope’’ and because the writer has found no presentation of evidence to prove this hypothesis. The products of erosion in the normal geographic cycle are practically all removed from the land by streams. The rock waste is moved downstream partly as débris and partly in solution, and, if one could compare the amount of load carried by the streams on any land form during two stages of its progress, when average slopes were known, a measure would be furnished of comparative changes in the rate of erosion as the geographic cycle progresses toward old age. The load consists of débris dragged along the stream bed, débris carried in suspension, and rock matter carried in solution, each of which will be considered in the order named. A series of experiments have been made by Gilbert? on the rela- tions between the load of débris that a stream can drag along its bed, and its slope. The experiments proved that the quantity of load dragged by a stream varies in a complex manner with a set of controlling factors—such as slope of stream bed, discharge of water per second, fineness of débris, and form of stream channel. The changes, in amounts carried, vary at a different rate for each of the factors con- cerned. Under the conditions of the laboratory, the load dragged along the stream bed varied with the slope, but at a greater rate. tW. M. Davis, “The Peneplain,’ Am. Geol., XXIII (January-June, 1899), R.S. Tarr, “The Peneplain,” ibid., XXI (January-June, 1898), 354, 365. 2G. K. Gilbert, ““The Transportation of Débris by Running Water,” U.S.G.S., Professional Paper No. 86, pp. 10-54, 120, 121. AVERAGE REGIONAL SLOPE | 4l If, for instance, the slope, expressed in percentage of fall to hori- zontal distance, was doubled, the load dragged was, in the experi- ments, increased three to more than tenfold. Conversely as the slope decreased, the load decreased, but at a greater rate. The load carried in suspension is partly a function of the stream’s velocity, and depends partly upon the fineness and amount of débris supplied. In experiments made on streams without load, the velocity was found to vary approximately as the 0.3 power of the slope, and the 0.25 power of the discharge." The size of pebbles which can be carried in suspension varies as the fifth power of the velocity; that is, if velocity were unaffected by the addition of débris, it would vary approximately as the 3 power of the slope. Velocity is diminished by suspended matter, but not enough to make the factor of 3 less than unity. It is prob- able that in the majority of cases the grading of débris supplied to a stream is such that, if the slope be increased, the maximum load of suspended material carried by a stream will increase at a rate comparable to the rate of increase of the size of débris carried; that is, it will increase at a slightly greater rate than the increase ‘in slope. Conversely if the slope be decreased, the maximum load carried in suspension will be decreased but at a greater rate than the slopes. If discharge and fineness of débris supplied remain the same, therefore, both the maximum load dragged along a stream bed and that carried in suspension decrease at a greater rate than the slope, and the difference in the rate of decrease of the two functions becomes greater as the slopes decrease. ‘This law of variation is applicable to natural streams as well as to those in the laboratory.’ But changes in discharge and fineness of débris as old age pro- gresses, both tend to reduce further the load carried. For the rain- fall on a land form, the size of a geographic unit is likely to decrease as the land becomes lower, and the proportion of runoff to rainfall will also decrease so that the discharge of the streams would decrease. The débris supplied to the streams, moreover, becomes tIbid., 225. Discharge is defined as the number of cubic feet of water passing a given point per second. 2 Gilbert, op. cit., p. 233. 42 LEOPOLD REINECKE finer with old age and its increasing depth of soil. But suspended matter added to a stream retards its velocity, and the rate of retardation becomes greater as the débris becomes finer.‘ Hence as the slopes became lower, both the factors of decreased discharge and increased fineness of débris would help to a further and more rapid rate of decrease of maximum load carried. Obviously also the amount of creep and wash of débris down hillsides into the stream beds is smaller on gentle than on high slopes. The rate at which a land form is worn down by mechanical erosion must therefore diminish very rapidly as the slopes decrease. On the other hand, lower slopes may aid chemical erosion, in that more rain water is absorbed and the solution of the rock materials near the surface is increased. Chemical erosion must, however, be a very small factor in the wearing down of a land surface, for the matter dissolved in river waters is derived from surface rocks in the process of weathering, and the greater bulk of the rocks at the sur- face lose on an average less than one-third of their original weight by the process of solution when weathering is complete. Moreover, part or all of this loss is compensated for, both in weight and in bulk, by gains in the form of oxygen, water, and carbon dioxide obtained from the atmosphere, and recombined as insoluble mineral products in the residual soil.’ If the average slopes of a land surface, therefore, be reduced in the progress of the geographic cycle from, say, 4 per cent to 2, the rate of reduction of the land surface by erosion will be less than one-half what it was before, and as the slopes decrease, the process becomes slower and slower. The final stages of old age in which the surface is reduced to slopes as low as one-tenth of 1 per cent must therefore represent a very much longer period of time than the stage of maturity or of early old age. Chronologically, therefore, ‘beveled hills” are probably closer to land forms in early maturity than to ‘“pene- plains,’”’ and for that reason alone the subdivision proposed in this paper should be justified. t Gilbert, op. cit., p. 228. 2F. W. Clarke, ““The Data of Geochemistry,” Bull. 491, U.S.G.S., pp. 462 and 465. AVERAGE REGIONAL SLOPE 43 OBJECTIONS TO A SUBDIVISION BY AVERAGE REGIONAL SLOPES The objections which may be urged to a subdivision of this kind are: (a) that it is an arbitrary one; (6) that the slopes on a land form vary widely, and two observers may come to different conclusions regarding it; (c) that it requires a greater amount of detailed field and office work than is necessary when the average slopes. are not taken account of. a) The subdivision is arbitrary, for as far as we know there is no distinct change in the cycle at either of the two limiting points of 2 and 3 per cent, which is placed on the two type forms proposed here. Moreover, it is probable that there are old erosion surfaces to represent all stages of average regional slope from 6 per cent to less than one-tenth of 1 per cent, and there may be as many examples lying between the two types proposed as within the limits of the peneplain type. A parallel might be drawn between the classification of igneous rocks and the subdivision proposed here. In 1886 or thereabouts, when geologists of the United States Geological Survey found large areas in the Sierra Nevada Mountains underlain by intrusive masses of approximately similar composition lying between the quartz-diorite and granite families, they suggested the name granodiorite for them. ‘Thirteen years later Lindgren’ proposed definite limits for the “granodiorite”’ family in regard to both its chemical and mineralogical composition. The bulk of the rock masses referred to in the Sierras, and later found to occupy great areas in Canada, fall within the limits proposed by him. His quantitative restriction of the term granodiorite is therefore justi- fied, for it represents a natural group of rocks. The term gains stability, moreover, because of the occurrence of this group within a definite and accessible region. This definition of granodiorite is of great value to petrographers because it furnishes a clear-cut standard of comparison and datum point within the scheme of classification. The occurrence of rocks intermediate in composition between granodiorite and granite on the one hand, or granodiorite and quartz-diorite on the other, does t Waldemar Lindgren, “‘Granodiorite and Other Intermediate Rocks,” Am. Jour. Sct., IX (1900), 269-82. 44 LEOPOLD REINECKE not detract from the usefulness of the quantitative definition of the family, but rather adds to the necessity of such a definition. In the same way the term ‘“‘beveled hills” proposed here represents a land form which actually occurs over the known portion of a large geographic unit, the Interior Plateaus of British Columbia, and the quantitative limits proposed for the type are those measured upon the land form in question. The value of the establishment of a subdivision centering about a quantitatively defined physiographic type should also not be seriously impaired by the occurrence of numerous intermediate forms. b) The slopes on a land form vary widely, and an accurate average is not easily obtained. Variation in slope will cause trouble only in old surfaces of fairly high relief, that is, in the “beveled hills” type. In the work in the Interior Plateaus, it was found that the greater part of the surface within areas of about 200 square miles lay between 3 and 6 per cent, and where slopes of a mile or more in length varied greatly from the general average they occurred over small areas. By estimating the relation of the size of these areas to the whole, irregularities of slope were cal- culated into the whole, and found to change the general average very slightly. If care is taken first carefully to separate forms due to different cycles, and then to note the frequency of the occurrence of irregularities varying from the average, the final results will be found to be fairly consistent. In land of lower relief, that is, in the peneplain division, the results will be found to agree much more closely for the variation in slope is very much less. c) The amount of field work is greater than is necessary when slopes are not measured. This is true even when the measurements are made on topo- graphic maps, for in order to appreciate the meaning of the forms shown on a topographic map it is necessary that one examine them closely at first hand. The extra time and energy which physi- ographers will of necessity have to spend in traversing old land surfaces before they can obtain data upon the nature and extent of their slopes is one of the best arguments for the adoption of this classification instead of an objection to it. The geologist knows AVERAGE REGIONAL SLOPE 45 that rocks must be cracked if results are to be obtained, and the physiographer who depends largely upon distant views will miss a great deal of the detail which helps to prove or disprove field hypotheses. SUMMARY In the study of the physiographic development of the Interior Plateaus of British Columbia, certain characteristics commonly accepted as criteria for peneplanation were found well developed upon an erosion surface with regional slopes of 3 to 6 per cent. Stress is laid on the value of the study and measurement of the regional slopes of old erosion surfaces, and a quantitative subdivision of old erosion surfaces on the basis of their average regional slopes is suggested. Regional slope is defined as the general slope of the land toward main drainage lines, and an outline is given of the methods of measuring regional slopes, as followed in the work on the Interior Plateaus of British Columbia. The determination of the regional slopes of a land form is of value in furnishing clews to the agencies which have affected its development, and in separating the products of the different cycles of erosion through which it may have passed. The measurement of regional slopes is essential if a land surface is used as a datum for the measurement of movements of the earth’s crust. For if such measurements are not made, an uplifted old erosion surface is very likely to be considered of much lower relief than is actually the case. Such an assumption leads to serious errors in estimates of the manner and extent of movements of the crust. It is caused partly by optical illusions, and partly from the fact that the char- acteristic flat horizon lines and the discordance of topography and structure which prevail over old erosion surfaces of plainlike form are also found well developed upon surfaces of much greater relief. It is proposed that old erosion surfaces be divided into two central types “‘peneplains” and “beveled hills.” Peneplains are to be characterized by average regional slopes up to 2 per cent by discordance between topography and structure, by the absence of local irregularities of slope, such as cliffs and flat areas, and by deep soil covering. 46 LEOPOLD REINECKE “‘Beveled hills” are old erosion surfaces with dominant dis- cordance between topography and structure, with a general absence of irregularities of slope, and with deep soil covering. In addition, one may expect to find on them the accessory characters of partial accordance between topography and structure and occasional cliffs and flat areas. It is proposed that the term be confined to forms with average regional slopes of from 3 to 6 per cent and that the upland portion of the Interior Plateaus of British Columbia be considered the type of this land form. The subdivision is desirable because it will stimulate the meas- uring of regional slopes and thus assist in working out the physio- graphic development of the surface and diastrophic movements of the crust after its formation. It is not contrary to the accepted hypotheses of the genesis of a land form through the normal cycle of erosion, for the degree of slope is a factor in the manner, as well as the rate of development, of an erosion surface. The objections to the subdivision are that it is arbitrary, that a true average of the regional slopes on a land form are hard to get, and that it involves more field and office work than are otherwise necessary. The objections are met in the following manner. The type form “beveled hills’’ is represented by an old erosion surface, which is found throughout the southern portion of a large geographic unit, the Interior Plateaus of British Columbia. The subdivision is therefore not entirely arbitrary. The difficulty of obtaining a true average of the regional slopes on a land surface can be met by taking account of the relative area occupied by slopes departing from the general average. It is thought, finally, that the extra field work involved in traverses over the region will be of advantage in calling attention to details of physiographic interest which can be obtained in no other way. It is well to repeat here that the object of this paper is first of all to point out the importance of the study and determination of regional slopes on old erosion surfaces, and that the particular form of subdivision proposed is not considered final. CAIMANOIDEA VISHERI, A NEW CROCODILIAN FROM THE OLIGOCENE OF SOUTH DAKOTA MAURICE G. MEHL University of Wisconsin In the summer of 1to11, Mr. S. S. Visher, then connected with the Geological Survey of South Dakota, collected some interesting crocodilian material from the Oligocene of Washington County, South Dakota. Some time ago, the attention of the writer was called to these remains by Dr. Visher, and recently through the courtesy of Mr. W. H. Over, director of the museum at the Uni- versity of South Dakota, the collection was loaned to the writer for study. The material herein described consists of a goodly portion of a skull, a nearly complete mandible, two femora and other limb bones, a nearly complete series of vertebrae, many dorsal scutes, and numerous fragments. According to Dr. Visher, the collection was made from the Titanothere zone of the Lower White River beds, perhaps 20 to 30 feet above their base. THE SKULL Of the skull, nearly the entire right half is preserved as well as portions from the left side including the quadrate region, the posterior half of the cranium roof from the median line to the middle of the orbit and supratemporal vacuity, and fragments of the maxilla along the alveolar margin. Of the base of the skull but little remains save the separate occipital condyle and portions of the exoccipitals. In general shape and appearance it is quite similar to that of the alligators. A lateral expansion of the maxilla in the region of the third to fifth maxillary teeth produces a marked break in the otherwise regular outline of the muzzle, more prominent, perhaps, than in most of the Crocodilia. The width of the skull in the region of this maxilla expansion is 72 mm. Immediately back 47 48 MAURICE G. MEHL Fic. 1.—Caimanoideus Visheri, dorsal view of skull, about three-fourths natural size. CAIMANOIDEA VISHERI, A NEW CROCODILIAN 49 of the expansion, the width is but 66 mm. The greatest width, across the quadrates, is approximately too mm. The length of the skull, along the median line from the posterior border of the supraoccipital to the tip, is some 180 mm. From the posterior edge of the quad- rates to the tip of the muzzle the measurement is 203 mm. Unlike the most of the Croc- odilia, the inter-orbital region is flat, or essentially so, rather than presenting the marked concavity. All the bones of the facial region of the skull are sculp- tured by more or less pro- nounced rounded pits or longitudinal grooves except — the posterior ends of the nasals. In the maxillary region, the markings are more or less ill defined and, for the most part, take the form of long and narrow longitudinal grooves. The pits of the posterior frontal region are round, well de- fined, and deep. They are crowded close together and separated by narrow ridges only. The pits average 2.5 mm. in diameter, per- haps, in this region. On the jugals and squamosals, the SSS RA Lt — aT LSS - x Fic. 2.—Caimanoideus Visheri, right lateral view of skull and mandible, three-fourths natural size 5° MAURICE G. MEHL pits are shallow and varying in size, for the most part minute, but well defined. The relations of the various bones of the skull are quite alli- gatoroid, as are their general proportions with few exceptions. The prefrontal is relatively longer than in the genus Alligator and extends at least half the length of the lachrymal.in advance of the latter element. The nasals are relatively broad, the two making nearly half the total width of that part of the skull. Unfortunately, part of the anterior border of the snout is miss- ing, but the portions present are sufficient to show some of the important characteristics. Unlike the condition in all the true alligators, the nasals, while projecting slightly into the external narial opening, do not form a more or less complete bony septum. Furthermore, the premaxillae do not form an arch over the anterior border of the opening as is the case in all modern and extinct alli- gators with one or two exceptions, perhaps. In this respect, the skull simulates that of Brachychampsa Montana, an alligatoroid form described by Gilmore from the Upper Cretaceous of Montana.’ To quote: ‘In the absence of a roof-like covering formed by the premaxillaries over the anterior part of the external nares, Brachy- champsa differs from all known alligators, both recent and extinct.” The anterior border of the nares in the described form differs from that of Brachychampsa, however, in that the premaxilla in the former are still further reduced till the nares are directed slightly forward and lack entirely the definite ridgelike anterior border. This is the condition pointed out by Loomis in a specimen described by him from the Oligocene of South Dakota and referred to the genus Crocodilus.2 Quoting Loomis on this point: “The undivided nasal opening is very far forward, and differs from that of the other crocodiles in the lack of a distinct anterior border, this portion of the nasal cavity having a smooth, rimless boundary on the premaxilla. The nostril opening would seem, therefore, to have been directed to the front, rather than upward on the snout. «A New Fossil Alligator from the Hell Creek Beds of Montana,” U.S. Nat. Mus., XLI (1914), 290. 2A New River Reptile from the Titanothere Beds,” Am. Jour. Sci., CLXVIII (1914), 420. ‘ CAIMANOIDEA VISHERI, A NEW CROCODILIAN ~° 51 (The lack of a rim gives the snout a distinctly mammalian appear- ance.) ”’ The teeth are well preserved for the most part and, with few exceptions, have the crowns preserved entire. On the right pre- maxilla, there are preserved four alveolae. On the portion of the premaxilla broken away, there is apparently space for two addi- tional alveolae, but there is possibly only one. In the latter case, the total number is five, the number in the specimen described by Loomis. Concerning this series but little can be said, as only the circular roots of four remain. Of these the next to the last is the largest, 5.5 mm. in diameter. In each maxilla there are appar- ently thirteen teeth, a smaller number than is usually found in the alligators. Of these the fourth is much the largest, fully 8 mm. in diameter at the base and approximately circular in section. The posterior maxillary teeth are all more or less laterally compressed. The roots are oval in cross-section and in measurements range from 2.5mm.X4 mm. to4mm.xX7 mm., the more posterior teeth being the larger in general. The crowns of the first four maxillary teeth are somewhat flattened on the inner side and present a slight trenchant anterior and posterior edge. They are sharply conical and very slightly incurved, perhaps. The crowns of the posterior teeth show a rapid transition from this type to those with swollen crowns, rather sharply conical in the first of the series and more blunt or even of a rounded form posteriorly. A brief description of the well-preserved eleventh maxillary tooth will serve well to characterize this type. The crown is subglobular and is flattened somewhat on the upper, inner surface. It is sharply set off from the root by a marked necklike constriction. The apex is marked by an indistinct, antero-posteriorly Bee carina that does not extend down on the sides. Of the posterior part of the shel but little remains except the basi-occipital. The condyle is more or less spout-shaped and is divided into two distinct surfaces by a well-marked median groove. The palate is almost entirely lacking except for a small portion along the maxilla-premaxilla union. At this point there is a deep, sharply outlined pit for the reception of the fourth mandibular 52 MAURICE G. MEHL tooth. The pit does not perforate the facial surface of the bone, however, as is occasionally the case in the alligators. The lower jaws are represented by an almost perfect right ramus and a goodly portion of the left, including the symphyseal and articular regions. The sculpturing of the mandible is much the same as in any of the alligators; the dentary is marked by small, deep pits and more or less prominent longitudinal grooves. In the region back of and below the external mandibular foramen the surface is dotted with deep rounded pits, irregular in size but dis- tinct. In the relation of the various mandibular elements there is nothing of particular interest except perhaps in the unusual for- ward extent of the splenial. This element extends far forward and takes an ample part in the symphysis. Each ramus bears nineteen teeth, very similar, in general, to the maxillary teeth. The fourth tooth is the largest in the series, round in section, and measures fully 6 mm. in diameter at the base. The crown is largely missing, but was apparently sharply conical, and the indi- cations are that there was a faint anterior and posterior carina extending down the sides. This was probably the condition of the anterior four teeth in each ramus. The root of the first tooth indicates that it was somewhat larger (about 4 mm. in diameter) than the two subequal teeth that follow. Immediately following the fourth tooth is a series of seven, all comparatively short and slender and averaging about 2 mm. in diameter at the base and 4.5 mm. in height. These small teeth have slightly swollen crowns and a faint suggestion of anterior and posterior carina. The posterior eight teeth are subequal in size and quite like the eleventh maxillary tooth described above. The nature of the bite is alligatoroid in that the teeth of the mandible all close within the teeth of the upper jaw and the fourth mandibular tooth fits into a deep socket in the palate surface. THE VERTEBRAL COLUMN Eight of the cervical vertebrae are present, all but the atlas. With the exception of the axis, they are all very pronounced pro- coelous. Although the union between the centrum and the arch is clearly distinguishable, the two parts are as a rule closely united. CAIMANOIDEA VISHERI, A NEW CROCODILIAN 53 There is nothing distinctive in this series that serves to set them off from the other Crocodilia save perhaps in the atlas. The odontoid process is very prominent, and in a superficial way resembles the spout-shaped process of some of the mammals. The total length of the series, allowing for the missing atlas, is approximately 18.5 cm. Of the dorsal series all are preserved. As the cervicals, they are of the pronounced procoelous type. The total length of the series is about 23.7 cm. The five lumbar vertebrae are well preserved. ‘The posterior.articular faces of the centra are all highly convex and round in outline except the fifth. In this the posterior face of the centrum is fully twice as wide as high. The lumbars together measure 11.7 cm. The two sacrals measure 4.7 cm. A noteworthy feature is the shifting of the sacral ribs. The posterior rib forms a small part of the posterior concavity into which the first caudal fits. The anterior rib is shifted forward to such an extent that it might be said to articulate intercentrally, for although it is solidly fused to the first sacral vertebra, nearly half of its diameter extends beyond the anterior face of that vertebra and articulates broadly with the last lumbar. Of the caudal vertebra, there are 23 preserved. There are prob- ably about 11 missing, almost entirely from the posterior end of the series. The vertebrae present measure about 23.7 cm. To this should be added perhaps 16.5 cm. for those missing, making a total of 40. 2 cm. for the caudal series. Of the numerous appendicular bones indiscriminately pre- served the femora alone, perhaps, deserve special mention. Both of these bones are nearly complete, but the right is the more nearly perfect (Fig. 3). The head is broad and much flattened. The articular surface extends entirely around the end and on the side of the pronounced rounded cone that rises from the concave lower surface close to the proximal end of the bone. The tro- chanteric ridge on the flexor surface is very pronounced, more so than usual. At this point, about 45 mm. below the head, the shaft is bent abruptly back and from there sweeps backward in a broad, anteriorly concave curve to the distal end. The dorsal armor consists of a large number of pitted plates, sub-rectangular in outline for the most part. Of these there are 54 MAURICE G. MEHL several distinct types and a great variation in size. All bear a more or less prominent antero-posteriorly directed carina on the dorsal surface and are concave below from side to side to some extent at least and often sharply so. The dorsal surface of all is deeply sculptured by small, rounded, closely crowded pits. While the anterior and posterior edges are smooth, the lateral edges of Fic. 3.—Caimanoideus Visheri, right femur; a, from below; b, from above; c, from behind; three-fourths natural size. most of them indicate a more or less firm union with an adjacent plate. In some plates, this union is indicated on one margin only, and a few apparently were entirely free. The plates were prob- ably arranged in more or less rigid transverse rows of five units at least, and, in much probability, seven units for certain regions of the back. Fig. 4 indicates the apparent arrangement of the plates. As pointed out above, the affinities of this form with the genus Alligator are marked. This is shown chiefly in that the fourth ‘CAIMANOIDEA VISHERI, A NEW CROCODILIAN 55 -mandibular tooth fits into a deep pit in the palate instead of closing in a notch, as in crocodiles. Furthermore, all the mandibular teeth close within those of the upper jaw. In the lack of a division of the anterior nares and the lateral union of the dorsal scutes, the form is quite similar to the genus Crocodilus. It apparently stands much nearer to Caiman, however, than to either of the genera mentioned above inasmuch as this group combines the alligatoroid bite, the un- divided anterior nares, and the lateral union of the dorsal scutes. One of the striking differences from Caiman is to be seen in the entire lack of the anterior border of the external narial opening in the form here described. A new genus, Caimanoeda, is proposed to include this form C. Visheri, which may be considered the type, and C. (Crocodilus) prenasalis, Loomis,‘ a form that a careful examination of the type skull has shown to be very similar to C. Visheri. , If one may depend on the figures of C. prenasalis there is a noticeable differ- ence between the dorsal scutes of this form and those of C. Visheri. In the latter, all the scutes are more or less strongly keeled, and are much more finely and regularly pitted than in the former. ae, A lation between academic scholarship and i our new search for social justice. Yet _ these movements have an extremely in- | timate and vital connection, which comes “into view as we go behind the scenes and “i carefully observe the working of the es 1 forces that control the evolution of hu- _ man society.”’—The Author’s Foreword. ‘ The Struggle for Justice the Funda- a - mental Theme in the Religion of the ‘ey | Bible and the History of the Church. P cial pasteboard; net 25 cts., postpaid 27 cts. a | The University of Chicago Press __ CHICAGO ce ?: ILLINOIS Do You Want to Know What Your Handwriting Reveals? WE have published a fascinating booklet of 32 pages on the sub- ject of graphology, by William Leslie French. Every reader of this publi- cation will be interested in comparing the examples of handwriting in judg- ing character and disposition of the writer. A copy of this booklet and 12 differ- ent patterns of The’ Spencerian Steel Pens will be sent postpaid on receipt of ten cents. If you can do good work with a poor pen you can do better work with a good pen. Silver plated Falcon pens, smooth points, wear long, will not corrode. SPENCERIAN PEN CO. 349 Broadway - - : New York City 3 he Astrophysical J ournal | The forty-third volume, January to June, 1916, among other papers Foreign pasties, 62 cents ~~ - -- —_—_-_~--—- — _} CHICAGO will contain important articles on ASTROPHYSICS On the temperature and radiation of the sun, by Felix Biske, of the University of Warsaw; on the albedoof the planets and their satel- lites, by Professor Henry Norris Russell, of Princeton University The Astrophysical Journal published monthly excepting in February and August Subscription price, $5.00 Canadian postage, 30 cents THE UNIVERSITY OF eee aa PRESS . ILLINOIS FINE INKS 4D ADHESIVES | For those who KNOW A Simplified Typewriter The Remington Junior Drawin Drawing iabs Ink ; ; Engrossing Ink © ~ ° 9 Taurine Mucilage __ H | x gi ns 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 © ill-smelling inks and adhesives and adopt the Higgins Inks and Adhesives. They will bea revelation to you, they are so sweet, clean, well put up, and withal so efficient, At Dealers Generally Small in size—but big in capacity Light in weight—but heavy in stability Remington-built Remington-guaranteed Price $50.00 Easy payment terms can be arranged if desired Write for Catalogue REMINGTON TYPEWRITER CO. (Incorporated) New York and Everywhere CHAS. M. HIGGINS & CO., Mfrs. Branches: Chicago, eceessh 271 Ninth Street Brooklyn, N.Y. Efficient for Classroom, Club, or Platform — The educator or lecturer requires of a oo wk ern accuracy, brilliancy, simplicity, and dur- a very feeiate ea definitely characterize all models of the Bausch’ [omb — BALOPTICON THE PERFECT STEREOPTICON They have been developed from our 60 years’ experience with the intricacies of lens-making, in connection with our unusual facilities for research in the optical field and our command of reliable and highly skilled workmanship. For whatever use you need a lantern, there is a model of the Balopticon which meets it exactly. Model C—ideal for use with slides. Has the new 400-watt automatic gas-filled Mazda lamp, giving more bril- liant illumination than the arc—with much less current. Price, complete, $35.00. New Combined Model—for both slides and opaque an (maps, photos, postcards, etc.) with instant interchange. Its 1000-watt lamp uses but 9 amperes. Price, Other models from $20.00 up. Write for our interesting illustrated circulars on Balopticons, with price lists of all models Bausch € lomb Optical G. 554 ST. PAUL STREET ROCHESTER, N.Y. Leading American Makers of Microscopes, Projection Apparatus, Photographic Lenses, Stereo-Prism Binoculars, and other high-grade optical products| a NUMBER 2 a Te “ : tae GEOLOGY A SEMI- tl-QUARTERLY \ EDITED BY THOMAS C. CHAMBERLIN AND ROLLIN D. SALISBURY ORY ae ang With the Active Collaboration of VE SAMUEL ) W. ‘WILLISTON ; Sy ALBERT JOHANNSEN Vertebrate Paleontology ; _ Petrology _ STUART WELLER ROLLIN T. CHAMBERLIN - Invertebrate Paleontology ~ : soviet Peclozy, ee te doce || ASSOCIATE EDITORS IBALD GEIKIE, Great Britain JOSEPH P.IDDINGS, Washington, D.C, t RLES BARROIS, France - net R : JOHN C. BRANNER, Leland Stanford Junior University T PENCK, Germany : f _ RICHARD A. F. PENROSE, Jr., Philadelphia, Pa. WILLIAM B. CLARK, Johns Hopkins University j / - _. WILLIAM H. HOBBS, University of Michigan -E GEWORTH DAVID, dasigtealiag _' FRANK D. ADAMS, McGill University j LLIS, Leland Stanford Junior Liven, CHARLES K. LEITH, University of Wisconsin SILBERT, Washington, D.C. .. s WALLACE W. ATWOOD, ‘Harvard University LES D. WALCOTT, Smithsonian Tnrstitiitior ; WILLIAM H. EMMONS, University of Minnesota | vs. _ WILLIAMS, ‘Cornell Cy, hd 2 it ARTHUR L. DAY, Carnegie Institution | ren ve FEBRU ARY-M ARCH oe i ah kD eS [Ee ACADIAN TRIASSIC. PART Leas - - - - - - Sheer POWERS 105 SMIBn MACE Se imi eres , VROPHYLLITIZATION, PINITIZATION, AND SILICIFICATION OF ROCKS AROUND K AB abe BAY, len ea - - - - - A. F. BUDDINGTON 130 NITED. STATES. “PART BE en ge ee aetna oh teen er aan ite ge N co 153 UDIES. IN HYDROTHERMAL ALTERATION. Pre: wees - E, A. STEPHENSON 180 WEATHERING 0 OF A HORNBLENDE GABBRO ALBERT. D. BROKAW: AND LEon Pp. SMITH. 200 = mecat -— Jaf ye se Se eh = ay = pice ES ot = x PS 206 THE UNIVERSITY OF CU ECA GO PRESS ~ CHICAGO, ILLINOIS, U.S.A. AGENTS , THE CAMBRIDGE UNIVERSITY PRESS, Lonpon AND EDINBURGH HV Dac : . KARL W. HIERSEMANN, LEtpzic & * THE MARUZEN-KABUSHIKI-KAISHA, Toxyo, Osaka, Kyoto The Journal of Geology Vol. XXIV CONTENTS FOR FEBRUARY-MARCH 1916 THE ACADIAN TRIASSIC. PART Il -- =) =) = = 02 (> Ss 02 = = Spey’ Powers (iam NOTES ON RIPPLE. MARKS¢01 i=. cei penetrate a eg rene mane Rr Ce Pe TIT SPIE ae PYROPHYLLITIZATION, PINITIZATION, AND SILICIFICATION OF ROCKS AROUND CONCEPTION BAY, — NEWFOUNDLAND - - - - - - - - - - - Eig o 7 A. F. BuppIncTton _ THE ORIGIN OF RED BEDS. A STUDY OF THE CONDITIONS OF ORIGIN OF THE PERMO- — CARBONIFEROUS AND TRIASSIC RED BEDS OF THE WESTERN UNITED STATES. PARTI C. W. TomMLinson§ 15 STUDIES IN HYDROTHERMAL ALTERATION. I - aie Na <4 = mle igen E. A. STEPHENSON ZONAL WEATHERING OF A HORNBLENDE GABBRO - - - ALBERT D. Brokaw AND LrEon P. SmuitH REVIEWS gsi im mes aba elim rage Se aah e D ince non cei nA SE Nace MNS TESpAah Cg St NR et et cera pret a 06 The Journal of Geology is published by the University of Chicago at the University Press semi- quarterly, on or about the following dates: February 1, March 15, May I, 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. 4] Postage is prepaid by the publishers on all orders from the United States, Mexico, Cuba, Porto Rico, Panama Canal a Zone, Republic of Panama, Hawaiian Islands, Philippine Islands, Guam, Samoan Islands, Shanghe ai {| Postage is charged extra as follows: 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 single copies, 11 cents (total 76 cents). {Patrons are requested to make) all remittances payable to the University of Chicago Press in postal or express money orders or ae c drafts. . ‘q The following agents have been appointed and are authorized to quote the prices indicated: For the British Empire: The Cambridge University Press, Fetter Lane, London, E.C.— ‘Yearly subscriptions, including postage, 18s. 8¢. each; single copies, including postage, 3s. 2d. each. 3 For the continent of Europe: Karl W. Hiersemann, Konigstrasse 29, Leipzig, Germany. ‘eat subscriptions, including postage, M. 22.65 each; single copies, including postage, M. 3.80 each. For Japan and Korea: The Maruzen-Kabushiki-Kaisha, 11 to 16 Nihonbashi Tori Sanchea Tokyo, Japan. Yearly subscriptions, including postage, Yen 9.00 each; single copies, including ~ * — postage, Yen 1.50 each. 7 Claims for missing amber’ should be made within the month following the regular month of publication. The publishers expect to supply missing numbers free only when they have been lost in” transit. ~~ Business correspondence should be addressed to The University of Chicago Press, Chicago, Ill. Communications for the editors and manuscripts should be addressed to the Editors of THE JOURNAL OF GEOLOGY, the University of Chicago, Chicago, IIl. a Contributors are requested to write scientific and proper names with particular care and 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 leading articles will be printed as ordered, of which 50 (without covers) will be furnished graizs, the actual cost of the remainde (and covers, if desired) to be paid for by the author. The actual cost may vary, and will depend upon the amount of work in re-making the pages into forms, presswork, paper, binding, etc. Separates containing half-tones may be expected to cost somewhat more, the increase depending upon the number of cuts and the amount of work required upon them. Entered as second-class matter, March 20, 1893, at the Post-office at Chicago, Ill., under the Act of March 3, 1879. VOLUME XXIV NUMBER 2 THE FOURNAL OF GEOLOGY FEBRUARY-MARCH 1916 THE ACADIAN TRIASSIC SIDNEY POWERS Troy, New York IPENRCIE JOE Gerrish Mountain.—Gerrish Mountain consists of a basalt flow capping Triassic sandstones (Figs. 20, 22). The sandstones on the west side of the mountain are horizontal, those on the east side dip northward at angles of 15° or less. Opposite Moose Island is a cross-section of the basalt flow (Fig. 22), showing sandstone faulted against agglomerate on the west, with the agglomerate overlain by sandstone in one exposure; a columnar basalt dike east of the agglomerate, and probably separated from it by a fault; and red sandstone, east of the dike, overlain by a 3-foot bed of ash, above which is columnar basalt. The dike is 50 feet thick. It extends northward for some distance, but it is cut off on the south by a fault between the mainland and Moose Island. In either the dike, or the basalt flow, magnetite has been mined near Lower Economy. The final exposure of Triassic basalt on the east, which may originally have been connected with that at Gerrish Mountain, is at Portapique Mountain, 12 miles distant, and north of Birch Hill (see Fig. 23). Gerrish Mountain—Truro—tThe broad lowland underlain by Triassic strata east of Gerrish Mountain is interrupted by a long Vol. XXIV, No. 2 105 106 SIDNEY POWERS strip of Carboniferous conglomerates and shales forming a high, rugged topography from Lower to Upper Economy, as shown in Figs. 19 and 24, rising 250 feet, or more, from the lowland on the south. The relation of the Triassic to the older rocks, as seen along the shore, is a fault west of Carr Brook, and an unconformity near Lower Economy (Fig. 25). At the fault, the Triassic consists of very calcareous sandstones, containing conglomerate lenses and cross-bedding. Calcite has been introduced into the sandstone, forming dark-red concretions. The unconformity near Lower Economy shows a basal con- glomerate composed of subangular pebbles 1 to 3 inches in length, resting on upturned and leveled Carboniferous red shale. The shale within a foot of the contact is weathered into a clay, but this Fic. 22.—The shore section along Gerrish Mountain, as seen from the eastern end of Moose Island. The Triassic sandstone, at the left, is faulted against a mass of agglomerate overlain by a capping of sandstone. The agglomerate is probably faulted against the dike of diabase, which has fed the basalt flow which caps the cliffs on the right. At the base of this flow is a bed of green ash, and beneath this is normal Triassic sandstone, with some shale. weathering is of recent date. The basal Triassic conglomerate pebbles consist of slate, schist, quartz, and igneous rocks from.the Cobequid Mountains. Above the basal conglomerate, which is 25 feet thick, are interbedded sandstones and conglomerates as seen in Fig. 26. It is 1,200 feet stratigraphically between the basal unconformity and the Gerrish Mountain basalt flow. From Economy to Truro, Minas Basin is fronted by a compara- tively low land underlain by practically horizontal Triassic sand- stones, above which, in places, are Pleistocene gravels. The northern contact is of importance because there is a ques- tion whether there is a fault or an unconformity. The exposures are confined to the steam valleys, and are quite unsatisfactory. It may be briefly stated that the relation is probably a fault as far as Chiganois River, with another fault from there to North River, 107 THE ACADIAN TRIASSIC worousof. sypedouuy[\\| f/esog TH) P. Ko, YNY. Es me SN | SSW L ONAIDIT \ 2, uIseg SVUIJ[ Jo pvay oy} Jo deyy—fe ‘ory SHIH 06845 04; ; 46 SQ ie tas aay Py 12¢%y AVG GNOFGOI SS 108 SIDNEY POWERS as shown in Fig. 23. The evidences which favor this conclusion are the apparently dragged dips of the Triassic in several cases; and the lack of pebbles of adjoining older rocks in the Triassic conglomerates near by, and the actual fault seen in Harrington River on the west." At Birch Hill and at Folly Village, older rocks project through the Triassic strata according to Fletcher, who mapped both Fic. 24.—The base of the Triassic at Minasville, showing the basal sandstones and conglomerates resting on horizontally trunkated Carboniferous shales. The character of the erosion surface indicates a peneplain, and the composition of the basal conglomerate indicates a lack of residual soil on this surface at the beginning of Triassic sedimentation. localities with exaggeration of their size. There are reasons for questioning the existence of older rock at Birch Hill, but the copper prospect where these rocks are supposed to occur was not visited in the reconnoissance. t Certain of the localities are described by H. Fletcher, Geol. Surv. Canada, Ann. Rept., V (1892), 142-43 P. THE ACADIAN TRIASSIC 10g At Folly Village, Mississippian fossiliferous limestone and gypsum, belonging to the Windsor group, appear on the northwest side of Debert River. The Triassic sandstones are very calcareous, and resemble the Windsor limestones, as both are red in color. It appears that the limestone is overlain conformably by the gypsum, and that these Mississippian strata are overlain disconformably by the Triassic red sandstones. With this view, the Mississippian is confined to a small area on the northwestern side of Debert River where fossils are readily found." Truro—W olfville—The end of the arm of Triassic in Minas Basin lies near Truro, and the relation of these to the older rocks Ss D Pe Ls T UR Bn Ln ie Ne Fic. 25.—Section EE. Structure section through Minas Basin near Bass River | to show the unconformity of the Triassic on an island of Permian strata, Parrsboro formation, just north of the Basin. SD, Cobequid group; UR, Union-Riversdale series (Pennsylvanian); Pe, Permian. The major Cobequid fault is shown at the south of the Cobequid group. The closely folded syncline of the Parrsboro formation is in part overlain by Triassic sediments (7), which appear to be down-faulted on the north. is an unconformity, as shown in Fig. 23. This unconformity is well exposed in Salmon River and in Victoria Park Brook (where there is also a fault). The underlying Carboniferous strata always show a beveled surface. This unconformity continues along the south shore of Minas Basin, and may be seen at Minasville, on the sides of Moose Brook (Fig. 24), at Tennycape, and at Walton. Over this area the Triassic sandstones show nothing unusual, except for calcitization north of Maitland. West of Cheverie, the first exposure of Newark rocks is at Oak Island, north of Avonport (Fig. 3). On the east side of this island, quartz-pebble conglomerate and sandstone are exposed, overlain by stratified Pleistocene gravels, 6 feet thick, above which is Wisconsin till. t This view differs from that of J. W. Dawson, Acadian Geology, 3d ed., 1878, p. 99. 110 SIDNEY POWERS North of Oak Island is Boot Island, which is separated from the mainland, called Long Island, by a narrow channel, formed within the last two centuries. North of this channel is a buried forest, exposed at low tide. North of Boot Island, and on the north side of Long Island, are exposures of red sandstones with occasional thin shales in which Dr. H. M. Ami reports the presence of Estheria ovata." Fic. 26.—Details of the Triassic sandstone and conglomerate 30 feet above the base of the Annapolis formation, one mile east of Lower Economy. At Wolfville the basal Triassic unconformity is exposed in a small brook west of the buildings of Acadia College, resting hori- zontally on upturned and beveled slates of the Meguma series.’ The basal unconformity is again exposed at Kentville, just below the mill on Black River. t Verbal communication. 2 The writer is indebted to Professor Ernest Haycock of Wolfville for pointing out this locality. 3 J. W. Dawson, op. cit., p. 92; L. W. Bailey, ‘‘Geology of Southwestern Nova Scotia,” Geol. Surv. Canada, Ann. Rept., IX (1898), 128 M. THE ACADIAN TRIASSIC TEE W olfville-Scots Bay.—The Wolfville sandstone outcrops on the exposed points between Wolfville and Pereau River, the most con- tinuous exposure being near Kingsport. ‘The sandstone contains occasional conglomerate beds and red shales. The proportion of shale to sandstone gradually increases toward Blomidon. The shale in this locality is largely a red clay, with occasional green bands, persisting horizontally throughout the exposure. The general dip of the strata is 5-10 northward. Small faults are numerous. About midway between Kingsport and Pereau River, Haycock found fragments of well-consolidated fossiliferous red shale in till, overlying the Triassic.‘ The fossils are Estheria ovata. Dipping under North Mountain are poorly consolidated Blomidon shales, with the characteristic thin green beds at distances of 1o~20 feet. The Blomidon shale continues around the hook of North - Mountain beyond Cape Blomidon, but not as far as Amethyst Cove. At the latter locality, basalt cliffs, partly columnar, rise abruptly from the sea to a height of 300 to 4oo feet. These cliffs are kept vertical by frost action on the vertical joint planes parallel to the shore. | Two basalt flows are visible at Amethyst Cove, dipping gently northward, with undulating folds. The collecting place for amethysts is in a greatly veined area about too feet below the top of the lower flow. Scots Bay—Bennetts Bay.—In the region around Scots Bay there are two points of especial interest: first, the origin of the curve in North Mountain at this point, and, secondly, the presence of the Scots Bay formation overlying the North Mountain basalt along the southeast side of the Bay. Furthermore, the structural evidence furnishes a clue to the former thickness of the younger formation. The curve in North Mountain is formed in a syncline pitching down to the west, and in the nose of this syncline Scots Bay has been eroded (see Figs. 27, 28). The basalt flows of North Mountain dip toward the Bay on all sides at angles of 3°--5°. The topographic tE. Haycock, “Fossils in the Boulder-Clay of Kings County, Nova Scotia,” Trans. N.S. Inst. Sci., X (1901), 376-78. » aaa L2 SIDNEY POWERS slope follows the dip slope closely, beveling it slightly. At the water’s edge there is no sea-cliff, but merely a sheet of basalt (where exposed) sloping upward from the shore. Farther southwest along North Mountain there are low sea-clifis, but the general dip-slope persists to the end of Brier Island. The crest of North SCOTS BAY LEGENO Sco7s Bay fermotion TRIASSIC (L | Werrn ar. Bosalt Fic. 27.—Map of the Scots Bay—Cape Blomidon region Mountain is a mile or two southeast of the Bay of Fundy shore. It is a uniformly rolling surface which bevels the tilted basalt flows and forms a remnant of the Summit peneplain. Around Scots Bay the top of the uppermost basalt flow is marked by a green amygdaloidal layer in which the amygdules are one-half to three-quarters of an inch long. This amygdaloid THE ACADIAN TRIASSIC outcrops on the shore from Bennetts Cove to Ells Brook (Fig. 27). From this point to the north shore of the Bay there are no outcrops. Along the north shore toward Cape Sharp, the amygdaloid has been eroded, exposing the solid basalt beneath. In places this basalt is colum- nar, and in other places it contains balls, a foot or more in length, composed of dense basalt surrounded by amygdaloidal rims. The balls are not sufficiently abundant for the basalt to be called a pillow lava. The younger, sedimentary formation, named for its occurrence on the south side of Scots Bay, the Scots Bay forma- tion, has a thickness of 25 feet. It was discovered by Ells in 1876, and described by him? and later by Haycock.? It con-_ sists of white, very calcareous sandstone, quite distinct in color from any other Triassic sandstone of Acadia, with some interbedded shale and normal sandstone. The Scots Bay formation outcrops in five small synclines between Scots Bay and Bennetts Bay (formerly called Wood- worth Bay). These small remnants rest conformably on the basalt, contrary to the opinion of Haycock, and are pre- served in small synclines in the basalt. They extend southeast only a few hundred feet, as the topographic slope of the The vertical scale from Cape Split southward is exaggerated to show the overlying Scots Bay formation which is only 25 feet thick. The level of the Summit peneplain is shown by the line with three dots and a dash, the top of the flows of North Mountain by the finely dashed line. The area between these two lines must have been filled with the Scots Bay formation at the time the Summit peneplain was developed. \Wueess any Z = el ES =a ss re} REE = WW yal tR. W. Ells, ‘““Notes on Recent Sedimentary Formation on the Bay of Fundy Coast,” Trans. N.S. Inst. Sci., VIII (1894), 416-10. Ernest Haycock, ‘Records of Post-Triassic Changes in Kings County, Nova Scotia,” zbid., X (1900), 287-302. Fic. 28.—Section DD. Geologic cross-section through Cape Sharp on the north; Cape Split, Scots Bay, and Cape Blomidon, on the south. C, Carboniferous; 7;, Triassic, Annapolis formation; T,, North Mountain basalt; 73, Scots Bay formation. 4 (=I OW II4 SIDNEY POWERS 29.—Structure sections of a syncline of the Scots Bay formation, lying conformably over the top of the North Mountain basalt. amygdaloidal. Fic. Above the Scots Bay formation is till, filling the pre-Wisconsin valley. The top of the upper flow is very The basal beds of sandstone have been replaced by chert. mountain cuts across them and the flows at a low angle. These synclines are shown in Fig. 27. The Scots Bay formation consists of calcareous white or gray sandstone, frequently replaced by chert, and green- ish sandstone or shale. The exposures are nowhere over 15 feet in thickness (Fig. 29) and are remnants of a forma- tion which once filled the syncline of Scots Bay up to the level of the Summit pene- plain, as shown in Fig. 28. The white sandstone, or chert, rests directly on the amygdaloid at the top of the basalt flows, and veins of chert run downward from the sandstone into the amygda- loid. The beds in any one syncline do not matchexactly with those in any other, but this condition is to be ex- pected in basal beds of which only a few feet are shown, on the irregular top of .a lava flow. Fossils have been found in the calcareous sandstones by Haycock. They consist of faint green markings, probably plant remains; worm burrows; fish scales, bones, coprolites, and other THE ACADIAN TRIASSIC II5 fragments of fish. The coprolites are from 1 to 13 inches in length, an inch wide, and half an inch thick. In 1913, Haycock found a portion of the head of a fish which has been identified by Mr. L.M. Lambe as the Triassic genus, Semionotus fulius (Agassiz). Digby Gut.—Between Kentville and Digby Gut there are few outcrops of the Annapolis formation and no cross-sections of the North Mountain basalt. The best cross-section of the latter is at YS (news me casait Gulliver’s Cgve Ss . TRIASSIC. [\Neniden stale : i SS \' QQ efile sandstone KS Sms les Fic. 30.—Map of the Digby-Rossway region Victoria Beach, on the east side of Digby Gut (Fig. 30). The exposures on the west side of the Gut are disturbed by faulting. The section of North Mountain basalt, except the lower flow, which appears at the side of Digby Gut, commences on the Bay of Fundy shore, where 7 flows may be seen (see Fig. 31), each hav- ing a thickness ranging from 2 to 45 feet. All the flows dip toward the Bay of Fundy at a low angle, as seen in Fig. 32, but this uniform slope is in places interrupted by minor folds. The presence of a low syncline and the lack of exposures make the exact thickness of the lower flow uncertain, but it is probably about 600 feet. 116 SIDNEY POWERS Faults occur at right angles to the axis of North Mountain, at Digby Gut and at Bay View (Fig. 33). The physiographic evidence of these faults is seen in the offset of the North Mountain basalt, and in the valleys along the fault-lines. On the shore of Annapolis Basin near Port Wade, and at Digby, there are exposures of slightly cemented sand, containing blocks Fic. 31.—Five lava flows of North Mountain at Digby Gut, as seen north of Victoria Beach, on the northeast side of the Gut. The upper flow is the third from the top of the series as exposed. of basalt, which were considered by Bailey to be of Triassic age, and to underlie the basalt flows of North Mountain.’ These beds are of post-Wisconsin age, because of (1) the lack of consolidation, except very locally; (2) the yellow color, lke ordinary stream gravels (unlike any Triassic deposit except the “Lae Bailey, ‘Triassic (?) Rocks of Digby Basin,” Trans. N.S. Inst. Sct., IX (1898), 356-60; also ‘Geology of Southwestern Nova Scotia,” Geol. Surv. Canada, IX (1898), 126 M. “uIeJUNOJ, YIION Jo apis YINos ay} uo yuoWIdIvdsa ay} sdNpoid 0} Yorq posoy}voM Uosaq SLY YOTYA opvYS pot SI 9pyspury sty} Jo YSII sy} 07 pue Mopog *(saysep Aq uMOYs sv) a1njoId dy} Jo 10799 dy} Iv9U UTeJUNOU oY} UO 9plspue] oy} YYeoueq ysnfl st SMOG oYyI Jo aseq sy, “Io WUIoI]x9 ay} 4 odojs-dip oy} ut Uses of8ue oy} Ayo}eUNLxoIdde ye YJI0U dy} 07 SuIddip smoy zeSeq Jo S}SISUOD UTe}JUNOT YON ‘aroys aytsoddo dY} UO US ST YOVIG VIIOJTA JO WBRT[IA oy, “ND AQSIC JO opIs }Svoy}AOU 9Y} UO UTLJUNOJY YON YSnory} uorjoos oy[—ze “Oly TL] THE ACADIAN TRIASSIC 118 SIDNEY POWERS Quaco conglomerate); (3) the lithological character and the lack of green laminae; (4) the delta character of the deposit at Digby; (5) the basalt fragments decreasing in number with increasing distance from the talus of North Mountain basalt; (6) the horizontality of the stratification (the red and green Triassic shales near Victoria Beach dip northward at an angle of 3°). Rossway—Brier Island—At Rossway on St. Mary’s Bay (Fig. 30), there are excellent exposures of red shale, with occasional green beds and red sandstone beds, comprising the Blomidon shale. Rossway and Gulliver’s Cove are connected by a valley which marks a north-south fault similar to that at Bay View. The dis- placement of the fault may be seen at Gulliver’s Cove, in columnar basalt. No thin flows are shown here, or along Digby Neck to Brier Island. The Blomidon shales at Rossway have a thickness of about 500 feet. They dip northward at angles up to 10°. No other shales in the Acadian Triassic are as well consolidated. Ripple marks, current undulations, cross-bedding, and rarely mud cracks are seen in the shales. On the shore of Digby Neck, west of Rossway, the shales are exposed for a mile, dipping under the basalt. Half a mile west of the first exposure are Pleistocene clays containing black lignite and basalt fragments. This clay may have been deposited in the post-Wisconsin submergence. Digby Neck, Long and Brier islands show in common a depres- sion in the center of the ridge, parallel to the strike of the lava flows. This depression marks the amygdaloid at the top of the lower flow, as portions of but two flows are shown above the sea. No sedimentary rocks are shown on the St. Mary’s Bay side west of the exposure near Rossway.* Cross-faults are shown between Digby Neck and Long Island, and between Long Island and Brier Island. Another fault prob- ably occurs between Brier Island and the submerged ledge on the : A. Gesner described red sandstone exposed off Brier Island at low tide, but, he probably mistook either the red seaweed or a hematite stain over basalt for sandstone (Remarks on the Geology and Mineralogy of Nova Scotia, Halifax, 1836). 11g THE ACADIAN TRIASSIC "JSIMYINOS VY} UO SI S}[NVJ-SSOID 9Y} JO MOIYJUMOP 9Y} asNed0q ‘IJo] VY} UO UILJUNO YON Jo Jopureuros oy} ueyy AqsIq Jorvou ATIATWLIAI SpULIS SMOIIY BY} UIIMJOq YO] oy J, ‘“oUul]-}][Ney TepWMIs ve UO powI0J ATqeqoid sem yoryM “ny AqSIC, 0} sjutod }yS11 oy} UO MOIIY OY, “}NeF MotA AVG oY} SYTeW YOIYA Yoou ay} 0} szutod yo] oy} UO MoIIe oyy, ‘“Urefdousd yruMUNS oY} SyIew “souIsIp oY} Ul “UIeJUNOPT YWON jo do} oy, ‘“urejUNOy YANO Jo Joo; oy} WoIy AqSIq pur yng AqsIq— Ef! “914 120 SIDNEY POWERS west. Beyond this ledge there is no evidence of the existence of any Triassic rocks. AGE The age of the Newark group must be determined by a compari- son of its fauna and flora with that of Europe where the Triassic system is well developed. As forms common to both countries are not abundant, there are slight differences of opinion as to the exact correlation. Table I gives a correlation scheme which is modified from one given by Eastman." Trias Great Britain Germany Eastern U.S. Rhaetic Rhaetic Keuper marl Upper Upper Keuper marl Middle + Keuper (Upper Keuper sandstone) | Lower ————._$ANN |__| Newark group Lettenkohle Middle | Lower Keuper sandstone Upper Middle + Muschelkalk | tower Upper variegated sandstone Lower Pebble beds Bunter sandstone Buntersandstein On the evidence of the fish fauna Eastman? concludes that the Newark is to be correlated with middle and upper divisions of the Alpine Trias (the Upper Muschelkalk—Middle Keuper of the German section). The plants indicate a similar age, and several forms have been cited as the equivalent of the Lettenkohle forms of Germany. In the Acadian area the fossils which have been found are: plant remains at Split Rock (Gardner’s Creek), Quaco, and Martin Head, New Brunswick; fish remains in the Scots Bay formation at Scots Bay; and impressions of the shells of bivalved crustaceans in drift material from the Blomidon shale, found near Kingsport. «C. R. Eastman, ‘‘Triassic Fishes of Connecticut,” Conn. Geol. and Nat. Hist. Surv., Bull. 18, t911, p. 26. 2 [bid., p. 29. THE ACADIAN TRIASSIC IZ The plant remains were described by Dawson’ from poorly _ preserved material as Dadoxylon Edvardianum. ‘They consist of silicified plant-stems and of lignite, showing pith-casts. At Gardner’s Creek and at Vaughan Creek (Quaco), the material is largely silicified, and appears to have been transported some distance. At Martin Head, lignite is exposed in several horizons and is quite abundant. Silicification has not replaced the plant tissues to such an extent as in the other localities. The Martin Head locality is the only one where material is available for study. Miss Holden has recently examined the lignite, and found two species of plants.*, The form which was assigned by Dawson to the genus Dadoxylon has been found to be identical with Voltzia coburgensis Schaur., from the Lettenkohle and Lower Keuper of Germany. The other form is Equisetum rogersii, Schimper, which has been described by Fontaine from the Virginia Triassic. The correlation of these forms is also considered by Miss Holden. The Voltzia is apparently the same as the form described by Newberry as Palissya from the New Jersey area,* and as the form Cheirolepis from New Jersey and Virginia. The Equisetum rogersit is probably identical with E. columnaris, described by Bronn, from the Lettenkohle. Fragmentary fish remains have been found by Haycock$ at Scots Bay in the calcareous sandstones overlying the basalt. Recently further collections have been made by Professor Haycock and the material has been identified by Mr. L. M. Lambe, of the Geological Survey of Canada, as probably Semzo- notus fultus (Agassiz), a form common to the other Newark areas. tJ. W. Dawson, Acadian Geology, 3d ed., 1878, p. 108; also, Notes and Addenda, Pp. 99- 2 Ruth Holden, “‘Fossil Plants from Eastern Canada,” Annals of Botany, XXVII (r913), 248-54. 3 W. M. Fontaine, U.S. Geol. Surv., Mono. 6, 1883. 4J.S. Newberry, U.S. Geol. Surv., Mono. 14, 1888. 5 E. Haycock, ‘‘ Records of Post-Triassic Changes in Kings County, Nova Scotia,” Trans. N.S. Inst. Sct., X (1900), 287-302. 122 SIDNEY POWERS In fragments of shale found in the drift near Kentville, by Haycock, are impressions of Estheria of two slightly different types, both of which must be provisionally called E. ovata. The shale fragments were evidently derived from the Blomidon shale, and they are very similar to some of the hard barren shale exposed at Rossway. The paleontological and paleobotanical evidence proves that the Acadian area is a part of the Newark system, and further shows | a pronounced similarity between the Newark and the Lettenkohle of Germany. [To be continued| NOTES ON RIPPLE MARKS J. A. UDDEN University of Texas, Austin, Texas In a paper on ripple marks, recently published in the Journal of Geology, by Dr. E. M. Kindle, the opinion is expressed that the size of ripple marks may bear some relation to the depth of the water in which they were formed. Entertaining the same idea, I have on various occasions taken notes on the size of ripple marks. That most ripple marks vary in size with depth of the water seems to me hardly to admit of a doubt. Ripple marks from 3 to 4 inches in width appear to be most common. They are often to be seen in thoroughly sorted beach sands of all ages, from the Cam- brian up to the Pleistocene: In the Lower Comanchean, in Pecos County, in Texas, I have found some ripple marks of very small size, the smallest I have seen, with one exception. ‘These were noted at several points in some thin-bedded layers of sandstone of fine texture. These sandy layers are interbedded with clays and limestones. A piece of this ripple-bedded rock is shown in Fig. 1, in natural size. Twelve ripples measure together 3 inches across, making an average of one-fourth inch for each ripple, from crest to crest. The depth of the troughs measures about one twenty-fifth of an inch. These ripple marks are symmetrical. A rough mechanical analysis of the sand in this rock is as follows: Diameter of Grains Percentages in Millimeters by Weight fh Sea) is a ae Aa eas Gen ts REE 80 sty/ aN Oty le thas 8 artes) apeen nlc tr Ae OAR er Fg MR CR 20 Two years ago I found ripple marks of the same size, or possibly slightly smaller, forming in some fine sandy silt in the Rio Grande, in Webb County. The silt had been washed up on some large blocks of sandstone, which were strewn in the channel of the river. It lay in shallow depressions on these rocks, and the water covered 123 124 IA. TD DEN the ripple marks from a half to one inch deep. The wind stirred the surface of the water gently into small waves, and the ripple marks in the sand were seen to be building, under the influence of these waves. Fig. 2 shows some ripple marks in a fine silty sand of the marine Jurassic, near Minnekahta, South Dakota. They measure 13 inches from crest to crest and have an average depth of fifteen-hundredths Fic. 1.—Ripple marks in Comanchean sandstone, from Pecos County, Texas. Natural size. of an inch. These ripple marks are unsymmetrical, their longer slopes bearing the average ratio of 152 to 100, to the shorter slopes. A mechanical analysis of the sand in this rock was found to be, roughly, as follows: Diameter of Grains Percentages in Millimeters by Weight F/O PTAA esa ore a ee Ae Oke ae ee Trace L/h pe 20 Bia tte cas tik ee a ee eee a A eee 55 1/8! SU TOn eae Seah ie ee OL eee 30 Tif NOS) SO ek teenth ce cde Stan Fea eas ere Roe RR I 15 Some large-sized ripple marks occur in the Ordovician dolomites at Utica, in Illinois. In the old entries where cement rock long ago NOTES ON RIPPLE MARKS 125 was mined for the Utica Cement Works, some ripples have been disclosed that measure from 4 to 5 feet across from crest to crest. This is in a somewhat thin-bedded dolomite, which contains some sand. Evidently this limestone was not a shallow-water deposit. The ripple-bedded layers lie some 100 feet below the base of the St. Peter sandstone. . The widest ripple marks that have come under my observation are in a crinoide! limestone in the lower part of the Burlington, Fic. 2.—Ripple marks on Jurassic sandstone, near Minnekahta, South Dakota. One-half natural size. in the southeast part of Louisa County, in Iowa. These ripples measure nearly 6 feet from crest to crest, and are at least 6 inches deep. ‘The presence of crinoidal remains in this rock, which con- tains some shaly material, indicates, if not proves, deep-water conditions. How deep? Higher up in the geological column I have seen some quite large ripple marks in the Comanchean, in Texas, in a horizon near the Kiamitia clay. About 17 miles west-southwest from Kerrville such ripple marks occur in the bed of Guadalupe River (see Fig. 3). They measure about 14 inches across and are about 13 inches deep. They are slightly unsymmetrical. The rock in this case is a mixture of calcareous and shaly material, which contains variable quantities of fine sand, so that some layers might more properly be called 126 J. A. UDDEN sandstone. The same horizon is exposed in the bottom of Bosque River, at Clifton, 155 miles northeast from the locality just men- tioned, and again some 6 miles north of Clifton in the same beds in the same stream. Some layers of limestone here show ripple marks that measure 4 feet across, near Clifton (see Fig. 4), and from 2 to 3 feet across at the northernmost locality (Fig. 5). The lime- Fic. 3.—Ripple marks in thin-bedded sandy limestone in the bottom of Guada- lupe River, about 17 miles southwest of Kerrville, Texas. stone layers here are compact and quite pure in composition, but are interbedded with marly shales. Perhaps it may be permitted to submit some general remarks anent the phenomena of ripple marks. They shall be brief. Ripple marks must be due to rhythmic variations in currents in the medium of sedimentation. They are in this respect kin to wavelike etchings, known to be caused by rhythmic movements of corrading currents. Perfectly symmetric ripple marks are probably the result of to-and- fro movements of equal extent in both directions, when these move- ments are such that the velocity of the motion happens to be NOTES ON RIPPLE MARKS E277 sufficiently strong to move material of the coarseness present where the rhythmic motion prevails. On the bottom of any billowy water, sufficiently shallow for the size of the waves, there must be a to-and-fro motion for each passing wave. For waves of the same size, the deeper the water the more slow and the more limited will this motion be. Hence the less will Fic. 4.—Ripple marks in Comanchean limestone in the right bank of Bosque River, near Clifton, Texas. be the diameter of the particles it will be able to stir. There must be a certain depth where the motion will be just speedy enough to stir particles of silt. Where the bottom lies at this depth, and where it is covered with silt, ripple marks will form. Should not their width be determined by the extent of the to-and-fro move- ment in each direction? ‘This decreases downward according to a known law. It is evident that the velocity of each to-and-fro movement on the bottom of an agitated body of water begins with zero, rises to a 128 J. A. UDDEN momentary maximum, and falls again to zero. For particles of different sizes, there must be different times of duration of speeds attaining and exceeding the respective limits effective for their transportation. This time, and hence the latitude of this effective translatory motion, will increase with the fineness of the stirred sediment. With waves of one and the same size, and with the Fic. 5—Ripple marks in Comanchean limestone in the bed of Bosque River, about 6 miles north of Clifton, Texas. same depth of water, the width of ripple marks should be greater in fine sediments than in coarse. The currents producing them will carry fine elements farther than coarse. With waves of the same size ripple-mark building in sand should then also take place in somewhat more shallow water than ripple-mark building in silt. Some ripple marks must be produced by a wavelike or rhythmic motion which results from a reaction by the transported material on translatory bottom currents in water and in the air. No surface NOTES ON RIPPLE MARKS 129 billows in the atmosphere can have anything to do with ripple marks in dune sands. Do dune-sand ripple marks vary in size with wind velocity and coarseness of the sand? They do not vary very much. May ripple marks be formed by a like reaction with bottom currents in deep water? If so, their variation in size may also be small. Such ripple marks, like those in sand dunes, should always be unsymmetrical. Their sizes are probably independent of depth of the water. PYROPHYLLITIZATION, PINITIZATION, AND SILICIFI- CATION OF ROCKS AROUND CONCEPTION BAY, NEWFOUNDLAND A. F. BUDDINGTON Princeton University INTRODUCTION This paper embodies the results of a study of the regional and local alterations which have affected a series of volcanic rocks lying around the borders of the head of Conception Bay, on the Avalon Peninsula of Newfoundland. The field work was accom- plished by the writer as a member of the Princeton Geological Expedition in Newfoundland during the summers of 1913 and 1914, in connection with a general study of the pre-Cambrian rocks of this region. The writer is indebted to the Geology Department of Princeton University for facilities for studying these rocks in the field and laboratory; to Dr. C. H. Smyth for supervision in the preparation of the report; to Professor G. Van Ingen for the photographs with which this paper is illustrated and for his interest in the work. The numbers used in this report refer to specimens deposited in the museum of Princeton University. LOCATION Regional alterations, such as silicification and chloritization, have affected the volcanic series wherever they outcrop, either in the area here under consideration (Fig. 1), or at other points to the north, such as Clarenville on Trinity Bay or Goose Arm on Bona- vista Bay. The local alterations, comprising pyrophyllitization and pinitization, have affected the volcanics only in limited areas; the former being exhibited in a long narrow strip of rocks south of Manuels and the latter in outcrops to the north of the mon- zonite stock at Woodfords and in minute amounts associated with the pyrophyllite rocks at Manuels. 130 HYDRON SILICATES IN ROCKS OF CONCEPTION BAY 131 STRUCTURE The volcanic series forms the lowest member of the Algonkian rocks in this region and has been mapped as Huronian by Howley (1907). The volcanics at the head of Conception Bay outcrop on CONCEPTION. BAY. Nn *Feip ssi." Gre, GY Ste, " -228F2 “ + = Ye, 4 Tale 2 he228r5 1 2 a a “ “4 . . 4 “ eee aos Sketch Map of Geology at the head of Conception Bay Newfoundland Tepagraphy Based en Admiral Cros! No 236 By AF Buddinglan iss Legend GlCambrian Shales & limestones Mone onile Ga Protas (cman | —— Railroad Avondale Veianics wee Aerial Tram Tale Pyrophyliile) Prospect Pa y Fic. 1. the core of a major anticline, the eastern limb of which has been intruded by a huge batholith of granite. On the west side of Holy- rood Bay they are also intruded by a stock of monzonite. : The rocks are in addition excessively disturbed by profound and intensive faulting, and usually dip steeply. There is a strong probability that the line of contact between the granite and volcanics at Manuels marks the approximate locus 132 A. F. BUDDINGTON ofa fault zone. For at talc prospects 3, 4, and 5 the volcanics are faulted against either granite or green slate beds, and at prospect 3 the granite adjacent to the fault plane is silicified and pyritized and the rhyolite is silicified and carries traces of pyrophyllite. The pyrophyllite veins in turn are offset by small cross-faults. CHARACTER OF VOLCANIC SERIES The volcanics comprise a thick series of rhyolite and basalt flows with corresponding interbedded breccias, crystal tuffs, and tufis, and a minor amount of waterworn material. The evidence with respect to their origin all points to their having accumulated under subaerial conditions. TOPOGRAPHY The topography developed on the volcanics at Manuels is that of a long, narrow, more or less barren plateau about 600 feet above sea-level. The volcanics at the head of Conception Bay are carved into a series of rugged isolated hills or ridges with differential eleva- tions of from 200 to 1,000 feet. Glaciation during the Pleistocene period had a marked effect on the superficial features of the country, and many of the outcrops were scraped and polished by this agency. WALL ROCKS OF THE PYROPHYLLITE VEINS The pyrophyllite is confined almost exclusively to the rhyolite flows. Occasionally, however, pockets are found in the rhyolite breccias and conglomerates, but none at all occurs in any of the other rocks. The rhyolite flows exhibit three characteristic struc- tures: flow or banded, spherulitic, and elliptical or lenticular. The spherulites may range in size from micro-spherulites visible only with the high powers of the microscope to huge spheroids as big as a man’s head or even larger. They are usually more or less replaced by quartz of chalcedony. The elliptical structure has been called such because of its appearance on the weathered surface of the rock, where it shows as an assemblage of rude ellipses, or as lenses surrounded by a more or less schistose material which may be pyrophyllitized (Fig. 2). The ellipses vary from several inches to a foot in the direction of their longest axis. At talc prospect 5 a rhyolite showing this structure also contains scattered HYDRON SILICATES IN ROCKS OF CONCEPTION BAY 133 spherulites. This structure is characteristic of the white rhyolites of this area and may owe its origin to primary flowage phenomena, or to secondary dynamic forces, or probably to the former accentuated by the latter. DESCRIPTION OF PYROPHYLLITE The pyrophyllite veins are of such an extent that they attracted attention as a source of talc; many prospects were opened in the Fic. 2.—Lenticular structure in rhyolite. The material surrounding the more massive portions is partially pyrophyllitized. deposits and a 24-mile aerial tram was built to the nearest mine. But, owing probably to the difficulty in separating the pyrophyllite from the admixed quartzose nodules, all work has been abandoned since 1904. , The pyrophyllite where it replaces rhyolite flows, as it does almost exclusively, is a soft cryptocrystalline, light greenish-yellow rock with a waxy luster and a good cleavage parallel to the schis- tosity. In one case where it replaces the matrix of a volcanic conglomerate it is a light brown and in another where it replaces the matrix of a volcanic breccia it is cream colored. 134 A. F. BUDDINGTON The pyrophyllite may occur either as single well-defined veins, or as a series of veins, pockets, and lenticels, which together con- stitute what may be called a pyrophyllitic zone. The former character is illustrated at talc prospect 5, where the pyrophyllite forms a vein about 500 feet long and varying from 6 to 15 feet in width in a white, densely spherulitic rhyolite. Near the one end of the vein which is exposed the pyrophyllite is full of nodules and stringers of the rhyolite, but becomes almost clear pyrophyllite in its central portion. The latter character (pyrophyllitic zone) may be illustrated by the character of the pyrophyllite deposits at talc prospects 1 and 2. The country rock of the pyrophyllitic zones may be so altered as to constitute a quartz-pyrophyllite schist consisting of micro- crystalline quartz and pyrophyllite, as at talc prospect 1, or it may be partially pyrophyllitized, as at talc prospect 2, or relatively unaltered as at talc prospect 4. At talc prospect 1 large masses of pyrophyllite occur in pockets from 1 to 15 feet in diameter containing more or less country rock, or as thin sheets incasing lenses of quartz-pyrophyllite rock oriented parallel to the cleavage. It frequently occurs as an interlacing network of films veneering lenses of the quartz-pyrophyllite rock or as lenticels replacing the matrix between adjacent quartzose nodules. The pyrophyllite (222 E 2 f x) usually serves simply as a matrix for these nodules varying from a fraction of an inch to several feet in diameter, and even hand specimens are infrequent which do not contain one or more of them. Ramifying stringers of country rock may wander aimlessly through the pyrophyllite (Fig. 3) and veins of pyrophyllite reticulate in the country rock. It is quite possible that the nodular structure originated through the total replacement by pyrophyllite of the sheared zones between lenses of a rhyolite like that shown in Fig. 2, and the alteration of the lenses themselves to a quartz-pyrophyllite rock. On the west side of the talc mine here small pockets of cream-colored pyrophyllite, weathering green or yellow, are found replacing the matrix of a very coarse rhyolite breccia. At talc prospect 2 there is a pyrophyllite zone about 30 feet wide in which pyrophyllite constitutes from a small percentage HYDRON SILICATES IN ROCKS OF CONCEPTION BAY 135 to one-half of the rock. Paralleling this zone at a distance of about 20 feet is a pyrophyllite vein 18 feet in width with a 3-foot stringer and many spheroids of altered rhyolite wall rock in its central portion. The spheroids vary from an inch to a foot in diameter, but average about 4 inches. A lenticular or elliptical structure (Fig. 2) characterizes the rhyolite adjacent to the vein and the schistose matrix of the lenses or ellipsoids is partially pyrophyllitized. A few hundred feet north of here a light brownish pyrophyllite is Fic. 3.—View of portion of pyrophyllite vein, showing intermingling of pyrophyl- lite and country rock. =pyrophyllite; g¢=quartz-pyrophyllite. found replacing portions of the matrix of a white rhyolite con- glomerate. DESCRIPTION OF PINITE At Manuels the pinite is a relatively rare constituent and is interesting only from the viewpoint of its origin. It is best exhibited at talc prospect 5 and at a point marked 228 D 1 on the map. Fig. 4 is a photograph taken at this latter locality and repre- sents the matrix of a spherulitic rhyolite (228 D 1 h) replaced by dark-colored pinite. The spherulites here average about 1 inch 136 A. F. BUDDINGTON in diameter and are so intermingled with smaller ones as to make up almost the entire bulk of the rock. The pinitized groundmass possesses a waxy luster, dark dirty-green in color, and is quite soft. The parting of the pinite is in general parallel to the cleavage of the rhyolite, although in detail it is a series of curving shell-like scales, owing to its parting following the circumference of the more resistant spherulites. An analysis of this matrix is given in this Fic. 4.—Pinite (p), replacing portions of the matrix of a spherulitic rhyolite Journal on p. 137 (No. 7). Very rarely pinite is found along the original contraction cracks of the spherulites or at the heart of a spherulite. Patches, lenticels, and minute veins of pinite are found throughout the rhyolite flows and agglomerates, often repla- cing the matrix of spherulitic zones (228 G 1 k) or certain flow lines (222 BE 2%). In the valley of Harbour Main Brook, among a series of rhyolite flows and tuffs, tuff beds up to 75 feet thick have been partially altered to pinite, and spherulitic rhyolite flows up to 30 feet thick are streaked and banded with pinite. These rocks have been / HYDRON SILICATES IN ROCKS OF CONCEPTION BAY 137 prospected to depths of 15 and 20 feet, probably under the misapprehension that they carried pyrophyllite. A chemical analysis of the pinitized groundmass of the spherulitic rhyolite is given on p. 137 (No. 8). CHEMICAL ANALYSES Chemical analyses of eight typical rocks showing the various types and some of the stages of the alterations and replacements which have affected the volcanics are here given. These analyses were recalculated to correspond approximately to the mineral composition of the rocks. Small amounts of water of absorption, iron oxides except in No. 7, excess alumina, etc., have been lumped together as such under ‘other constituents.”’ This involves of course a slight but inappreciable error in the proportions of the other minerals. It is probable that some sericite is also present in rocks Nos. 3 and 4, but owing to the difficulty of distinguishing sericite from pyrophyllite under the microscope, and because of the fact that the decrease in potash with an increase in water TABLE I SHOWING CHARACTER OF ALTERATIONS OF RHYOLITE I 2 3 4 5 6 7 8 SIO ree sto: 76.24 | 80.60 | 74.51 | 72.10 | 65.04 | 88.09 | 54.47 | 61.07 AUG O ese nce 22 Ss USE OF | ln 7 ek fl2) e240 51/1 20.4G) © 19253) |) 27. LAs" 22.00 GOs Sea aeoees 0.89 | 0.89 T.28 | 0.54! 0.28 | 0.20 2.58 1.56 EO Ras. 0.13 | 0.08]. 0.08] n.d. n.d. n.d | 0.47 |) 0.42 IMIG OP.) OsDy OFS Ll OnO4 AMON OA MVOO4N I O05) |) 12244) ONOS CAO Ree. TOT OHOOM ONT H OSA GIN OsTON || LOSS) | ovor 1.63 Naz OR elie: ASS TOM On4Onlow42h in O.33.|)) ned.) | OnOS qin o.03 IK AO) sco ie temp ; 4.05 | 4.68 | 3.68 202 Th 1O233 | nid.) Shon 7.58 Oa 0.15 0.49 2.44 | 3.09 | 4.84 1.68 | 3.44 2.93 AO ae ere 0.03 O.1I 0.09} 0.23 OXOZ sie OOM Oni On 28 EVEN O eat sy Trace | Trace | Trace | n.d. Gly) mk, |) G0 |) Gury | 100.22 |100.25 | 99.85 |100.63 |100.48 |100.17 |100.40 | 99.86 SDRG ree. DV | BOA | QL7e | Beaty 2EOBH 2270 teen Se 2.70 ! I. Unaltered dark-gray flow rhyolite (228 D1 c). 2. Silicified drab spherulitic rhyolite (228 F 2). 3. Pyrophyllitized rhyolite. Matrix of lenses; talc prospect 2 (228 F 3). 4. Pyrophyllitized rhyolite. Matrix of spherulite; tale prospect 1 (222 E 2 j). 5. Pyrophyllite. Light greenish-yellow waxy pyrophyllite from talc prospect 1 (222 E 2 /). 6. Quartz-pyrophyllite schist. Nodule in pyrophyllite (5); talc prospect 1 (222 E 2 ). 7- Pinite; dark dirty-greea waxy, matrix of spherulitic rhyolite. Manuels (228 Dt h). 8. Pinite schist. Matrix of spherulitic rhyolite, Harbour Main (232 E 2 6). 138 A. F. BUDDINGTON indicates that the replacing mineral is pyrophyllite, it has been calculated as that alone. Owing also to the lability of error involved in assigning the elements of the pinitic rocks to the correct minerals in the right proportions, only a rough estimate of their mineral composition is given. The iron oxides and magnesia of No. 7 must be present as an integral part of the white mica molecule, as no other mineral except quartz can be distinguished in thin section. TABLE II RECALCULATED ANALYSES © | Quartz Orthoclase Albite Anorthite |Pyrophyllite oe eae White Mica 1 SEENON c 39.6 29.4 21.4 5 ed oul: eee oe, eee Fy Wed Weer oe er ae Ore Sees 5367 27.8 9.9 Ea Gi lea ee Sho. eeieuate sayemts By Botte 28.3 OTe 4.2 0.6 43.6 bay OPEN FE A AS te TS Be beg cc 18:7 Tae3 3.6 Zs 61.1 0.8 “bean ee Since Teg) |Secte aceck Stee seta vetoes 91.8 1.6 5-3 ORI LST P7 HEN Penta Stan mies ME Oe ol Oh A RBI, este 33-4 O.9' reece Teas Present - [hss hinvacua|aacher ieee tell okt Stes Sei ae lee oes About 75 per cent Ss ae Present’ || Present. || Present 71). 2's see] ses ee eee About 60 per cent PETROGRAPHY No. 1 (228 D 1c). This specimen was taken near the top of a 50-foot banded reddish-gray felsite flow. In thin section the tex- ture varies from microfelsitic to very minutely microcrystalline and the flow lines are marked by hematite dust. The flow lines are sharply curved and crenulated and several are replaced by quartz, especially in the loops of the curves, so that the rock analyzed as representing the composition of the original rhyolite only approxi- mates such an unaltered condition. No. 2 (228 F 2). This rock is a drab to fawn-colored micro- spherulitic rhyolite with secondary iron oxide in veinlets and specks. In thin section the groundmass is a finely microcrystalline aggregate of quartz and orthoclase, with fan-shaped microspherulitic areas. Secondary quartz is present as grains and lenses, as well as replacing portions of the spherulitic aggregates. A minute amount of sericite and quartz occurs along fractures. HYDRON SILICATES IN ROCKS OF CONCEPTION BAY 139 No. 3 (228 F 3). This specimen, a pyrophyllitized rhyolite, was taken from the slightly sheared matrix surrounding lenses of white rhyolite (Fig. 2) adjacent to pyrophyllite veins at talc pros- pect 2. In thin section the rock shows as a microcrystalline aggregate of granular quartz and feldspar and of scales and fibers of pyrophyllite in about equal amounts. Fic. 5.—Perlitic structure preserved in pyrophyllitized rhyolite. Ordinary light, X60. No. 4 (222 E 27). This is the matrix, a pyrophyllitized rhyo- lite, in which a 6-inch spherulite was found essentially unaltered. In thin section the rock consists of an aggregate of very minute microscopic scales of pyrophyllite, complete except for a remark- ably well-preserved perlitic structure, outlined by microcrystalline quartz with probably some orthoclase (Fig. 5). The rock is in a much more advanced stage of alteration to pyrophyllite than the preceding specimen. No. 5 (222 E 2f). Light greenish-yellow pyrophyllite with a fair cleavage. In thin section the rock is seen to be composed of 140 A. F. BUDDINGTON a homogeneous felt of exceeding minute microscopic scales and fibers of pyrophyllite, with a strong tendency toward a very good alignment parallel to the cleavage in a section at right angles to it and with long, parallel, fibrous shreds in a section approximately parallel to the cleavage. No. 6 (222 E 2g). A white quartzose nodule or lense of quartz- pyrophyllite schist about 1 foot in diameter taken from a pyro- phyllite vein at talc prospect 1. In thin section the rock presents what might be called a micro-blotchy groundmass composed of aggregates of either microcrystalline quartz or of scales of pyro- phyllite. Some fibers of pyrophyllite also occur interstitially in the quartz areas. No. 7 (228 D 1h). This is the dark grayish-olive, waxy- lustered matrix of the spherulitic rhyolite illustrated in Fig. 4. In thin section the rock is seen to consist of an aggregate of extremely fine shreds and scales of white mica with lines of partially replaced microcrystalline quartz which are probably replacements of certain of the original rhyolite flow lines not yet entirely replaced by the white mica. No. 8 (232 E 2 6). Grayish-olive pinite schist with small unaltered spherulites or spherulites partially replaced by quartz. In thin section the material appears as a perlitic microcrystalline groundmass of quartz and orthoclase partially replaced by sericite. The perlitic cracks are outlined by threads of sericite fibers, as illustrated in Fig. 6, and they often form the boundaries of sero- citized areas which present the appearance of eyes, sometimes with a reticulating network of sericite veins connecting two adjacent eyes. Within the sericitic material, isolated microspherulites, clusters of microspherulites, and long axiolites are often preserved intact. A few phenocrysts of orthoclase are present and are remarkably fresh, although occasionally flecked with sericite. The secondary material consists of an aggregate of microscopic sericite scales and fibers associated with grains and areas of secondary quartz. Minerals originating through decomposition at the sur- face are completely absent except for a trace of iron oxides in the groundmass and a slight cloudiness in the feldspars, probably due to kaolin. ee, HYDRON SILICATES IN ROCKS OF CONCEPTION BAY 141 PRELIMINARY SILICIFICATION The first process of alteration which operated on the Avondale Volcanics was that of regional silicification. On these preliminary silicified rocks a local series of alterations, those of pyrophyllitiza- tion, pinitization, and further local silicification, were superimposed. Fic. 6.—Perlitic structure preserved in pinitized rhyolite. Ordinary light, X35. m=microspherulites; p=pinite. This is evidenced by the following data: (1) few later quartz — veins are found traversing the pinite, quartz schists, or pyrophyllite; (2) fragments of breccia in the volcanic breccias often exhibit quartz veins which stop abruptly at the contact with the matrix, and (3) under the microscope aggregates of sericite scales are found replacing granular quartz which had previously replaced the heart of a spherulite, and these sericite scales finger into and inclose unreplaced fragments of quartz, proving definitely their later origin. 142 A. F. BUDDINGTON The spherulitic rhyolites are the rocks which exhibit most clearly the manner in which the silicification has taken place. The silica here is present for the greater part as a milky-white chalcedonic quartz, but vitreous, granular, and white vein-like quartz as well as quartz crystals are common. The chalcedony usually forms the outer borders of the concentric crescent-shaped areas, and of the hearts of the replaced zones of the spherulites, while the inner portion may be recrystallized to form comb struc- ture through the interlocking of quartz crystals, or little geodes with terminated crystals projecting into a small cavity, or granular vitreous quartz. When only one form of the silica is present it is very generally of a chalcedonic nature. It is interesting to note that while at Manuels it is the spherulites of the rhyolites which are most generally replaced, at Clarenville it was the groundmass which was replaced instead of the spherulites, because of the perlitic structure of the former offering the most favorable surfaces for attack. The banded rhyolites are often lined or streaked with quartz veins parallel to the planes of flow, which in some cases are a result of replacement and in others of vein filling. In thin section, lenses, lines, and granules of secondary quartz are found to be a common characteristic of the slightly silicified banded flows. Fig. 7 illustrates the preservation of the perlitic structure in the quartz which is replacing the groundmass of a spherulitic rhyolite from Clarenville. The perlitic cracks are outlined by sericite. That the silicification of the rhyolites has been due to secondary metasomatic processes and is not a primary phenomenon is indi- cated by (1) the interruption of fluxion lines by the replacing quartz, (2) by the presence of unsupported fragments of unreplaced rhyolite in the quartz areas, and (3) by the preservation in the quartz of original structures of the rhyolite, such as the perlitic structure. The first stage then in the alteration of these volcanics has consisted in the silicification under relatively static conditions by hot siliceous waters of rhyolite flows which may be represented as having had a similar chemical composition to the present relatively unaltered gray felsites. Analysis No. 1 may be taken as the com- HYDRON SILICATES IN ROCKS OF CONCEPTION BAY 143 position of the original rock from which the silicified rhyolite represented by analysis No. 2 has been derived. Chemically this has resulted in a decrease in the percentages of potash, soda, lime, and alumina, an increase in the percentage of silica, and a relative decrease in the percentage of sodium with respect to potassium. ‘The process operated through the replace- ment of the feldspars by quartz and a relatively more rapid Fic. 7.—Perlitic structure preserved in quartz replacing the groundmass of a spherulitic rhyolite. Ordinary light, X35. s=spherulite. replacement of the soda feldspars than of the potash feldspars. The solutions which effected this alteration doubtless belonged to the same general period of volcanic activity as the extrusion of the lavas themselves. PYROPHYLLITIZATION, PINITIZATION, AND SILICIFICATION As has been remarked before, a later series of local alteration processes has been superimposed on the already widespread slightly siliciied volcanics. ‘The origins of the pyrophyllite, the pinite, and 144 A. F. BUDDINGTON the quartz-pyrophyllite or quartz schists are so intimately inter- woven that all may be treated together. The proof that these rocks have originated through replace- ment is based on the following data: (1) the preservation of the structures of the primary rock in the secondary rock, (2) the pres- ence of unattached and unsupported portions of the country rock within the replacement products, (3) the introduction of large quantities of some elements and the solution of others without any notable change in volume or porosity, (4) gradational contacts, and (5) the massive homogeneity of all the rocks, and especially of the pyrophyllite, which does not show the foliated crystalline structure so characteristic of pyrophyllite veins which fill pre- existing fractures. To quote examples which belong to the first category, we find the following structures preserved in pyrophyll- itized rhyolite: (1) flow structure, (2) spherulites, (3) pebbles of a partially replaced conglomerate, and (4) perlitic structure (Fig. 5); these in the quartz-pyrophyllite rocks: (1) spherulites and (2) breccia structure; the following in pyrophyllite: (1) fragments of volcanic breccias, (2) pebbles of conglomerates, and (3) spheru- lites; while in the pinite and pinite schists we have preserved spherulites, traces of flow structure, axiolites, microspherulites, and perlitic structure (Fig. 6). Additional evidence of replace- ment is found in the inclusions and stringers of country rock within the pyrophyllite veins and the intimate manner in which the two are often intermixed. Not only this most convincing field evidence but chemical considerations prove almost conclu- sively that these rocks must have originated through replacement of rhyolite or rhyolitic volcanics. From a study of the foregoing field and chemical evidence, con- clusions have been drawn as to the genetic relationships of the eight different rocks described under ‘‘Chemical Analyses”? and ‘‘Petrography,”’ and as to the succession of processes which pro- duced them. This relationship is graphically represented by the following diagram, in which the numbers refer to the chemical analyses given under ‘‘ Chemical Analyses,”’ which may be taken as typifying the composition of the respective rocks: HYDRON SILICATES IN ROCKS OF CONCEPTION BAY 145 (z) Gray Rhyolite Dp “101 uoly Roy Vv (2) Silicified Rhyolite do1Ag lox Va (3) Pyrophyllitized Rhyolite (4) Pyrophyllitized Rhyolite UOI}eZ V \ (6) Quartz-pyrophyllite | (5) Pyrophyllite (7 and 8) Pinite Schist TEMPERATURE OF FORMATION OF PYROPHYLLITE Clapp (1914, 120) assumes that alunite and pyrophyllite are probably developed only under moderate conditions of pressure and temperature such as exist near the surface. Although this is a common mode of origin for both alunite and pyrophyllite, it is certainly not the only set of conditions under which the latter forms. For instance, pyrophyllite is noted by Dana (1909) as a mineral often forming the base of schists and gneisses, and by Lacroix (1895) as a mineral of the crystalline schists and Paleozoic meta- morphics. Artificially, K. von Chrustchoff (1894) obtained what he believed to be pyrophyllite by heating gelatinous silica, gelatinous alumina, and gelatinous zirconium hydrate in a platinum tube at 146 A. F. BUDDINGTON increasing temperature for six days. The product obtained was a zirconia-bearing pyrophyllite. Its specific gravity was 2.87 and it appeared in thin hexagonal plates, which he states are not true hexagonal plates unless there be optical anomalies. ALO; ZrO2 | H.20 | Total 23.76 14.54 | 7.86 | 99.81 Le Chatelier (1887) determined the points at which pyrophyllite loses its water by noting the points at which the temperature remained constant with absorption of heat and found two such points, the first at 700° and the second at 850°. From the foregoing data it is evident that pyrophyllite is a mineral which may form under conditions varying from the high temperatures of dynamic metamorphism to the near-surface tem- peratures and pressures of solfataric agencies. TEMPERATURE OF FORMATION OF PINITE Pinite, if considered as an impure sericite, as suggested by Clarke (1911), has a varied range of conditions under which it may form. Clarke states, however, that ‘‘the alteration [to sericite] is most conspicuous in regions where dynamic metamorphism has been most intense, high temperature, the chemical activity of water and mechanical stress all working together to bring it about.” A green micaceous mineral described as mariposite by Silliman, and whose composition, shown by two analyses, as suggested by Hillebrand (1895), resembles pinite, is characteristic of the mother lode in Tuolumne and Mariposa counties, California. Crosby (1880) describes pinite as a product of surface decom- position of petrosilex and felsites in the vicinity of Boston, Massa- chusetts. Bell (1887) found it at Ballater Pass interspersed through granitic rocks and along their joint planes, and ascribes its origin to the decomposition or alteration of orthoclase feldspar, an inter- mediate stage in its conversion into kaolin. HYDRON SILICATES IN ROCKS OF CONCEPTION BAY 147 Cole (1886) describes pinite occurring as an alteration product of spherulitic rhyolites, in conjunction with silicified spherulites. He suggests that thermal waters are responsible for the origin of both the pinite and quartz. From these references it is evident either that there is a differ- ence of opinion as to the conditions under which pinite forms, or that it is stable under widely variant temperatures and pressures. It is probable, however, that it demands higher temperatures and pressures than exist at the surface as conditions for its most favor- able development, and such is doubtless the case with respect to the pinite of Conception Bay. ALTERNATIVE DEVELOPMENT OF PINITE OR PYROPHYLLITE Since sericite or pinite is the usual product of hydrothermal alteration it is pertinent to inquire if any reason can be found why in certain cases pyrophyllite should be the product formed. A possible equation (1) representing the formation of sericite from orthoclase is quoted from Clarke (1911), and a similar possible equation (2) representing the formation of pyrophyllite from ortho- clase is given below: Gp) 56 ST Or HO 2A Or 2 Oe (2) 6 KAISi,0Os+3H.0s 6HAISi.0.+3K.Si0+3510, (pyrophyllite) From these equations three factors are suggested as the possible elements influencing the alternative development of sericite and pyrophyllite: (1) the effectiveness of hydrolysis, (2) the mass action of the excess silica in solution, and (3) the mass-action effect of excess potash in solution. The dominance of the first two factors would be conducive to the formation of pyrophyllite, and the dominance of the third factor would be favorable to the pro- duction of sericite. This may be illustrated more graphically, without however implying anything as to the actual mode of opera- tion, by writing the equation for the formation of pyrophyllite from sericite as a balanced reaction: . 2KH,AI,Si,O,.+6Si0.+H.O = 6HAISi,05+K.Si0, (sericite) (pyrophyllite) 148 A. F. BUDDINGTON If now the quantity of silica is present in the solution in large enough excess and the effectiveness of hydrolysis is relatively stronger, the reaction will produce pyrophyllite: 2KH.ALSi,O,.+-6Si0,+ H.0 > 6HAISi,05--K.SiO, (sericite) (pyrophyllite) while if K.SiO, or potash in some other form is present in large enough quantity the alternative reaction will take place and sericite will be produced: 6HAISi,06+K.SiO, > 2KH.AI,Si,0;,+6Si0,.+-H.O0 (pyrophyllite) (sericite) CHEMICAL PHENOMENA CONNECTED WITH ORIGIN OF PYROPHYLLITE From a comparison of the analysis (No. 2) of the silicified rhyo- lite with that of the pyrophyllite (No. 5), it will be seen that the change in composition has been such as might have been brought about essentially through three processes: (1) the introduction of alumina, (2) the replacement of the alkalies by hydroxyl], and (3) the solution of silica. The analyses 2, 3, 4, and 5, recalculated into their mineral composition, show a direct transition from the country rock (the silicified rhyolite) into pyrophyllite through a decrease in the quantity of quartz, feldspars, and impurities and a simul- taneous increase in the content of pyrophyllite. In order that the original rock may be so altered as to give the mineral analyses shown by the transitional rocks, it is necessary that metasomatic replacement of both the quartz and the feldspars should have pro- ceeded synchronously and at a much faster rate with respect to the quartz than with respect to the feldspars. This process would involve the introduction of large amounts of alumina, the gradual replacement of the alkalies by hydroxy] at a more rapid rate in the case of the soda than of the potash, and the solution of portions of both the silica existing in combination with other elements in the rock and that present as free quartz. The heterogeneous, blotchy character of the quartz-pyrophyllite rock when seen in thin section suggests that the rock may have been in the condition of a more or less homogeneous glass when acted upon by the silicifying and pyrophyllitizing solutions, in view of the fact that its chemical composition and mineral arrangement HYDRON SILICATES IN ROCKS OF CONCEPTION BAY 149 would involve the simultaneous replacement of the feldspars of a crystalline rock by silica and pyrophyllite, and of its quartz by pyrophyllite. ORIGIN OF PINITE AT MANUELS From a further study of the chemical analyses, it becomes evident that while the silicified rhyolites, pyrophyllitized rhyolites, and pyrophyllite have all decreased in their content of iron, mag- nesium, potassium, and soda, the pinite analysis (7) shows a decided increase in the first three of these elements. During the formation of pyrophyllite vast quantities of potash must have been liberated and carried in solution in the circulating waters. Is it not possible that away from the main channels these waters deposited their load as pinitic replacements of the rhyolite under the control of lower temperatures and pressures and the mass-action effects of the excess potash in solution? It seems reasonable to suppose that the pinite here was an essentially contemporaneous formation with the quartz-pyrophyllite schists and pyrophyllite, receiving some of the magnesia, potash, and iron released by the formation - of the pyrophyllite, as the quartz schists have received some of the silica originating at the same time. ORIGIN OF THE PINITE SCHIST AT HARBOUR MAIN The chemical analysis of the pinite schist recalculated for sericite gives the rock a mineral composition of about # sericite and ? quartz, feldspar, and other constituents. Examination of thin sections shows that this result has been brought about through the replacement of both the feldspar and quartz by sericite. If we consider the rock previous to pinitization to have had the com- position of the silicified rhyolite (No. 2), then the process cited involves the substitution of potash for soda in the feldspars, the addition of potash and alumina, and the subtraction of silica and soda. There is considerable evidence that this rock was not formed at the surface. There are no secondary products of decomposition, such as kaolin or limonite, associated with the pinite. It occurs in quantity only in certain zones and extends to a considerable depth, as exposed in a prospect pit for talc, south of Harbour Main. 150 A. F. BUDDINGTON Where it occurs as patches or specks it bears no apparent relation to the surface. Furthermore, the pinitic material has undergone dynamic metamorphism and is sheared and cleaved, while chemi- cally its origin involves the introduction of alumina and potash through replacement. GENERAL OBSERVATIONS It is quite probable that the foregoing processes operated under conditions of some dynamic movement, as their characteristic development is along shear zones, and their products assume a lenticular structure which is as characteristic of the minuter struc- tures of the rocks as of the veins themselves. Moreover, the rocks themselves have been sheared and possess a more or less prevalent cleavage, conditioned by the growth of their constituent minerals in more or less parallel arrangement. It is probable that the factors determining which of these three rocks—pyrophyllite, pinite, and quartz-pyrophyllite—shall form are to be found in the temperature, pressure, and chemical content of the solutions themselves, in the relative effectiveness of hydrolysis, and in the mass-action effects of the compounds in solution. The concentration of the potash in the pinite, and the silica in the quartz-pyrophyllite rocks may be accounted for on the theory of a redistribution of the elements, but the tremendous contribu- tions of alumina represented by the pyrophyllite, and to a minor extent by the pinite, must be accounted for otherwise. The close connection between the pyrophyllite deposits and the granite- Avondale Volcanics contact south of Manuels and between the pinitized rhyolites and the Woodfords monzonite stock is hence of significance. It may be that there is no genetic connection between these minerals and the intrusives, and that their formation was entirely dependent on the locus of fault zones in those localities. But there are numerous other profound fault and shear zones in this area with no exceptional alterations. Hence it seems probable that when faulting took place between the volcanics and the intrusive granite or monzonite, the still hot magmatic waters were released and found their way upward along these fault zones, either contributing the alumina directly, or per- HYDRON SILICATES IN ROCKS OF CONCEPTION BAY 151 haps indirectly through mingling with already aluminous solutions. _ Such solutions, if present, would be those which had accomplished the silicification of the volcanics with their attendant solution of alumina under static conditions and hence before the period of orogenic movement, which folded the volcanics and witnessed the intrusion of the plutonic batholith and stock. The evidence bear- ing on the exact dates of the periods of faulting and folding, however, is not conclusive. COMPARISONS WITH OTHER DEPOSITS Comparisons with other deposits show that, according to the descriptions, the pyrophyllite deposits of the Pambula goldfield, New South Wales, and of Chatham and Moore counties, North Carolina, are essentially similar to those in Newfoundland, and it is here suggested that possibly they have had a similar origin. © Clapp (1914) described quartz-pyrophyllite rocks from Kyuquot Sound, Vancouver Island, which may be taken as a type of pyro- phyllite deposits developed by solfataric agencies under conditions of temperature and pressure existing near the surface; while those of Newfoundland are a type originating under intermediate con- ditions of temperature and pressure. In the first case the pyrophyllite rocks are associated with alunite and in the latter case with pinite. In the Kyuquot deposits the original quartz of the replaced dacite has not suffered any loss except in one doubtful case, while the distinctive feature of the Newfoundland rocks has been the replacement of quartz in rhyolites by pyrophyllite. The two deposits are similar in that, in both cases, the rocks are ~ associated with intrusive batholiths, the one with a feldspathic quartz diorite and the other with granite. Both are metasomatic replacements of acid volcanics, while in the zone of alteration there seems to have been some transfer of material, and soda, lime, magnesia, and iron oxides have been lost in each case. CONCLUSION From the foregoing evidence the conclusion may be drawn that: the pyrophyllite, pinite, and quartz-pyrophyllite schists of the 152 A. F. BUDDINGTON Avondale Volcanics owe their formation to metasomatic replace- ment and alteration of previously silicified rhyolites or rhyolitic volcanics by thermal waters, under conditions of dynamic stress and intermediate temperatures and pressures, operating along channels primarily determined by fault or shear zones. Chemi- cally, the salient features of these alterations have been the intro- duction of alumina, the more or less complete substitution of the hydroxy! element in place of the alkalies, and the solution of soda. The solutions instrumental in causing these alterations may have been to a greater or less extent juvenile waters emanating from the intrusive granite batholith and monzonite stock at some period subsequent to the time of their injection. May, 1915 BIBLIOGRAPHY Bell, W. H. ‘“‘New Localities for the Mineral Agalmatolite with Notes on Its Composition,” Miner. Mag., VII (1887), 28. Chrustchoff, von K. ‘Artificial Zircons,” abstract in Miner. Mag., X (1894), 250. Clapp, C.H. “The Geology of the Alunite and Pyrophyllite Rocks of Kyuquot Sound, Vancouver Island,” Summary Rept. of the Geol. Surv., Dept. of Mines, for 1913 (1914), pp. 109-26. Clarke, F. W. “Data of Geochemistry,” U.S. Geol. Surv., Bull. 491 (1911), p..567. Cole, G. A. J. “On the Alteration of Coarsely Spherulitic Rocks,” Quart. Jour. Geol. Soc., XLII (1886), 186-87. Crosby, W. O. “Pinite in Eastern Massachusetts, Its Origin and Geologic Relations,’ Am. Jour. Sci., 19 (1880), 116. Dana, E.S. A System of Mineralogy (1909), p. 601. Hillebrand, W. F. Quoted by H. W. Turner (1895). Howley, J. P. Geologic Map of Newfoundland (1907). * Lacroix, A. Minéralogie de la France (1895), I, 471. Le Chatelier, M. H. ‘De l’Action de la chaleur sur les argiles,”’ Bull. Soc. Min. France, X (1887), 207. Powers, F. D. “The Pambula Gold Deposits,” Quart. Jour. Geol. Soc., XLIX (1893), 233-35. Pratt, J. H. ‘Talc and Pyrophyllite Deposits in North Carolina,’ V.C. Geol. Surv., 1900, Economic Paper No. 3, pp. 23-29. Turner, H.W. “Further Contributions to the Geology of the Sierra Nevada,” U.S. Geol. Surv., 17th Ann. Rept., 1895. Part 1, p. 520. THE ORIGIN OF RED BEDS A STUDY OF THE CONDITIONS OF ORIGIN OF THE PERMO- CARBONIFEROUS AND TRIASSIC RED BEDS OF THE WESTERN UNITED STATES: C. W. TOMLINSON University of Minnesota PART I PREFACE SUMMARY DESCRIPTION OF THE FEATURES CONSIDERED List of Formations Colors of Gypsum and Limestone Color in Clastic Strata Distribution of Greenish Colors Nature of the Coloring Matter Is THE FERRUGINOUS MATERIAL AN ORIGINAL CONSTITUENT OF THE SEDI- MENTS, OR A LATER INTRODUCTION ? Hypothesis of Introduction of Iron from Igneous Magmas Later Introduction of Iron by Meteoric Waters Deposition of Coloring Matter Contemporaneous with Sedimentation Microscopic Evidence Extent of Known Secondary Redistribution of Coloring Matter Is THE FERRUGINOUS MATERIAL IN THE SAME FORM AS AT THE TIME OF SEDIMENTATION ? Variations in Degree of Oxidation Origin of Mottling Cause of Gray and Green Bands in Red Beds: Barrell’s Hypothesis Organic Matter the Controlling Influence in the Case of the Western Red Beds Variations in Hydration of Ferric Oxide WAS THE COLORING MATTER A CHEMICAL OR A MECHANICAL SEDIMENT ? t A revision of a thesis presented in partial fulfilment of the requirements for the degree of Master of Arts at the University of Wisconsin, in June, ro14. 153 154 C. W. TOMLINSON PART If CONDITIONS OF DEPOSITION OF RED CLASTIC SEDIMENTS: MODERN TYPES Red Clay of the Deep-Sea Bottom Stream Deposits Derived from Pre-existing Red Beds Arkosic Stream Deposits Stream Deposits Deriving Their Coloring Matter from Ferruginous Residual Soils Terrigenous Marine Clastics Deposits of Desert Lakes or Playas Red Dune Sands EVIDENCE OF FEATURES OTHER THAN COLOR AS TO THE CONDITIONS UNDER WHICH THE RED BEDS WERE DEPOSITED Evidence of Conglomerates as to the Sites of Land-Masses Significance of Non-clastic Sediments The Clastics: Minor Structural Features The Clastics: Mineral Composition Evidence Supplied by Fossils Summary RELATION OF OROGENIC HISTORY TO RED BEDS SEDIMENTATION SUMMARY RAR SI PREFACE Certain general facts concerning the origin of the western Red Beds have been known for some time and have become incorpo- rated into current textbooks. There has been, however, much difference of opinion in the interpretation of some features of this remarkable group of sediments, especially as to the significance of the color itself. This paper is devoted to an investigation of the causes and history of the coloring matter, which, more than all other features put together, distinguishes the Red Beds from other sedimentary series. The investigation on which this paper is based’has been chiefly a study of the literature to which reference is made in the footnotes, together with all available published descriptions of the western Red Beds, and much miscellaneous literature dealing with related subjects. The writer’s first-hand acquaintance with the Red Beds has been gained from two summers of field work in Wyoming and Idaho (under the direction of Eliot Blackwelder, of the United THE ORIGIN OF RED BEDS 155 States Geological Survey), and from laboratory study of thin sections. For valuable suggestions and criticism and for review of the manuscript the writer is indebted to Messrs. Eliot Blackwelder, A. N. Winchell, W. J. Mead, and F. T. Thwaites, of the University of Wisconsin; to Messrs. W. H. Emmons, C. R. Stauffer, F. F. Grout, C. J. Posey, and A. W. Johnston, of the University of Minnesota; and to Dr. R. D. Salisbury, of the University of Chicago. SUMMARY DESCRIPTION OF THE FEATURES CONSIDERED List of formations.—Clastic sedimentary strata of reddish color outcrop or constitute the uppermost part of the bedrock over about 4 per cent of the area of the United States and exist beneath a cover of younger sediments under an additional area probably twice as large. Much the greatest volume of such strata is included in the single group of rocks which forms the subject of the present study: namely, those Red Beds which outcrop in many areas from Kansas and Texas to Arizona and Montana (Fig. 1). The forma- tions included in this group are related closely in age, ranging from Pennsylvanian to Triassic, or possibly Jurassic, and are prob- ably for the most part physically continuous. They comprise the Cimarron series of Kansas and Oklahoma, and the Wichita, Clear Fork, Double Mountain, Greer, Quartermaster, and Dockum beds of Texas; the Wyoming, Fountain, and Maroon formations of central Colorado; the Cutler and Dolores of southwestern Colorado; the Aubrey, Shinarump, Vermilion Cliff, and Moencopie of the Colorado Plateau, with the Saliferous and Zuni of the Zuni Plateau in New Mexico; the Spearfish, Opeche, and Chugwater of Wyoming and southern Montana; and the Ankareh and Nugget of south- eastern Idaho and northeastern Utah. Similar Red Beds constitute a large part of the sediments of the Newark series (Triassic) of the Appalachian Piedmont region— typified by the Stockton and Brunswick formations of New Jersey." They make up also, among others, much of the Catskill formation (Devonian) of eastern New York and Pennsylvania; the Medina and Clinton (Silurian) of New York; the Vernon shale (Silurian, a «See H. B. Kiimmel, Ann. Rept. of the State Geologist of New Jersey, 1896. TOMLINSON W. C. Sprg pay W19}SeM IY} Jo UOTNALYsIp Surmoys deyy—'t “ony "uo1s04q fq penouay Fea | suclyouso} sabunch yjveusg pring lm | aovfang ye pesedxy | ss “S$d3g G4u NYSLSAM SHL SPW 49 2189S uoy1d2foig 21U05 243 UD (sm) VOIMAWV | mae hie WE | is ose ieee THE ORIGIN OF RED BEDS Ts5 7, member of the Salina beds), the Bedford shale (Mississippian), and the Dunkard series (Permian) of Ohio; the Lake Superior sand- stone (probably Algonkian) of northern Michigan, Wisconsin, and eastern Minnesota; and the Belt series (Algonkian) of Montana and British Columbia. Occasional reference will be made in this paper to these formations, and to Red Beds in other countries as well; but this discussion applies especially to the western group, with which the writer is most familiar. ' Colors of gypsum and limestone.—The beds of gypsum occurring in many areas of Red Beds strata are described everywhere as remarkably pure and white, except where stained by a red coating washed down from overlying clastic sediments. The same is true of a majority of the limestones and dolomites occurring in the red series, most of which are gray or drab or bluish on fresh fracture. Even the famous Redwall limestone (2,500 feet thick) underlying the Aubrey Red Beds of the Grand Canyon section is gray on fresh fracture; its surface color is due to concentration of iron oxide by weathering, to wash from above, or to both of these causes. Exceptions are found in limestone or dolomite bands in the lower Chugwater in Wyoming,’ and in the Minnekahta limestone? of the Hartville quadrangle, which are characteristically purplish or rosy gray; and locally in certain limestones in northern Oklahoma, where they are in transition to sandstone.* These limestones are in most localities nearly or quite barren of fossils. Color in clastic strata.— Another and perhaps a yet more signifi- cant fact is the variation of hue among the clastic strata. Inter- bedding of greenish, gray, or buff beds with red sediments is found in a majority of Red Beds sections; notably in the Saliferous of the Colorado and Zuni plateaus, in the Dockum group of Texas, and in « G. K. Gilbert, ‘Geology of Portions of Nevada, Utah, California, and Arizona,” U.S. Geog. Surveys W. of the tooth Meridian, III (1875), 177-78. 2 Eliot Blackwelder and C. W. Tomlinson, “Field Notes on Work in Western Wyoming, roto and 1011,” unpublished; property of the United States Geological Survey. 3W. S. T. Smith; Hartville Folio (No. 91), Geol. Atlas of the U.S., U.S. Geol. Survey, 1903. 4J. W. Beede, ‘‘The Neva Limestone in Northern Oklahoma,” Okla. Geol. Survey Bull. No. 21, 1914, p. 24. 158 C. W. TOMLINSON the Red Beds of the San Juan region and of the Anthracite, Crested Butte, and Tenmile quadrangles of Colorado. Alternation of paler with darker and with brighter shades of red is remarked in almost every occurrence of Red Beds. In all these instances it is significant that the color boundaries tend to follow bedding planes, LEG INES Put ple Plates an apr ee e IS Blark |Slates 2 aks SI f o Be aN nh . ‘ ef Aves *, < oS e he SSK = es Red ate c H Ba RS ‘ ¢ i: Be ie ° aN VS = cy £ NG o 7\. (2) xx x Swe RAS w Be : Lf y 4 \ Ee oe »/ \4AY we i ee Ve Sa ~ ry x \x Ne *9 d ™ < ES eA an SSS ee EN “ rf c < bs Af a S x x x of \e ss ver > N mane \ ‘ ‘ £3¢ \ N . x 5 = sg i a 3 Iv a 7 N x X \ N ‘ ‘ \ Beery lt eh 2 3 Percent of Ferrous Iron (Fein FeO). Fic. 2.—Diagram to illustrate the relation between color and the proportion of ferric to ferrous iron in ferruginous slates. Analyses (from Dale) are lettered to correspond with text: and usually accompany changes in coarseness of grain. Many of the color boundaries are distinct, even planes, but they may be rendered irregular by downward migration of coloring matter. Where alternations of light- and dark-red strata occur, the more deeply colored beds are in most cases of finer grain than the THE ORIGIN OF RED BEDS 159 others. The occurrence of coarse-grained massive buff sand- stones in a series of maroon or chocolate shales has been noted by many writers. ‘This association holds true in many other Red Beds besides the group here under consideration. Thwaites' reports it as an almost unfailing relation in the Lake Superior sandstone series of northern Wisconsin, and Geikie? mentions its existence in _ the Triassic New Red Sandstone of Great Britain. The respective tints are understood to be uniform throughout the strata in which they are noted, and not to be merely the surface staining of the beds, which might have a very different origin. Distribution of greenish colors.—Greenish tints occur in the Red Beds as a fairly even color through continuous beds interstratified more or less closely with red strata, in streaks and blotches in red strata themselves, and occasionally in strips following joint planes. The completely greenish strata, and likewise the mottled beds, include both shales and sandstones, and are described from many districts. Green spots in red strata are as a rule irregularly ellipsoidal in shape, somewhat flattened parallel to the bedding, and indefinite in outline. They may vary in diameter from a few millimeters to several inches. Variations in color along the strike are as common and as much to be expected as similar variations in texture, cross-bedding, or any other stratigraphic feature. The Red Beds of the Southwest are noted for their inconstancy and frequency of change along the strike; those of Central Wyoming and the Black Hills are fairly continuous in lithologic characters for considerable distances. Where other features are variable, the color is variable also; and where other features are constant, color likewise is constant. Nature of the coloring matter.—Available data as to the chemical composition of the green bands and spots and other variations in color in the western Red Beds are very meager; but many of the same phenomena occur in roofing slates, whose commercial value has been the cause of painstaking investigation of their occurrence and character. The close dependence of color upon chemical 1F, T. Thwaites, ““Sandstones of the Wisconsin Coast of Lake Superior,” Buil. Wis. Geol. and Nat. Hist. Survey No. 25, 1912, p. 31. 2 Archibald Geikie, Text-Book of Geology (London: Macmillan, 1903), p. 1064. 160 C. W. TOMLINSON composition is brought out strikingly by the following analyses, which are at least sufficient to show that the differences in color in the Red Beds are caused by the same differences in chemical compo- sition as those which cause corresponding differences in color in slates. The analyses are plotted graphically in the diagram (Big. 2). The following analyses are taken from Dale:’ Fe.0; FeO Net Fe |Ratio Fe’”’: Fe’ Free Carbon bau (a) yc eu sn chtiane: oe 5-30 I.20 4.69 | 4.04 :1.00 il BaP rc 5.61 1.24 4.89 | 4.09 (No free carbon found fais Sys Wee ester 3.48 I.42 3.54 2.29 in any of the red or 3 NOP eee 3.86 T44 3.83 | 2.42 purple slates) 1 lad es eee 7.10 1.00 Saas 6.37 Average...| 5.26 it 4.62 giclose” Gitatole) 22 G, eae: 4.10 Dont 4.99 Te e{0) Qamatere) pies baeoee coe 5.16 2.54 550). lb 84 To 8 Erect oe? 6.63 1.20 5-58 | 5.00 Average...| 5.30 Dy ils Suelo |) Qewe Sige} AT aa ere 0.81 4.71 Ano 3teal OnD5 05100 S| Biba secre 1.34 5-34 x00)" 102.220 Trace 2 Cece at 1-23 4.73 4.54 ©. 234 Trace rep WEG tocevare ci Sietecs Ter 6.58 Han. || Ce uss 5 Hin Sechaceal 1.24 6.81 6.16 | 0.165 co Olam arcterae 1.47 3.81 3.99 | 0.348 DG) OB Bo Miers eee Te33 5.64 agi 0.213 Trace DIVES ae 2.24 4.07 4.73 ©.406 AVETALCs se lsat ss aoe AROOWMEON232n TROO | et a | 0.52 4.87 4.15 | 0.095:1.00 | 0.46 TE INU cee see ode oe 9.03 7.02 | 0.000 1.79 ao | Re eas 1.98 3.65 4.23 | 0.490 0.77 OU rseter 0.53 3.52 Seu 0.135 1.54 Average...| 0.75 Seo AGO2 | NOnt20).0OOmleteetA Se ES ric 3.86 1.44 3L03) 242) S100 sa OF esi igare 1.79 1.19 De Sw a | 3 4 le OO Gans | Ree 1.09 1.06 1.58 | ©.027:1.00 * K2 is same slate as K, but finer grained. M-=red slate near green spot. Q=purple rim of spot. R=green part of spot. ™T. N. Dale, “The Slate Belt of Eastern New York and Western Vermont,” Ann. Rept. U.S. Geol. Survey No. 19, 1899, Part 3, pp. 232, 246-53, 257, 264; and “The Roofing Slates of the United States,” U.S. Geol. Survey Bull. No. 275, 1905, PP. 34-36. THE ORIGIN OF RED BEDS 161 The following analyses are taken from W. J. Miller: VERNON FORMATION, CENTRAL NEW YORK Fe.0; FeO Net Fe | Ratio Fe’” : Fe” IRGol Gk es 5-6 o ole emule wmminloe 2.25 0.75 2.16 D2 OO Green spot in red shale........ 0.00 1.19 0.93 0.00:1.00 The following analyses were furnished by Eliot Blackwelder? CHUGWATER FORMATION, WIND RIVER VALLEY, WYOMING Fe.0; | FeO Net Fe Ratio Fe’” : Fe” Nedisandstone we ee eneee 3.50 T.04 3.26 3.02:1.00 Greenish sandstone*.......... THOR 1.04 T53 0.89:1.00 *The greenish sandstone of the Chugwater in this case is a strip of the ordinary red sandstone, leached along a joint crack. The following analyses are taken from Richardson :3 SPEARFISH FORMATION, BLACK HILis Fe.0; FeO Net Fe Ratio Fe’” : Fe” (Creentshalesars cesar eee: 1.85 1.04 2.02 I.78:1.00 Red shale, adjacent to green... 4.601 1.24 4-55 BE 721 OO Redyshal eee aati m seine natal: 3.64 0.65 3.05 5.04:1.00 Wedkshalewieie yee Meee 2.04 0.18 1.57 10. 20:1.00 The black color in the slates is due, not directly to any peculi- arity of the iron content, but to the presence of carbonaceous matter, which incidentally brings about the reduction of iron oxide to the ferrous form. Black shales are very rare in the western Red Beds, but highly carbonaceous and even coal-bearing strata occur in the Newark series of the Atlantic Piedmont. The typical Newark clastics are quite intensely red, many of them with a purplish tone, but the carbonaceous strata are always gray or black. The color of the prevailing red strata in the Red Beds series is due to the presence of ferric oxide. The iron of the coloring matter tW. J. Miller, ‘Origin of Color in the Vernon Shale,” Bull. N.Y. State Museum, No. 140, in 63d Ann. Rept. N.Y. State Museum, 1909, 1, 150-56. 2U.S. Geol. Survey, Division of Chemistry, Analysis No. 2530. 3G. B. Richardson, ‘‘The Upper Red Beds of the Black Hills,” Jour. Geol., XI (1903), 365-93. 162 C. W. TOMLINSON is chiefly in the ferric state in light- and dark-red, buff, and yellowish strata. A gray or green color, on the contrary, signifies a low pro- portion of ferric oxide, and usually a preponderance of ferrous over ferric compounds. The mineral composition of the coloring matter is difficult to determine accurately because of its fine division. A green color is “common with silicates in which ferrous iron is prominent,’ and silicates may be important where the ratio of ferrous to ferric iron is high. IS THE FERRUGINOUS MATERIAL AN ORIGINAL CONSTITUENT OF THE SEDIMENTS, OR A LATER INTRODUCTION ? Hypothesis of introduction of iron from igneous magmas.—lt the ferruginous matter was an integral part of the original sedi- ments, we have no more difficulty in explaining its presence than in accounting for any other common mineral constituent of sedi- mentary rocks. Its source was most probably in the same rocks which gave rise to other materials of the series, such as quartz, feld- spar, and calcite, and the agencies of transportation were the same as those which were responsible for the entire series. If, however, we postulate that the iron has been introduced since the comple- tion of sedimentation, after the series of Red Beds was otherwise complete, we shall find it very difficult to reconstruct in imagina- tion any natural agency which might have brought about such a result, and we shall be puzzled to find an adequate source for this vast amount of iron. There is one hypothesis of this kind which received credence and vigorous support in America for several decades of the nine- teenth century, in explanation of the red stain in the Newark series of the Connecticut Valley. The close association, in that series, of clastic red sediments with contemporaneous extrusive and intrusive basic igneous rocks of great thickness and extent made it a natural suggestion that the exceptional color of the sedimentary members was connected directly with igneous action. ‘The absence of any 'E.S. Dana, A Text-Book of Mineralogy (New York: John Wiley & Sons, i909)» 10s 2A 2See J. D. Dana, Manual of Geology, ed. 1880, p. 764; and also his review of Russell’s bulletin on the “‘Subaerial Decay of Rocks” (U.S. Geol. Survey Bull. No. 52, 1889), in Am. Jour. Sci., 3d Ser., XX XIX (1890), 310. THE ORIGIN OF RED BEDS 163 strong evidence in support of this hypothesis, aside from that of association with igneous rocks, and the discovery of various lines of evidence (among those outlined below) in direct opposition to it, however, caused it to fall into disrepute. In the case of the western Red Beds there is not even the association with igneous rocks to suggest such an explanation. There are no contemporaneous igneous rocks in any part of the Red Beds group, and in the greater part of the area in which the group occurs there are no later igneous rocks known. In no case is there a relation comparable to that in the Connecticut Newark. Where the effect of later intrusions upon Red Beds has been observed carefully, that effect is not to heighten, but to destroy, the red color. This action has been noted in the Tenmile district* and in the Anthracite-Crested Butte district,? Colorado. Later introduction of iron by meteoric waters——There is no apparent reason why the Red Beds should have been favored by post-sedimentational iron-bearing solutions while other clastic series in the same region, both older and younger, were not stained. It might be expected that the coloring matter would be most abundant in the most pervious strata, especially along the lower surface of such strata, where they are in contact with less pervious rocks. It is quite true that color boundaries in the series follow bedding planes very closely; but, unfortunately for the hypothesis of later introduction, the more highly ferruginous strata of the Red Beds are as a rule, and with few exceptions, the more impervi- ous. As noted above (pp. 158-59), in a series of alternating sand- stones and shales it is almost invariably the shales which are deeper in color. Ferric hydrates are not being introduced into the Red Beds strata along the present joint planes. The analyses of Chugwater sandstone given on p. 161 (see also footnote) show a marked leaching of iron along a joint. Nor is iron commonly concentrated along joints in paler sediments underlying the Red Beds. «S, F. Emmons, Tenmile District Special Folio (No. 48), Geol. Atlas of the U.S., U.S. Geol. Survey, 1896. 2G. H. Eldridge, “Description of the Sedimentary Formations,” Anthracite- Crested Butte Folio (No. 9), Geol. Atlas of the U.S., U.S. Geol. Survey, 1894. 164 C. W. TOMLINSON Deposition of coloring matter contemporaneous with sedimentation. —Tf the ferruginous material which furnishes the color was con- centrated in the sediments chiefly at the time of their deposition, this association of high color with fine sediment is explained readily, as follows: In the weathering of igneous rocks, ferruginous material is separated out chiefly by the chemical decomposition of iron-bearing silicate minerals, and is therefore at the time of separation in a very fine state of division. At the same time ferrous salts of iron usually are altered to ferric oxide or hydrate. During surface transporta- tion and sorting it is segregated in whole or in part, by reason of this fine division (if it persists) and in spite of its high specific gravity, along with other finely divided materials constituting muds and “clays.” This does not apply to iron occurring in the parent rock in the form of magnetite or other very stable minerals, which in most cases are concentrated with the coarser products of mechanical disintegration, such as sands and sandy shales. Ferruginous materials firmly cemented to sand grains during weathering may also be transported and deposited with the sand. Iron taken into solution will have yet a different history. In the weathering of sedimentary rocks, the behavior of their ferruginous content is dependent upon the behavior of that material in a former sedimentary cycle, and thus ultimately upon the con- ditions already outlined for the weathering of the igneous rocks. Ferric oxide, the form in which iron occurs most abundantly in sediments, is, because of those conditions, usually finely divided, and in a second cycle of transportation and deposition will be con- centrated again chiefly with the muds. In so far as assortment is imperfect, the ferruginous material may be deposited with any type of sediment. Because of the usual absence of any commercial value in the series, there is an unfortunate dearth of analyses of Red Beds shales and sandstones. The fact of the concentration of ferruginous matter in the fine-grained sediments is well illus- trated, however, by composite analyses of shales and sandstones THE ORIGIN OF RED BEDS 165 from all sources. The following figures from Clarke’ are to the point: Percentage | Percentage €:03 FeO Teen Average of 78 shales (mean of two composites) . . 4.03 2.46 6.40 Composite analysis of 253 sandstones........... 1.08 0.30 1.38 See also analyses K and K’, on p. 160, supra. Microscopic evidence.—Russell? has presented evidence to show that the red coloring matter in sandstones of the Newark series in Virginia occurs as a coating on the sand grains, and that it existed in that form even before the transportation of the sand from the residual soils from which it was derived. Since much of his argu- ment would apply equally well to the red sandstones of western Red Beds, it is worth investigation in this connection. In a few thin sections of Newark sediments from Virginia which have been available for study by the writer, it is shown clearly that ferruginous matter exists there, both in the form described above and as a later interstitial filling between grains. Many of the sand grains, both of quartz and of feldspar, are well rounded. Most of them are surrounded by a thin coat of hematitic material, whose outer surface is smooth and even, but whose inner border may show slight irregularities penetrating into the body of the grain. This red film may not be entirely continuous around the grain; and this fact, together with the smoothness of its outer surface, suggests, as Russell maintained, that the coating was acquired by the sand grain before its final deposition, and that the coating suffered wear during transportation, without being completely removed. In a number of instances it was found that part of the cementa- tion of the rock had been accomplished by enlargement of the original grains of quartz and feldspar in optical continuity, outside of the red coating; and that outside of the enlargements there existed other bodies of hematitic matter, irregularly scattered 1. W. Clarke, ‘‘The Data of Geochemistry,” 2d ed., U.S. Geol. Survey Bull. No. 491, 1911, chaps. lil, v, vi. 21. C. Russell, ““Subaerial Decay of Rocks,” U.S. Geol. Survey Bull. No. 52, 1880. 166 C. W. TOMLINSON through the interstitial spaces. This material may have filtered in during the process of cementation; or it may have been present at the time of sedimentation and have been thrust out of the way by the growing crystals during enlargement. Occasional specks of red-stained material are found within the area of the enlargements and distinct from the inner coating of the original grains; and where the enlargement is missing, the interstitial hematitic matter may be difficult to distinguish from that of the primary coating. The stained areas probably do not represent pure ferric oxide, but ferric oxide in such ratio to silica or clayey material as to render the mixture nearly opaque and quite uniformly red or reddish brown; for chemical analyses of the same rocks show the percentage of ferric oxide in the rock to be much lower than the percentage of red- stained area in the sections. A similar microscopic investigation has been made by Richard- son’ in his study of the Spearfish formation of the Black Hills. He says: ‘Amorphous red pigment is prominent in the slides. It irregularly coats and spots the minerals, and . . . . constitutes the chief interstitial substance.”’ He also presents an analysis to show that the ferric oxide of the pigment is essentially anhydrous. Thin sections of Red Beds from all parts of the West are not available at present, and it has not been practicable, therefore, to make a thoroughgoing microscopic study of the group. Richard- son’s description suggests a relation of pigment to cementation similar to that found in the Newark sandstones: a twofold relation, indicating that part of the pigment was transported and deposited as a coating on sand-grains, and that the remainder constitutes an important part of the cementing material of the rock and is coeval with the rest of that material. But what was the time of cementation? Was this process com- pleted before the exposure of the Red Beds series to erosion, or is it still going on, or did it cease at some intermediate time? In so far as this question affects the iron content of the Red Beds we really have answered it already; for if any considerable part of the iron had been introduced as a cement later than the time of sedimenta- tion it would not be found most abundantly in the strata which r Op. cit. 2 Tbid., p. 379. THE ORIGIN OF RED BEDS 167 afford the least ready passage to water, as is actually the case. The cementing material of the Red Beds, in so far at least as it is ferruginous—and the remainder of it has little or no bearing upon the point in question—was therefore for the most part present in those beds at the time they were deposited as sediments. | Extent of known secondary redistribution of coloring matter.— Although the evidence is convincing that the ferruginous matter in the Red Beds was an integral part of the original sediment, and that it was deposited originally in the series in practically its present distribution and arrangement, it is equally certain that there have been some later modifications of that primary distribution. A large majority of all variations of color within the Red Beds are limited by bedding planes; but there are numerous minor varia- tions in color due to the migration of coloring matter along the lines of movement of ground-water." Reducing solutions locally extract ferruginous matter along joints, or cause a general down- ward movement of iron. In the western Red Beds, as in Barrell’s section of the Catskill formation in Schuylkill County, Pennsyl- vania,” the lower boundaries of many deeply stained bands are drawn less sharply than the upper. All of these are very minor features, however, in the distribution of the coloring matter of the series as a whole. The analyses (p. 161) furnished by Blackwelder illustrate the effect of leaching along a joint; the color changes from red to greenish, the net iron content is reduced from 3.26 per cent to 1.53 per cent, and the ratio of ferric to ferrous iron drops from BHO2 ICO LOOs60)1 200: IS THE FERRUGINOUS MATERIAL IN THE SAME FORM AS AT THE TIME OF SEDIMENTATION ? Variations in degree of oxidation.—Is the color due to recent oxidation in weathering? It has sometimes been asserted that the color of the Red Beds is a superficial matter, due to the weathering of originally dull-colored sediments. In support of this idea drill records have been quoted, showing that beyond a certain depth *Cfi. Richardson, op. cit., p. 376. 2See Joseph Barrell, “The Upper Devonian Delta of the Appalachian Geo- syncline,” Am. Jour. Sci., 4th Ser., XXXVI (1913), 437 ff. 168 C. W. TOMLINSON below the surface the rocks are no longer red. S.F. Emmons states, respecting the Maroon formation of the Tenmile district, Colorado," that ‘‘in depth, as shown in underground workings, the red color generally gives way to a greenish gray.’ In another paragraph of the same folio, Emmons states that igneous or hot-water alteration destroys the red color. Inasmuch as the ores of this district are related intimately to igneous and hot-water action, it is natural to suppose that mine workings would be the most likely of all places in which to find such effects. The relation here described is probably a local and abnormal phenomenon. In his description of the sedimentary rocks of the Anthracite and Crested Butte quadrangles, where the general situation is similar to that in the near-by Tenmile district, Eldridge’ states that “the upper division [of the Maroon conglomerate] is of a peculiar red or chocolate color, except in regions of local metamorphism”’ of the kind mentioned also by Emmons. Drilling explorations in the oil regions of Oklahoma and Texas recently have given us some additional information on the behavior of the color with depth. Gould? tells of a well some 3,300 feet deep in southeastern Oklahoma, in which ‘‘the last of the typical Red Beds was encountered”’ at a depth greater than’ 2,000 feet. The underlying strata were dull-colored sediments like those outcropping along the eastern (lower) margin of the Red Beds. Nowhere does he mention any change in color due to depth. ° A well sunk 3,095 feet at Ashland, Wisconsin, passed through typical red sandstones of the Lake Superior group all the way, without any suggestion of a change in color with depth.‘ It is to be remembered that most of the western Red Beds series are not colored uniformly throughout, but include many lighter-colored strata; and that many of the Red Beds successions include gray and green members. Any drill cutting through such a series would find, of course, changes in color with depth, but they would not be progressive, and they would have no causal connection with depth whatsoever. OPN Gis 2 Op. cit. 3C. N. Gould, “Petroleum in the Red Beds,” Economic Geology, VIII (1913), 768-80. 4 Data from F. T. Thwaites, op. cit. THE ORIGIN OF RED BEDS 169 This variegation itself constitutes one of the most unanswerable arguments against weathering in the present cycle as the cause of the red color. The distribution of gray and green beds among the red does not appear to have any definite relation to coarseness of grain, as in the case of intensity of the red color, nor to any other single stratigraphic feature; and it certainly bears no constant relation to the topographic surface. Origin of mottling —The green spots present in some otherwise red strata have been explained, on the hypothesis that the red color is due to recent oxidation in weathering, as remnants of the original color of the beds; but, as will be shown presently, it is much better in accordance with fact to explain them as spots in originally red sediments, deoxidized by the agency of some particle of organic matter which was present in the original sediment. Dale? discusses the origin of the green spots in red and purple slates in part as follows: The difference in color from the green to purple to red is manifestly due to the differences in the amount of hematite. [See analyses, p. 160.] The green fossil impressions in purple slate may throw some light on the origin of these spots. In this case the effect of organic matter, whether the carbonaceous matter of the lining of an annelid boring or from a marine alga, has been to diminish the quantity of Fe.O, in the slate. The increase of the carbonates may be directly connected with the production of CO, by decaying organisms and the consequent decrease of the Fe.O;. In view of all these facts and indica- tions, the spots may be safely regarded as probably produced by chemical changes consequent upon the decay of organisms. The same conclusion is reached by Miller with respect to green spots in the red Vernon shale (Silurian, central New York). Miller’ finds dark organic centers in many of the spots, and attributes the color of green shales associated with the red strata to more abundant dissemination of organic matter. It is probable that a very small quantity of organic matter may reduce or prevent the oxidation of a considerable amount of iron. The ferruginous matter necessary to stain a sediment is so small in amount that the quantity of organic matter necessary to accom- plish the reduction of a patch less than an inch in diameter, like «T. N. Dale, The Slate Belt of Eastern New York and Western Vermont, Ann. Rept. U.S. Geol. Survey No. 19, 1899, Part 3. 2 Op. cit. 170 C. W. TOMLINSON most of the green spots in Red Beds, is so small that it could easily be effaced or removed. A tiny fragment of vegetable fiber or the remains of a few minute organisms of any kind probably would suffice. Cause of gray and green bands in red beds: Barrell’s hy pothesis.— Barrell’ states that in the Catskill formation of eastern Penn- sylvania gray and green colors are typical of sandstones, and red colors of shales. He therefore suggests a causal relation between coarseness of grain and condition of the iron content, as follows: These relations show that there was a tendency toward deoxidation during the formation of the beds of sand, of oxidation during the deposition of the Catskill muds. Where the clay and iron oxide were sparing in quantity, the deoxidation was effective. The conditions which accompanied the deposition of clay and iron oxide also permitted oxidation to dominate over deoxidation.? The lack of oxidation of the iron in the sandstones, in spite of its lesser quantity, suggests that more abundant ground-waters in the sands may have kept out the air and permitted the organic matter to accomplish its effects, or perhaps that here the ratio of organic matter was in excess of the ferric oxide.s A few rare carbonaceous streaks have been observed in the Catskill and the plant impressions are in places found in deoxidized shales. Coaly and pyritiferous plant fossils are also preserved in some of the olive sandstones.4 No actual remnants of organic matter are reported to have been found in red strata, though markings interpreted by Barrell as rootlet marks are noted by him in certain horizons of red shales. That physical Conditions of deposition alone should have favored oxidation in the finer-grained sediments and retarded it in the coarser seems highly improbable. Where meteoric water moves most freely, there the most oxygen is carried in solution. The more rapid the circulation of ground-waters, the more effective are those waters as oxidizing agents, instead of vice versa, as suggested by Barrell. For the lesser quantity of iron in the sandstones we already have offered an explanatory hypothesis (p. 164). The smaller the amount of ferric oxide present in a sediment, the smaller, of course, is the amount of organic matter necessary to bring about its reduc- tion—or the reduction of a sufficient part of it to mask the color of the remainder. If organic matter were distributed equally among T Op. cit. 2 Tbid., p. 458. 3 Tbid., p. 460. 4 Ibid. THE ORIGIN OF RED BEDS 17/1 all the sediments of the series, it would cause gray and green colors in the sands rather than in the muds, by reason of the mechancial concentration of ferric oxide in the latter. The ratio of ferruginous materials to organic matter seems to be the dominating factor in determining the colors of ferruginous beds. Barrell’s second suggestion in the foregoing quotation, dealing with this point, deserves more careful investigation. Much of the organic matter carried as sediment in surface waters tends to be concentrated with the muds, for reasons analo- gous to those already suggested in the case of ferruginous material (i.e., fine division); but the proportion so concentrated is probably smaller in the case of organic matter than in that of ferruginous material, because of the indiscriminate distribution of driftwood, the growth of vegetation in most regions of clastic deposition, and other similar factors. In any of these cases organic matter in large quantities may enter into the composition of sands and muds alike, and thus bring about a higher ratio of organic to ferruginous matter in the sands than in the muds. These controlling factors are so complex that no constant rela- tion between coarseness of grain and distribution of organic matter is to be expected. It is evident from the quotations on p. 170 that definite organic remains in the section there under consideration are confined chiefly to gray and green strata; and it is also evident that such strata include both shales and sandstones. An examina- tion of the detailed section! of the Catskill formation published by Barrell in this paper reveals a somewhat less uniform relation between coarseness of grain and color of beds than one would understand to be the case from reading the text or the labels on the graphic columnar section. Of the 3,864 feet of beds definitely described as of one or the other type, 2,179 feet (56.4 per cent) bear out Barrell’s generalization that gray and green colors are typical of sandstones and red colors of shales, gor feet (23.3 per cent) are noncommittal, and 754 feet (20.3 per cent) are in opposi- tion to the rule. This variability is in better accord with the com- plexity of the factors controlling the distribution of organic matter in sediments than a more constant relation would be; and suggests t Barrell, op. cit., pp. 451-50. 2 [bid., p. 457- 11s C. W. TOMLINSON that even here the distribution of organic matter may be a con- trolling influence in determining the colors of the individual strata. Dawson,’ in discussing the Triassic (?) Red Beds of Nova Scotia, says of the gray sandstones and shales interstratified with them, that ‘‘where thick, they always contain either fossil plants, bituminous matter or thin seams of coal, or all of these. The fol- lowing sentence from Geikie,? relative to the Old Red Standsone of the British Isles, is also interesting in this connection: ‘‘It may be observed also that where gray shales occur intercalated among the red sandstones and conglomerates they are often full of plant remains, and may contain also ichthyolites and other fossils which are usually absent from the coarser red sediments.”’ Organic matter the controlling influence in the case of the western Red Beds.—Nowhere in the literature on the western Red Beds is there suggested such a definite and relatively constant association of green and gray colors with sandstones, and of red with shales, as that which Barrell sees in the Catskill formation, and as that which is described as occurring in the Siwalik formation of India.’ In the foregoing quotation from Geikie, the opposite relation is implied in the Old Red Sandstone series of Great Britain. In the Red Beds of the western United States variegation is perhaps more common in shales than in sandstones, though it occurs to a marked extent in both. The distribution of gray and green colors in the Red Beds coincides very closely with the distribution of organic remains in the same series, in so far as such remains are present; and this close association, together with the chemical probabilities of the case, suggest that organic remains now obliterated explain at least the greater part of the remaining gray and green areas and strata. The decolorization of these sediments may, therefore, have been complete before their burial under later strata. The «J. W. Dawson, ‘‘On the Colouring Matter of Red Sandstones and of Greyish and White Beds Associated with Them,” Quar. Jour. Geol. Soc. London, V (1848), 25-30. Quotation from p. 26. 2 Geikie, op. cit., p. 1003. 3 Medlicott and Blanford, A Manual of the Geology of India, II (1879), 524-26. Quoted by Barrell, op. cit., pp. 463-64. 4 Cf. Permian of the Pecos Valley and of the Zuni and Colorado plateaus; and the Jura-Trias Painted Desert sandstone of the latter plateau. THE ORIGIN OF RED BEDS 173 occurrence of traces of organic matter in red strata may be explained by an unusually high content of ferric oxide in those strata, or by later reoxidation of iron at one time in the ferrous state. The fact that much the greater part of the occurrences of colors characteristic of ferrous compounds are in distinct beds with definite boundaries, indistinguishable in most other characteristics from other beds which are not so colored, is good evidence that this distribution oi ferrous and ferric compounds, or of the substances.responsible for — these compounds, was accomplished for the most part at the time of sedimentation. It is well to remember that the gray and greenish strata are very subordinate in the Red Beds of the western United States. The general conclusion to be drawn from the preceding dis- cussion is that there has been in the western Red Beds no extensive change of ferrous to ferric iron, or vice versa, since the time of sedimentation; and also that the original distribution of these compounds in the series was influenced most largely by the dis- tribution of organic matter. Variations in hydration of ferric oxide.—That various degrees of hydration exist in the ferric oxide of the Red Beds today is clear from the variety of red, brown, and yellow hues which appear in some members of the group. The major part of the ferric oxide in the Red Beds is no doubt but poorly hydrated.’ The bright- and deep-red and red-brown colors (which are most common in the western series) may be attributed to hematite (anhydrous Fe.O;) or to turgite (2Fe.0;-H.O). The lighter browns, yellow-browns, and yellow tints are referable to gothite (Fe,0O;-H.O) or limonite (2Fe,0;-3H.O), or possibly in some cases even to xanthosiderite (Fe.O,- 2H.0). The freer passage of water through the sandstones as com- pared with the shales makes the constituents of the former, after consolidation, more susceptible of hydration than those of the latter. It is entirely probable that in some cases at least this factor of porosity heightens the contrast in color between coarse and fine sediments; but it apparently has not affected the greater part of the Red Beds, in which the ferric oxide is relatively anhydrous. Van Cf, Richardson’s investigation, discussed on p. 166. 174 C. W. TOMLINSON Bemmelen has shown that chemically prepared ferric hydrate, after being partially dehydrated, if placed in a medium saturated with water vapor, at ordinary temperature, takes up again part of the lost water.t. Brescius went so far as to say that “when nearly dry, ferric hydrate has almost as great a tendency to take up water as oil of vitriol itself.’”” This does not appear to be true, however, of hematite found in nature. Once completely dehydrated, ferric oxide becomes a stable compound. It remains to inquire into the means by which dehydration of the more hydrous compounds might have been accomplished to produce the low hydrates, in case these were not originally in the same condition. The first agent of dehydration which presents itself is that of heat. Elsden makes the following statement: The influence of temperature and moisture upon the iron hydrates is well known. In the case of laterite in India, the yellow xanthosiderite soon weathers to reddish-brown turgite, owing to dehydration. In the hot and arid regions ot South California the soils are dark red in colour, the iron being in the form of hematite instead of the hydrous forms, géthite or limonite. Dehydration also takes place in the hot regions of the Southern Appalachians, where the air is comparatively humid. It is only the deeper portions of the soil which retain the iron in a hydrated form. In all of the above mentioned cases, the source of the heat which produces the reaction in question is the sun’s rays. Its action in the soils is limited to a superficial stratum rarely more than 15 feet in maximum depth.? Obviously, the direct influence of insolation cannot be responsible for any extensive dehydration in the Red Beds, whose characteristic colors are known to extend to depths of more than 2,000 feet. «J. M. Van Bemmelen, “Sur le colloide de oxyde ferrique,” Recueil des travaux chimiques des Pays-Bas et de la Belgique, VIL (1888), 112. > 2E. Brescius, ‘Researches on Ferric hydrate,” abstract in Jour. Chem. Soc. London, XXIV (1871), 407. 3 J. V. Elsden, Principles of Chemical Geology (London: Whittaker & Co., 1910), Pp. 97. 4W. O. Crosby, ‘‘On the Contrast in Color of the Soils of High and Low Lati- tudes,” Amer. Geologist, VIII (1891), 74; E. A. Smith, Geol. Survey of Alabama, Reft. Jor the Years r88r and 1882, p. 186. 5 See p. 168. THE ORIGIN OF RED BEDS 175 Another source of heat which might be suggested is that of igneous intrusions. The absence of igneous rocks in the greater part of the western Red Beds, including the regions where drilling has shown the colors to be unchanged at depth, proves that this has not been a factor of widespread importance. Heat due to regional metamorphism or to structural deformation of any sort must like- wise be discredited here, as the Red Beds are substantially flat- lying over vast areas, and are nowhere intensely deformed or metamorphosed, except locally in the immediate neighborhood of igneous bodies.t That compression due to the weight of overlying sediments may have created sufficient heat for the accomplishment of extensive dehydration must be recognized as a possibility, although in many areas where the Red Beds occur the requisite overlying sediments are not known ever to have existed, and the uppermost members of the series in these localities are as brilliantly red as any below them. Furthermore, pressure creates heat only by the performance of mechanical work, and microscopic study of the Red Beds reveals no evidence of internal deformation. Crosby’ concludes that the process of dehydration of ferric oxide is largely a spontaneous one, which goes on independently of any outside influence whatsoever, though aided by high temperatures. He states as evidence in support of this hypothesis that the red sedimentary formations and the red iron ores of the world occur in the older systems chiefly, while in the younger systems ferruginous formations and ores are commonly yellow. ‘There are, however, many exceptions to this rule, such as the buff Cambrian sandstones of the Mississippi Valley, the modern red residual iron ores of Cuba, and the dark-red hematitic bog ores in Sweden and elsewhere; and furthermore, the dehydration of the pre-Cambrian red sandstones and argillites may be attributed in some cases partly to regional metamorphism, which has not affected the younger beds. Richardson; has given much weight to this “essentially spon- taneous’” process. He states that ‘‘the dehydration of ferric hydrates tends to go on under ordinary conditions without any unusual cause.”? ‘This has been repeatedly demonstrated by experiment.’”’> But none of the experimenters to whom Richardson t See p. 168. 2 Op. cit. 3 Op. cit., Pp. 392. 4 [bid. 5 [bid. 176 C. W. TOMLINSON ‘ refers' appears to have considered the observed dehydration as truly ‘‘spontaneous”’; and certainly they have shown the process to be closely dependent upon external conditions. The chemically precipitated colloidal ferric hydrates possess, when first formed, a higher content of water than any of the forms known to occur as minerals. According to Van Bemmelen,? on standing in a dry medium at ordinary temperatures these colloids gradually approach the composition 2Fe,0;-H.,0, beyond which the percentage of combined water is not reduced without the application of much higher temperatures. Under water, or in air of moderate humidity, they have not been shown to lose water beyond the composition Fe.O;- 2H.O except at temperatures above 50° C.,3 and heating at 50-60° C. for 2,000 hours failed to bring about dehydration beyond the composition 2Fe,0;-H.O0.4. Temperatures as high as these cannot be assumed to have existed, except locally, in the Red Beds sediments since burial. The so-called spontaneous dehydration observed in the laboratory is probably subject to the terms of Van Bemmelen’s conclusion: ‘‘The red-brown substance, which has been considered to be a hydrate, is a colloid . . . . which has no stable composition; it maintains an equilibrium with the tension of the water vapour in the surrounding medium.’’> Since burial, the great mass of the Red Beds sediments, except in the most arid dis- tricts of the West, have been saturated with ground-water, a con- dition decidedly unfavorable to dehydration at the temperatures there existing. Yet another agent of dehydration is mentioned by Elsden: The presence of any substance in solution which lowers the vapour tension of water will lower the inversion temperature of gypsum. ... . Even solid ‘J. M. Van Bemmelen, ‘‘Sur le colloide de l’oxyde ferrique,” Recueil des travaux chimiques des Pays-Bas et de la Belgique, VII (1888),106-14; Edward Davies, ‘“ Action of Heat on Ferric Hydrate in Presence of Water,’ Jour. Chem. Soc. London, XIX (1866), 69-72; G. C. Wittstein, ‘Uber das Verhalten des Eisenoxyhydrates unter Wasser,” Vierteljahreschrift fiir praktische Pharmacie, I. Band (1852), 275-76. ? Op. cit., pp. I10-11. : ’T. Carnelly and James Walker, ‘‘The Dehydration of Metallic Hydroxides by Heat,” Jour. Chem. Soc. London, LUI (1888), 89; D. Tommasi, ‘“ Ferric Hydrates,”’ abstract in Jour. Chem. Soc. London, XLIV (1883), 24. 4 Davies, op. cit., p. 70. 5 Van Bemmelen, of. cit., p. 114. Translated from the French. THE ORIGIN OF RED BEDS 177 gypsum ... . can be changed into anhydrite by a concentrated solution of sodium chloride. ... . These facts are of interest as pointing to the possibility of dehydration of minerals in rocks, in contact with salt solutions, at a tempera- ture considerably below their normal inversion point.! . .. . The occurrence of red ferruginous sandstone in conjunction with layers containing brown hydrated ferric oxide is less readily explained [than the superficial dehydration of soils, mentioned in the quotation on p. 174, supra], but the dehydration of certain beds may have been effected by contact with salt solutions, as in the case of gypsum already referred to above.” Dehydration by contact with salt solutions presumably would affect the more porous beds first; whereas in the western Red Beds, as already described,’ it is generally in the more porous beds that the lighter colors occur, and the shales are in general of deeper hue than the sandstones. Just how far the action of salt solutions may have been effective, both during and since sedimentation, in accom- plishing dehydration in the neighborhood of such saline deposits as occur in the Red Beds, it would be difficult to say; but for the group as a whole it appears that this agency cannot have been of general importance. In summation, it may be said that while widespread dehydration of iron oxide in the Red Beds since sedimentation cannot, at present, be proved not to have taken place, the greater weight of evidence now at hand is opposed to it; that the opposite process, hydration, may well have been active in the more pervious beds of the series; and that, therefore, the probabilities are quite as much in favor of a lower degree as of a higher degree of hydration, on the average, in the Red Beds at the time of sedimentation than at present. WAS THE COLORING MATTER A CHEMICAL OR A MECHANICAL SEDIMENT ? Having determined as nearly as the available evidence permits the condition of the coloring matter of the Red Beds at the time of their deposition, we may proceed to inquire as to the geographic conditions which gave rise to sediments so colored. The. first t Elsden, op. cit., pp. 85-86. See also H. Stremme, “‘Zur Kenntnis der wasser- haltigen und wasserfreien Eisenoxydbildungen in den Sedimentgesteinen,” Zeit. fiir prakt. Geol., January, 1910, pp. 18-23. 2 Elsden, op. cit.? p. 97. P. 158-50. 178 C. W. TOMLINSON question to be answered in this connection is: Was the ferruginous matter deposited as a mechanical or as a chemical sediment ? The general absence of coloring matter from the non-clastic members," which are in very large part inorganic chemical precipi- tates, indicates quite clearly that the conditions favoring free chemical deposition of calcium and magnesium carbonate or of calcium sulphate were not those under which the coloring matter was usually deposited. The close limitation of ferruginous material to the clastic sediments proves that the conditions under which clastic sedimentation took place favored the deposition of iron oxide also, and strongly suggests that that material itself was carried and deposited as a mechanical sediment, for the most part at least. In this connection it is of interest to note that Dawson? drew a similar conclusion as early as 1848 with reference to the contrast in color between the clastic and non-clastic strata of the Red Beds of Nova Scotia. The condition of the ferric oxide in the Red Beds sediments, as revealed by the microscope, is one of very fine division. Since fine division is to be expected from the mode of origin of the ferric oxide in soils,’ this cannot be taken as evidence that it is a chemical pre- cipitate, in place, in the rocks. If it were the latter, definite orien- tation of crystals of hematite with respect to peripheries of grains might be looked for. I never have seen this phenomenon in thin sections of Red Beds sediments. The microscopic evidence is therefore rather noncommittal as regards the present question. The processes of weathering leave much the greater part of the iron content of all types of rock as a residue in the soil, subject to mechanical transportation. All of the other common chemical constituents of rock, with the probable exception of alumina, are leached from the soil more rapidly than ferric iron. The scarcity of iron in any form in the surface waters of the continents, abun- dantly shown by the analyses of lake and river water published by Clarke, and its even greater scarcity in the ocean,’ testify to the ~See p. 157. 2 Op. cit. 3See p. 164. 4F. W. Clarke, “‘The Data of Geochemistry,” 2d ed., Bull. U.S. Geol. Survey No. 491, 1911, chap. iii. 8 Ibid., chap. iv. THE ORIGIN OF RED BEDS 179 fact that chemical deposition of salts of iron is only exceptionally an important process in earth metamorphism. Glauconite, the most common of such deposits at the present time, is of such peculiar nature as to be readily recognized where it occurs in older sediments: it certainly is not involved to any appreciable extent in the origin of the Red Beds. Bog iron ore, the only other common type of ferruginous chemical precipitate at present, is connected intimately, in origin, with abundance of vegetation and with peculiar and limited topographic conditions; and although deposits of this type are scattered over many parts of the world, none of them is comparable in extent or in thickness to even the smallest of the Red Beds areas. Furthermore, no deposits similar in textural character to bog ores are known in the western Red Beds. In view of the facts above stated, it seems a safe conclusion that the coloring matter of the Red Beds was transported and deposited almost if not quite wholly as a mechanical sediment; and, therefore, without danger of serious error, we may limit investigations of possible conditions of origin of this coloring matter to those which would produce it as a mechanical sediment purely. This applies to the gray and green members of the Red Beds series as well as to those in which the iron is present chiefly as ferric oxide; for if the ferrous iron in the former be explained as the result of the action of organic matter deposited in those strata,’ the ferruginous matter may have been in the ferric form during transportation, quite as well as in any other. = SEO oh 27C—72- [To be continued] STUDIES IN HYDROTHERMAL ALTERATION PART I. THE ACTION OF CERTAIN ALKALINE SOLUTIONS ON FELDSPARS AND HORNBLENDE EK. A. STEPHENSON University of Chicago In the study of ore deposits from a genetic standpoint the subject of attendant wall-rock alteration has received deserved at- tention from geologists. Profound changes of a chemical and min- eralogical character have been recorded at many places, and the relation existing between the various types of wall-rock alteration accompanying ore deposition led to the suggestion that a knowledge of the conditions which bring about such alteration would throw great light on the problem of ore genesis. The data obtained in regard to the temperatures, pressures, and nature of solutions would also be of value in interpreting the geologic history of such occur- rences. The most important of these alteration minerals are kaolin, sericite, and chlorite, and the knowledge concerning their origin is chiefly confined at present to speculations based on their modes of occurrence and their associated minerals. There is especially great difference of opinion’ as to the origin of kaolin. By some writers kaolinitic alteration is attributed to the action of meteoric waters rich in carbonic acid, by others to meteoric waters which have made a cycle of underground courses, and by still others to the emanations from a cooling magma while these are possibly yet in a gaseous state. In regard to sericite and chlorite associated with ore bodies there is nearly general agreement that these have been formed by the action of the solutions which deposited the primary minerals, whether these solutions be magmatic or meteoric, upon the feld- t Bibliographies of the literature on kaolin are given by Résler, Neues Jahrb., Beil. Bd. XV (1902), 231, and by Lazarevic, Zeit. prakt. Geol., XXI (1913), 345. An introduction to a discussion of the origin of kaolin was initiated by Lindgren, Econ. Geol., January, 1915. 180 STUDIES IN HYDROTHERMAL ALTERATION 181 spars and the ferromagnesian minerals. Calcite and quartz are also prominent products of these reactions. Further, the relations of these minerals to the ores indicate that the processes of altera- tion and ore deposition have gone on contemporaneously. This is shown by the progressive decrease in the intensity of the alteration laterally from the veins, and by changes in the character of the alteration in the same direction. For example, the following extract from the report by F. L. Ransome! on the “Economic Geology of the Silverton Quadrangle”’ illustrates these points. At points 150 feet east of the lode the country rock is fine-grained and faintly mottled, showing only a few pale phenocrysts of feldspar and an occa- sional tiny grain of quartz. Under the microscope the rock reveals the char- acter of a much-altered andesitic tuff or fine breccia. The feldspars have been completely altered to aggregates of sericite and calcite, while areas of calcite and chlorite probably represent former phenocrysts of augite. The ground- mass is a rather indistinct aggregate of secondary quartz, sericite, and chlorite with a little apatite and rutile. The rock is wholly recrystallized into a sec- ondary aggregate while retaining the gross structure of the original. At a distance of 100 feet from the vein the . . . . chlorite and calcite are abundant, but much of the plagioclase is still recognizable. Sericite and quartz are not such prominent constituents. ... . At 50 feet from the vein .... the feldspar phenocrysts have been changed to aggregates of calcite and sericite, while areas of chlorite and calcite with sometimes rutile are all that remain of the phenocrysts of augite or biotite. The groundmass also, while preserving the outlines and in small part the substance of former lath-shaped feldspars, is now an aggregate consisting chiefly of quartz, chlorite, sericite, and a little rutile and apatite. .... At 2 feet from the vein .. . . it is seen that alteration has been more thoroughen) oy.) The forms of the phenocrysts are preserved by pseudomor- phous aggregates of sericite with some chlorite, calcite, and rutile apparently after biotite, and quartz, sericite, and chlorite in varying proportions after augite and plagioclase. The groundmass is entirely recrystallized . . . . and the dominant minerals are quartz and sericite. A specimen taken from the wall of the vein showed more evident alteration . . and the rock is wholly recrystallized. The former phenocrysts of feld- spar are replaced by pseudomorphus aggregates‘of quartz and sericite. .... Of the augite no trace remains, but some sericite inclosing rutile is apparently pseudomorphous after biotite. The groundmass is a finely crystalline mosaic of quartz and sericite. The notable feature of this wall rock is the absence of calcite and chlorite. .... tF. L. Ransome, Bull. 182, U.S. Geol. Survey, pp. 116-18. 182 E. A. STEPHENSON To sum up then, the alteration involves the change of the feldspars to sericite, calcite, and quartz; of augite to calcite and chlorite; and of biotite to chlorite, sericite, and rutile. Chloritic alteration appears to precede sericitic alteration and to require less intense or less prolonged action or solutions of a different character, so that sericite is found closer to the veins than chlorite, and chlorite dominates farther from the veins. It may be in place here to note that many consider the solutions which ema- nate from the magma in its final stages acidic in character while others hold that they are alkaline. A study of volcanic gases and of the sublimates present in craters indicates that the volcanic vapors are quite certainly acid, and Day and Shepherd’ found that the magmatic waters which they collected from the small dome within the crater at Kilauea were acid. However, to conceive of these solutions as remaining acid for a long journey through rock masses after their escape from the magma requires a high degree of acidity, and geologists have been loath to accept such a view. It is clear then that the experimental formation of these altera- tion products from the feldspars and ferromagnesian minerals will give some clew as to the nature of the solutions and the tempera- ture and pressure conditions that obtained during the deposition of the associated primary ore bodies. NaAlISi,0¢,H,O+Si0O.+Na’. The analcite was identified optically and also by sifting out the ~ unchanged feldspar with its adhering analcite, gelatinizing the remaining crystals with hydrochloric acid, and allowing a small portion of the solution to crystallize out under the microscope as sodium chloride; the remaining portion of the solution was tested for alumina with ammonia water. After drying at 110° C., the crystals yielded water in a closed tube. Group IIT: Sodium carbonate solutions on hornblende.—A series of experiments, Nos. 11, 12, and 13, exactly like those in Group I, were then tried with hornblende as the mineral (Table III). In no TABLE III | Mineral Time Solution Concen: Vol. Sol. Nob Temp. ies Results aa Tees este) | .| Hornblendel3 yrs. 1 mo.| NazCO;| N/2 |1,200 c.c.|...... a (C5 ae 1 | Nochange c 9o days : e BOOe Hlniacee Loon |G: wy « 82 t s 60 85 c.c.|183° II ‘ 192 E. A. STEPHENSON case did there appear, under the microscope, to be any change. No gelatinous silica separated from the solutions after long standing. Group IV: Potassium fluoride solutions on feldspars and horn- blende.—F luorine has long been looked upon as one of nature’s important mineralizers;' it occurs in small quantities in the emana- tions from Kilauea,? is abundant as fluorite in metalliferous veins, is a constituent of many minerals such as apatite, amblygonite, lepi- dolite, topaz, and cryolite, and is considered by Spurr? an essential constituent of muscovite. Table IV shows the experiments con- TABLE IV No.| Mineral Ti Solution |CPS™Iyo1. Sol] VO | oT ues. Result No. era | ime olution |} pation - Sol.) Tube emp. Aton, esults a | 14..| Adularia 82 days} KF* | N/ro]| 60c.c| 85 c.c.| 183°C.| 311z |No change Hse c “ 18 “ “ 50 80 233° 30 “ “ 16..| Microcline | 18 mes en A Omen 715 233° 30 ~— | Minute rods 17..| Albite 18 Le oC eine | 30 |7o | 233° 30 No change 18..| Hornblende) 82 Femara tA &) 60 SSaeEOse 11 | Brown iron oxide iQ z 55 283 30 15 | ESO = Tube burst | * This salt contained a small amount of the acid salt HKF:2. ducted with potassium fluoride. In experiments Nos. 14, 15, 16, and 17 the feldspars showed no change other than a possible slight etching. In No. 16 there appeared some minute bacteria-like rods that were not further identifiable. In No. 18 the hornblende was vigorously attacked. The product is an amorphous brown mass resembling limonite and consisting of hydrated iron oxide together with grains of partially altered mineral. Some of these grains are bleached to isotropic transparency, others have a rim of isotropic matter surrounding ellipsoidal grains in the interior. No gelatinous precipitate appeared in the decanted solution on standing in paraffin lined bottles for several months. In No. 19 the tube burst and the mineral appeared unchanged. Group V: Mixtures of carbonate and fluoride solutions on feldspars and hornblende.—Since the feldspars were not visibly attacked by the fluoride solutions but had been attacked by the alkaline carbo- ™See especially C. Doelter, Allgem. chem. Min., p. 207, and W. Bruhns, Neues Jahrb. Min. Geol., II (1889), p. 26. 2A.L. Day and E. S. Shepherd, Bull. Geol. Soc. Am., XXIV, 592. 3 J. E. Spurr, Professional Paper 42, U.S. Geol. Survey, p. 233. STUDIES IN HYDROTHERMAL ALTERATION 1)5) nate solutions, it was deemed advisable to try a mixture of the two solutions to see if the traces of fluoride present would modify the results in any way (Table V). The feldspars in Nos. 20, 21, and 22 No. DOM Die g DP 0 RE 24.. r DS 20). DG DSi ZO 30.. Bite TABLE V Mineral Time Solution Concentration |Vol. Sol. Adularia | 90 days) NazCO,;+KF) N/2+N/r10|300c.c “ 4 I “ “ “ (3 54 “ 82 “ « “ « Es « 18 « « « « 16 « Is « « « « 65 Orthoclase | 15 “ “ « « 50 Microcline |} 18 “« « « « 48 Albite 18 “ « « « 48 Hornblende go “ “ “ « 300 Bus 41 « « « « 60 “ 8 2 “ (3 “ “ 60 e 3 “ « “ “ 60 Ae. Temp. apt la ae too C. 80c.c.|183° Som |n8an 85)" |233, 100. = |280" 100. = |280° LOOM al2sem Sommlescn sre 100° Soo alr8az 85 183° Som |235u Press. Atm. 30 I It It 30 Results No change “ “ “ “ Analcite Needles Twins of unknown mineral, also needles No positive change Analcite No change ¢ were not visibly attacked and the decanted solutions gave no pre- cipitate on standing. No. 23 showed crystals of analcite. No. 24 showed some needles with parallel, or nearly parallel, extinction and pos- sibly some isotropic forms, though these were not positively identified. No. 25 contained many needles like those in No. 24 with no analcite. These needles have an extinction angle of less than 2°; index in the direction of elongation is 1.490, and at right angles to this 1.517; elonga- tion is negative. The crystals after drying at 110° C. yield no water in a closed tube. Good terminations at both ends are common. No. 25 also Fic. 7.—Twinned crystals pro- duced by experiment No. 25. contained beautifully twinned crystals, illustrated in the accom panying sketch (Fig. 7) made with a camera lucida. The crystals 194 E. A. STEPHENSON are perfectly transparent, have an index close to but greater than 1.565, an extinction angle of 43°, and are probably monoclinic. In experiment No. 26 no evidence of any change appeared. In experi- ment No. 27 icositetrahedra of analcite appeared, some free and some including fragments of the feldspar. The hornblende in all of this group of experiments was unchanged. It is concluded that the presence of the fluoride had practically no influence upon the alteration. Group VI: Sodium bicarbonate solution upon feldspars and hornblende.—These experiments embody an attempt to increase the pressure both by raising the temperature and by increasing the concentration of the carbon dioxide through dissociation of bicar- bonates (Table VI). In experiment No. 32 scarcely any feldspar TABLE VI Concen- Vol. Sol. Vol No Mineral | Time Solution peatinn Tube Temp. eae Results 32..| Adularia 15 days NaHCO, N/2 | 45 c.c.| 85 c.c.| 233°C.| 30 | Analcite Boer “ iis . Selso. 85 280° 65 é 34. .| Albite 15 & sn ic 85 2335 30 | Analcite and needles aie _ 5 A KOO") ss 280° 65 | Tube burst, but needles formed 36..| Hornblende 15 £ eg 85 233° 30 | No change « 15 “ “ 45 85 280° 65 “ « remained and much analcite appeared as free crystals and as aggre- gates. Gelatinous silica appeared in the decanted solutions after some time. Similar results with decided etching of the feldspar grains appeared in No. 33 where the pressure was practically doubled. Possibly less analcite formed in this than in the previous experiment at the lower temperature. In No. 34 with albite, anal- cite also formed with many needles like those in experiment No. 25, Table V. In experiment No. 35 the tube burst where poorly sealed, but the mineral was nearly all altered to needles as in experiments Nos. 24, 25 and 34. These were analyzed qualitatively and found to consist of soda, alumina, and silica and yielded no water in a closed tube. The writer was not able to find any natural sodium alumi- num silicate whose properties agree with these. Needles having STUDIES IN HYDROTHERMAL ALTERATION 195 the same optical properties as these also appeared in experiment No. 46 and are undoubtedly the same thing. The hornblende was unchanged in experiments No. 36 and 37; the tube in No. 37 burst some time during the course of the heating. Group VII: Potassium bicarbonate solutions on feldspars and hornblende-—These are similar to the experiments of Group VI, except that potassium bicarbonate solutions were substituted for the sodium bicarbonate solutions (Table VII). In no case either TABLE VII Concen- Vol. Sol. Vol. Press. No Mineral Time Solution aan Tube Temp. Nn. Results 38..| Adularia 15 days) KHCO,| N/2 | 45c¢.c.| 75c.c.| 233°C.| 30 | No change 30. i (73 15 “ “ 50 90 280° 65 (13 “ 40. .| Albite 15 FAS 750) \233)0 04 3 Ol | aaraana HD s 5 Y is 60 =|I05 280° 65 Y - 2..| Hornblende} 15 bs 45‘ |I00 Pyeles 30 be iS with the soda or potash feldspar or with hornblende did any change appear in the minerals. The writer is unable to give an adequate explanation of this fact, but it may have some bearing on the question whether potash is introduced or not in hydrothermal processes. Group VIII: Sodium tetraborate solution upon feldspars and hornblende-—A few experiments were tried with borax solutions with results very similar to those produced by the alkali carbonates (Table VIII). In No. 43 well-formed crystals of analcite as rhom- TABLE VIII No. Mineral Time Solution Concent; Vol. Sol. Nel Temp. Eres : Results 43..| Adularia 15 days| Na2B,0,| N/4 | s50c.c.\100c.c.| 233°C.| 30 | Analcite 44..| Albite 15 - i 50 95 233° 30 | No change 45..| Hornblende| 15 g . 50 90 233° 30 ee is bic dodecahedra appeared. No alterations occurred in the other experiments. Group IX: Sodium sulphide solutions upon feldspars and horn- blende.—The presence of the metallic sulphides indicates that at 196 E. A. STEPHENSON certain phases of the vein-forming process sulphide solutions must be present. Though the solutions are undoubtedly very complex, their efficiency as hydrothermal agents is very probably due to a few components. The physical state of these solutions has been recently shown by Tolman and Clark’ to depend decidedly upon the composition, at least at ordinary temperatures, and it may also have some decided influence upon the character of the alteration. In the following experiments the copper tubes were vigorously attacked and chalcocite crystals lined the walls of the tubes (Table [X). In No. 46 needles of the anisotropic crystals like those TABLE IX No. Mineral Time | Solution | | | = Concen -|;; Vol. Press. é tration Vol. Sol. Tube Temp. Aten Results .| Adularia 15 days) NaS | N/2 | 5o0c.c./120¢.c.| 233° c.| 30 +| Analcite and | | needles ..| Albite }) sets ng lea 50. ~=—|100 | 2320 30 | Analcite .| Hornblende| 15 | - | 50 105 | 23 3° | 30 | No change obtained in experiments Nos. 25 and 34 were formed, together with well-formed analcite crystals. Very little of the original feldspar remained. In No. 47 perfect analcite crystals appeared as icosite- trahedra and as combinations of the cube and rhombic dodeca- hedron. These vary in size from one-half to one millimeter in diameter. The hornblende in experiment No. 48 was not attacked and no pyrite could be identified in the product. No sulphur was obtained by heating the mass in a closed tube. The results with the feldspars are quantitatively greater than in any other experi- ment; this was probably due to the fact that the hydrogen sulphide is a weaker acid than is carbonic acid and the hydrolysis therefore produces a more strongly alkaline solution. Group X: Aluminate solutions on feldspars and hornblende.—In the previous experiments the loss of silica from the minerals resulted in an apparent rise of the alumina content of the new minerals. With the thought that possibly an increase in the concentration of the alumina in the solutions might cause the solubility product for «C. F. Tolman and J. D. Clark, Econ. Geol., TX (1914), 550. STUDIES IN ‘HYDROTHERMAL ALTERATION 197 compounds richer in alumina to be exceeded, a few experiments with aluminate solutions were tried. These solutions were pre- pared by taking a weighed quantity of aluminum sulphate, precipi- tating the aluminum as hydroxide, washing the precipitate and then adding it to normal sodium or potassium hydroxide and diluting TABLE X Concen- No. Mineral Time Solution of Kor Vol. Sol. nce Temp. ies: Results a 49..| Adularia 15 days} Sodium N/2 | 40c.c.| 80c.c.| 280°C.| 65 |Analcite aluminate 50. S 15 Potassium | “ —| 4o 85 280° 65 |Hexagonal aluminate plates and | needles 51..| Orthoclase | 15 Potassium # 60 ~=|120 280° 65 |Hexagonal aluminate, | plates and needles 52..| Albite 5 Potassium S 30 70 280° 65 |Hexagonal aluminate} | plates 53--| Hornblende| 15 Potassium = 35 75 280° 65 |No change aluminate one-half. After this had been allowed to stand for several hours it was filtered from its slight precipitate (Table X). In No. 49 anal- cite crystals formed, which were identified chemically and micro- scopically. In Nos. 50, 51, and 52 twinned hexagonal plates, with anomalous division into fields under polarized light, resulted, as shown in Fig. 8. These gelatinize with hydrochloric acid and contain sodium, but no aluminum could be detected in them by a microchemical test. The hornblende was not attacked. Fic. 3—Twinned Group XI.—Albite and hornblende were hexagonal plate result- heated for 15 days at 280° C., with saturated Oe solutions of calcium bicarbonate. The tubes ~~ ’ A burst in each case and the minerals showed no change. GEOLOGIC BEARING Though the alteration of feldspars to analcite has not been com- monly described, in many cases it is very possible that some of the 198 E. A. STEPHENSON determinations of isotropic minerals as glass may be incorrect, and that analcite has been overlooked. In examining the slides in the University of Chicago collection a slide from a trachyte of Bannber- scheid, Westerwald, Nassau, was found which showed the soda feldspars altered to analcite though the original crystal boundaries remained sharp. The alteration does not follow cleavage cracks but appears in irregular patches. Mr. K. F. Mather in a forth- coming paper will describe an eruptive cone of Quaternary age in the canyon of the Mancos River ten miles southwest of Mancos, Colorado—locally known as the “ blowout’’—which is cut by dikes of augite minette. These dikes carry fragments of granite xenoliths which are deeply corroded and partially assimilated. The feldspars are altered to analcite—identified microchemically. A careful study of rock specimens would probably show that this type of alteration is much more common than has been supposed. SUMMARY 1. Alkaline solutions of different characters dissolve the feld- spars with separation of silica and crystallization of compounds less rich in silica. The solutions are probably hydrolyzed since the reactions are accelerated by the presence of alkalies, by increased concentration of the alkalies, and by higher temperatures. 2. Feldspars and hornblende are not appreciably attacked by pure water at temperatures up to about 300° C. adularia to at least 350 C., showing that the dissolved substances rather than water alone must cause the differences in the nature of the alterations. 3. Albite and orthoclase feldspar seem to respond to the action of the alkalies in nearly identical ways, and hence the conclusion is patent that they have very similar chemical structures. 4. The influence of small amounts of fluoride and borates as mineralyzers has not been found important, at least in the presence of the other substances. This leads to the suggestion that possibly the mineralyzing effect is merely that of causing solution at tem- peratures where the silicates in question would otherwise be much less soluble. 5. It is notable that no kaolin or kaolin-like substance forms from alkaline solutions at temperatures up to 280° C. The sugges- STUDIES IN HYDROTHERMAL ALTERATION 199 tion seems necessary that since pure water has practically no effect on the feldspars, and that since the alkaline waters produce minerals other than kaolin, kaolin probably forms by the action of acid solu- tions upon the feldspars. The literature bearing on the field occur- rences of kaolin shows a striking lack of references to association of carbonates and kaolin, though this would be expected if carbonated waters are the cause of the formation of kaolin from the feldspars. 6. The general agreement of the data obtained throughout the range of temperatures used shows that the silicates may be studied with the apparatus described, up to 300° C., with gratifying results. and without great mechanical difficulties, and also without the necessity of contamination from undesirable sources such as glass tubes. The writer is indebted to Professor W. H. Emmons for suggest- ing the problem and to Professor A. D. Brokaw for sincere interest and many suggestions during the progress of the work. Further work of a similar sort is in progress and the next paper will deal with the action of various acid solutions, especially hydrofluoric acid on the same group of minerals. ce ZONAL WEATHERING OF A HORNBLENDE GABBRO ALBERT D. BROKAW anp LEON P. SMITH University of Chicago In connection with a study of the alteration of the so-called trap or diabase dikes at La Grange, Georgia,’ one of the writers collected some interesting specimens showing a weathered zone in which the decomposition is extreme, with an abrupt transition into fresh rock, practically free from the effects of weathering. ‘This paper is con- cerned with the description of a typical specimen, and with the results of chemical analyses of the fresh rock, the partly altered material, and the extreme phase of alteration present. La Grange is in the extreme western edge of the state, not many miles from the southern end of the belt of crystalline rocks which extends from Maine to Alabama. The country rock is for the most part gneiss, cut by granitic intrusions. Both gneiss and granite are cut by pegmatites and basic dikes, the latter usually called diabase, and supposed to be of Triassic age. At La Grange there is a series of dikes, only a few feet apart, ranging up to forty feet in width, with an average of about four to six feet. The dikes have a general north-south trend, and are abundant over an area of a little more than half a square mile. The older crystalline rocks are very deeply weathered. The basic dikes have been considerably altered at the surface, and in many instances may be traced by the strong iron stain they have imparted to the soil. In some excavations, however, material showing no appreciable effects of weathering may be obtained. In some case: weathering has been strongly marked along joints, changing the character of the material for an inch or less, beyond which the rock is fresh. Fig. 1 shows a specimen in which the zone of alteration is about an inch thick. Analyses were made of the tL. P. Smith, Alteration of Diorite by Weathering, Dissertation, University of Chicago, 1915. ZONAL WEATHERING OF A HORNBLENDE 201 outside, much-altered portion, and of the part of the weathered zone adjacent to the fresh material. A specimen from an entirely fresh portion of the same dike was analyzed for a comparative study. The fresh rock is finely grained, holocrystalline, nearly black in color, with small white feldspar crystals evenly distributed through Fic. t—Specimen showing zone of weathering X2. Circle shows approximately the part from which the slide shown in Fig. 3 was taken. the mass, which is dominantly hornblende. Small garnets, irregu- larly distributed, and small crystals of pyrite are present. Micro- scopically, the rock is found to consist of hornblende (65 per cent), which may be secondary, labradorite (32 per cent), with small amounts of orthoclase and augite, and accessory magnetite, pyrite, titanite, and apatite. One of the typical slides is shown in Fig. 2. The rock is a hornblende gabbro. | ALBERT D. BROKAW AND LEON P. SMITH to fe) iS) The partly altered material is notably lighter in color than the fresh rock. Most of the hornblende has been weathered, leaving brown iron stains and a small amount of chlorite, but some particles of appar- ently fresh hornblende are tobeseen. ‘The feld- spar has been changed to a chalky white material in an iron-stained matrix. Microscopical study re- veals some unchanged hornblende, but for the most part only alteration products are discernible. Fic. 2.—Section of fresh hornblende gabbro 2 : ‘ : X20. (Two rather large cracks appear as light Limonite, white mica bands.) Ordinary light. (probably also gibbsite), chlorite, zoisite, and a small quantity of mag- netite make up the mass. The weathered zone is sharply set off from the fresh material even in thin section, In Figs 92> the fresh portion is shown on the left, changing to altered on the right. The most completely weathered portion its a fri- able, earthy, non-plastic mass, strongly iron stained, and so thoroughly disintegrated that the - : Frc. 3.—Slide showing transition from fresh to original texture of the altered portion of specimen. X25. Crossed nicols. ZONAL WEATHERING OF A HORNBLENDE 203 rock is practically lost. No microscopical study of this material was undertaken. The results of chemical analyses, made in duplicate in every case and in triplicate in a number of determinations, are given in Table I. The following recognized effects of weathering are well illustrated: loss of silica, apparent increase in alumina, increase in ferric iron, decrease in ferrous iron, loss of bases, increase of combined water, and decrease in specific gravity. The very small development of carbonates may be due to the fact that most of the available carbon dioxide has formed soluble bicarbonates in removing lime, mag- nesia, and alkalies." TABLE I r* 2t 3t 1 2t 3r SIO ee aay ae ASH LOM 20K02) 2334) lMiOsa. = asses .31 .18 gills AIO; ee ee es 17634 || ASCO |) BAcGO. WIEN 505 aco 47 5 Sit tr KeOe si osess: Qowy || Wie ss | Dies IlCOhs osotousoc none eI 382 LOO eps aia nis ea G.@0) || AoCe) ||| wen NSoseouaceonec .10 .04 | none IN OF ale ies! 4.67 3-03 585) CaO Re eran! 17.50 | 7.96 B75 ee MO talennmense 100.10 |100.23 |100.20 INE aetna A 6X0) |). wg Gy 39 KOR a ee it 4 BO 81 1.19 ||Total Fe...... 72O2\4\ PL. 78 | 15.24 Oe Sa 04 RS 7 esi 77anl|iSps Gres 3.020| 2.813] 2.340 BOR sin occ aoe .46 | 15.17 | 15.05 * Fresh Rock. } Altered near the fresh rock. { Altered, most decomposed portion. § Pycnometer method. On the assumption of constancy of alumina? the analyses may be recalculated, and gains and losses estimated, as shown in Table II. The change in total iron is worthy of note, in that there is a slight loss shown in 2, with a marked increase in 3. A suggested explana- tion is that during the early stages of alteration some soluble ferrous compound was formed, which migrated, perhaps by capillary action, to the outer zone before it was precipitated by oxidation to t Tn this connection it may be noted that wells in or near the dikes are said to yield water containing considerable amounts of lime, while those in the gneiss, farther from the dikes, yield comparatively soft water. Unfortunately, no quantitative data on these waters are available. 2G. P. Merrill, Rocks, Rock Weathering and Soils, p. 208. 204 the ferric condition. ALBERT D. BROKAW AND LEON P. SMITH That the removal of ferrous iron may take place in the early stages of alteration has been commonly recog- nized.* TABLE II I 2 Gain Loss 3 Gain Loss SiOpe sce 5m ewer eee Aisa Koye |e atyaceyel |e. sas 2OR 22.3) ate Gi) || eats eeae 32.65 AOS se ee cee 7S Dia! | 7 en 2a | ee ll oe RS T7:..5 2) || 2 ey HeOheiocs +c cee By 6.97 St ately ate 11.66 8: 54clia se eee PeOe coos oan eh eee 6.99 Qt AIS Paar een 2 AMS Ase NONE |p eee 6.99 MaOcrnudie? iaesieeere 4.67 Tx GOR h eae ne 2.81 S30) (Sarees 4.36 CaO visio sone eee E750: | A Oia = Ae aie 12.63 A 34 all< eee 17.16 INGO. cid eee 2.39 BCOVI heats onder 1.45 SR Ol ees 5c 2.18 KO. 37 5 OMe eeness .87 HOAs Sets sae 73 H,0— OA Gi anes 3 PAO tescros 2.36 2:32: eae H,O+ .46 9.29 SFO iste sie yets 8.06 7, OO hadeneene iO. choc ne eee AB ses o | Gach create 20 COS alc hity Nene a23 MnO... cdiecusnoseeeee .47 COTeN oilers ae § FiO) EE? wiyeccueeee .47 CO} ee eee eee none 577 ROP Ale corse 28 /28) |) eee Sa cis pa era rors Hist: .10 HODwilaicer ence JOS!s| WAMONE! |e eee 10 Totalcc «Ss he eee eee LOOMIOs | WOTS3 Of il ose teal hse 53.295) leone 2s Cee Total Bey xs. yess 7.62 GA2Te en eee cine 4.05 9.33 1. 707 Essentially the same relations are shown by the ‘“‘straight line”’ In Fig. 4 the analyses of the altered portion diagrams of Mead.? are compared with that of the fresh rock. The full line represents analysis 2, and the broken line analysis 3. In general, 3 is merely an accentuation of 2, except for potash and total iron, both of which show a change of sign in the direction of change. It is apparent that in the later stage of alteration the removal of bases continues, and that it is out of proportion to the further removal of silica as compared with 2. The retention of potash is by no means unusual. In Table III the analyses are recast to molecular proportions to emphasize some of the chemical and mineralogical features. It is to be noted that even in analysis 2 the amount of silica is insufficient to combine with all of the alumina to form kaolin. The alumina: silica ratios areas follows: 1, 524.3632, 371257; 3, 1:1.10, Farka. olin 1:2 is required. On the extreme assumption that all of the silica in 3 is present in the form of kaolin, the analysis may be said to tC. K. Leith and W. J. Mead, Metamorphic Geology, p. 22, Henry Holt & Co. (1915). 2W. J. Mead, Econ. Geol., VII (1912), 141-44. ZONAL WEATHERING OF A HORNBLENDE 205 represent (by weight) kaolin, 50 per cent; limonite, 25 per cent; bauxite, 18 per cent; other substances, 7 per cent. It is highly probable that part of the silica is present in some compound not containing alumina, and if so the amount of bauxite is correspond- ingly greater. This lends a distinctly bauxitic-lateritic aspect to the alteration. Fic. 4.—“ Straight line diagram”’ showing changes in composition with alteration. The full line represents analysis 2, the broken line analysis 3, compared with the analysis of the fresh rock. TABLE III Per cent+mol. wt. I 2 3 I 2 B SiOz sae 747 431 387 CaOuan eae: .312 .142 013 INGOR 6 sods c a 7i2 . 280 1320 |\Na.O).-. 3) - 039 025 006 WE oo dec .020 .070 eX, WLRAD Se bo0 3 O15 .008 013 EO se Becks .097 .056 ooo |{/H.O+..... .026 845 .836 MgO...... . 116 .075 .OT4 COR oes . 000 .003 .O12 Specimens similar to the one studied are not uncommon in the weathered part of the dikes, and the changes shown are believed, to be fairly typical for the rock in question. It is to be noted that the fresh rock is much richer in iron than rocks which yield bauxite deposits, and richer in alumina than those yielding high grade laterite deposits. REVIEWS Lockatong Formation of the Triassic of New Jersey and Pennsyl- vania. By A. C. Hawkins. Annals N.Y. Acad. Sci., XXIII. 145-76, Plate 1, January 27, 1914. The Lockatong formation is the middle member of the Newark series of the Triassic, extending from a point just west of Phoenixville, Pennsylvania, to Princeton, New Jersey. The rocks of the formation are dense, fine-grained, massive argillites, with some shales. The forma- tion as a whole has a decidedly lens-like character. On the basis of the general structure, lithologic character, and type of fossils, which include estheriae, fish-scales, ostracods, and plant remains, it is concluded that the sediments were laid down near the center of an inland basin. The particles of the argillite are for the most part cemented by silica, which renders the formation very hard and a pronounced ridge-maker. The color of the beds is due to iron in various states of oxidation. The boundaries of the Lockatong are very uncertain, owing to the fact that it passes by a series of transitional dovetailing strata into the other formations of the Newark. Since part or all of the Lockatong may be contemporaneous with portions of the Stockton and Brunswick forma- tions elsewhere, it seems that as a definite geological time unit the Lockatong is valueless. There are three principal joint directions in the Lockatong formation, the most important of which is remarkably con- stant, and extends into the borders of a diabase mass near Rocky Hill, which is interpreted as an extension of the Palisade sill. Titanium minerals, brookite and ilmenite, are found in this major joint series, apparently far removed from the diabase. Analcite and barite also occur. That these minerals are derived from the igneous rocks is indi- cated by similar occurrences in New Jersey elsewhere. Parts of the Lockatong argillite are very well adapted for commercial use. R: C. Me Geological Map of Tennessee. Compiled by Otar P. JENKINS, A. H. PurpvueE, State Geologist. This map represents Archean, Cambrian, Ordovician, Silurian, Devonian, Mississippian, Pennsylvanian, Cretaceous (Upper), Eocene, Pleistocene, and Recent formations. Few states have so wide a range 206 REVIEWS 207 of systems, Proterozoic, Permian, Lower Cretaceous, Miocene, and Oligocene only being absent. Pliocene (Lafayette) is indicated in the legend of the map but. not shown on the map itself, and the legend seems to be intended to throw doubt on the validity of the formation in Tennessee. Under the designation “Columbia Formation,” loess, loam, and loose sand are grouped. This seems to us an unfortunate classification. The “Terrace Deposits” of the map are quite as appropriately classed as “Columbia” as the loess and loam which are so classed. We are of the opinion that the use of the term ‘Columbia Formation” should be dis- continued (though possibly the term “‘Columbia Series”? may be useful to include all Pleistocene non-glacial formations). What was originally grouped under the name Columbia included several formations of which the probable equivalents of the Terrace Deposits of this map were a chief member. ‘Loess’? would seem to be an adequate designation of the deposits included under that term, without classing them as Columbia. Their classification as Pleistocene seems. altogether adequate. The loess, of many regions at least, is of very different ages, and all of it does not belong to one formation in the chronological sense. The map is distinct and represents sufficient change from its prede- cessors to be welcome. It is accompanied by elaborate explanatory legends and by four cross-sections which represent well the structure of the formations in the state. The map may be had by application to the State Geologist, Nash- ville, Tennessee. Postage, 8 cents. ReaD aS: Cretaceous Deposits of the Eastern Gulf Region, and Species of Exogyra from the Eastern Gulf Region and the Carolinas. By L. W. STEPHENSON. U.S. Geol. Surv., Prof. Paper 31, 1914. Pp. is, OS Ait, (Goetecs,'o). In eastern Alabama and Georgia a terrane, previously regarded as forming the eastward extension of the Tuscaloosa series of western Alabama, has been shown by its unconformable relations with overlying formations, lithologic character, and contained plant fossils to be of Lower Cretaceous (Comanchean) age, though probably somewhat younger than the Patuxent. Belonging to the Upper Cretaceous (Cretaceous) of the eastern Gulf region are four formations, Tusca- loosa (regarded as Lower Cretaceous), Eutaw, Selma chalk, and 208 REVIEWS Ripley. The first, consisting of irregularly bedded sands, clays, and gravels, has an estimated thickness of 1,000 feet and rests unconformably on a basement of Paleozoic metamorphics, and in the east on Pre- Cambrian crystalline rocks and in part on Lower Cretaceous. The Eutaw formation, somewhat similar to the Tuscaloosa in lithologic character, is believed to be entirely marine, though much of the formation was doubtless laid down in very shallow water. It is 400-500 feet thick, rests conformably on the Tuscaloosa, and is overlain conformably in part by the Selma chalk, and in part by the Ripley formation. The Selma chalk consists mainly of more or less argillaceous and sandy limestones rendered chalky by their large content of foraminiferal remains. It is abundantly fossiliferous in certain portions, yielding large numbers of the Exogyra described in the latter portion of the paper. The Selma grades into the sandy member of the Tuscaloosa, and the clastic beds of the Ripley formation when followed along the strike. A thickness of 930 feet of the chalk formation has been measured in western Alabama. The Ripley formation, 250-350 feet in thickness, consists typically of calcareous and glauconitic sands, sandy clays, and impure limestones and marls of marine origin. It extends through parts of the Gulf states from southern Illinois to Georgia. A study of the faunas of the various formations is detailed, and correlations with other Cretaceous regions indicated by chart. A description of the genus Exogyra, which includes three species with two varieties, constitutes the second portion of the paper. RC. Me The Jurassic Flora of Cape Lisburne, Alaska. By F.H. KNowLton. U.S. Geol. Surv., Prof). Paper 85; Part D, 1914: Pp. 25; plsyze The Jurassic of the Cape Lisburne area is estimated to have a very great thickness, 15,000 feet, and contains from 40 to 50 coal beds which range in thickness from 1 or 2 feet to over 30 feet. Plant collections from this area show 17 species of well-defined Jurassic types. The close similarity or identity of a number of forms with species from eastern Siberia and Mongolia is noteworthy. The flora indicates a warm- temperate or subtropic climate and the geographic range, especially into the Arctic and Antartic, is suggestive of the uniform mildness of the Jurassic earth-climate. RCE | Outlines of Geologic History with Especial Reference to North America Symposium Organized by BAILEY WILLIS Compilation Edited by ROLLIN D., SALISBURY ; = and all readers of geologic literature will welcome the publication, in book form, of an important series of essays and discussions on the subject of geologic 1 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, 810, cloth; price $1.50, postage extra (weight 2 Ibs.) 2 The University of Chicago Press | Chicago - - Illinois S— \ a in correcting the proofs of Vol. IT. or $15 net in half-leather. A HISTORY OF ARCHITEGTUR Vol. 1: Antiquity. Vol. il: Romanesque and Oriental. Vol. Ill: Gothic in Europe. Vol. IV: Gothic in England, Renaissance, Modern Architecture This is the most extensive study of architecture in the English language, being the first attempt to cover exhaustively the development of architectural form from the earliest monuments of Egypt and Chaldea to the buildings of the present day. The work was originally planned by Russell Sturgis more than ten years ago to be a complete survey of the architectural expression of the various peoples throughout theages. Mr. Sturgis unfortunately died before the publication of the second volume in rg09. The completion of the work was thereupon undertaken by A. L. Frothingham, who had assisted Mr. Sturgis THE BAKER & TAYLOR CO FINE INKS 482 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 ill-smelling inks and adhesives and adopt the Higgins Inks and Adhesives. They will bea revelation to you, they are so sweet, clean, well put up, and withal so efficient. At Dealers Generally Higgins’ CHAS. M. HIGGINS & CO., Mfrs. Branches: Chicago, London 271 Ninth Street Brooklyn, N.Y. Daus Improved Tip Top Duplicator will reproduce them _ easily—quickly—neatly—cheaply. The simplest and quickest duplicator on the market. Makes rso handwritten copies or 50 typewritten copies in afew minutes. No expensive supplies to buy. Lasts a lifetime. No printer’s ink. No smudged copies. No soiled hands. Ready instantly. Negatives require no washing out after using. Complete outfit contains a con- tinuous roll of Daus Non-Peeling Anti-Stick Negative Roll, which can be used over and over again. Cap size 83x13 inches. Price complete $5.00. Ten days’ free trial without deposit LIX K. DAUS DUPLICATOR CO. Daus Building, 111 John Street, NEW YORK Volumes | and I by RUSSELL STURGIS Volumes Ill and IV by A. L. FROTHINGHAM Description of the Volumes : The volumes are royal octavo, size 73 X to}, and contain about 400 pages each. The volumes are bound in two styles —one a serviceable blue cloth with gold stamping and tops; the other a brown half-morocco with cloth sides for those who require a more enduring form. ‘The paper is heavy coated, of fine quality, to reproduce the illustrations to the best ad- vantage, deckle edged. The type used is large and clear. A complete index will be found at the end of Vol. IV. 4 Price and Conditions The price of A History of Architecture, complete in four volumes, cloth, is $20 net; half-leather, $30 net. For the advan- tage of those who purchased Vols. I and II when they were published, Vols. III and IV will be sold for $ro net in cloth, Wholesale Dealers in the Books of All Publishers e 354 Fourth Ave., NEW YORK, at Twenty-sixth St. A Simplified Typewriter The Remington Junior Small in size—but big in Capacity Light in weight—but heavy in stability Remington-built Remington-guaranteed Price $50.00 Easy payment terms can be arranged if desired Write for Catalogue REMINGTON TYPEWRITER GO. (Incorporated) New York and Everywhere Do You Want to Know What Your Handwriting Reveals? WE have published “a fascinating booklet of 32 pages on the sub- ject of graphology, by William Leslie French. Every reader of this publi- cation will be interested in comparing the examples of handwriting in judg- ing character and disposition of the writer. A copy of this booklet and 12 differ- ent patterns of The Spencerian receipt of ten cents. If you can do good work witha poor pen you can do better work with a good pen. Silver plated Falcon pens, smooth points, wear long, will not corrode. SPENCERIAN PEN CO. 349 Broadway - - New York City Steel Pens will be sent postpaid on | MICROSCOPE FS2 Ki 7.5X Eyepiece; two objectives, 16 mm. (3 in.) and 4 mm. (4 in.) in dustproof revolving nosepiece. Price $34.00 Unexcelled for all-round Laboratory uses Several of our notable improvements are involved in this example of the superiority of the Bausch loa Microscopes It has our lever type of side fine adjustment—simple, practical, and durable. The rate of movement is correct for both high- and low-power objectives. The stage is large and is completely covered with vulcan- ized rubber. Its diaphragm is operated by a milled ring, from any point of its circumference. ‘ The long handle arm is curved to give graceful appearance and proper balance—also to afford more than usual space for manipulation of the object. Optical equipment and mechanical construction of the highest standard. Special Terms to Educational Institutions Write for illustrated price list of all models from $2.50 up Bausch £5 lomb Optical ©. 412 ST. PAUL STREET ROCHESTER, N.Y. New York Washington Chicago Leading American Makers of Photographic and Ophthalmic Lenses, Projection Lanterns (Balopticons), Stereo-Prism Binocu- lars, and other high-grade optical products. Spit San Francisco Go MOLUME XV". NUMBER 3 — THE ; JOURNAL orf GEOLOGY A SEMI-QUARTERLY iin pride : EDITED By THOMAS C. CHAMBERLIN AND ROLLIN D. CHa Surtees: With the Active Collaboration of SAMUEL W. WILLISTON, Vertebrate Paleontology ALBERT JOHANNSEN, Petrology pelea WELLER, Invertebrate Paleontology ROLLIN T. CHAMBERLIN, Dynamic Geology : ALBERT D. BROKAW, Economic Geology ’ ASSOCIATE EDITORS ‘SIR ARCHIBALD GEIKIE, Great Britain JOSEPH P.IDDINGS, Washington, D.C. CHARLES BARROIS, France JOHN C. BRANNER, Leland Stanford Junior University ALBRECHT PENCK, Germany ; RICHARD A. F. PENROSE, Jr., Philadelphia, Pa. HANS REUSCH, Norway WILLIAM B. CLARK, Johns Hopkins University GERARD DEGEER, Sweden WILLIAM H,. HOBBS, University of Michigan “TT. W. EDGEWORTH DAVID, Australia FRANK D. ADAMS, McGill University - BAILEY WILLIS, Leland Stanford Junior University CHARLES K, LEITH, University of Wisconsin GROVE K. GILBERT, Washington, D.C. WALLACE W. ATWOOD, Harvard University CHARLES D. WALCOTT, Smithsonian Institution WILLIAM H. EMMONS, University of Minnesota ' HENRY S. WILLIAMS, Cornell University ARTHUR L. DAY, Carnegie Institution = es APRIL-MAY 10916 on aE A. DERBY BR Pa PF yee: Semen 8b a ith en canmaneanine SOHN Ca BRANNERS 0/300 gcieslgges) OF PRISMATIC STRUCTURE IN IGNEOUS ROCKS - - Ropert B. SOSMAN = 215 _ ELLIPSOIDAL LAVAS IN THE GLACIER NATIONAL PARK, MONTANA LANCASTER D. BuRLING 235 THE ORIGIN OF RED BEDS. A STUDY OF THE CONDITIONS OF ORIGIN OF THE PERMOCARBONIFEROUS AND TRIASSIC RED BEDS OF THE WESTERN UNITED STATES: PART II - - - - - - - C.W. Tomitnson 238 ‘THE ACADIAN TRIASSIC. PART II - - - - - - - SIDNEY POWERS 254 THE LOMBARD OVERTHRUST AND RELATED GEOLOGICAL FEATURES Wintoror P. HAYNES 269 THE SKELETON OF TRIMERORHACHIS - .-. - .- .-.--...8..W. Wituston. — 201 JEL PL POE aR nat eel an Naar pa ea ey ines eR PUETTCA TIONS ree PET sh ween me gor 6 kaa ode Sunes See Sarg MeN Bol TY OR CHICAGO.-PRESS CHICAGO, ILLINOIS, U.S.A. AGENTS \ THE CAMBRIDGE UNIVERSITY PRESS, Lonpon anv EDINBURGH KARL W. HIERSEMANN, Letpzic THE MARUZEN-KABUSHIKI-KAISHA, Toxyo, Osaka, Kyoto ' ‘THE MISSION BOOK COMPANY, Suancuat THE JOURNAL OF GEOLOGY EDITED BY THOMAS C. CHAMBERLIN AND ROLLIN D. SALISBURY With the Active Collaboration of , SAMUEL W. WILLISTON ALBERT JOHANNSEN Vertebrate Paleontology ee" Petrology STUART WELLER - ROLLIN T. CHAMBERLIN Invertebrate Paleontology Dynamic Geology ALBERT D. BROKAW Economic Geology The Journal of Geology is published by the University of Chicago at the University Press semi- quarterly, on or about the following dates: February 1, March 15, May I, 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. {J Postage is prepaid by the publishers on all orders from the United States, Mexico, Cuba, Porto Rico, Panama Cana} Zone, Republic of Panama, Ilawaiian Islands, Philippine Islands, Guam, Samoan Islands, Shanghai | Postage is charged extra as follows: 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 single copies, 11 cents (total 76 cents). [Patrons are requested to make all remittances payable to the University of Chicago Press in postal or express money abe or bank drafts. The following agents have been appointed and are authorized to quote the prices indicated: For the British Empire: The Cambridge University Press, Fetter Lane, London, E.C. Yearly subscriptions, including postage, 18s. 8¢. each; single copies, including postage, 3s. 2d, each. For the Chinese Empire: The Mission Book Company, 18 Pekin Road, Shanghai, China. Yearly subscriptions, $4.00; single copies, 65 cents, or their equivalents in Chinese money. Postage extra, if mailed direct outside of Shanghai, on yearly subscriptions 53 cents, on single copies 11 cents. For the continent of Europe: Karl W. Hiersemann, Konigstrasse 29, Leipzig, Germany. Yearly subscriptions, including postage, M. 22.65 each; single copies, including postage, M. 3.80 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, including postage, Yen 1.50 each. ; Claims for missing numbers should be made within the month following the regular month of publication. The publishers expect to supply missing numbers free only when they have been lost in transit. \ Business correspondence should be addressed to The University of Chicago Press, Chicago, Ill. Communications for the editors and manuscripts should be addressed to the Editors ‘oF THE JOURNAL OF GEOLOGY, the University of Chicago, Chicago, Ill. Contributors are requested to write scientific and proper names with particular care and 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 leading articles will be printed as ordered, of which 50 (without covers) will be furnished gvaéis, the actual cost of the remainder (and covers, if desired) to be paid for by the author. The actual cost may vary, and will depend upon the amount of work in re-making the pages into forms, presswork, paper, binding, etc. Separates containing half-tones may be expected to cost somewhat more, the increase depending upon - number of cuts and the amount of work required upon them. Entered as second-class matter, March 20, 1893. at the Post-office at Chicago, IIl., under the Act of March 3, 1879. VOLUME XXIV NUMBER 3 THE JOURNAL OF GEOLOGY APRIL-MAY 1916 ORVILLE A. DERBY JOHN C. BRANNER Leland Stanford Junior University Orville Adelbert Derby, for many years one of the associate editors of this Journal, was born at Kelloggsville, New York, on July 23, 1851, and died by his own hand at Rio de Janeiro, Brazil, on November 27, 1915. After graduating at the high school, Derby entered Cornell University in 1869, taking what was then called the scientific course. While he was yet a Freshman, however, he became so interested in geology and was such a promising student that he was selected by Professor Charles Fred Hartt, then professor of geology at Cornell, to accompany him on a trip to Brazil in the summer of 1870. That was the first trip made to Brazil by Derby; it deter- mined both his career and the whole course of his life. On his first voyage he visited Pernambuco, and made the first considerable collection of fossils ever made at Maria Farinha, a locality that has since been looked upon with especial interest by students of the Mesozoic history of South America. In the summer of 1871 he went to Brazil with Hartt again, this time visiting the Amazon valley and making an important collec- tion of Carboniferous fossils from the limestones at Itaitiba on the lower Tapajos River. Vol. XXIV, No. 3 209 210 JOHN C. BRANNER In 1873 he graduated from Cornell University with the degree of Bachelor of Science, and the year following he continued his geological studies for the Master’s degree, which he received in June, 1874. His thesis was “On the Carboniferous Brachiopoda of Itaitiba, Rio Tapajos,’’ and was published as No. 2 of Vol. I of the Bulletin of Cornell University, Ithaca, 1874. That was Derby’s first publication on the geology of Brazil, and it is not only a valuable paper in itself, but it is especially interesting in view of subsequent developments. The Itaitiba fossils were in com- pact limestone, but as they were silicified they could be obtained in satisfactory form only by dissolving away the surrounding rock —a long and tedious process which would have thoroughly dis- couraged most young men of Derby’s age. The spires of many of the specimens of these brachiopods have seldom been surpassed for delicacy and perfection. The art of illustration was far from being so well developed in those days as it is now, and we thought ourselves very fortunate in being able to make and use the crude photographs with which that paper was illustrated. In 1873 Derby was appointed instructor in geology at Cornell, and in the summer of 1874 Professor Hartt made arrangements to go to Brazil again. Leave of absence was obtained, Derby was placed in charge of the work of instruction in the department, and in Sep- tember, 1874, Hartt went to Brazil again, taking Branner with him as his only assistant and going by way of Europe. It-is often said that Hartt went to Rio on the invitation of the Brazilian government or of the Emperor D. Pedro II. As a matter of fact he went entirely on his own responsibility and without invitation from anyone, but with the idea’ of inducing the Brazilian govern- ment to establish a geological survey under his direction. Arriving in Rio de Janeiro, he at once devoted all his energies to interesting the leading men in a geological survey of the empire, and by the end of the year the survey was authorized and provided for, and O. A. Derby, Richard Rathbun, and E. F. Pacheco Jordao were named as assistants of the new “‘Commissao Geologica do Imperio do Brasil.”’ In December, 1875, Derby reached Rio de Janeiro and began his work under the government. He held this ORVILLE A. DERBY 25E position less than two years, for through a change of ministry the survey was abolished in 1877, and Hartt died in Rio that same year. Shortly after the suspension of the survey, however, Derby was given a position in the National Museum at Rio as curator in charge of geology, a position which enabled him to continue his studies on the geology of Brazil, and, to a certain extent, to pre- serve the results of the work of the extinct survey. He remained in the museum until 1886 when he was made state geologist of the Brazilian state of Sao Paulo. The establishment of the Sao Paulo survey was a step of great importance to geological science in Brazil, for Derby’s knowledge of and interest in the geology of the country as a whole enabled him to grasp more firmly the geological problems of that particular state, and at the same time he became and remained, up to the time of his death, the leading authority on the geology of Brazil. He was state geologist of Sao Paulo until 1904, when he resigned. In 1907 a new federal geological service was provided for, and Derby was made its chief, a position he held during the rest of his life. The first edition of Branner’s Geologia Elementar, a work pre- pared especially for Brazilian students of geology, was thus dedi- cated: ‘To Orville A. Derby, who has devoted his life to the study of the geology of Brazil, and has done more than anyone else to solve its many problems, this work is affectionately dedicated.”’ This is a brief and mild statement of Derby’s great services to Brazil and to the science of geology, without mentioning his many other services to science and to that country. First and last Derby was a paleontologist. He had no fond- ness for administrative work; he was but little interested in struc- tural geology or in its methods; he was forced by circumstances into some acquaintance with microscopic petrography; but his interest in paleontology was genuine, deep, and all-comprehensive. From all the cares of office and the worries of life he found relief and happiness in boxes of poorly preserved fossils that most paleontologists would have put away as not worth while. It was chiefly to this interest of his in paleontology that we owe Dr. C. A. White’s Contributions to the Paleontology of Brasil, 212 JOHN C. BRANNER published at Rio in 1887; John M. Clarke’s Trilobites of the Ereré and Maecuru Sandstones, Rio, 1896; Upper Silurian Fauna of Rio Trombetas, Rio, 1899; Devonian Mollusks of the State of Parad, Rio, 1899;. and Devonian Fossils of Paranda, Rio, 1913. Besides these excellent works there are many smaller papers on paleontology that cannot be mentioned here, and there still remains unpublished an important volume by D. S. Jordan on the Cretaceous fossil fishes of Ceara. During the last eight years Derby gave much of his time to the study of Psaronius and its relationships. The last of his published papers was on the stem structure of T7zetea singularis, and appeared , in the American Journal of Science for March, 1915, pp. 251-60. Because he had to undertake work in regions but poorly supplied with maps, one of his first and most important duties, when he became state geologist of Sao Paulo, was the inauguration of topo- graphic work. This work was intrusted to Horace E. Williams, an able and energetic young American to whom the state of Sao Paulo and the scientific world are indebted for an excellent series of topographic maps on a scale of one to 100,000, to say nothing of his explorations of the western portions of Sao Paulo, his work on the Serra da Canastra, etc. Derby’s own list of publications on the geology of Brazil num- bers 125 papers. Naturally they embrace a wide range of sub- jects. Ten of his papers relate to the geology and genesis of the Brazilian diamonds. One of these, on the geology of the diamond and carbonado region of the state of Bahia, was the first publication to give an idea of the geology of that little-known district. He became interested in the early cartography of Brazil, and published a number of papers on that subject. As an author and as a scientific reasoner he was an extremely cautious man, so much so that the word “‘hedge”’ was constantly on his lips both for his own guidance and as a warning to his assistants. The last evening I spent in his rooms at Rio de Janeiro he referred to this personal trait, and remarked that it had prevented his marrying—that he was too cautious to take the risk. This cautiousness of his was probably the real reason for some of the ~ ORVILLE A. DERBY 213 long delays in publishing his results, delays which led to the tying up of his own results and those of his assistants. Without doubt “he hoped that the delays would enable him to put everything beyond question and to make his reports final and complete instead of preliminary and tentative. But the delays were prolonged from year to year until his assistants became discouraged and the gov- ernment more or less exasperated at the lack of practical results for such great and long-continued expenditures. It was probably this long delay that finally led to his resignation as state geologist of Sao Paulo. 5 Derby never felt obliged to show results. After he had been state geologist of Sado Paulo for ten or twelve years, and had pub- lished next to nothing on the geology of that state, I asked him point blank, and with some feeling, where his results were. He replied: ‘‘They are in my head.” We had to change the subject. But the important fact behind his delays is that the geology of Sao Paulo is difficult and involves problems that he had not been able to settle to his own satisfaction, and he was unwilling to commit himself definitely to paper and thus lay himself open to adverse criticism. It seemed unfortunate for Brazil, for himself, and for the cause of science that he was unable to bring himself to take an active interest in the economic geology of the country. But his first and only interest in geology was in geology as a pure science. To him a fossil was a thing of beauty, of interest and value, and a joy forever, but a mine or an industry was, after all, only an industry whose main object was money-getting. Derby was a man of unlimited grit. When once he decided upon a course of action nothing turned him to the right or to the left. His whole life is a demonstration of his power to make good in spite of obstacles that would have been insurmountable for most men—his determination to be the leading authority on the geology of Brazil, cost what it might. How many of us would have lived for forty years, in a foreign country, cut off, as he was, from all personal contact with the geologists of the world at large and from the people of his own race and from his own family? And yet, from the time he went to Rio 214 JOHN C. BRANNER in 1875 to the day of his death he visited the United States only twice. One of these visits was in 1883 when he spent three months at Washington; the other was in 1890 when he attended the meeting of the American Association for the Advancement of Science at Indianapolis. When the Commissao Geologica was abolished in 1877 the rest of us took to our heels. Not so Derby; he was not to be stampeded by a simple lack of funds or of employment; he meant to save the results of the work of Hartt and of his colleagues, and, in so far as it could be done, he did it. Personally Derby was one of the kindest-hearted and most affectionate men I have ever known. His last dollar was at the service of his friends, and his right hand knew nothing of the kind deeds done by his left. The beggars in the streets found him their easiest victim. He was held in the highest esteem in the community in which he lived. He stood for uprightness and honorable dealing, and he was never the willing tool of designing adventurers. For many years he has been justly regarded as the leading geologist in South America, and his standing is due, not to the fact that there are but few first-class geologists in South America, but to his ability and to his excellent work. In 1892 he was awarded the Wollaston prize of the Geological Society of London, while his distinguished services led to his being made one of the associate editors of the Journal of Geology and to his election to membership in various learned societies in different parts of the world. He was a frequent contributor to the American Journal of Science and to this Journal. A list of his papers on the geology of Brazil up to 1go09 is given in the Bulletin of the Geological Society of America, XX, pp. 36-42. To that list should be added thirteen additional titles of papers that have appeared since its publication. TYPES OF PRISMATIC STRUCTURE IN IGNEOUS ROCKS? ROBERT B. SOSMAN Geophysical Laboratory, Carnegie Institution of Washington The question of the cause of columnar or prismatic jointing in igneous rocks was thought to have been satisfactorily settled by the writings of Thomson, Mallet, Bonney, Iddings, and others, until it was reopened recently by the investigations of several French physicists. As the subject seems to be in need of further discussion and experimental study, I have brought together observations on several hypotheses of prismatic jointing, hoping to show that the study of these structures may yield much more precise information than is now available as to the original conditions of occurrence of the igneous rocks in which such structures are found. CRYSTALLIZATION HYPOTHESIS The first hypothesis as to the origin of prismatic structure which had any experimental or observational basis was that of Gregory Watt2 and may be entitled the “crystallization hypothesis.”’ Watt, in 1804, observed that a large mass of basalt which he had melted down in a reverbatory furnace crystallized radially from centers which were fairly regularly spaced in a horizontal plane; the intersections of these radially growing fibrous bundles formed a network of hexagonal partings through the mass, leading Watt to the conclusion that this manner of crystallization, by its vertical extension upward from the base of a mass of basalt, must have been the cause of the prisms found in the Giant’s Causeway, Fingal’s Cave, and elsewhere. x Presented before the Geological Society of Washington, April 28, 1915. 2 Gregory Watt, ‘Observations on Basalt, and on the Transition from the Vitreous to the Stony Texture,” etc., Phil Trans., 1804, pp. 279-314- Watt also explains clearly the contractional origin of such structures as mud and starch prisms. 215 ‘ 216 ROBERT B. SOSMAN This explanation seems to have been satisfactory to many of the earlier authors of geological treatises,’ but before many years had passed doubts began to arise as to whether this process could have been an efficient cause of the numerous cases of columnar structure which began to accumulate in geological literature as travel became more extensive and observations multiplied. James Thomson? in 1863 urged that contraction of a homogeneous mass was a sufficient cause for all columnar structure, and that the hypothesis of crystallization from centers was unnecessary and improbable. Mallet’ discussed the contraction hypothesis in detail, showing how it would account, in his opinion, for all of the struc- tures found in columnar rocks. Bonney,’ Iddings’ and others have followed the same lines of argument. CONTRACTION HYPOTHESIS The radial-contraction hypothesis is still the explanation gen- erally accepted by the textbooks, and perhaps applies in the majority of cases of prismatic structure. But a much more com-. plete discussion of this hypothesis than has yet been published could be profitably made, for there has been no attempt at any quantitative application of it to actual occurrences. It has served hitherto simply as a qualitative explanation: The relation of the size, shape, curvature, jointing, and other properties of the columns to the original temperature, viscosity, and rate and manner of cool- ing of the rock is capable of more exact definition. For instance, the time factor in cooling in its relation to the elastic properties of the rock does not seem to have been considered * More or less vaguely associated with this definite hypothesis was the idea of a “‘concretionary force” which is frequently referred to. The idea that columns might be due to the mutual compression of actual spheroids of lava (now understood as ‘‘pillow”’ lava) was also more or less confused with the crystallization hypothesis. Watt’s idea of the matter seems to have been perfectly clear, but Mallet, for instance, misunderstands Watt’s “‘mutual compression of spheroids” to mean actual compres- sion (Phil. Mag., L [1875], 221-24); the words “mutual interference of radially growing spheroids” state Watt’s meaning more clearly. 2 Brit. Assoc. Rep., 1863, Abstract, p. 89. 3 Phil. Mag. (4), L (1875), 122-35, 201-26. 4 Quar. Jour. Geol. Soc., XXXII (1876), 140-54. 5 Amer. Jour. Sci., XX XI (1886), 321-31. PRISMATIC STRUCTURE IN IGNEOUS ROCKS —_.217 in previous discussions. If the mass is cooling slowly, the crystal- lized shell may be able to adjust itself by a slow movement to the stress produced by contraction, so that the strain does not for some time pass a given value. If the cooling is rapid, on the other hand, the strain may be rapidly raised through the inability of the mass to flow as rapidly as the stress is applied. Under conditions of rapid cooling, therefore, the temperature at which the stresses become sufficient to produce rupture will be higher than under conditions of slow cooling." : - Another point concerns the conditions of rupture. Published discussions of formation of columns by contraction have tacitly assumed that the condition of rupture is that the extension shall exceed a certain limiting value. This is only one out of several possible conditions of rupture. Various hypotheses have been pro- posed by physicists (limiting tension, limiting positive or negative strain, limiting shear), of which the best founded experimentally is that of Tresca and Darwin, according to which rupture occurs when the maximum difference of the greatest and least principal stresses . reaches a certain limiting value.? Although the acceptance of this condition of rupture as the fundamental one does not simplify the problem of calculating the actual physical magnitudes of tempera- ture, temperature gradient, stress, and strain in any given case, yet it should permit a more complete analysis of the kinds of structure that will result from different conditions of cooling. Such an analysis is; however, beyond the scope of the present article. CONTRACTION OF PHYSICALLY HETEROGENEOUS MATERIAL Prismatic structure is very common in materials which are heterogeneous as regards their state of aggregation (such as mud and wet starch), that is, in which solid matter is suspended in or mixed with a liquid. It is a question whether the formation of a prismatic structure in such materials is strictly comparable with most cases of contraction prisms in igneous rocks. The pr-ncipal The above-mentioned effect of the rate of cooling is quite distinct from the commonly recognized effect, which appears in the temperature gradient away from the surface of the cooling mass. 2 Love, Theory of Elasticity, 1906, p. 119. \ 218 ROBERT B. SOSMAN difference is in the strength of the materials. Very considerable stresses may accumulate in a glassy or crystalline rock before rup- ture occurs, and when it does occur, the crack extends suddenly a considerable distance into the mass. A layer of wet mud, on the other hand, accumulates practically no stresses, as the forces of cohesion and liquid surface tension to be overcome are very small. The cracks therefore form much more gradually, and grow little by little as desiccation proceeds. ‘They have even been observed to form under water,’ probably as a result of freezing and melting.” It is possible that some basalt prisms have been formed in the same way as the slowly formed mud cracks, by the slow shrinkage of a material which is partly solid and partly liquid, for the normal course of crystallization of an igneous rock consists in the separation of certain portions as crystals while the remainder stays liquid until a lower temperature is reached. It has been commonly observed, however, that the boundaries of contraction columns frequently cut across the crystals of the rock, showing that solidification was practically complete before the crack formed. ; An example of prismatic, although not columnar, structure pro- duced in this manner is probably to be found in the ‘‘apparent sun- crack structure in diabase,” described by Wherry as occurring in the upper surface of the great diabase sill of Pennsylvania, west of Philadelphia. He explains it as due to contraction jointing fol- lowed by the penetration of still liquid material into the cracks from below. At first sight this occurrence has some of the characteristics of prismatic structure due to liquid convection accompanied by segregation, but a re-examination of the structure by Dr. Wherry and the author in May, 1915, showed that the angles and polygons t Moore, Am. Jour. Sci., XX XVIII (1914), 101-2. 2 Mud cracks may also belong to the other types of columnar structure. Where the deposit is very fine grained and homogeneous, the walls of the columns may show the feathery patterns characteristic of a fractured solid, resulting from breaks (either sudden or slow-growing) which occurred when the material was nearly dry, and indicating the existence of tensional stresses. On the other hand, a prismatic structure of apparently convectional origin has been observed by Guillaume (Soc. Franc. Phys., Bull. Seances, 1907, pp. 50-51) in mud flows in sub-Arctic regions. 3 E. T. Wherry, “‘Apparent Sun-Crack Structures and Ringing-Rock Phenomena in the Triassic Diabase of Eastern Pennsylvania,” Acad. Nat. Sci., Philadelphia, Proc., LXIV (1912), 169-72. PRISMATIC STRUCTURE IN IGNEOUS ROCKS 219 are those produced by contraction, not by convection (see p. 22 ie A photograph of the occurrence is shown in Fig.1. An examination by Wherry of the cross-section of one of the small ‘“‘dikes’’ shows that it has an irregular boundary, that it grades off without a sharp break into the surrounding rock, and that it is more coarsely crys- talline than the surrounding material. It appears to be, therefore, a case of prismatic structure due to contraction in physically hetero- geneous material, and quite distinct from the usual type of con- traction prisms. Dr. N. L. Bowen, of this laboratory, informs me that he has seen a similar structure in the upper surface of a diabase sill north of Lake Superior.* CONVECTION HYPOTHESIS E. H. Weber? described in 1855 a phenomenon observed by him on microscope slides on which a solid was being precipitated from alcohol-water mixtures. The liquid was observed to circulate and to divide itself up into regular polyhedral cells. A similar phenomenon was observed by James Thomson! in 1882, in a soap solution. It remained for the French physicist Bénard,* in 1900, to make a really thorough study of the subject, and his experiments have brought out a number of new and interesting facts. A polygonal structure is easily produced in a layer of liquid which is shallow in comparison with its horizontal extent, and which is losing heat from its upper surface or is gaining heat from its lower surface. If the top surface is cooler than the bottom, then the colder and denser liquid at the top tends to sink and the warmer bottom layer to rise, and convection currents must be set up: if the conditions are uniform and constant, a steady state of flow of some kind must ultimately be set up. In a flat liquid sheet of indefinite extent this state of flow must take the form of parallel rising and descending currents, and these will flow with minimum «Canada, Bur. Mines, Ann. Rep., XX (1911), 125-26. 2 Pogg. Ann., XCIV, (1855) 452-59- 3 Phil. Soc. Glasgow, Proc. XIII, (1882), 464-68. Thomson recognized the simi- larity of the pattern to that of the Giant’s Causeway. 4H. Bénard, Les tourbillons cellulaires dans une nappe liquide, etc., thesis, Paris, 1901; Rev. gén. Sci., XI (1900), 1261-71, 1309-38. ROBERT B. SOSMAN 220 “eluvA[ASUUO ‘erydjapeiyd JO }SoM [IIs aseqetp Jo doy, “[erayeuI snoauasosoyoy Aqpeoisyd ut uorjseI}UOD 0} onp sinjonI}s OeUSTIG—I “OI PRISMATIC STRUCTURE IN IGNEOUS ROCKS 221 friction only when they divide the liquid into hexagonal cells, as can be shown by the same line of argument as is used to prove that a uniform shell, under tension due to its own contraction, breaks with minimum energy expenditure when it divides into hexagons. Fig. 2 is a cross-section of one of these hexagonal cells, showing how the currents rise in the middle of each prism and flow down at FREE SURFACE ! { | | | | | | | | | | | eee we we ea U/ Fic. 2.—Cross-section of a hexagonal cell, showing how the currents rise in the middle of each prism and flow down at the boundaries. the boundaries. The contour of the surface of the liquid is exagger- ated in the figure, but the relief is quite sufficient to permit the structure of the circulating liquid to be observed by various optical methods. Fig. 3 shows three examples of these structures in a melted wax, taken under different conditions of temperature and thickness and before the final steady state of circulation had been attained. A state of subdivision into irregular cells of from four to seven sides is attained in a few minutes, even in a viscous oil. These cells then join and subdivide repeatedly until finally, if the condi- tions are constant, a perfect system of hexagonal cells is produced. Even when the liquid is originally in motion, convection cells form which show little or no trace of the original direction of movement of the liquid as a whole. ROBERT B. SOSMAN bo bo bo Waxes and oils were used for most of Bénard’s experiments, because at his working temperatures of 100° and lower the requisite conditions as to viscosity and low volatility could best be obtained with these materials. By suspending in them finely powdered sub- stances such as graphite or lycopodium, Bénard was able to show visually and to photograph the cells produced, without the aid of special optical devices. As is to be expected, the dimensions of the cells depend upon the thickness of the liquid layer, the temperature difference between top and bottom, and the viscosity and temperature of the liquid. Fic. 3.—Three examples of hexagonal cells formed in a melted wax, taken under different conditions of temperature and thickness and before the final steady state of circulation had been attained. In a given liquid at a given temperature, and at a constant tem- perature difference, the ratio of diameter to height is found to be constant. Other laws governing the form and size of the cells were found by Bénard, but it is unnecessary to discuss these in detail. Following Bénard, Dauzére* in 1907 showed that crystallization in salol and wax mixtures begins on the boundaries of the convec- tion cells. A mixture of beeswax and stearin, on solidifying, sepa- rates spontaneously into hexagonal prisms coinciding with the original convection cells. In pure stearin, also, crystallization begins at the corners of the cells. In every case the cells leave a permanent record of their existence in the crystallized solid, although in some cases the structure is quite invisible, and only *C. Dauzére, Jour. physique, VI (1907), 892-99; VII (1908), 930-34; Assn. franc. av. sci., 1908, pp. 289-96. PRISMATIC STRUCTURE IN IGNEOUS ROCKS 223 appears when the wax is bent. Dauzére pointed out the strong probability that certain symmetrical columns in Auvergne have been due to convection in the basalt in which they are formed.” In a horizontal sheet of molten rock which has come to rest after extrusion or intrusion it is obvious that we have some at least of the conditions necessary for the formation of convection cells. Hi the cells succeed in leaving any permanent record of themselves when the sheet solidifies, then subsequent contraction may bring out the structure by cracking the rock along the boundaries of the cells. In general there are two ways in which the convection cells might impress themselves on the crystallized rock. In the first case the axes of the liquid convection cells and of the solid prisms are coincident. Bénard found that a finely powdered substance which is heavier than the liquid tends to gather on the bottom of the vessel in little heaps situated on the axes of the convection cells, giving an appearance from above of uniformly spaced round spots. A floating substance, on the other hand, gathers along the boundaries of the cells at the surface. A substance in suspension gathers within the interior portion of the closed curves of Fig. 2, so that the liquid shows transparent both on the axes and along the boundaries of the cells. In a mixture, therefore, in which different crystalline phases are separating at different temperatures, a Cer tain amount of segregation is to be expected, and the solid prisms will coincide with the convection cells. In a substance which crystallizes as a unit, on the other hand, whether it be a pure substance or a considerably undercooled mix- ture, prisms may be formed without segregation. Bénard observed that in spermaceti the crystallization began at the corners of the cells. In pure stearin Dauzére found that crystallization beginning at centers on the cell boundaries extended uniformly in all directions until the growing cylindrical groups intersected to form prisms. It is evident that in this case the prisms will not coincide with the convection cells, but will nevertheless be symmetrical and regularly spaced. 1C. Dauzére, Assn. franc. av. Sci.; 1908, pp. 436-38; also Longchambon, Bull. Soc. Geol. Fr., XIII (1913), 33-38- 224 ROBERT B. SOSMAN It is of interest to note that this convection-crystallization hypothesis explains the original observation of Watt on the forma- tion of columns in a cooling artificially melted basalt mass (see p. 215). He accounted for his columns on the assumption that they were produced by the mutual interference of radially.growing crystal bundles, uniformly spaced in a horizontal plane. Why the crystal- lization centers should be uniformly spaced he was unable to say. The existence of convection prisms in the still liquid basalt provides the missing link in the series of phenomena. Crystallization may have begun at the axes of the convection prisms where a few early separating crystals had collected, or at the corners as observed by Bénard; in either case the crystallization centers would be uni- formly spaced horizontally.' If liquid convection is really the cause of all or any of the familiar naturally occurring basaltic columns, then it is important to know what criteria will help to decide the question in a given case. Furthermore, a systematic examination of natural columns will throw light on their history, whatever may be their mode of origin. What are the important characteristics of a given occurrence which should be observed in the field ? CHARACTERISTICS OF CONTRACTION AND CONVECTION PRISMS 1. Altitude-——The original attitude of columns formed by con- vection should be vertical, or very nearly so. Contraction columns, on the other hand, are usually perpendicular to a cooling surface; irregular conditions of cooling, furthermore, may cause them to curve in a great variety of ways. 2. Dimensions—Convection columns should be much wider, in proportion to their length, than contraction columns, which are commonly very longand narrow. The columns at Murols described by Dauzére are 1.5 to 2 m. wide and 5 to 1o m. high; those measured by O’Reilly in the Giant’s Causeway are 0.4 to 0.5 m. in width; the Causeway columns vary from 3 to 25 m. in total height. Scrope describes columns near La Queuille as much as 5 m. in diameter, and ro m. or less in height.2, The common contraction «See Longchambon, Bull. Soc. Geol. France, XIII (1913), 33-38. 2 Volcanoes of Central France, London, 1858, p. 136. PRISMATIC STRUCTURE IN IGNEOUS ROCKS 225 columns, on the other hand, are usually about o.2 m. or less in diameter; their length is often 20 m. without a joint, and their total length may be over 40m. It should be noted, however, that the composition of the rock may have a considerable effect on the size of columns under given conditions of cooling, the more salic rocks forming larger columns than the more femic rocks. 3. Shape of cross-section.—Convection columns, if perfect, should all be hexagonal. The more uniform the conditions have been, the greater the proportion of hexagons; in any case, the _ hexagonal sections will be in the majority. Seven-sided figures will be common, produced by the trunkation of one angle of a hexagon; pentagons will also occur frequently, by the elimination of one side of a hexagon. But three- and four-sided figures will be very rare. : In contraction columns, on the other hand, pentagons are likely to be the prevailing type, and four-sided figures are fairly numerous, while hexagons become less important. This distribution of poly- gons arises from the fact that a mass cracking under the stresses of its own thermal contraction, although theoretically it should break into perfect hexagons of equal area, actually tends to yield by the formation of master-cracks which are then joined up by the forma- tion of shorter cracks.* An example of thermal contraction prisms on a large scale is seen in the soil polygons of Arctic regions; a map of a set of these polygons, in a recent article by Leffingwell, shows clearly the contraction-type fissures described above. The relative frequency of polygons in some of Bénard’s artificial convection cells,? in the Giant’s Causeway,‘ and in a columnar dike® is Shown in Table I. t “The rock may rather be said to be divided into numerous perpendicular fissures, than to be prismatic, although the same picturesque effect is produced.”—Lyell, description of Torre del Greco. 2E. De K. Leffingwell, Jour. Geol., XXIII (1915), 653. 3 The photograph used for this computation was one taken while the liquid was cooling and the polygons were undergoing gradual changes, leading to the formation of 5- and 7-sided figures. Under steady conditions of heat flow the cells were hexagons almost without exception. 4J. P. O’Reilly, Roy. Irish Acad. Trans., XXVI (1879), 641-734. 5 A. Geikie, Ancient Volcanoes of Great Britain, illustration, p. 459. 226 ROBERT B. SOSMAN Bénard’ has recently shown photographically the identity of pattern between his convection cells and the cross-section of the basalt columns of the flow of Estreys (Haute-Loire), and has also pointed out the qualitative differences between this pattern and that produced by contraction. TABLE I COMPARATIVE FREQUENCY OF POLYGONS (PERCENTAGE) Grant’s CAUSEWAY : 4 ARTIFICIAL COLUMNAR No. or SIDES Gane Tee Alonga 50-Meter Within Line Measured Area Sil will dised sare ee ° ° ° 5.2 ye aE a eI POR ene ks S GAG 230 Bu5 28.4 Biola nistnin/s = 0a cies erm sasaki ai 36.3 30.7 | 24.8 ASCE CRIS On mirtcrets crs oc 45.2 Agen 50.5 20.7 Do shosdjs ose Ost aed een ae ee 12.7 -19.6 19.2 2.6 Steves SU a rhe 0.3 0.6 20 ° Total number counted. . 292 | 153 | 206 116 4. Frequency of angles——The angles of convection columns should approximate to 120°, while contraction columns will have a large proportion near go’. While the frequency of angles is a much more logical criterion than the frequency of different polygons, it is much more difficult to apply on account of the large number of angular measurements to be made. Such a series was made with great care by O’Reilly on the Giant’s Causeway, and I have sum- marized his results in Table II. O’Reilly’s deduction from his TABLE II FREQUENCY OF ANGLES IN 206 POLYGONS OF THE GIANT’S CAUSEWAY Range No. of Occurrences Range No. of Occurrences 64° to “Sse a meee 9 ETS 0) COWL 5 ww. 2 2383 E> gor tO! JSG ua ee 19 mac sOntOiras clas. cu 2363 $5430 HtO OS oc 563 T35; ee .to waste, eras Q5° 30) tO Tose. tee, 1033 TAG BON EO TOS ocho 183 EOS” 30" LOVERS uae ees 215 ROGeGetO Esso. . Jee 5 measurements was that the form of the columns had been governed by the principal angles of the constituent minerals of the basalt, a view which has not met with general acceptance. 2 Compt. rend., CLVI (1913), 882-84. PRISMATIC STRUCTURE IN IGNEOUS ROCKS 227 5. Difference in composition and texture between the axis and the periphery of the columns.—Obviously, no variation whatever should appear in contraction columns. If the columns are due to convec- tion, however, there might or might not be a differentiation, depend- ing upon whether the rock crystallized practically as a unit, or whether it crystallized in stages which permitted of segregation in the convection cells (see p. 223). In 1914 Dr. H. S. Washington, of this laboratory, examined, in the museum of the University of Catania, a polished section of a column from one of the prehistoric basaltic flows of the Mount Etna region, and observed no variation of texture across the section. From their shape and manner of occurrence, these columns at Etna would seem to be due to pure contraction, and no variation is to be expected. On the other hand, evidence is not lacking in geological litera- ture of what seems to be a differentiation between the border and axis of some basalt columns. Scrope, in his description of the vol- canoes of central France, states that “occasionally (as for example at La Tour d’Auvergne, in the Mont Dore), the columns show a cylinder of compact black basalt within a prismatic case of lighter colour and looser texture, a segregation of dissimilar matter having accompanied the concretionary action.’’* Delesse? made in 1858 an interesting comparison of the density of the interiors and exteriors of a variety of columns, the results of which are shown in Table III. Here again a difference between the interior and exterior is indicated in some of the columns, though not in all. Unfortunately the source of the samples which showed small differ- ences is not stated; it may be that they are columns of the narrow contraction type. Delesse took care to assure himself that the differences were real and were not due to weathering of the columns, but it is not impossible that the differences are really due to weather- ing, since he had not the modern microscopic facilities for examining the individual minerals in thin section. t Volcanos, 1862, p. 100. In speaking of ‘‘concretionary action” Scrope seems to be referring to the rather vague hypothesis of columnar structure which prevailed at the time (see note, p. 216). 2 Delesse, ‘Variations dans les roches se divisant en prismes,’’ Compt. rend., XLVII (1858), 448-so. 228 _ ROBERT B. SOSMAN The regularity and symmetry of the columns of the Giant’s Causeway suggests the convectional origin. It seemed of interest, therefore, to examine a polished cross-section of one of these columns for evidence of differentiation. Through the kindness of Dr. G. P. Merrill, of the United States National Museum, a polished section was cut for us from a Giant’s Causeway column in the Museum, and also one from a column from near Bonn on the Rhine. TABLE III DIFFERENCE IN DENSITY BETWEEN AXIS AND SURFACE OF BASALT COLUMNS (DELESSE) DENSITY DIFFERENCE OF DENSITY Center Outside Per Cent rachyte, Icelande- eras 2.404 2.478 0.64 Mrachyte, isle Bonces. 424° eee 2.469 2.439 ea Phonolite, Isle Lamlash.....:.... 2.541 2.500 1.26 Trap yAntrims 294 s2-4 ae aoe 2.QI1 2.857 1.85 Basalltc.2 2 ciscshie oes see 2.930 2.933 —o.10 Basalt: Saenotecticn = oer Gee eee 3.030 3.030 0.00 Basalticnk 2 As pac etn cee eee 2.924 2.916 One Basalt: 5/2 8i\\acc ce cee ee 3.053 3.030 0.75 Basal tsi. hcl aks tae ge eee 3.044 3.008 1.18 The Bonn column was five-sided, with a maximum cross dimen- sion of 18 cm. The cross-joint near which the section was cut showed fracture lines radiating from one corner, and the joint passed straight across. What appeared to be an inclusion about 16 mm. in diameter showed near the center, and another of similar size seemed to have been cut in two by one face of the column. A sharp weathered zone 3 mm. wide showed clearly, but no other difference between center and border appeared. The Giant’s Causeway column was also five-sided, with a maxi- mum cross-dimension of 37 cm. ‘The section was cut near the convex side of a shallow ball-and-socket joint; the fracture of this joint seemed to have radiated from the center, not from any point of the border. The rough surface gave an appearance of finer grain at the border than at the center. On the polished face, how- ever, no such gradation was visible. There was a sharp weathered zone 3 mm. wide, inside of which was a zone varying from 6 to PRISMATIC STRUCTURE IN IGNEOUS ROCKS 229 22 mm. in width, with an ill-defined wavy border. This also may have been due to weathering. Within the central eighth of the area appeared 5 amygdules from 2 to 4 mm. in diameter, and filled with a greenish opalescent mineral. Five others, from 1 to 2 mm. in diameter, appeared in the remaining seven-eighths of the area, none being closer than 25 mm. to the border. The section therefore offers no decisive evidence of a differentiation, although meiesealy different in character from the Bonn column. 6. Types of cross-jointing in the columns.—A differentiation due to convection might be expected to affect the cross-joints of the columns. The peculiar convex-concave or cup-and-ball joints are seldom found in irregular narrow columns of the typical contraction type, and might have some direct connection with a convection structure. Another type of cross-jointing of columns is the “platy” variety, which is sometimes very regular; its origin has not been satisfactorily explained from the physicist’s standpoint. Certain special peculiarities of the cross-jointing may also have to do with the mode of origin of the columns. For instance, James Thomson! observed that the symmetrical concave-convex joints of columns from the Giant’s Causeway have their origin in a small spot or knob which lies at or near the axis of the column, and differs in texture and hardness from the rest of the rock; from this origin the crack has spread outward, as shown by the radial fracture lines. This same form of fracture has just been described above, as occur- ring in the National Museum’s specimen from the Giant’s Cause- way. Dauzére mentions the same peculiarity in the columns at Murols, and compares it with the core (noyau) which forms in the convection prisms of his wax-salol mixture. There seems to be good foundation for the opinion that some sort of original structure is responsible for the spheroidal weathering of columns, and that it is not due solely to the rounding off of jointed blocks by weathering, as some have claimed. ‘Thus Bonney cites numerous examples of spheroids formed from columns which showed no cross-joints whatever.2. Whether these latent spheroids have any connection with the manner of growth of the column it is t Belfast Nat. Field Club, Ann. Rep., VII (1869), 28-34. 2 Quar. Jour. Geol. Soc., XXXII (1876), 140-54. 230 ROBERT B. SOSMAN as yet impossible to say. LLongchambon’ suggested that the super- imposed spheroids are due to a breaking up of long liquid convec- tion cells into a number of shorter ones, each with its own local circulation, but there is no experimental evidence to support this. 7. Irregularities of faces of prisms.—Some basalt prisms show the ‘“‘feather-patterns”’ characteristic of fractures in homogeneous solids. Their occurrence points strongly to a purely contractional origin. They have been observed in the joint planes of slates, and have been made the subject of an interesting study by Woodworth.” SURFACE STRUCTURE PRODUCED BY INTERNAL EXPANSION In addition to the prismatic structures produced by contraction and convection or by convection combined with crystallization and contraction, still another type needs to be considered, namely that due to expansion. The accompanying photograph of a polygonal structure in a cement briquette (Fig. 4) is an illustration of the formation of this structure by internal expansion. This sample, which was kindly furnished us by Mr. A. A. Klein, of the Bureau of Standards in Pittsburgh, was made from a cement which contained free lime; this by its hydration and absorption of carbon dioxide from the air has expanded and destroyed the briquette. It is possible that the ““weather-crack”’ structure on the surface of diabase bowlders is likewise caused by internal expansion. Wherry? has shown that there is no visible difference in texture underlying these weather-cracks. Expansion of the surface by hydration has been assumed as the cause of the structure; but this would produce compression in the surface, accompanied by the formation of shells (as indeed often occurs), whereas the ‘‘weather- crack”’ structure is one indicating tension. It is necessary for hydration to proceed into deeper portions of the rock before tension t Soc. Geol. France, Compt. rend. somm., 1912, pp. 181-83; Bull., XIII (1913), 33-38. 2 Proc. Boston Soc. Nat. Hist., XXVII (1896), 163-83. For an extended study of these feather fractures in glass and metals see Ch. de Fréminville, ‘“‘Recherches sur la fragilité; L’éclatement,” Rév. metallurgie, 1914; also Mallock, Proc. Roy. Soc., A, LXXXII (1909), 26-20. 3 Loc. cit. PRISMATIC STRUCTURE IN IGNEOUS ROCKS 231 is set up in the surface; the cracks then produced are soon widened by solution. A photograph of an excellent example of this type of structure in diabase is given in Fig. 5. Internal expansion may also account for the prismatic surface structure of ‘‘bread-crust bombs,” although this remains to be proved. Fic. 4.—A polygonal structure in a cement briquette, caused by internal expansion. SUMMARY From the physical standpoint, several types of prismatic struc- ture in igneous rocks can be distinguished. The first and most common is due purely to thermal contraction in the crystallized rock; examples are numerous and familiar. A subordinate type of contraction structure is produced when the contraction and separation occur while the magma is still partly crystalline and partly liquid; this type is illustrated by an occurrence in a diabase sill in eastern Pennsylvania. ROBERT B. SOSMAN Jop[Moq asequip ¥ JO aov]INs ay} UO aINJONAYs ,, YovId-19y}9\\ ,,—'S “OI ci PRISMATIC STRUCTURE IN IGNEOUS ROCKS 233 The second general type is produced by convectional circulation of the magma while still liquid. The cells so produced persist until solidification begins, and may leave a record in the rock either by causing segregation in the cell walls and axes, or by originating regularly spaced centers of crystallization. The experimental and observational data on the occurrence of this type in igneous rocks are suggestive, but cannot yet be said to amount to decisive proof. A third type of prismatic structure is produced by internal expansion. It has been produced artificially, and is offered as the explanation of the “weather-crack”’ structure seen in diabase bowlders. In the study of these structures, the following field observations are those which will be of greatest interest in the further study of the problem: (1) attitude of prisms, (2) their diameter and length, (3) frequency of four-, five-, six-, and seven-sided polygons, (4) fre- quency of angles (especially 90° and 120°), (5) variation, if any, of composition and texture in the cross-section, (6) types of cross- jointing (platy, concave or convex, spheroidal), (7) spacing of cross- joints, (8) peculiarities of cross-joints (e.g., whether cracked from center or from borders), (9) degree of irregularity in sides of prisms, (10) other peculiarities, such as tapering, partial longitudinal joint- ing, etc. The primary object of this discussion is to call attention to the possibilities of the prismatic structure of a given rock body as an index of its conditions of formation. Quantitative data on col- umnar structures are very scarce; yet quantitative measurements must precede quantitative deductions. We wish to know the tem- perature of the rock when it was intruded or extruded; its viscosity when it began to cool and when it began to crystallize; the amount and kind of gases which it released; if extrusive, the climatic con- ditions under which it cooled; if intrusive, the properties of its inclosing strata at the time of intrusion. These and other facts are deducible only from the present properties of the rock, among which its prismatic structure will prove of great importance. Equally necessary with the field data are experimental studies of the structures produced in a cooling magma under conditions ‘ g 234 ROBERT B. SOSMAN that can be controlled and measured. Such experiments will require the melting and handling of larger quantities than it has been customary to use for laboratory experiments, but the diffi- culties ought not to prove serious. Even in the absence of such experimental data, much can be learned from a careful field exami- nation of prismatic and columnar structures. ELLIPSOIDAL LAVAS IN THE GLACIER NATIONAL PARK, MONTANA! LANCASTER D. BURLING Geological Survey, Ottawa, Canada The paper by Capps on “Some Ellipsoidal Lavas on Prince William Sound, Alaska,” recalls to my mind a similar occurrence which [ visited in 1907. The locality is now so accessible and the flow is so clearly subaqueous in character that a brief description of it may be of interest. Its outcrop appears in the ridge (Shepard Mountain) northeast of Flattop Mountain, Glacier National Park, Montana, and the features described are in that portion of the bed which overlooks the Shepard Glacier. The lava, to which the name Purcell lava has been commonly applied, interrupted the sedi- mentation of a flat-lying, greenish argillite which forms the upper- most part of the Siyeh formation of the pre-Cambrian. This argillite lies in normal position, and the portions above and below the lava bed are macroscopically identical. The Purcell lava is approximately 150 feet thick on Shepard Mountain and can be traced for miles to the southeast, north, and northwest. It is composed of six or more successive flows, each -of uneven and more or less ropy surface, separated by small and more or less local accumulations of shale. The lower 25 or 30 feet of the flow is composed of a conglomeration of dense, homo- geneous, spheroidal masses averaging 1 to 2 feet in diameter. They preserve their shape in the lower layers, being separated from each other by chert or drusy cavities, and many individuals have displaced considerable portions of the mud upon which they were rolled or shoved, even to the extent of complete burial. The bottom of the flow is therefore exceedingly irregular. Toward the top of this bed the individual spheroids yield more or less to the ‘Published by permission of the Deputy Minister of Mines. . 2 Jour. Geol., XXIII (1915), 45-51. 235 236 LANCASTER D. BURLING pressure of their fellows, and they unite to form an upper surface of moderate unevenness. The upper part of the entire flow is composed of a bed about 20 feet thick, which, though massive in character, is very porous. Vesicles are common near the base of several of the individual flows in the lower portion of the lava. On Mount Grinnell, to miles to the southeast, Finlay’ gives the thickness of the lava bed as 42 feet, but does not mention the ellip- soidal masses which Daly later describes from the same locality.” Finlay records the discovery of five genetically connected dikes on Flattop Mountain close to the localities where the ellipsoids are present. Elsewhere, though the lavas reached the surface through numerous widely scattered dikes, ellipsoidal structure has not been recorded. This period of igneous activity has been described’ as having genetically connected extrusive and intrusive phases, and it is interesting to note that strata, upon whose upper subaqueous surface lava was being extruded, should have been able, at a depth of only 600 feet, to accommodate themselves to the essentially contemporaneous intercalation, along single planes, of intrusive sills scores of square miles in extent. This flow seems to afford an excellent opportunity for determin- ing the value of certain criteria for distinguishing (1) subaerial from subaqueous flows, and (2) the top from the bottom of sub- aqueous flows. Here the normal attitude of the flow and its includ- ing sediments is unquestionable, and the bottom of the bed in which the ellipsoidal structure is developed is far more uneven than the top, an observation which lessens the importance of one of the criteria advanced by Capps. Furthermore, the silting up of cracks in the surface of the flow would seem more natural than the upward penetration, into cracks several feet in height, of mud sufficiently resistant to flatten the bases of individual ellipsoids. That the latter is true for the Prince William Sound locality? merely illus-_ trates the difficulty of obtaining competent and unconflicting criteria. t Bull. Geol. Soc. America, XIII (1912), 350. 2 Memoir Geol. Survey Canada, No. 38, Part I (1912), p. 217. 3 Daly, ibid., pp. 218-20. 4 Capps, Jour. Geol., XXIII (1915), 40. ELLIPSOIDAL LAVAS IN GLACIER NATIONAL PARK 237 So far as the Glacier National Park exposure of the Purcell lava is concerned, the following criteria would seem to indicate the bottom of a subaqueous flow: (1) discreteness of the basal spheroids and their relative competence to resist mashing; (2) comparative unevenness, with reference to the top; (3) the irregular displace- ment of the underlying shale by the basal spheroids; and (4) the presence of vesicles near the base of the individual flows. Criteria indicating the top are: (1) the common ropy structure; (2) more or less complete fusion of the individual spheroids; (3) comparative evenness, with reference to the bottom; (4) silting up of hollows in the top by strata whose laminae parallel those of the adjacent strata; and (5) the absence of vesicles in the upper portions of the individual flows. The flow under discussion covers an area hundreds of square miles in extent, and while its extrusive character has been recog- nized by various observers, the ellipsoidal structure has only been found at the localities described. It may have been sub- aqueous in places, subaerial in others (the Siyeh argillites are abundantly ripple-marked and sun-cracked in places), but many lines of evidence seem to prove its subaqueous character at the locality described, and indicate that ellipsoidal structure is a com- petent criterion of subaqueous extrusion. — THE ORIGIN OF RED BEDS A STUDY OF THE CONDITIONS OF ORIGIN OF THE PERMO- CARBONIFEROUS AND TRIASSIC RED BEDS OF THE WESTERN UNITED STATES C. W. TOMLINSON University of Minnesota PART II CONDITIONS OF DEPOSITION OF RED CLASTIC SEDIMENTS: MODERN TYPES Seven distinct types of partly or wholly clastic modern red sediments have come to the attention of the writer, some of which, however, are closely related. The occurrence of each one of these types and its application to the Red Beds of the western United States is discussed in the following paragraphs. Red clay of the deep-sea bottom.—This material is invariably very fine-grained, it contains little or no terrigenous matter of any kind, and it accumulates very slowly indeed, so that a thickness of it comparable to the total thickness of shales in any series of the western Red Beds is practically inconceivable. Nearly every part of all the series included in the Red Beds group exhibits incon- testable evidence of shallow-water deposition, while the oceanic red clays are exclusively abysmal deposits. Such arguments could be multiplied almost indefinitely; it is quite clear that deep-sea red clay is not related to our problem in any way whatsoever. Stream deposits derived from pre-existing Red Beds.—This type of deposits is illustrated by the flood-plain deposits of the Red River of the South, in Texas, Oklahoma, and Louisiana. This type cannot be dismissed so easily, for there are yet in existence masses of pre-Cambrian red sediments within the possible drainage areas tributary to some of the areas of the Red Beds. One objection to this source for the ferruginous matter of the Red Beds is that there were other sources for the sediments in question, nearer than 238 THE ORIGIN OF RED BEDS 239 the pre-Cambrian series, and that much of the material of the Red Beds is known to have been derived from other rocks. Arkosic stream deposits.—These are illustrated by deposits of limited extent in and downstream from the region occupied by the Sherman granite of Wyoming, which weathers to a coarse pink gravel, owing its color to a high content of undecomposed pink orthoclase. It is obvious that much the greater part of the color of the Red Beds is not due to pink feldspar; but in some of the very arkosic sediments of the Cutler and Dolores formations, and probably elsewhere, this is an element not to be ignored.” Stream deposits deriving their coloring matter from ferruginous residual soils.—The fourth type of modern red sediments is exempli- fied by the continental portion of the deposits of the lower Amazon, and by smaller deposits in some of the rivers of the United States. Russell says: Each grain [of sand in residual soils left by the decomposition of crystalline rocks in the southern Appalachian Piedmont] is coated with a thin shell having a brownish or red color. Prolonged washing fails to remove this superficial coating, a fact which is well illustrated by the color of the sands deposited by the streams of Virginia and the Carolinas in the regions underlaid by crystalline rocks.? Russell appears to assume that all of the ferric oxide produced by the decay of the crystalline rocks of this area is attached to grains of other minerals in this way. That which fills interstices between grains of sand in the final deposit,3 as distinct from that which occurs in coatings on the grains, probably persisted inde- pendently, however, and was transported as a fine sediment like clay. The relation between surface weathering of the Piedmont crystalline rocks and the color of the Newark clastics in the neigh- boring areas, as developed by Russell, is very much the same as a relation recently advocated by Beede* between weathering of lime- t Cf. Whitman Cross, Telluride Folio (No. 57), Geol. Atlas of the U.S., U.S. Geol. Survey, 1899, p. 2. 21. C. Russell, ‘“Subaerial Decay of Rocks,” U.S. Geol. Survey Bull. No. 52, 1889, 5 Wl 3 See p. 164, this volume. 4]. W. Beede, “Origin of the Sediments and Coloring Matter of the Eastern Oklahoma Red Beds,” abstract in Bull. Geol. Soc. America, XXIII (1912), 723-24. 240 C. W. TOMLINSON stones and the color of the Red Beds of eastern Oklahoma. He Says: The coloring matter is thought to have been derived from the solution of the 7,000 or 10,000 feet of pre-Carboniferous limestone which formerly covered the Arbuckle-Wichita Mountains and much of the surrounding region. The solution of the limestone furnished optimum conditions for the oxidation of its iron content, as it does at the present time in the limestone regions of the Mississippi Valley, southern Europe, West Indies, and elsewhere. Moreover, the solution of the pre-Carboniferous limestones and the conglomerates of the Arbuckle-Wichita region now in progress produces a red residuum practically indistinguishable from Red Beds sediments. ‘The red granites, red porphyries, and other crystalline rocks of the region under discussion contributed their shares of material to the Red Beds.* The Red Beds of the Grand Canyon section are underlain by the famous Redwall limestone, and limestones underlie the Red Beds practically throughout the Plateau province and in the San Juan region. Areas in Colorado of history similar to that of the Arbuckle-Wichita region, in that highlands existed there after the earlier Paleozoic limestones were deposited, and during Red Beds times, may well have played the same part in the central Rocky Mountain region that Beede assigns to the Arbuckle-Wichita uplift in Oklahoma. The existence of such highlands is demon- strated by the great conglomerates in the Colorado Red Beds. Various other land-masses which contributed material to the sedi- ments of the Red Beds’ may have been quite as efficient as the Arbuckle-Wichita highlands in producing residual soils stained by ferric oxide. It is evident from the foregoing discussion that stream deposits deriving their coloring matter from ferruginous residual soils are probably of no little importance in the Red Beds, and may con- stitute a major part of the series of sediments included under that term. Terrigenous marine clastics—The fifth type of modern red sediments is illustrated by deposits in the Atlantic Ocean off the mouth of the Amazon River.‘ This is an exceptional occurrence, * Beede, op. cit. * See pp. 244-245. 3 See pp. 245-246. 4 John Murray, Challenger Reboris: Deep Sea Deposits, 1891, p. 234. THE ORIGIN OF RED BEDS 241 as most terrigenous red muds lose their color on entering the sea. A vivid description of this process of loss of color in the case of certain rivers in Nova Scotia is given by Dawson, as follows: This harbour [Pictou] receives the waters of three rivers and several smaller streams, which in times of flood carry into it large quantities of reddish mud, which sometimes discolours the whole surface. This mud, with similar sedi- ment from the shore of the harbour, is deposited in the bottom, and there undergoes a remarkable change of colour. A portion of old mud recently taken from the bottom is of a dark grey colour, and emits a strong smell of sulphuretted hydrogen... . . The iron of the red clay has entered into combination with sulphur, and this is probably obtained from the sulphates contained in the sea-water, by the deoxidizing influence of decaying vegetable matter . . . . which grows abundantly on the mud flats. ... . In some parts of the deposit forming in Pictou harbour, the vegetable matter which caused the change of colour is so completely decomposed that no visible fragments of it remain.t The chemical action of marine organic matter is summarized by Clarke in part as follows: ‘‘ Decomposing organic matter reduces the sulphates of sea-water to sulphides, which by reaction with carbonic acid yield sulphuretted hydrogen. Bacteria also assist in the process.’” The interbedding of marine limestones with Red Beds shales in the Texan section is consistent with an origin for the shales similar to that of the semi-oceanic sediments from the Amazon River. The interbedding of gray and green strata with red ones in certain of the Red Beds series indicates an oscillation of domi- nance between oxidizing and deoxidizing conditions, such as might be caused at the margin of the sea or in the waters of an inclosed basin by variations in the rate of sedimentation or in the abundance of organic matter. The marine type of deposition of red sediments is not to be neglected, therefore, in an attempt at reconstruction of the conditions of origin of the Red Beds; though the complete absence of marine fossils from other parts of this group of sediments, together with independent proof of the continental origin of most of the group, shows that the marine type cannot be of more than subordinate importance. tJ. W. Dawson, “On the Colouring Matter of Red Sandstones and of Greyish and White Beds Associated with Them,” Quar. Jour. Geol. Soc. London, V (1848), 29. 2F, W. Clarke, “The Data of Geochemistry,” 2d ed., U. S. Geol. Survey Bull. No. 491, 1911, pp. 136-137. 242 C. W. TOMLINSON Deposits of desert lakes or playas.—For an example of the sixth type of present-day red sediments, deposited under water in desert lakes or playas, we turn to Chinese Turkestan. The following description, by Huntington, relates to the northern extremity of the bed of Lop Nor, near the southern base of the Kuruk-Tagh or Dry Mountains: ‘Beyond the fatiguing plain of salt [dry bed of the dwindled lake Lop Nor] we found easy traveling for a time. A fantastic red plain, the soft, dry bed of an older expansion of the lake, glittered with innumerable gypsum crystals.” Here we have a recent deposit of gypsum (or, more properly, selenite), in which the crystals presumably are imbedded in a red clay or mud. The relations here described could be duplicated by many minor deposits of gypsum in the Red Beds of the western United States. Farther out toward the center of the lake floor occur the purer non-clastic sediments, which in the case of Lop Nor are described as salt beds. The coloring matter of the clays probably was derived, as in the two preceding types of modern red sediments, from the decay of rocks on the neighboring uplands. Red dune sands.—The seventh and last type differs from all the others in being an eolian deposit. Red dune sands are excep- tional rather than the rule in the desert regions of today, but they occur in sufficient abundance to warrant attention. Their most striking occurrence is in the Nefood or Red Desert of North- western Arabia. The following quotation is from Palgrave’s narrative of a journey taken in 1862: ‘‘We were now traversing an immense ocean of loose reddish sand, unlimited to the eye, and heaped up in enormous ridges, . . . . undulation after undulation, each swell two or three hundred feet in average height.’” The extreme breadth of the Nefood is about 150 miles, its greatest length about 400 miles.$ Huntington mentions ‘an almost absolutely barren area of reddish or yellowish sand dunes, from ten to a hundred or more feet ‘Ellsworth Huntington, The Pulse of Asia (Boston and New York: Houghton Mifflin Co.), p. 254. 2W. G. Palgrave, Central and Eastern Arabia (London: Macmillan, 1908), pp. 62-63. 3 See J. A. Phillips, ‘‘The Red Sands of the Arabian Desert,” Quar. Jour. Geol. Soc. London, XX XVIII (1881), 110-13. THE ORIGIN OF RED BEDS 243 high’? between Karakir and Keriya River, in the southern border of the Takla Makan Desert, Chinese Turkestan. Some 40 or 50 miles farther north is the district described in the following passage: ss . ridge after ridge of sand, fifty to one hundred feet high. .... Their gently sloping backs to windward were gray with a cover of rather coarse sand, while their steep fronts to leeward were pale brick-red with the fine sand of the main desert.’” _ There are in the Red Beds of the western states no sandstones of this type of such great thickness as that of the Nefood sands, yet the possibility must be recognized that there may be local sandstone members of this origin in the series. A region dry enough to admit of the production of great bodies of gypsum might easily be transgressed by shifting sands; or the two types of deposi- tion might exist side by side, as they do today in the region of Lop Nor and Takla Makan. The coarsely cross-bedded sandstones of the Chugwater formation along the eastern base of the Wind River Range in Wyoming, for instance, will bear further inves- tigation with this possibility in mind. EVIDENCE OF FEATURES OTHER THAN COLOR AS TO THE CONDITIONS UNDER WHICH THE RED BEDS WERE DEPOSITED The wide range in grain shown by the Red Beds of various parts of the West, and the varying quantity of non-clastic sedi- ments in the group, show that a variety of conditions existed in this region during the time the Red Beds were accumulating, as is to be expected from the great extent of the group. What the varying relations were will be pointed out as accurately as possible in the following pages. Evidence of conglomerates as to the sites of land-masses.—Con- glomerates, by the pebbles which they contain, display more clearly than other sediments the source of their component materials. We can therefore determine with some confidence the sites of the land- masses which gave rise to the Red Beds, where these are con- glomeratic. In southeastern Oklahoma, Beede* has presented facts to show that the lower Red Beds sediments were derived from Ree Huntington, op. cit., pp. 183-84. 2 [bid., pp. 184-85. 3 Op. cit. 244 C. W. TOMLINSON the Arbuckle-Wichita region. Near the border of the present Arbuckle and Wichita mountains, limestone conglomerate and conglomerate of crystalline rocks dovetail into Red Beds sediments of finer grain. The limestone undoubtedly was derived from earlier Paleozoic formations, and the crystalline fragments from more ancient rocks, both of which are known to have been exposed in the uplifted region just named. Increasing thickness of sediments toward the mountains testifies to the same relation. To what distance sediments from this isolated highland may have been distributed is largely a matter of conjecture. The entire Red Beds series thins northward from this area to the northern limit of their outcrops in Nebraska, and in the same direction clastic sediments give place in part to limestones: both of which facts signify increasing distance from the source of terrigenous material. The small outlier of Red Beds in central Iowa, which probably is to be correlated with the Cimarron series of Kansas," is composed chiefly of red shale and gypsum, likewise indicating relatively clear-water conditions. It would be an unwarranted assumption, however, to assume that the Kansas and Iowa Red Beds were derived wholly from the Arbuckle highlands. The greater areas of upland which probably existed in the region surrounding and including the pre-Cambrian areas of Minnesota and Wisconsin may have been important sources of material, as may also the ancient Appalachian continent of the East; but it may safely be said that their influence is not exhibited in the strata now available for study, as the influence of the Arbuckle highland so clearly is. Similar criteria may be applied to the great conglomerates of the Fountain and Wyoming formations of the Front Range, and to those of central and southwestern Colorado. The conglomerates of the Front Range Red Beds and of the Maroon formation of the Anthracite, Crested Butte, and Tenmile districts are made up chiefly of fragments from pre-Cambrian crystallines,? which still «Cf. F. A. Wilder, “The Age and Origin of the Gypsum of Central Iowa,” Jour. Geol., XI (1903), 723-48. 2 See Whitman Cross, Pikes Peak Folio (No. 7), Geol. Atlas of the U.S., U.S. Geol. Survey, 1894; G. K. Gilbert, Pueblo Folio (No. 36), 1897; and G. H. Eldridge, “Description of the Sedimentary Formations,’’ Anthracite-Crested Butte Folio (No. 9), 1894. THE ORIGIN OF RED BEDS 245 outcrop in wide areas in various parts of the ranges. The Maroon conglomerates include also fragments of quartzite and limestone from older sediments. The composition of these formations renders it certain that their materials were not carried far from their sources; it is therefore certain that there were highlands, and there may have been mountain ranges of no insignificant relief in various parts of Colorado during Red Beds time—which here probably was included for the most part within the Pennsylvanian and Permian periods. The coarser beds of the Cutler and Dolores formations in south- western Colorado show by their composition that they also were derived very largely from igneous and metamorphic terranes.* The studies of Cross? have shown that the northern part of the San Juan region itself, as well as the neighboring Uncompahgre Plateau, was exposed to erosion between early Cutler and Dolores time—at or near the beginning of the Mesozoic era. This upland may have furnished sediment to a considerable part of the plateau province to the west and southwest. Cross is of the opinion that the absence of the Red Beds on the Uncompahgre Plateau is due to post-Dolores erosion; but the conglomeratic character of the Red Beds in the San Juan Mountains demands that a source for those sediments be found close at hand. In the absence of any conclusive evidence that the Red Beds ever were deposited over the plateau in question, it may be regarded at least provisionally as the probable site of that source. The sediments of the Plateau province, on the whole, do not indicate mountainous topography in the vicinity; but their great thickness (maximum more than 5,000 feet, excluding non-clastic beds) calls for the existence of a land area contributing sediments to this region for a long period of time. Such an area may well have existed toward the south and southwest, in Mexico, southwestern «See Whitman Cross and others, in the following folios of the Geol. Altlas of the U.S., U.S. Geol. Survey: Telluride (No. 57), 1899; LaPlata (No. 60), 1899; Silverton (No. 120), 1905; Needle Mountains (No. 131), 1905; Ouray (No. 153), 1907; Engineer Mountain (No. 171), 1910. 2 Whitman Cross, “Stratigraphic Results of a Reconnaissance in Western Colorado and Eastern Utah,” Jour. Geol., XV (1907), 648-49, 654-56. 3 [bid., pp. 648-49. 246 C. W. TOMLINSON Arizona, and southeastern California, or farther north in the Great Basin, where no sediments contemporaneous with those in question are known to occur.t The absence of sediments between the Mississippian and the Cretaceous in the El Paso quadrangle? in western Texas may be due altogether to erosion following deforma- tion at the close of the Jurassic period, but this gap in the record makes it possible that land may have existed even here during Red Beds times. | Richardson has concluded from his studies of the Black Hills Red Beds? that those sediments were derived chiefly from the Rocky Mountain area to the southwest and west. West of central Wyoming, the Red Beds group thickens, and the quantity of lime- stone and gypsum in it diminishes, continuously westward across the Idaho border, suggesting a source of sediments in that direction. The great thickness of the group all along the Wasatch Range, wherever it is exposed, extends this suggestion to include a con- siderable land area trending north and south from southern Idaho into central Utah. Significance of non-clastic sediments~—The more important limestone members of the Red Beds record the existence of exten- sive bodies of clear and not excessively salty water during parts of the Pennsylvanian period in central Texas, in the Plateau Province, in the San Juan region, and in southeastern Wyoming; during the Permian, in the region north and east of Great Salt Lake (in the early part of the Permian, marine deposition throughout much of Wyoming, prior to the initiation of Red Beds sedimentation there), and in western Texas; and in the Triassic, in northeastern Arizona. «Cf. paleogeographic maps by the following authors: T. C. Chamberlin and R. D. Salisbury, Geology (New York: Henry Holt & Co., 1909), II, 545; ILI, 3 and 62; W. B. Scott, An Introduction to Geology (New York: Macmillan, 1909), pp. 616, 662; Charles Schuchert, ‘“‘ Paleogeography of North America,” Bull. Geol. Soc. America, XX (1909), Pls. 84-88 inclusive. 2G. B. Richardson, El Paso Folio (No. 166), Geol. Atlas of the U.S. U.S. Geol. Survey, 1909. 3G. B. Richardson, ‘‘The Upper Red Beds of the Black Hills,” Jour. Geol., XI (1903), 365-93. THE ORIGIN OF RED BEDS 247 The extraordinary development of gypsum in the Permian Red Beds deserves more than passing comment. In association with salt, deposits of gypsum are interpreted as indicating aridity of climate at the time of deposition, and of formation in at least partially inclosed basins by the evaporation of bodies of water not freely connected with the open sea.t. This occurrence of gypsum, supported by independent evidences of continental origin for the Red Beds, has been the one strongest influence in establish- ing the idea that the red color itself is an indication of aridity. The absence of gypsum from many series of Red Beds, and its occurrence in series free from Red Beds, make it necessary to investi- gate the two problems on their own independent merits. Rock salt is of relatively rare occurrence in the group of sedi- ments under discussion. Since the saturation point of gypsum in aqueous solution is much lower than that for common salt, it is logical to suppose that the deposition of gypsum unaccompanied by rock salt signifies.a condition of aridity and of continuous or ' intermittent supply of normal sea-water or of fresh water such as to maintain a degree of salinity more moderate, for example, than that of Great Salt Lake at present, but sufficient to cause the con- _ tinued precipitation of gypsum. Such a condition might be kept up by a limited or intermittent connection between the open sea and the basin of deposition. The relation of the gypsum and salt deposits of the West to the Red Beds proper suggests a relation similar to the relation between marine limestones and terrigenous sediments. The Rustler dolomite and the Castile gypsum of the Texan portion of the Pecos Valley give place northward to typical Red Beds with a few interbedded strata of dolomite and gypsum. May not the gypsum, as well as the dolomite, be but the complements of the red clastic sediments, deposited in the clear central waters of an inland sea, or in lagoons near or at sea-level but partly or wholly cut off from the sea; while clastic sedimentation went on nearer the shores and on river deltas or flood-plains yet nearer to the sources of sediment ? t Cf, Wilder, op. cit. 248 C. W. TOMLINSON The clastics: minor structural features —Returning to the clastic sediments, we may draw still further inferences regarding the con- ditions under which they were deposited, from their structural characteristics and mineral composition. Ripple-marks and mud- cracks in the majority of Red Beds sections testify to the prevailing shallowness of the water in which these sediments were laid down. Mud-cracks repeated in layer after layer, as in some parts of the Red Beds, mean complete emergence and at least partial drying, after the deposition of each stratum and before that of the next following. Shallow water means shifting currents, and these too are recorded clearly by cross-bedding in most sandstones of the series, and by rapid variation along the strike in the shaly members as well. We do not have, in the clastic portions of the Red Beds, and seldom do we find in the non-clastic members thereof, the continuity of a single type of sedimentation over wide areas and through long periods of time, which are to be expected in truly subaqueous or marine deposits. Furthermore, imperfect assort- ment, which is one of the universal characteristics of fluviatile deposits, is the rule in the Red Beds. The sandstones are earthy, the shales sandy, the conglomerates gritty, etc. Each of these characteristics is suggestive of subaerial conditions, and the occur- rence of all of them together in the same series, and widely dis- tributed through that series, is conclusive testimony to such an origin. The clastics: mineral composition.—From the mineral composi- tion of the clastic sediments we may infer something of the con- ditions of weathering and transportation which preceded their deposition. The high proportion of feldspar in many of the Red Beds shales and sandstones indicates a preponderance of mechanical disintegration over chemical decomposition. The abundance of undecomposed mica flakes in most Red Beds confirms this inter- pretation. ‘Transportation may precede complete decomposition because of exceptional rapidity of disintegration, exceptional slow- ness of decomposition, or both. Rapid disintegration may be caused by such factors as high relief and great daily or seasonal range of temperature; slow decomposition by low rainfall or low temperatures. Low temperatures throughout the year explain THE ORIGIN OF RED BEDS 249 the slowness of chemical decomposition in polar regions; aridity explains it in the desert, where disintegration is accelerated by great daily range of temperature; and disintegration is acceler- ated on mountain peaks by all of the factors mentioned. Absence of vegetation, which itself is dependent chiefly on climatic factors, is unfavorable to rapid decomposition of rocks because of the important part played by organic acids in the chemical processes of weathering. As we have seen, the occurrence of gypsum indicates aridity and high temperatures, so that we may rule out the hypothesis of Arctic conditions as applicable to the Red Beds. The coarse conglomerates in certain parts of the Red Beds are indications of high relief in certain areas and at certain times. The occurrence of limestone conglomerate in the Red Beds of the Arbuckle-Wichita region emphasizes the predominance of disintegration over decomposition in that area, as limestone is one of the most readily decomposed of rocks. If the limestone con- glomerates of the Cutler formation were detrital, it would have the same significance concerning the processes of the San Juan region; but it has been interpreted otherwise. Evidence supplied by fossils——It has been stated that marine limestones. carrying abundant faunal remains occur in central Texas interbedded with the Permo-Carboniferous Red Beds. The significance of the relations found in this region is well sum- marized by Chamberlin and Salisbury, as follows: The oldest part of the Permian system (Wichita formation) indicates that the critical attitude which characterized the surface farther east during the Pennsylvanian period now affected Texas, for the beds are partly of marine and partly of fresh-water origin. These beds are succeeded by a formation of limestone (the Clear Fork) of marine origin, which overlaps the Lower Permian. The Upper Permian (Double Mountain formation) which follows indicates a reversal of relations, for much of Texas was again cut off from the ocean, and converted into an inland sea, or into inland seas, in which the phases of deposition common to such bodies of water took place. Occasional beds of limestone with marine fossils point to occasional incursions of the sea, while deposits of salt and gypsum point with equal clearness to its absence, or to restricted conditions, and to aridity of climate.! tT, C. Chamberlin and R. D. Salisbury, College Geology (New York: Henry Holt & Co., 1909), p. 661. 250 C. W. TOMLINSON The large and varied vertebrate fauna which has now been described from the clastic members of the Red Beds series in various parts of the Southwest' includes no forms requiring other than a land or fresh-water habitat, with the exception of fish remains in some of the marine beds just mentioned. In general, then, it is true that the paleontological evidence corroborates the purely stratigraphic and lithologic evidence for a continental origin for at least the greater part of the Red Beds. In certain places and at certain horizons the fossil remains, both plant and animal, are sufficiently abundant and of such types as to eliminate the possi- bility of extreme aridity as a continuously prevalent condition. The bone beds and petrified forests of northeastern Arizona, for instance, prove that during the time of deposition of the Shinarump group,” at least, there was a water supply in that vicinity sufficient to permit the support of abundant land life, both vegetable and animal. It is possible, however, that this water supply was derived from precipitation in a distant region, like the waters of the Nile delta. In the underlying Permian, which contains more salt and gypsum than the other members of the series in this district, vertebrate remains are absent. The Permian of the neighboring Kanab Plateau has yielded an extensive invertebrate fauna} sug- gesting brackish-water environment.‘ Summary.—The most salient of the facts and inferences brought out by the foregoing discussion of the significance of features other than color as to the conditions of deposition of the Red Beds may be summarized as follows: (1) rapid erosion on land-masses of considerable relief; (2) decomposition not complete in advance of transportation; (3) sediments diminishing in thickness and in coarseness of grain away from sources of material, and clastic « See especially various publications by S. W. Williston in the Journal of Geology, 1903-13. 2L. F. Ward, “Geology of the Little Colorado Valley,” Am. Jour. Sci., 4th Ser., XII (1901), 401-13. On p. 405 is the following statement: “‘The Shinarump constitutes the horizon of silicified trunks and there is no part of it in which fossil wood does not occur in great abundance.” 3See C. D. Walcott, ‘The Permian aid Other Paleozoic Groups of the Kanab Valley, Arizona,” Am. Jour. Sci., 3d Ser., XX (1880), 221. 4 Interpretation by Eliot Blackwelder (personal communication). THE ORIGIN OF RED BEDS 251 sediments giving place to non-clastic in the same direction; (4) flu- viatile deposition most important; (5) all deposits in relatively shallow water, or subaerial; (6) oscillating marine and non-marine conditions at edge, non-marine in most of region of deposition; (7) moderate aridity long continued in some parts of the region of deposition, alternating with less arid conditions in other parts. These conditions coincide to a remarkable degree with those inferred from the study of modern red sediments. RELATION OF DIASTROPHISM TO RED BEDS SEDIMENTATION It may well be asked at this juncture why it is that earlier Carboniferous clastics underlying the Red Beds of Oklahoma and other states are not similarly colored. The difference in color, since this has been shown to be a feature dating from the time of sedimentation,’ must be due to differences of some sort in the geo- graphic conditions of the times when the successive series were deposited. One of these differences is the emergence of the plains of deposition. Another may be found in the fact that some, at least, of the highlands from which the Red Beds derived their materials were not in existence in the earlier part of the Paleozoic era. The Arbuckle-Wichita uplift probably dates from the later part of the Pennsylvanian period; and there may have been mountain-building in Colorado at the same time, as the strati- graphic relationships of the scattered Paleozoic sediments in that state seem to indicate. Dr. Blackwelder’? calls attention in this connection to the fact that the general trend of the Arbuckle- Wichita folding is directly in line with the suspected areas in Colorado. Local climatic changes influencing the type of sedi- mentation may have been brought about by these changes in topog- raphy. The general continental expansion of North America from Pennsylvanian to Jurassic times leads one to expect extreme types of continental climate, including aridity, the localization of which would depend largely on the configuration of the continent. The deposition of red sediments derived from ferruginous soils means either the development of red soils and the transportation of the material thereof without hydration, or the development of t See discussion of this point, pp. 162-67, this volume. 2 Personal communication. 252 C. W. TOMLINSON limonitic ferruginous soils and the dehydration thereof during transportation. The development of ferruginous soils is the chief prerequisite to the deposition of Red Beds of the western type. The areas from which the Red Beds derived their materials certainly included uplands, and in part at least they are known to have been possessed of fairly rugged relief; they were therefore in all probability the sites of more abundant rain than fell upon the plains or delta flats upon which the Red Beds were in large part deposited. The combination of well-watered highlands with less humid or semi-arid lowlands furnishes the conditions for the devel- opment of red soils, and at the same time provides for the trans- portation and deposition of the sediments derived from them. without extensive hydration or reduction of the ferric oxide con- stituent during the transfer. An unusually extensive development of red soils during the time of deposition of the Red Beds might have been due, in some part at least, to the higher proportion of oxygen inferred by Chamberlin and Salisbury’ to have existed in the atmosphere at this time. SUMMARY The several steps which have been followed in the interpreta- tion of the color of the Red Beds, and the results obtained, may be summarized as follows: 1. The ferruginous matter which gives the Red Beds their color has been present in the series in very nearly its present dis- tribution and arrangement since the time of sedimentation. 2. This material has suffered no extensive change of ferrous to ferric iron, or vice versa, since the time of sedimentation; the proportion and present distribution of these compounds in the series were influenced most largely by the original distribution of organic matter. 3. Changes in the degree of hydration of the ferric oxide in the Red Beds since sedimentation probably have not been of great importance; and hydration probably has been quite as active as the reverse process during this time. tT. C. Chamberlin and R. D. Salisbury, Geology (New York: Henry Holt & Co., 1909), II; 665. THE ORIGIN OF RED BEDS 253 4. The ferruginous matter of the Red Beds was transported and deposited almost, if not quite, wholly as a mechanical sediment, both independently and as a coating upon grains of other material. 5. The types of sediments probably most important in the Red Beds group are stream deposits, submarine fluviatile deposits, and playa deposits, all predominantly of red color, and all deriving at least the greater part of their ferric oxide from ferruginous residual soils. Of these types the first is by all odds the most important. 6. The study of characteristics of the Red Beds other than color bears out the conclusion stated in No. 5. 7. The inauguration and cessation of Red Beds sedimentation probably were connected closely with climatic and topographic changes involved in the orogenic history of the continent. The colors which distinguish Red Beds from other series are due to a combination of lithologic, topographic, and climatic fac- tors in the regions of denudation and in those of deposition, which have not been reproduced over so great an area in more recent times. It is apparent that, in accordance with Barrell’s view,’ ‘‘red color in sediments is not in itself an indication of aridity’’; for the material of red ferruginous soils may be transported and deposited in regions of high rainfall, or even under the sea, with- out change of color; and red soils themselves develop in regions of heavy rainfall. But since the dehydration of the limonitic material of non-red ferruginous soils, as well as the continuance of the relatively anhydrous condition of the hematitic material of red soils, is favored by aridity in the regions of transportation and deposition, therefore red sediments should form a larger part of the sediments of arid than of humid regions. 1 Joseph Barrell, ‘Upper Devonian Delta of the Appalachian Geosyncline,” Am. Jour. Sct., 4th Ser., XXXVI (1913), 437. THE ACADIAN TRIASSIC SIDNEY POWERS Troy, New York PART III STRUCTURE OF THE ACADIAN TRIASSIC The Newark rocks in the Acadian area exhibit a monoclinal structure, with a prevailing northwesterly dip, interrupted by broad, low folds. The monocline is broken by numerous faults with a small displacement and by occasional faults with a displace- ment of hundreds of feet. The other areas of Newark rocks have undergone deformation of a similar nature, but the direction of the monoclinal tilting differs in the various areas. In the case of the Connecticut Valley, the Pomperaug Valley (Connecticut), the Deep River (North Carolina), and the Wadesborough (North Carolina) areas, the dip is southeast, where as in all the other areas it is northwest. The two structural features, the folds and faults, will be treated separately and finally some attention will be given to the theories of origin of this structure. FOLDS The most important and the most conspicuous fold in the Acadian area is that shown by the hook in North Mountain wHich incloses Scots Bay. The point of the hook forms Cape Split, and the back of the hook, Cape Blomidon. This syncline pitches down on the north side and is cut off on the north by a fault shown in cross-section DD, Fig. 28. The syncline is shown principally in the North Mountain basalt which dips toward Scots Bay on all sides of the Bay at angles of about 5°. Under the basalt flows the Blomidon shale is seen following the erosional escarpment, on the south side of North Mountain, around to Cape Blomidon, near which point it disappears under the waters of Minas Basin, as 254 THE ACADIAN TRIASSIC 255 shown in Fig. 27. Above the North Mountain basalt comes the Scots Bay formation, the youngest formation in the Newark group of the Acadian area. The Scots Bay formation is exposed on the south side of the Bay, as shown in Fig. 27. The eastern extremity of Minas Basin, east of Economy Point (the area shown in Fig. 23), also forms a syncline which has been disturbed by faulting at various points. The beds of red sandstone on either side of Cobequid Bay dip toward the bay at angles of about 3°—-5° except where they have been tilted by faulting. This gentle dip must simulate that of the strata when they were first deposited in the slowly subsiding geosyncline. A syncline, which is well shown in a shore section, is found at Quaco, between West Quaco and Melvin’s Beach, on the north side of the Bay of Fundy (see cross-section BB, Fig. 7). The sediments of the Quaco section are readily identified by the Quaco conglomerate in the center. This conglomerate is exposed on the shore near Vaughan Creek with a dip of 30° to the north and again a mile inland (northwest) with a corresponding dip to the south. The syncline is cut off obliquely on the north by a fault in such a way that the axis of the syncline is shown in the shore section near Melvin’s Beach, but the Quaco conglomerate of the northern limb does not reappear. At Split Rock a low anticline is shown, at Martin Head a syn- cline, and at Waterside an anticline and the adjoining syncline. In each of these cases the folds are cut off by faults. The folds are at a low angle with broad arches or troughs. Cape d’Or shows a small syncline in the basalt flows where the basalt ridge turns, from its east-west course, to make the Cape on the south. Horseshoe Cove has been formed at the axis of the syncline. The basalt is also faulted as is shown in Fig. 12. Everywhere in the sea-cliff exposures there are minor flexures in the Acadian Triassic, both in the sediments and in the igneous rocks. In Fig. 19, an example of the folds in the sediments west of Five Islands is given. In the North Mountain basalt, gentle folds are shown at Scots Bay, where the Scots Bay formation is preserved in synclines (Fig. 29), and at Digby Gut, where a long syncline is shown at Victoria Beach. 256 SIDNEY POWERS FAULTS The disturbance at the close of the Newark sedimentation threw the rocks of this group into fault-blocks with a monoclinal tilting toward the northwest. With such a structure, the major faults would tend to assume a northeast-southwest trend, and some of the more important faults should bound the formation on the north and northwest. The faults at the margin of the Triassic area are confined to the northern and western sides. ‘Thus the basalts of Grand Manan are faulted down on the west side, while the pre-Triassic rocks on the east side of the island are probably tilted up. The older forma- tions against which the basalts were downthrown have since been eroded away, because they were less resistant than the basalts, and Grand Manan Channel has been formed in them. The northern and northwestern sides of the Triassic areas at Split Rock, Quaco, Martin Head, and Waterside are all dropped down as fault-blocks against older rocks. At Martin Head the pre-Cambrian rocks form Martin Head itself, which is south of the exposure of the Triassic sediments. This exposure of older strata may be explained either as a horst or as the basement upon which the southern limb of the Triassic syncline rests. The latter view is favored, making the Triassic and the exposure of pre- Cambrian part of one fault-block, with a fault south of the pre- Cambrian. There also appears to be a minor fault in the axis of the Martin Head syncline. The fault of greatest displacement in the Fundy region is the Cobequid fault (shown on the general map of the region), which stretches from West Advocate, north of Cape d’Or, to a point northeast of Truro, a distance of go miles. On the north side of the fault is the Cobequid group of sedimentary and igneous rocks which composes the Cobequid Mountains. On the south side of the fault are Triassic sandstones at West Advocate and Advocate Harbour, and Pennsylvanian rocks east of Advocate Harbour. The displace- ment of this fault is probably 2,000-3,000 feet. South of the Cobequid fault is another east-west fault which bounds the Triassic on the north from Cape Sharp to the Chiganois River (northeast of Truro). The displacement of this fault appears THE ACADIAN TRIASSIC 2 to be greatest on the west, with a downthrow of 1,500 feet or less. _ Parallel to this fault is another at Clarke Head which has brought the Triassic down on the north against older rocks on the south, forming a small graben shown in Fig. 17. All the rocks at Clarke Head are intensely faulted. The remnants of North Mountain basalt at Cape Sharp and at Partridge Island appear to be faulted off on the south side. The throw of this fault is uncertain in direction, but it may be a continuation of the southernmost fault at Clarke Head. The exposure of North Mountain basalt at Cape d’Or exhibits several faults in a north-south direction, as shown in Fig. 12. The end of Cape d’Or is probably on an east-west fault line. This same fault may extend eastward. The Five Islands region exhibits complex block-tilting with blocks of relatively small size. Besides the fault bounding the Triassic on the north, and the Cobequid fault farther north, an east-west fault is shown at Gerrish Mountain (Figs. 20, 22). The Five Islands are each separated by faults and are each tilted in different directions. These faults on the north become lost in a_ greatly slickensided region shown in detail in Fig. 21. The slicken- sided surfaces are usually vertical and have a north-south direction. The major movement appears to have been in a horizontal plane, but the stratification shows that there also has been vertical move- ment. Many other north-south faults are shown along the shore from Clarke Head to Five Islands, and a typical section is shown in Fig. ro. Near Lower Economy a strike (east-west) fault brings the Triassic down into contact with a mass of Pennsylvanian strata on the north on which the Triassic rests unconformably. The hook of North Mountain, at Cape Split, is cut off by a northeast-southwest fault which gradually cuts across this limb of the Scots Bay syncline. North Mountain is composed of basalt flows tilted to the north- west so that an erosion escarpment is produced on the south side ef the mountain and a gentle dip-slope on the north side. The sea-cliffs on the north side are never very high for this reason. With a continuation of the dip-slope, the erosion top of the flows 258 SIDNEY POWERS appears to extend under the Bay of Fundy. The coast charts do not show any pronounced submarine ridges parallel to North Mountain, such as some authors have referred to, and therefore there is a lack of evidence of any major fault parallel to North Mountain. Moreover, no geological structure under the Bay oi Fundy appears to be deducible from the submarine topography. Cross-faults in North Mountain are readily shown by offsets in the ridge of basalt flows because the flows are dipping at a low angle northwest. The offsets are at Digby Gut, Bay View, Gulli- ver’s Cove, Petit Passage, Grand Passage, and southwest of Brier Island. The line of these faults is north-south. The displacement of the flows by these faults, with the exception of the first and last faults, is to the north on the west side of the fault. These offsets are shown on the accompanying general map of the region. The offset at Digby Gut is shown on Fig. 30, and that southwest of Brier Island is shown by the position of a short submarine ridge on the coast chart. As shown by Daly" and by Haycock,? these fault lines across North Mountain, and also the depressions at Parker Cove and Sandy Cove were occupied by rivers at the time that the Summit peneplain was being developed over the region. When the pene- plain was uplifted the rivers became rejuvenated and persisted in their courses until the present valleys were cut. Headward erosion up the valley which is now St. Mary’s Bay diverted the streams flowing across the basalt south of Bay View, and the more rapid erosion in Digby Gut caused the abandonment of the Bay View and Parker Cove valleys. THEORIES OF ORIGIN The faults which traverse the rocks of the Newark group are of deep-seated origin, extending into the older formations. The character of the underlying formations varies with the different areas. Thus the Acadian Triassic is underlain in part by Carbonif- erous folded sediments, in part by Silurian and Devonian slates and tR. A. Daly, “The Physiography of Acadia,” Bull. Mus. Comp. Zool., Harvard College, XX XVIII (1901), 92. 2 E. Haycock, “Records of Post-Triassic Changes in Kings County, Nova Scotia,:’ Trans. N.S. Inst. Sci., X (1900), 297. THE ACADIAN TRIASSIC 259 Devonian granite, and in part by pre-Cambrian slates (the Meguma series) and other metamorphic rocks (the pre-Cambrian complex of New Brunswick). The Connecticut Valley area is underlain by gneisses and schists, the New Jersey area by gneisses and some Paleozoic sediments, and the Richmond area by gneisses and granites. A theory which accounts for the structure of the Newark beds must therefore suit the various basement rocks. Davis,’ in studying the Connecticut area, reached the con- clusion that the origin of the monoclinal fault structure was the slipping of blocks of the underlying crystalline rocks on each other along cleavage planes. As pointed out above, although the Con- necticut area is underlain by gneisses and schists, the other Newark areas are not. Suitable cleavage planes would therefore not be expected in the other areas. In the Minas Basin region, the crystalline rocks are several thousand feet below the base of the Triassic. Furthermore, the planes of slipping in these crystallines are parallel to the main struc- tural lines of the formation. These lines run at an angle to the axis of Minas Basin, as is seen In the nearest exposures of the crystallines (the Meguma, or Gold-bearing series). The theory proposed by Professor Davis does not seem, therefore, to apply to the Acadian area. Hobbs? considers that Professor Davis’ theory does not suit the facts in the Connecticut Valley or in the Pomperaug area. For the latter area, Hobbs proposes another theory to account for the peculiar system of quadrangular block-faults. As this detailed faulting is not typical of all the Newark areas, the theory is of limited application. Professor Barrell’ has recently ascribed the origin of the Con- necticut Valley Triassic area to the gradual development of a fault on the east side of the geosyncline, contemporaneously with the tW. M. Davis, “The Structure of the Triassic Formation of the Connecticut Valley,” U.S. Geol. Surv., 7th Ann. Rept., 1888, pp. 486-89. 2 W. H. Hobbs, ‘‘The Newark System in the Pomperaug Valley, Connecticut,” U.S. Geol. Surv., 21st Ann. Rept., Part 3 (1901), pp. 122-33. 3J. Barrell, “‘Central Connecticut in the Geologic Past,” Proc. Wyo. (Penn.) Hist. and Geol. Soc., XII (1912). 260 SIDNEY POWERS filling of the basin with sediments. This fault is supposed to have been initiated after sedimentation commenced, and to have increased in displacement with the accumulation of the sediments. In the Acadian area a corresponding fault is found on the north and west, but there is no evidence that this fault developed until sedimentation ceased. No completely satisfactory theory to account for the structure has yet been presented. IGNEOUS ROCKS DISTRIBUTION A description of the igneous rocks in each locality has been given in the description of the general stratigraphy of the region, and therefore merely a summary is attempted here. The flows at Cape d’Or have been especially studied, and will be considered in a separate paper by Professor Alfred C. Lane and the writer. All of the igneous rocks associated with the Acadian Triassic are of a basaltic composition. From the form of occurrence, they are grouped into dikes and flows. According to the time of forma- tion, they are classified as the Five Islands volcanics and the North Mountain basalts. Dikes are so rarely exposed that it is necessary to consider the rocks from the point of age, rather than form. In Nova Scotia, outside of the Triassic area there are some dia- bases and basalts which are probably of Triassic age. At Cheverie, near the Avon River, there is a sill of diabase intruding Pennsyl- vanian. strata." Again, in Guysborough County, near Guys- borough, Fletcher has mapped on the sheets of the Geological Survey of Canada masses of diabase cutting the Union-Riversdale series. The nature of these masses is described by Fletcher? as partly amygdaloidal, partly dioritic. Dikes of Triassic age occur in a number of places between Nova Scotia and the Connecticut Valley. The large majority of them are of diabase composition. « Verbal communication from Mr. W. A. Bell, of the Geological Survey of Canada. 2H. Fletcher, Geol. Surv. of Canada, Annual Report, 1886, pp. to1-3 P; also Geol. Surv. Canada, Maps, Nova Scotia, Nos. 30, 31, 35, 36. THE ACADIAN TRIASSIC 261 FIVE ISLANDS VOLCANICS Under the heading Five Islands volcanics are included the tuffs, agglomerates, and basalt flows in the vicinity of Swan Creek and the Five Islands.. The thickness of the volcanics is estimated as at least 350-400 feet. One associated dike is exposed at Gerrish Mountain. The Gerrish Mountain diabase dike is almost vertical and about 20 feet or more in thickness. The diabase shows marked columnar jointing, the columns being rather short and largely horizontal or dipping at a low angle to the horizontal. The dike is connected with the basalt flow which caps the sandstones of Gerrish Mountain, and it has evidently furnished the material for this flow and perhaps for a large part of the other igneous rocks for the vicinity. The basalt flows associated with the Five Islands volcanics are found at Gerrish Mountain, on four of the Five Islands, on Two Islands, and at Portapique Mountain (east of Gerrish Moun- tain). It is noteworthy that the relation of these flows to the agglomerates is unknown, and that there is no proof that they are not connected with the North Mountain basalt instead of with the Five Islands volcanics. The structure of these flows is in large part columnar, and the base and the top of each individual flow is marked by amygdaloid. The basalts are the usual fine-grained, dark-gray, heavy rocks composed of augite and plagioclase with accessory amounts of magnetite and occasionally olivine. A more detailed petrographical description will be given below for the North Mountain basalt, which will apply equally well to these flows. Only one flow is exposed in Gerrish Mountain. This has a thickness of over 75 feet. Three flows are exposed on the north side of Moose Island, the upper one being agglomeratic. A portion of a single flow is exposed on Diamond Island and on Long Island. Two flows are seen on Pinnacle Island. The northern of the Two Islands consists of three flows, the southern of probably only one. The base of the series of flows is exposed on the eastern side of Moose Island and on Gerrish Mountain. Under the amygdaloid which marks the base of the flow is a layer of green ash 2-3 feet in 262 SIDNEY POWERS thickness. A similar ash-bed is exposed west of Swan Creek under the agglomerate flow mentioned below. The thickness of the flows on Gerrish Mountain may be considerable, as the basalt covers a large area. The agglomerate beds, with associated tuffs, are exposed from Greenhill eastward to Five Islands, in disconnected areas. The relation of these remnants of flows and volcanic ejectamenta to the sandstones is a problem only partly solved because of the faulted contacts, with possibly minor thrust-faults, and the landslides which are especially abundant in the tuff. The tuff underlies the agglom- erate in most cases. The thickness of the tuff varies from a few feet to 50 feet, and that of the agglomerate flows from 20 to 150 feet or more. Exposures show that the agglomerate is overlain by red sandstone, and is therefore older than the North Mountain basalt. The agglomerates consist of a mass of angular fragments of basalt and amygdaloid in a dark-green matrix of a basaltic com- position. The exact character of the matrix is difficult to determine because it is everywhere so badly weathered that a solid specimen could not be procured. The field evidence, however, indicates that this matrix is in part tuffaceous and in part a normal basalt. At the sides of some of the masses of agglomerate are blocks of angular basalt and amygdaloid imbedded in a red sandstone matrix, showing that the breccia was either blown out into the area where sandstone was being deposited, or washed out from a bed of tuff and breccia. The cross-cutting contacts at one side of the masses of agglomerate in two instances give them the appearance of intrusive bodies rather than of flows. If the agglomerates are intrusive, rather than extrusive, they probably fill volcanic necks. NORTH MOUNTAIN BASALT Under the term North Mountain basalt, used in a generic sense, are included the basalt flows of Grand Manan, Isle Haute, Cape d’Or, Cape Sharp, Partridge Island, and North Mountain. The series of flows at these localities are correlated either for structural reasons or because they are underlain by shale correlated with the Blomidon shale. THE ACADIAN. TRIASSIC 263 In each locality there are several flows, indicating successive extrusions within such a short time of each other that no sediments were deposited between the flows. It is impossible to state whether any single flow originally covered the geographical area over which the remaining exposures indicate that the formation once extended. The Palisade diabase formes one sill too miles long on the out- croping edge, while North Mountain is 120 miles long. In the former case the igneous material was intruded at some distance below the surface and had to push up this great weight of rock, which, however, acted as a blanket over the feeder. In the latter case the igneous material was extruded at the surface, with no roof to sustain, but the feeders were constantly subjected to the great heat loss by radiation at the surface, which would tend to freeze them up. Dikes associated with the North Mountain basalt are rare. Several were reported on Grand Manan by Bailey,’ but they were not observed by the writer. The largest of these is 50 feet wide, and occurs at Flag Cove, near Swallow-Tail Light. Other narrow dikes occur on the south side of Scots Bay, just east of the Scots Bay formation exposures. These dikes cut the basalt within 25 feet of the top of the upper flow. From the other exposures of this flow it is judged to be at least roo feet thick, and, if so, it is quite evident that the dikes cut the upper flow and are not the feeders. With the dikes are many fissures filled with vein material which is seen under the microscope to consist largely of silica stained red with hematite. The width of both the veins and the dikes varies from one to ten inches, and in the field they look very much alike. In thin-section the dikes are seen to consist of a very fine-grained diabase, greatly altered and stained with limonite. The rock is similar to that of the flows near the center, but shows some glass. From the field evidence of the dikes and veins side by side in the upper part of this thick flow, and from the microscopic evidence, it is concluded that the dikes were formed from the basalt of the upper flow after the crust of the flow had solidified and while the L. W. Bailey, Geol. Surv. Canada, Report of Progress, 1870, pp. 216-21. 264 SIDNEY POWERS center of the flow was still liquid. The crust appears to have become fissured, with some of the fissures reaching down to the still molten rock, and other of the fissures having no great depth and therefore being filled with quartz from above at a later stage. The structure of the flows is similar to that of all basalt flows. The individual sheets are clearly distinguished by a relatively thin amygdaloidal base and a relatively thick amygdaloidal top. Flows composed entirely of amygdaloid were observed only at Cape d’Or. The basalt is closely jointed and columnar joining is frequently developed. The angle at which the columns and planes between the sets of columns stand with respect to the vertical and horizontal, respectively, indicates the dip of the flow. Faulting in the sheets is frequently obscured by jointing. In North Mountain, from Cape Blomidon to Cape Split, and along the Victoria Beach shore of Digby Gut, the thickness of the flows may be estimated. A partial section is exposed at Sandy Cove and at Freeport, on Long Island, and at Tiverton, on Brier Island. In most of the sections the lowest flow is the thickest, and at the top of the series are several thin flows. The section from Cape Blomidon to Cape Split shows two and probably three flows, each with an estimated thickness of 150-300 feet. The top of the upper flow is exposed around the edge of Scots Bay. It exhibits the small folds into which all the basalt flows have been thrown. No other sections of the North Mountain basalt are exposed until Digby Gut is reached, because the sea- cliffs are low and expose only the upper flow or flows. At Victoria Beach the best section is found. There is some doubt if the lower flow, as here estimated, is not composed of two separate flows, but the microscopic examination of slides made from the first exposures above and below the blank in the section indicates a coarseness of grain which characterizes the center of a thick flow. Erosion has probably removed several flows from the top of the section. The section consists of: Top. Six flows 2-45 feet in thickness...... 160+ feet Bases (Main flow’ :.:rcd Ss .tneete ts eae 600+ The upper flows of the Victoria Beach section are absent from the exposures at the end of Digby Neck. They have either been THE ACADIAN TRIASSIC 265 removed by erosion or were never deposited there. The thicknesses of the portions of the flows remaining between the waters of St. Mary’s Bay on one side and the Bay of Fundy on the other are estimated as: Sandy Cove Tiverton (Ujoyovere wilon s's\ais Seeeeadinin UBI6 Wor 300+feet 150+ feet MO werploware Mia ctje cic « oluscsr 3 ths TSO ie te (Satie On Grand Manan the section is quite similar to those given above. The number of thin flows on the top of the series was not counted accurately. The section is: Top. Ten(?) thin flows averaging 10-15 feet in AV EKTIESS Naty nic eres efits august senenecey Set 100 feet SCEOMAPMOW asl eas Monta cemenns Bee) Base ues tnt Ow ar enee a slvars/ a) on eiclicveus beleuclietn ces sole ASO ts The number of flows exposed on Isle Haute is unknown. The section at Cape d’Or consists of 5 flows, of which the lower one (556 feet) is the thicker. At Cape Sharp and at Partridge Island two flows appear to be shown. Only one petrographic description of the basalt of North Mountain has been published. On account of the similarity of the basalts associated with the Newark group little attention has been paid to those of the Acadian area. The basalt is a dark-gray or dark-greenish fine-grained rock composed of plagioclase feldspar and augite with accessory amounts of magnetite, olivine, and glass. The feldspar is a labradorite, varying slightly in composition. The texture of the rock is ophitic, laths of feldspar inclosing augites, or masses of augite inclosing small feldspar laths. Chlorite, magnetite, limonite, hematite, and serpentine are present as alteration products. The proportion of glass to crystalline matter, of labradorite to augite, and the presence of olivine each depend on the proximity of the section to the top or bottom of the flow. The top of the flow is always quickly chilled in contact with the atmosphere, and solidifies with a large amount of glass and a large number of gas cavities. These cavities later become filled with quartz, calcite, 1\V. F. Marsters, “Triassic Traps of Nova Scotia,” Am. Geol., V (1890), 140-43. 266 SIDNEY POWERS or some other mineral to form amygdules. At the base of the flow, rapid chilling also takes place; less glass is developed, but well- crystallized magnetite is found. Alteration, however, soon com- mences in the base of the flow because of the reaction of heated waters on the basalt. The glass, characteristic of the top and the bottom of a flow, frequently contains most of the feldspar in laths already formed, showing that the feldspar had commenced to crystallize before the augite. In other cases the glass is accompanied by both augite and feldspar. The glass always has a cloudy appearance. Gravitative adjustment takes place in all flows which are sufficiently thick, and which remain hot sufficiently long for a movement of the crystallizing magma to take place without being recorded in flow structure. As in the case of the Palisade sill, olivine tends to form near the base of the flow and in the quickly chilled top. Gravitative differentiation is also shown in the relations of the labradorite to augite. The augite settles toward the base of a flow as in the case of a sill, and the feldspar rises. The chemistry of the Cape d’Or flows will be treated in a sep- arate paper, but it may be stated here that those basalts show a normal composition, averaging about 52.5 per cent silicia, 14.3 per cent alumina. 9.8 per cent lime, 2.5 per cent soda, and 1 per cent potash. Rosiwal measurements on thin sections from the center of a 556-foot flow show a mineralogical composition of 40 per cent plagioclase feldspar, 56.5 per cent augite, and 3.5 per cent iron ores. All the basalts show more or less alteration and disintegration except where rapid marine erosion exposes fresh rock. The amygdaloid, even where fresh, is always altered. In the drill- cores at Cape d’Or, the same character of alteration was shown in each amygdaloidal layer. A certain amount of hematite, with limonite, is developed, giving these rocks a reddish color. Veins are very common in the dense basalts as well as in ithe amygdaloids. The veins are formed of jasper or quartz, with either reddish (hematite) or greenish (malachite or chlorite) walls. THE ACADIAN TRIASSIC 267 ORIGIN The basalt unconformity of the Acadian Triassic always shows upturned and beveled rocks overlain by Newark sandstones or conglomerates with bedding parallel to the underlying erosion surface. This fact indicates that the Newark sediments were deposited on a peneplain, as has been found the case in the Connecticut? and Richmond? areas. On this peneplain, an orographic basin was formed, and into the geosynclinal area sediments were brought from all sides. An equilibrium between the rate of sedimentation and of subsidence _of the geosyncline appears to have been reached when the Blomidon shales were deposited at the top of the Annapolis formation. The Wolfville sandstone at the base of the Acadian Newark shows red sandstones and occasional conglomerates and shales, in general evenly bedded. ‘The pebbles in the conglomerates are stream-worn, but are frequently subangular. The character of the Quaco conglomerate has been sufficiently treated. The Wolfville sandstone indicates stream transportation, with deposition in flood-plains, and perhaps in past in broad alluvial fans. The Blomidon shales are generally evenly bedded, but show occasional ripple or current marks, and rarely mud cracks. The presence of Estheria indicates temporary bodies of water. Flood- plains of mature rivers would furnish the necessary conditions for the deposition of shales, with cut-off lakes in which the crustaceans could live. The red color of the Annapolis formation evidences long oxida- tion of the iron during transportation and deposition. The white or gray color indicates a lack of hematite, and the green color is caused by the presence of chlorite. The climate during the deposition of the Annapolis formation was apparently hot and dry, with occasional floods. The presence of calcite in nearly all the sediments, and the scarcity of arkose, tW. M. Davis, U.S. Geol. Surv., 18th Ann. Rept., 1898, p. 20. ?N.S. Shaler and J. B. Woodworth, U.S. Geol. Surv., roth Ann. Rept., 1899, p. 408. 3 J. Barrell, “Relation between Climatic and Terrestrial Deposits,” Jour. Geol., XVI (1908), 159-90, 255-95, 363-84. 268 SIDNEY POWERS and of plant and animal remains, all favor long oxidation of the sediments in a dry tropical climate. The Scots Bay formation was deposited in sie at least, in a lake, because fish remains occur in the strata. This lake came into existence soon after the extrusion of the North Mountain basal flows, as is indicated by the lack of erosion in the upper amygdaloid. The Five Islands volcanics are interpreted as representing a phase of igneous activity slightly earlier than that in which the North Mountain basalt flows were extruded. The volcanics may have come from central vents as well as from fissure eruptions. The North Mountain basalt must have come from fissure eruptions, and spread out over a large portion of the Triassic geosyncline, as is indicated by the widely separated areas at North Mountain and at Grand Manan. The geographical extent of any individual flow is impossible to determine, but it appears that the earliest flow, or series of flows, was the thickest. The physiographic conditions accompanying the formation of the Five Islands volcanics and the North Mountain basalts are poorly shown. The base of the North Mountain basalt is exposed only on Grand Manan, and there it is greatly weathered. No evidence of contemporaneous lakes over which the lava flowed has been found. THE LOMBARD OVERTHRUST AND RELATED LOGICAL FEATURES WINTHROP P. HAYNES Harvard University’ CONTENTS TOPOGRAPHIC AND STRUCTURAL FEATURES Folds Faults The Lombard Overthrust A Normal Fault STRATIGRAPHIC GEOLOGY Pre-Cambrian Spokane Formation Empire Shale Paleozoic Cambrian Absence of Ordovician and Silurian Strata Devonian Jefferson Limestone Three Forks Formation Carboniferous Madison Limestone Quadrant Formation Mesozoic Tertiary Pleistocene IcNEous Rocks Granite Diorite Diabase SUMMARY TOPOGRAPHIC AND STRUCTURAL FEATURES GEO- The region involved in this discussion lies near the head of the Missouri River in Montana. The chief topographic features are hilly dependencies of the Little Belt Mountains. 2690 270 WINTHROP P. HAYNES = = “Das da Formato > z z 3 = = AER te Fide hou Rormcioe eS | etic 7 a | Beit Fremetion HGNEOUS ROCKS di + of insler wen ng val 2eofert + ntowe Inter Fic. 1.—Geological map of the region about Three Forks, Montana THE LOMBARD OVERTHRUST AND RELATED FEATURES 271 The principal structural features are shown on the map (Fig. 1) and in the structure sections (Fig. 2). The dynamic features consist of folds and faults. FOLDS There is no indication that any marked deformation took place in this region during the Paleozoic and Mesozoic eras. At the close of the Cretaceous period, probably, the great series of sedi- ments which had been accumulating began to be deformed. In this region they were compressed into a series of closed folds with a general northeast-southwest trend. These folds are usually overturned to the southeast and pitch to the southwest. Two of these folds were named by Dr. Pealet the Horsehoe anticline and the Cottonwood isocline, both situated north of Logan, Montana. East of Lombard, in the vicinity of Crane Station, there is a northward-pitching anticline. In the long ridge west of the Mis- souri River there is an elongate domal structure (Fig. 3) the western side of which is interrupted by a normal fault and obscured by an extensive overthrust. The southern part of this elongate dome is overturned to the east and pitches steeply to the south (Fig. 1; Fig. 2, section D-D; and Fig. 4). FAULTS The Lombard overthrust—The most important feature of the structural geology of the region is an extensive overthrust fault which has its southern end in the ridge north of Three Forks, and extends a distance of at least 13 miles along the ridge to the north- ern border of the map. The writer proposes the name ‘‘ Lombard -overthrust” for this feature, because it is well exposed in the canyon of the Missouri River near Lombard. Here the fault plane dips about 40° to the west. This fault has brought strata of the Belt Series over strata of Cretaceous age in the north, near Lom- bard, and has brought the upper member of the Cambrian into contact with the Carboniferous Madison limestone in the southern end of the ridge (Figs. 2 and 5). The maximum displacement on the fault plane near Lombard cannot be very closely estimated, t A.C. Peale, Bull. U.S. Geol. Survey, No. 110, 1893. WINTHROP P. HAYNES N Nn SUOTJDIS [BOISO[OIH—'z ‘OI AO ee STINET SO ee e el ee 3 GOS Sham cN LP ROKK EEE Si 4 ats ; NS ae EEE a Higgs : oS ps ee eS er * ROK SONG Sire 1 mm lNee oats ee {egeNa - ikea ah au eu | YS ais! We - Zy~ (St THE LOMBARD OVERTHRUST AND RELATED FEATURES 273 but it is approximately two miles, and strata which are strati- graphically about 6,800 feet apart are here in contact. The age of the Lombard overthrust cannot be definitely deter- mined, but it is certainly younger than the Cretaceous strata ex- posed near Lombard, and probably older than the Lower Oligocene deposits which occur near the southern end of the ridge and are apparently undisturbed. It may therefore be assigned with some uncertainty to very late Cretaceous or early Tertiary time. A normal fault.—The only normal fault observed in this region appears in the highest part of the ridge north of Three Forks and west of the Missouri River. ‘This fault cuts across the western limb of the elongate dome already noted, and has caused a repetition of the upper part of the Gallatin formation and the base of the Jefferson limestone (see Fig. 2, section C-C). This fault has a length of about two miles and a diminishing throw to the south. It could not be traced to its intersection with the overthrust fault, but the displacement apparently dies out in that direction also. The age of the normal faulting is considered to be the same as that seen farther south in the Three Forks quadrangle, which is dated as probably Pliocene. STRATIGRAPHIC GEOLOGY PRE-CAMBRIAN The oldest rocks exposed in this region are a series of somewhat altered sediments which occur below the base of the Cambrian, and are considered to be part of the Belt Series, which are typically exposed in the Little Belt Mountain region to the north and north- east. The exposures of the Belt formation occur along the Gallatin River east and northeast of Logan, and also north of Three Forks, in a widening strip which trends northeastward and crosses the Missouri River at the double horseshoe bend west of Lombard (Fig. 1). The exposures north of the Gallatin River are of rather coarse micaceous sandstones and shales with thinly bedded siliceous lime- stones. They are not divisible on the basis of lithological char- acters into the various formations which characterize the Belt Series at the type localities. 274 WINTHROP P. HAYNES The extensive exposure of the Belt Series north of Three Forks, which, so far as the writer was able to ascertain, has not been described before, consists of two fairly distinct formations which are considered to be equivalent to the Spokane and Empire forma- tions of the Belt Series. Spokane formation.—In the vicinity of the double horseshoe bend of the Missouri River there is a fine section through the Fic. 3.—Domal structure in ridge west of the Missouri River Spokane formation. The formation at this place consists of a thick series of well-stratified red and green slates with frequent layers of ripple-marked and mud-cracked sandstone. The finer beds are mostly very hard and siliceous and may be called argillites or even metargillites. At several places in the section distinct folds are visible, and also some faults. The minimum thickness of the formation in this section is 1,650 feet, but the average thickness is probably considerably greater than these figures. Empire shale-—This formation, which overlies the Spokane formation, 1s exposed in a long strip west of the Missouri River, THE LOMBARD OVERTHRUST AND RELATED FEATURES 275 extending from near the southern border of the Fort Logan Quad- rangle near latitude 46° to the double horseshoe bend. It consists of evenly bedded, pale greenish shales with a few bands of quartz- ite. The quartzite occurs in beds from 1 to 25 feet thick. The formation is in apparent conformity with the overlying Cambrian Fic. 4.—Folded Madison limestone near southern end of ridge, north of Three Forks. quartzite near the southern end of the exposure, but inasmuch as the contact was traced for only a short distance an unconformity with very slight angular discordance may have been overlooked. Although the complete section of the Empire shale was not seen, it is probable that 800 feet is a conservative estimate for the thick- ness of the formation in this area. 276 WINTHROP P. HAYNES There is still much disagreement among the various geologists who have worked in the parts of the Cordillera where the Belt Series is exposed, in regard to the age of the series and the corre- lation of the different formations in it. The writer is disposed to agree with the correlation in a recent report on the Philipsburg quadrangle in Montana," in which strong evidence is shown for a rather long erosion period between the Belt Series and the over- lying Cambrian quartzite. The two formations identified by the writer as the Empire and Spokane formations are therefore con- sidered to be of Pre-Cambrian, Algonkian, or Proterozoic age. PALEOZOIC The Paleozoic formations recognized by the writer in this region are for the most part continuous with those described by Dr. Peale in his report on the “Paleozoic Section in the Vicinity of Three Forks, Montana,’” and later in the Three Forks Atlas Folio. His descriptions of the formations are very good and apply equally well to the exposures In the region to the north, on the Fort Logan Sheet. There are some additional facts concerning the thicknesses and ages of the formations and a few changes in the nomenclature which will be discussed under the following headings: Cambrian.—A comparison of sections made by different geolo- gists in the neighboring quadrangles shows that the seven lithologic divisions noted by Dr. Peale in the Three Forks quadrangle are persistent throughout southwestern Montana and the neighboring part of Wyoming. It seems advisable to have but one name for each of these divisions, and since locality names are preferable to descriptive names the writer suggests that the nomenclature used by Dr. Weed‘ in the Little Belt Mountains Folio be adopted for the Cambrian throughout the whole region where these seven lithologic divisions are recognized. For purposes of mapping it seems best to keep the broader divisions used by Dr. Peale, the two lower members forming the Flathead formation and the upper five the Gallatin formation. t Prof. Paper 78. U.S. Geol. Survey. 2 Bull. U.S. Geol. Survey, No. 110. 3 Atlas Folio, U. S. Geol. Survey, No. 24. 4 Atlas Folio, ibid., No. 56. THE LOMBARD OVERTHRUST AND RELATED FEATURES 277 Ko Be JSNIYJAIOAO pIVGwO'T IY} JO UOIQDes IVRUTWTLIBeIGa— S$ “O1T Be 278 WINTHROP P. HAYNES The following section of the Cambrian northeast of Logan, Montana, was measured by the writer. Dr. Peaie’s Nomenclature Dr. Weed’s Nomenclature Thickness 1. Pebbly limestone = Yogo limestone 75 feet 2. Dry Creek shale = Dry Creek shale 20 3. Mottled limestone = _ Pilgrim limestone 300 4. Obolella shale = Park shale 280 5. Trilobite limestone = Meagher limestone 175 6. Flathead shale = Wolsey shale 450+ 7. Flathead quartzite = Flathead quartzite 200 Total 1,500+ feet Fossils from the Yogo limestone have been submitted by the writer to Dr. Walcott, who considers them of Upper Cambrian age, while those from the Meagher limestone are regarded by him as of Middle Cambrian age. Apparently Lower Cambrian strata are entirely absent in sections in this region. Although the bound- ary between the Middle and Upper Cambrian strata has not been definitely ascertained, it is likely that it comes between members 2 and 3. Absence of Ordovician and Silurian strata.—In all of the sections studied by the writer in the Three Forks quadrangle and the neighboring district to the north, the Jefferson limestone lies in apparent conformity on the Yogo limestone without any inter- vening formations. The lower portion of the Jefferson limestone has been considered by Dr. Peale and others as probably of Ordo- vician and Silurian ages, although no fossils of those periods have been found in it. Dr. Kindle’ has described the Jefferson lime- stone and its fauna and established its age as chiefly Middle Devo- nian with the lower part probably Lower Devonian. In one or two good sections studied by the writer some rather poorly preserved corals were found within 25 feet of the base of the formation. These were identified as Favosites cf. limttaris Rom., which is rather common in much of the Jefferson limestone. The presence of these fossil corals is regarded as indicating the Devonian age of all of the Jefferson limestone, and since the gray Yogo lime- tE. M. Kindle, Bull. Amer. Pal. No. 20, 1908. THE LOMBARD OVERTHRUST AND RELATED FEATURES 279 stone immediately below the brown Jefferson dolomitic limestone contains Upper Cambrian fossils, the writer believes that at this contact there is a disconformity involving a hiatus in the sedi- mentary record of this region from the close of the Upper Cambrian to Lower Devonian time. Further evidence in favor of this disconformity and strati- graphic overlap is brought out by the presence in sections in neigh- Fic. 6.—Cliff of Jefferson limestone north of Crane Station boring regions to the west and southwest of intervening strata of . different lithologic character between the Yogo limestone and the Jefferson limestone, which in some cases contain fossils of Ordo- vician and Silurian ages. One very complete section from the Randolph quadrangle’ in northeastern Utah, with 3,000 feet of Ordovician and Silurian strata between the Upper Cambrian lime- stone and the Jefferson limestone, shows very clearly the hiatus in the sections in the Three Forks quadrangle and the neighboring region to the north and northeast. «G. B. Richardson, Amer. Jour. Sci., XXXVI (1913), 406-416. 280 WINTHROP P. HAYNES Devonian.—The strata of Devonian age in this region are divided into two distinct formations, the Jefferson limestone and the Three Forks formation. Jefferson limestone: The Jefferson limestone is well described by Dr. Peale’ as a massively bedded brown to dark-gray or black crystalline magnesian limestone with the composition of a dolomite. It is well exposed in the region under discussion in the form of brown cliffs too to 200 feet high (Fig. 6). In a few of the ridges Fic. 7.—Valley in Three Forks formation, near Rekap Station north of Three Forks the limestone is black in color, but shades of brown are the customary colors. In this region the Jefferson limestone has a thickness of about 500 feet, but it diminishes in thickness to the north and northeast, ,as noted in the sections in adjacent quadrangles. Three Forks formation: Lying upon the Jefferson limestone is a series of shales and limestones which have been described by Dr. Peale? and named the Three Forks shales. The writer has made a careful study of this formation in all of this region, and has measured numerous sections and made extensive collections of A.C. Peale, Bull. U.S. Geol. Survey, No. 110, 1893, pp. 27-28. 2 [bid., pp. 29-30. THE LOMBARD OVERTHRUST AND RELATED FEATURES 281 fossils from certain of the members. A detailed account of the formation and a description of some of the fauna is in process of publication elsewhere,’ so that only the more important points will be mentioned here. In all of the region included in Fig. 1 the Three Forks formation shows seven fairly distinct lithologic divisions. These members are well shown in the following section of the formation made northeast of Logan near the Gallatin River. Base of Gray Madison Limestone 1. Yellow arenaceous limestone............. 30 feet 2. Pale-yellow arenaceous shale............ 30 Bar lunplemissilershalernss amc uve wo. 8 0.5 4. Dark bluish-gray nodular limestone....... 9.5 Eemrssileyoreemsialeny. sau etes..c0 iia. 4 hae Os 47 (Yellow crystalline limestone............ 15 oe \ Gray MINES TOME He Peete Acca ean Tushar 12 7. Yellow and orange shales............... 78 Top of the Jefferson limestone. Total...... 222 feet Another section farther north along the Missouri River at Rekap Station (Fig. 7), shows the variation in thickness of the different members. 1 and 2. Yellow sandy limestone and shale . 74 feet Bepplackncoalivgshal emia arise caine elec ars 6 Ae Nodwlanerayalimestone yo) 0.0 2 rt 7 5. Fissile green shale | SEH GVGL, Sepsis ula wl” A oeea ea goers Ae mee 120 6. Gray and yellow limestone 7. Pebbly yellow and reddish limestones and ESN AVE WUSIS) 5a hae) Gay cee 80 SUroy eral heer Ae en een nn EL a 287 feet It will be noted from these two sections that the members con- sist of limestones as well as shales, so that the term “Three Forks formation” is preferable to Dr. Peale’s name “‘ Three Forks shales.”’ In the region north of Three Forks and west of the Missouri River there are numerous good exposures of the Three Forks t Annals Carnegie Museum, Pittsburgh. 282 WINTHROP P. HAYNES formation, whose erosion has formed some rather prominent valleys, as shown on the map and in Figs. 7,8,andg. These valleys extend in a general north-south direction and are nearly parallel with one another. This repetition of the formation is due partly to folding and partly to faulting. The easternmost valley eroded in the Three Forks formation is very narrow and shallow and extends northward along the eastern Fic. 8.—Great valley in Three Forks formation. Ridge north of Three Forks slope of the range of hills for five or six miles. The exposures are poor because the strata are vertical or overturned and much crushed by close folding. This valley, at its southern end, swings around to the west and opens into a much larger valley, which extends to the north for about two miles. The structure which is the cause of this curious arrange- ment of the valley is that of a southward-pitching anticlinal fold which is overturned to the east. The strata in this very large valley are in the western limb of the anticline (Figs. 8 and 9). THE LOMBARD OVERTHRUST AND RELATED FEATURES 283 West of the. overthrust fault there is another valley formed in the Three Forks formation. Numerous good sections of the forma- tion were obtained in the small tributary gullies which cut across’ the dip of the strata on the western sides of these valleys. The fossiliferous members of the formation are Nos. 1, 2, 4, and 5. The general conclusions from a study of the fauna are that the formation is very late Devonian in age, as reported by Dr. Fic. 9.—View north from southern end of valley, at apex of southward pitching anticline. Raymond in 1907,’ and probably represents a transition into the Mississippian in its upper part in members 1 and 2. Carboniferous.—Throughout the mountainous part of south- western Montana the Carboniferous formations are very prominent and form conspicuous and precipitous cliffs. In the region about Three Forks the Carboniferous strata attain a thickness of from 1,500 to 2,000 feet. tP. E. Raymond, Amer. Jour. Sci., XXIII (1907). 284 WINTHROP P. HAYNES Madison limestone: The lower formation has been named by Dr. Peale’ the Madison formation and was subdivided by him into three members; (1) the Laminated limestones at the base; (2) Massive limestone in the middle, and (3) Jaspery limestone at the top. The thickness of the Madison formation near Logan is about 1,300 feet. Although it forms conspicuous gray cliffs along the ridge west of the Missouri River, its best exposures ate seen where Fic. 10.—Missouri River in canyon in Madison limestone the river has cut a deep canyon through it near Lombard, and also in the smaller canyon along Sixteenmile Creek, east of Lombard (Figs. ro and 11). - A large collection of fossils was made by the writer from the Madison formation in all parts of the region. ‘These fossils all pointed to the general Lower Mississippian age of the Madison limestone. Quadrant formation: Lying in apparent conformity upon the Madison limestone in this region is the Quadrant formation which TAe CuLeale vOpNGl-433¢ THE LOMBARD OVERTHRUST AND RELATED FEATURES 285 forms the upper part of the Carboniferous system. The Quadrant formation consists of two members, as noted by Dr. Peale.t The lower is a red arenaceous limestone overlain by bands of shale and limestone. The upper member is thinly bedded cherty limestones alternating with quartzite layers. The top of the formation is somewhat arbitrarily placed by Dr. Peale at the base of a very massive and persistent quartzite layer which: is Fic. 11.—Double horseshoe canyon of the Missouri River. View east showing Lombard Station, and mouth of Sixteenmile Creek canyon. considered to be the basal member of the overlying Ellis formation of Mesozoic age. The writer obtained a thickness of about 400 feet for the Quad- rant north of Logan and 674 feet near Lombard. The exposure of the Quadrant formation in the canyon near Lombard is excellent, and a section was measured straight up the side of the canyon from the top of the massive cliff of the gray Madison limestone to the TAS ©. Reale, opi icit., p. 30: 286 WINTHROP P. HAYNES top of the massive quartzite layer which forms the rim of the canyon. The strata here strike N. 40° E. and dip 30° west (Fig. rr). Massive pink and yellow quartzite (base of Ellis formation ?)....... 16 feet Quartzite and arenaceous limestone in alternating layers........... 60 Massive white quartzite, limonite stains.................... Jon ee Limestone .brécelan ie 2. 7 ee ae aA ce ee ee ee 2 Brown: arenaceous limestones. +) seer ee ee eee 62 Grayish-brown arenaceous limestone and talus................... 62 Pink ‘arenaceous limestone anyelitie ee eye ee ee ee 36 Yellowish-red arenaceous dimestonecee tee es a ee 47 Gray limestoneuni clifis /shalyyamibase pears eae. it ee eae 10 Reddish'shale:, 2 :oia Ge. Bee ee eee ioe a ans 5 ee 30 Greenish shale 22:4. 2.597) oyn 6 ee Pee os Se ee en ae 57 Bufi. shaly-limestoneand talus) eee eee ee ee ee 100 Gray bituminous limestone in cliff, with black shale layers......... 45 Compact gray and yellowish-brown limestone.................... 2 Black coaly shale with calcareous bands and gypsum veins......... 20 Brown-crystalline limestoneswaa nee ea eet es ee 4 Coaly ‘black ‘shales very tossiliferousicmrcme ome tee ante ce ee 50 Yellow arenaceous limestone in cliff, some quartzite bands......... 46 Red. shaly lmestomess" ter ae goer oe ee noe eI ete ee 10 Potak sere se et ee ee Sete une ER SAY RU SEER 674 feet The fossils collected from the Quadrant formation indicate that it is probably of Lower Pennsylvanian (Pottsville) age. The absence of any strata referable to the Tennesseic suggests the presence of a disconformity between the Madison and Quadrant formations, although no other evidence of such a hiatus was observed by the writer. MESOZOIC Mesozoic formations are rather poorly exposed in this region and were not studied in detail by the writer. They consist of shaly limestones and sandstones which are generally much less resistant than the Paleozoic limestones and therefore usually occupy lowland areas. These Mesozoic strata border the Missouri Valley on both sides, and the more resistant layers form low ridges which are parallel with the trend of the higher Paleozoic hills. THE LOMBARD OVERTHRUST AND RELATED FEATURES 287 In this region the Ellis formation, consisting of sandy shales and limestones with numerous layers filled with pelecypod shells, lies on the Quadrant formation with no observed discordance of dip. In the region to the south there is a well-marked reddish sandstone formation of probable Triassic age intervening between the fossiliferous Ellis and the Quadrant. Since the Ellis fossils are considered to be Jurassic in age, it seems clear that there is a dis- conformity in this part of the sections of this region. Above the Ellis formation is a series of sandstones and con- glomerates which have been called the Dakota sandstone by Dr. Peale, but they have recently been shown to be more probably the equivalent of the Kootenai formation of the region to the north. These sandstones are therefore of probable Lower Cretaceous age. Strata of Montana and Colorado age were identified by Dr. Peale in the hills north of Logan, but there is now some doubt as to whether they can be referred to a horizon as high as that. TERTIARY All of the Mesozoic and Paleozoic strata were involved in the extensive orogenic movements which began at the close of the Cretaceous in this region. The type of folding and the associated overthrust faulting has already been described in this paper. Extensive erosion reduced the region to comparatively low relief in Tertiary times. . The great lowland areas were filled in by sedimentary deposits of sand- stone, limestone, and volcanic ash to a great depth. The major features of the present drainage were established on this late Ter- tiary surface and gradually, through uplift and erosion, they were brought into discordance with the underlying structure, as is well shown by the double horseshoe canyon of the Missouri River west of Lombard. This whole series of Tertiary valley sediments has been grouped under the heading of the Bozeman formation for convenience in mapping. Dr. Peale’s name ‘‘Bozeman Lake Beds” seems no TW. R. Calvert, Bull. U.S. Geol. Survey. No. 471—-E, 1912, p. 53- 288 WINTHROP P. HAYNES longer applicable, since they have been shown to be due to sub- aerial and fluviatile deposition rather than to lakes." The Bozeman formation here is chiefly of Miocene age, but in some parts of the region strata of Oligocene (White River) age have been identified. PLEISTOCENE The hills in this region were evidently too low for local glaciation and no signs of regional glaciation have been observed as far south as this in Montana. Gravel terraces. along the rivers indicate greatly increased stream action in Pleistocene times. IGNEOUS ROCKS The igneous rocks in the region north of Three Forks are rela- tively unimportant, and are in the form of rather small intrusions of three different rock types. GRANITE About two miles west of Lombard, in the double horseshoe bend, the Missouri River flows for a short distance through a gorge cut in an intrusive mass of granite. Only the eastern boundary of this granite could be accurately mapped, but the approximate western limits are noted on the map. The granite:is of a light-gray color, with a medium fine texture and a somewhat porphyritic structure. The minerals recognized in a megascopic examination are white and grayish feldspar some- what kaolinized, quartz in small amounts, and hornblendes mostly altered to chlorite. Under the microscope the feldspars are seen to be deeply kaolinized, but are chiefly orthoclase with some albite. There is a considerable amount of hornblende which is altered in part to chlorite and epidote. Some biotite and magnetite are also present. In places this rock is almost entirely without quartz and there- fore grades into a syenite. It seems to correspond closely with the description of the syenite of Yogo Peak’ and vicinity in the Little Belt Mountains, which is noted as grading into a granite-syenite- tH. F. Osborne, Bull. U.S. Geol. Survey, No. 361, 1909, p. 28. 2 Atlas Folio, U.S. Geol. Survey No. 56, 1890. THE LOMBARD OVERTHRUST AND RELATED FEATURES 289 porphyry. The granite has a well-developed set of joints which strike northeast and dip 80° east, and are about parallel with the contact with the Belt Series. The age of the granite cannot be definitely determined at this place, but it is probably about the same age as the granitic and syenitic intrusions of the Little Belt Mountains, which are post- Cretaceous and probably early Tertiary in age. DIORITE Small irregular intrusive masses of diorite and diorite porphyry occur in the vicinity of Dunbar’s mine, north of Three Forks. These intrusions cut the white Tertiary limestones which at this locality are considered to be of Lower Oligocene age. The diorite was observed to have nearly vertical contacts with the limestone and to occupy a much smaller area than is indicated on the geologic map of the quadrangle. The diorite porphyry seems to be a local variation in the normal diorite and its distribution can be shown only on a detailed map of the district. Specimens of fresh diorite were obtained from the dump at Dunbar’s mine. The rock from the main shaft is of medium fine texture and evenly crystalline. It consists of an even mixture of black hornblende and gray feldspar. Under the microscope the rock is seen to consist of greenish-brown to dark-green pleochroic hornblende and labradorite feldspar. Apatite, olivine, and mag- netite occur in small amounts as accessory minerals. Specimens of diorite from a shaft about a half-mile to the south show a small amount of pale-pink orthoclase feldspar scattered through the rock. Some of the diorite from a small intrusion which cuts the Cambrian formations a few miles north of the mine is distinctly porphyritic and consists of hornblende phenocrysts in rather slender crystals about a half-inch long in a gray ground-mass of plagioclase feldspar and hornblende. Magnetite and apatite occur in small amounts scattered through the ground-mass and are visible under the microscope. The rock is deeply weathered at the surface and the hornblende is mostly altered to chlorite, and the feldspar is kaolinized. 290 WINTHROP P. HAYNES There are zones of altered rock along the contacts of the diorite and the Tertiary limestone which are well exposed about Dunbar’s mine. In this contact zone are many secondary minerals which include garnets, and several copper-bearing minerals, chiefly chrysocolla, with some malachite and azurite. It is the presence of these minerals which has caused the development of Dunbar’s mine. This mine was not in operation during the summers of 1912 and 1913 when the writer visited the region. DIABASE A rather large intrusion of diabase was observed by the writer in the extreme northern part of the region, about a mile west of Lombard. This somewhat irregular dikelike intrusion follows the plane of the thrust fault across the double horseshoe bend of the Missouri River and varies in width from roo to 500 feet. The intrusion has produced a noticeable contact effect on the country rocks, particularly on the Cretaceous rocks on the east side, which are indurated near the contact. The diabase is deeply weathered near the surface and has a rusty brown color. It forms a very conspicuous massive wall on the north side of the Missouri canyon, northwest of Lombard. The rock shows the ophitic structure well and is composed of augite and labradorite with some olivine, magnetite, and apatite. The age of this intrusion cannot be very definitely placed but it is clearly post-overthrusting, and therefore of Tertiary age. SUMMARY The contributions of this article may be summarized as follows: rt. A new geologic map of a portion of the Fort Logan region, and a revised geologic map of a part of the Three Forks region. 2. The recognition of an extensive overthrust in the north- western part of the region, ‘‘the Lombard overthrust.”’ 3. New facts relative to the stratigraphy of the region mapped, including the identification of a portion of the Belt Series, and the recognition of a disconformity between the Yogo limestone and the Jefferson limestone. 4. Detailed sections of some of the Paleozoic formations. 5. The igneous rocks and their manner of occurrence. THE SKELETON OF TRIMERORHACHIS S. W. WILLISTON University of Chicago A year ago, in a paper on the structure and habits of Tri- merorhachis, I said, that “‘it will only be by the fortunate discovery of a connected skeleton that the tail, ribs, and feet will be made known.’* Such a specimen has been discovered and _ skilfully worked out by Mr. Paul Miller, a photograph of which, as pre- pared, is shown in Fig. 1. The specimen came from the pale- ontologically famous Craddock Ranch, near the town of Seymour, Texas, from the same horizon as that of the skeleton of Seymourza, described by me a few years ago, and within a stone’s throw of its locality. Its horizon seems to be nearly the same as that of the Craddock bone-bed, from which so many remarkable specimens have come. When found, the specimen was inclosed in a large, irregular nodule of bright red claystone; nothing was visible of it except the extreme tip of the nose and the base of the tail, as shown by a fracture. The under side of the nodule was smoothly convex both longitudinally and transversely; its upper side was irregular and gnarly. With this specimen, and in immediate rela- tion with it as it lay upon the surface, were found a number of pieces, which, when fitted together, formed a block about one foot in length, which seemed to be a continuation of the tail end of the larger block. When fully prepared, however, the smaller block proved to belong to a second specimen of Trimerorhachis, includ- ing about twenty chiefly precaudal vertebrae, with their ribs and an imperfect femur. The larger specimen, that figured herewith, is a nearly complete skeleton as far back as the sixth or seventh caudal vertebra. The 1 Cope, Proc. Amer. Phil. Soc., XVII (1878), 524; XIX (1880), 54; Amer. Nat- uralist, XVIII (1884), 32; Case, Revision of Amphibia and Pisces of North America (1911), 39, 106; Huene, Bull. Amer. Mus. Nat. Hist., XXXII (1913), 372; Broom, Anatom. Anzeiger, XLV (1913), 73; Bulletin Amer. Museum, XLV (1913); Williston, Journal of Geology, XXI (1913), 625; XXII (1914), 160; XXIII (1915), 246. 291 S. W. WILLISTON bones, as usual in such nodules, are white and rather soft, rendering Fic. 1.—Trimerorhachis. in original matrix. Specimen No. 1271, About one-fourth natural size. their preparation in the hard matrix difficult. It has been worked out so far as was possible without going below the surface of the bones. The come to rest in a prone cadaver had position, apparently, with the head and tail directed obliquely up- ward, its vertebrae con- nected throughout in a sinuous curve, and the ribs nearly all in im- mediate connection with their articulating The right humerus had been dis- diapophyses. located, and lay near the posterior end of the right mandible, with its radius a little distance The right femur lay nearly from its distal end. in apposition with the acetabular part of the ium; its distal part had been eroded away. Doubtless both the front and the hind legs of the left side are buried some- where in the matrix. The wholly unexpected fact remarkable and disclosed by these speci- THE SKELETON OF TRIMERORHACHIS 293 mens is the presence of a thin bony skin or armor closely sheathing the whole body, with the exception of the skull and clavicular girdle. As the cadaver came to rest it was immersed in the soft mud to near its middle. The skin lining the cavity thus made retains its original position. On the decay of the body, the bones fell to the lower part, closely covered everywhere by the skin of the upper part of the body. On the right side the skin had bulged outward near the middle. When first uncovered the bones were concealed everywhere by the skin. It has been removed on one side or the other to expose the bones, and between them, in a few places, to show the skin of the under side of the body, which in some places lies in juxtaposition with that of the upper side, in others séparated by a thin layer of the matrix. A dermal covering of peculiar type in Trimerorhachis has been several times observed by Cope, Case, and myself, but it was assumed that it covered the ventral region only, and its nature was ill understood. The present specimens show very conspicu- ously that it covered the whole body, with the exceptions men- tioned; in the preparation of the skull not a trace of it was seen, but it is closely connected with its hind margins. In no place in these specimens does it appear to have been more than a milli- meter in thickness. It is composed of slender and delicate bony fibrillae, in short pieces, and apparently in several layers. In another specimen (Fig. 3, B, C,) transverse sections show that the bony rods were in numerous layers. As these fibrillae lie in this specimen they extend through a thickness of 6 or 8 mm., and are separated from each other by intervals greater than their own thick- ness. It seems hardly possible that postmortem causes could have separated them so uniformly, and one must conclude that they were imbedded ina considerable thickness of integument, at least a fourth ofaninch. How long any of the rods were I cannot say; the longest connected piece that I trace is scarcely a fourth of an inch. It is still possible that the sections represent the ventral skin, since nothing of their character is visible in the connected skeleton. Notwithstanding this thickness, the skin must have been flexible to have followed every inequality of the bones below it. It was doubtless covered by a smooth epidermis. 204 S. W. WILLISTON Thirty-one precaudal vertebrae are visible, all in close articulation, the first one apparently with the condyle. That the vertebral column was very flexible is evident from its sinuosity as it lies in the matrix, with- out break. The general structure of the verte- brae is well known from isolated specimens. The pleurocentra are very small, and the intercentra are more or less U-shaped, indicating a large amount of carti- lage. The spines are of nearly uniform height throughout the column, curving backward and upward, and slightly dilated at their extrem- ities. Their het@he above the plane of the zy gapophyses is nowhere more than 14 mm.; their width at the ends from 8 to 10 mm. The ribs are preserved very completely in posi- tion; all have been ex- posed on one side or the other except two, Fic. 2.—Trimerorhachis. Skeleton, from above, th ‘ddl Th as restored from specimen shown in Fig. 1. One- ai - ven &. = fourth natural size. first eight -arew te THE SKELETON OF TRIMERORHACHIS 205 longest, measuring about 52 mm. They also have a considerable curvature and are more or less expanded at their extremities. With the tenth or eleventh they have decreased in length to 45 mm., are less curved, and not dilated at their distal ends. Thence to the tenth precaudal they are of uniform length, more slender, and pointed. When seen from above these are all slender and nearly straight, with a moderately expanded proximal end; when turned upon their sides they are broader, somewhat curved, and with a more dilated head. Apparently some of them at least have a capitular prolongation in articulation with the intercentra. The first two precaudal ribs are slender, pointed, and entirely free, and about one inch in length. The right humerus lies near the angle of the right mandible, as will be seen in the photograph, with its head directed forward, and with the radius somewhat removed from its distal end. There are no remains of the skin either above or below these bones. Some half-dozen skulls of Trimerorhachis have now been obtained with the peculiar clavicular girdle in position or nearly so, lying more or less between the mandibles posteriorly. In the drawing (Fig. 2) I have outlined in interrupted lines a clavicular girdle of another skull of the same size as the present one, in position, placing it farthest back of any of the connected specimens. The angles of the clavicles with their ascending process must indicate the position of the scapula and articulation of the humerus. The scapula is hidden and not certainly determinable in this specimen, though the edge of a protruding bone at the inner side of the distal end of the humerus, as it lies in the matrix, is probably that of the scapula. The scapula, ilium, and limbs of the left side are doubt- less preserved in this specimen covered by the skin and matrix, but it has not been thought wise to sacrifice so much of the speci- men as might be necessary in the search for them. In all the material from the bone-beds so far examined not a trace has been found of hand or foot bones, so that nothing can be said of their structure. The right ilium lies in place in the matrix, in a vertical position opposite the ends of the first two presacral ribs. The bone is relatively very small, as will be seen from the figure. Among the 206 S. W. WILLISTON numerous ilia in the collections (I have seen about two score) there is no indication of roughening for attachment to sacral ribs. So far as was prudent the matrix has been removed about the ilium; there are no indications of ossified pubis or ischium. The proximal end of the right femur lies nearly in place opposite the Fic. 3—Trimerorhachis Integument. Properties of Matter, 1902, p. 181. 3 The chief conclusions reached by Mare are contained in the following brief citations (Zeitschr. f. phys. Chem.): “Tt was found that the rate of crystallization, so far as it could be determined, of all the substances investigated was proportional to the square of the supersatura- tion”’ (op. cit., LX VII [1909], 500). a . that a very rapid change precedes the crystallization proper, which is interpreted to be an adsorption phenomenon. Support is given to this view by the THE LINEAR FORCE OF GROWING CRYSTALS BBY] again, ““On Adsorption and Saturated Surfaces” (1913),’ without developing any single fact in support of the hypothesis advanced by Bruhns and Mecklenburg. | Neither capillarity nor adsorption exerts any upward pressure on the loaded disks of porcelain in the experiments under discus- sion, while adsorption does not prevent the exercise of the very great downward pressure due to the surface tension of water. Yet the alum crystallized and the disks were raised. In the opinion of the observers it was essential to the elevation of the disks that evaporation should be complete.t Was the eleva- tion, then, produced after the crystallization was complete and the mass solidified? ‘The observers make no such statement, which, indeed, would seem absurd. But if the raising was not effected after solidification, it must have been produced before solidifica- tion, or while the underlying film was liquid and while crystalliza- tion was in progress, in opposition to capillary force as well as to the weight of the disks and their load. Liesegang appears to have appreciated this anomaly in Bruhns and Mecklenburg’s statement, though accepting their conclusion, for he sought to relieve it by the following explanation (referring to the experiments of Bruhns and Mecklenburg): ‘‘ Nicht ein Wach- stumsdruck der Krystalle sondern Capillar- und Adsorptionskraite bewirkten hier also die Hebung. Das heisst die Leistung war schon volbracht ehe Kristalle auftraten.”? ‘‘The lifting was done before the crystals formed.” This is not claimed by Bruhns and Meck- lenburg, nor supported by any experimental evidence which they fact that this preliminary phenomenon is particularly sensitive to slight impurities upon the crystal surface” (ibid., LX XIII [1910], 718). “No relation could be established between concentration and the quantity of adsorbed material” (ibid., LX XIII [1910], 686). “Tn all cases the rate of crystallization is diminished by the addition of substances which are adsorbed by the crystal, eventually even to the point of becoming practically zero”’ (tbid., LX XIII [1910], 718). “Tt was shown that the substances chiefly adsorbed by crystals are colloids, while the crystalloids are adsorbed only very slightly” (zbid., LXXXI [1913], 692). t Bruhns and Mecklenburg, op. cit., p. 106: “‘Es sei aber ausdriicklich betont dass der Versuch nicht gelang, wenn wir nicht die Masse bis zum Grunde trocken werden liessen.”’ 2R. E. Liesegang, ‘“ Kristallisationskraft,’ Naturw. Rundschau der Chem. Zig., Zweite Jahrg. 1913, p. 183. 328 GEORGE F. BECKER AND ARTHUR L. DAY offer, and indeed would seem to be without any foundation whatsoever. We fail to see any reason for connecting the rise of the porcelain disks with capillarity or with adsorption. These could only obstruct the elevation, and must have been overcome by a linear force attending the crystallization of the alum, as in our own experiments. It is not expedient, however, to rely on reasoning alone in matters of physics if experimentation is practicable, and we accord- ingly made the effort to separate the forces to which Bruhns and Mecklenburg appeal, through evaporation of solution of a colloid (gum arabic) in which was immersed a block of glass replacing the alum crystal between the two plates of glass (Table VII). Evap- oration to dryness caused no rise of the upper glass plate as it TABLE VII BLock OF GLASS REPLACING THE ALUM CRYSTAL (Fic. 1). Loap (Grass PLATE) = 24 Gm. GLAss BLock AND LOAD COMPLETELY IMMERSED IN 2 PER CENT SOLU- TION OF GuM ARABIC IN WATER. ROOM TEMPERATURE Time (Hours) Sean ‘Thickness Notes ORiii. eee eee | 37.2609 | o mm. i fa pen aarti ic | 37.2649 +o.004 Donte ee WRe nee 37.2654 +0.0045 PA Est ie eRe ghey Pein 37.2621 -++o.001 [OCR aces ao Bis ecko | 37.2615 | +o.001 Evaporated to dryness Refilled with 2 per cent gum arabic solution; all conditions unchanged CO Be has Ce te | 37.2669 +o .006 ci ee eee | 37.2670 +0.006 WORT vo cisucntess ane | 37.2670 +o.006 TLORO weiss ace Or | 37.2653 +o.004 Evaporated to dryness should have done were capillarity and adsorption the source of energy. A saturated solution of alum added to the colloid (Table VIII) starts crystal formation and growth at once, but at a rate much slower than in the cases where no colloid was present. This is in full accord with the experiments of Marc.t In addition to confirming the results of Marc, Table VIII offers independent and explicit experimental proof that the “linear force”’ appears here also in spite of the action of the colloid in retarding t See footnote, p. 326. THE LINEAR FORCE OF GROWING CRYSTALS 320 diffusion through increased viscosity and in interposing an adsorp- tion film at the crystal surface. Conditions in ore deposits appear to correspond very well with those in the laboratory, for crystallization may be found accom- panied by local evidence of linear thrust or not, according to the magnitude and distribution of the opposing forces. Its failure is most often manifest in comb structure, found: in crevices whose walls are each lined with tightly adhering crystals which either interlock and extend quite across the crevice or grow together near the central plane and mutually exclude further development. Such comb structure is common in veins, but far from universal. TABLE VIII SAME PLATES, BLOCK OF GLASS, AND CONDITIONS, EXCEPT THAT THE 2 PER CENT Gum ARABIC SOLUTION HAs BEEN SATURATED WITH POTASSIUM ALUMINIUM SUL- PHATE Time (Hours) Reine in Thickness (Oh yee nei Nae ee 37.2653 ° mm. AP OR aurea tae ee 37.2694 0.004 Das Ne cslgiat iad ipsa err ee 37.2721 0.007 QO} wetieepaiese ee aval caen: 37.2812 0.016 MOB eG ieine nets see meastyee 37.2818 0.017 TO ath eh a, deni yee ee 37.2842 0.019 OG ete ears vatiois eae es 37.2857 0.020 TAO agi ites eared enero tac 37.2873 0.022 DEG Ae eect er Cr eae 37.2916 0.026 Diamante a eek Anan ae 37-3079 0.043 DSi te a eee see RET cE Bo QUge) 0.053 3 OOPaae es nsec: Sensi BY] sf) 0.052 AOOii te corset ees as 37.3185 0.053 It may be inferred, further, that linear pressure plays a subordi- nate part in much more complex occurrences. Messrs. Bruhns and Mecklenburg seem to have misunderstood the last paragraph of our paper in which we called attention to the fact that the linear force of growing crystals cannot be disposed of as a mystery comparable to the growth of plant roots. It is a sharply defined physical process open to quantitative experimental investigation. It may not be fully understood, but it is no mystery. 330 GEORGE F. BECKER AND ARTHUR L. DAY . The conclusion of these authors seems to be that during growth, material is added only to the upper and lateral faces of the crystal, so that a molecule once added remains at its original level. This was Kopp’s contention in opposition to Lavalle, whose conclusions, however, were confirmed by Lehmann and others, including our- selves. ‘This is in fact the root of the matter. Ifa given increment of the mass after deposition remains at its original level throughout the subsequent growth of the crystal, this exerts no linear force; while if the motion of the particle has a vertical component in consequence of the vertical extension of the lateral faces of the crystal, linear force is exerted. . On the other hand, if several crystals are immersed, one or more of them being loaded while others are not loaded, the loaded crystals grow only when the concentration of the solution in contact with them exceeds the saturation concentration for each crystal. Pres- sure, of course, increases solubility or raises the point of saturation for most salts.‘ Hence in such circumstances the unloaded crystals, or, more strictly, the less loaded crystals, usually are the only ones to exert lifting power, but in this case, also, growth raises the weight of each crystal. Thus Bruhns and Mecklenburg’s results with loaded porcelain disks are readily explicable. They experimented with solutions containing many small crystals, some of them weighted, others free. The disks did not rise measurably until the liquid was low and its surface (and consequent rate of evaporation) greatly increased by protruding solid matter, or until the crystals reached from the bottom of the dish to the disks, after which the disks were lifted. Repetitions of this operation, extending over a few days, pro- duced aggregate displacements of 5.o mm. If to this be added our original measurement, twice confirmed in the course of the present control tests, that this linear force, because of the narrow rim « As is well known, if the solution of a solid at constant temperature is attended by a diminution in total volume and a liberation of heat, pressure increases solubility. Such is the case for most crystalline solids including the alums. If the change in volume accompanying solutions is an increase, as in ammonium chloride, pressure decreases solubility. THE LINEAR FORCE OF GROWING CRYSTALS 331 through which it acts, actually exerts a pressure of the same order of magnitude as the breaking load of the solid crystals, need there be further hesitation in assuming that this is a force to be reckoned with in engineering" or in geology ?? SUMMARY In 1905 we showed by appropriate experimental evidence that a single crystal immersed in its own saturated solution, and growing by reason of the potential supersaturation of the solution resulting from evaporation will lift a weight placed upon it. This observa- tion has been confirmed in the present paper. In 1913 Bruhns and Mecklenburg placed two crystals in a similar saturated solution, one loaded and the other free, and noted that the load upon the one crystal was not raised, although the free crystal grew rapidly. From this experiment they were led to deny the power of a crystal to lift a weight of foreign substance, although admitting the power of the unloaded crystal to lift its own substance. They appear to have overlooked in this conclusion the fact that the solubility of the loaded crystal is for most substances greater than that of an unloaded one, and also that this is a difference in degree only, for the unloaded crystal also supports weight (its own). In consequence of this greater solubility, with an unloaded and a loaded crystal in the same solution, the necessary condition of potential supersaturation will be reached in the liquid adjacent to the unloaded crystal before it is reached in the other, and the growth of the unloaded crystal thereafter may keep the concentra- tion below that necessary for the growth of the loaded crystal. This appears to be the condition reached in Bruhns and Mecklen- burg’s experiment. If it happens, however, that the rate of growth 1Cf. the investigations of Dr. Hans Kiihle, ‘“Die Ursache des Treibens der Zemente,” Tonindustrie Zig., XXXVI (1912), 1331-34; and of Klein and Phillips, “Hydration of Portland Cement,” Technologic Papers of the Bureau of Standards No. 43 (1914), pp. 50, 56, 57. 2Cf. the recent observations of Stephen Taber, Virginia Geol. Survey Bull., No. VII (1913), p. 222; also G. D. Harris, ‘Rock Salt, Its Origin, Geological Occur- rences and Economic Importance in the State of Louisiana,” Geol. Survey of Louisiana, Bulletin No.7 (1907), p- 75. 332 GEORGE F. BECKER AND ARTHUR L. DAY of the unloaded crystal is insufficient to take up all of the excess concentration provided by the continued evaporation, then super- saturation will increase. It is entirely possible under these condi- tions that the potential supersaturation necessary for the growth of the loaded crystal may then be attained or even exceeded, and © that the loaded crystal will also grow and lift its load. This condi- tion was attained experimentally without difficulty in the observa- tions recorded in this paper. If concentration increases still more rapidly, and exceeds the ability of both unloaded and loaded crystals to take up, through their continued growth, all the matter in excess of the saturation concentration, then additional nuclei may form upon which excess matter may be deposited. This appears to have been the condition attained in the last series of Bruhns and Meck- lenburg’s observations in which the solution was evaporated to dry- ness. Here six disks of porcelain loaded with weights were all raised a millimeter or more in the same solution, but Bruhns and Mecklen- burg attribute this result to the action of capillarity and adsorp- tion, and deny the competence of the “linear force of growing crystals”’ to effect such mechanical displacements. A simple analysis suffices to show that capillarity in a solution evaporating to dryness can have no other effect than to press the crystal down upon its base with a force equal to 27 V/d?, where T is the surface tension, V the volume of the drop of liquid between the crystal and its base, and d the distance separating the two, and that the lifting action observed by Bruhns and Mecklenburg has occurred in spite of this opposing force and not because of it. Adsorption delays diffusion and diminishes the rate of growth, but does nothing to promote it. These forces therefore cannot be appealed to in explanation of the lifting observed by Bruhns and Mecklenburg and by us. We therefore return to the original thesis that the growth of crystals in saturated solution develops a linear force in the direction of the load, and that neither the magnitude of the load (up to the breaking load) nor its character (whether exclusively crystal sub- stance or partly foreign substance) has any other effect than to THE LINEAR FORCE OF GROWING CRYSTALS 333 increase solubility and so to raise the concentration necessary for potential supersaturation and growth upon the loaded crystals. This degree of supersaturation is readily attainable through evapo- ration or otherwise, and when attained the loads are lifted. With this thesis established, there is no conflict between the observations of Bruhns and Mecklenburg and our own, and all the experimental evidence offered is perfectly correlated. CARNEGIE INSTITUTION OF WASHINGTON GEOPHYSICAL LABORATORY February, 1916 THE CLASSIFICATION OF THE NIAGARAN FORMATIONS OF WESTERN OHIO" CHARLES S. PROSSER Ohio State University CONTENTS INTRODUCTION DESCRIPTION OF SECTIONS Ludlow Creek Sections Section of Ludlow Falls and the Smith Quarry Section of Western Wall of the Smith Quarry Section of the Otto Ehlers Quarry Section of the Maxwell Quarry General Section along Ludlow Creek Correlation of the Brassfield Limestone Other Sections of Western Ohio Sections in and near Covington Section of the Jackson Stone Co. Quarry Section of the J. W. Ruhl Quarry at Covington Section near Lewisburg Section at the Lewisburg Stone Co. Quarry Section near Laurel, Indiana Derbyshire Falls Section INTRODUCTION There has been more or less uncertainty concerning the names which ought to be used for the Niagaran formations of the Silurian system in western Ohio together with their correlation with the formations of the same series in eastern Indiana. Field work in the summer of 1914 in this area has cleared up some of this uncertainty and part of the results are deemed of sufficient importance to war- rant their early publication. t Presented at the Ohio Academy of Science meeting in Columbus on November 28, 1914, and at the American Association for the Advancement of Science meeting in Columbus on December 28, 1915. Published by permission of the State Geologist of Ohio. 334 THE NIAGARAN FORMATIONS OF WESTERN OHIO — 335 A series of sections at Piqua, along the Stillwater River between Covington and West Milton, and along Ludlow Creek in Miami County, and in the vicinity of Lewisburg, Eaton, and New Paris in Preble County, has furnished the writer the complete section from the Ordovician to the highest Silurian rocks of this area. DESCRIPTION OF SECTIONS The contact of the Ordovician and Silurian systems is clearly shown at Ludlow Falls, and the succeeding rocks as high as they extend in this region are admirably exposed at the falls and in the series of quarries which border the creek for some distance above the falls. Sections at other localities agree with the ones along this creek and show that the general order of succession is essentially the same for these counties. LUDLOW CREEK SECTIONS Four of the series of sections measured along Ludlow Creek will be given, which were checked by several other sections along the same stream. From these a general section of the rocks shown along this creek can be compiled. The following section is based on the outcrops in the north- eastern corner of the Colonel Samuel B. Smith quarry and the bank at the northern end of Ludlow Falls: SECTION OF LUDLOW FALIS AND THE SMITH QUARRY ToTaL : THICKNESS TD SiSSESE No. Feet Inches Feet Inches 6. Dayton limestone——Northeast corner of the Colonel Samuel B. Smith quarry. The rock varies from light gray to somewhat darker gray on fresh fracture and some as weathered is bluish-gray. Other layers on the weathered faces are buff to brownish or rusty color from disintegrated iron pyrite. The rock splits into even-bedded layers; but the surfaces of the bedding planes are frequently rather rough and show stylolites structure. The majority of the layers vary in thickness from 2 to ro inches, most of them ranging from 4 to 6 inches. The lowest layer is from 3 to 4 inches thick, and the 336 CHARLES S. PROSSER THICKNESS eno No. Feet Inches Feet Inches next one above 10 inches, which in places splits into 2 layers with hackletooth structure at con- tact. The lower layers contain iron pyrite, which stains them on weathering. There are also calcite crystals, but no fossils were noted 8 2 AI 5 . Brassfield limestone.—Crystalline and crinoidal light-gray to pinkish limestone, with very ir- regular bedding planes. Specimens of corals and Stromatopora are rather common, and there is an occasional Brachiopod shell. In places there are irregular masses of blue shale which contain a good many corals. This is the bottom of the quarry at this corner, and the barometer with an interval of only 7 minutes read the same at the highest outcrop of the Brassfield limestone on the northern bank of Ludlow Creek, just above the Dayton, Coving- ton, and Piqua traction bridge and Ludlow On 4. The upper 5 feet of the bank at the northern end of Ludlow Falls varies in color from light gray to pinkish and greenish, while some of it when weathered is brownish. Most of it is coarsely crystalline, part of it is very crinoidal, and it contains large numbers of Bryozoa, corals; and"Stromatopora, .4-- see ee oa: 5 ° 28 2 3. Rather massive, more or less crystalline, lime- stone, which is of light-gray color, somewhat brownish as weathered; and some of it is very light gray, almost white, and is locally called a “marble.” It contains very few fossils, if any. At the center of the falls there are 11 feet of the Brassfield limestone undercut by the water of the Stream: 02 foc. aik ss a ee 16 6 23 2 2. Belfast bed*—Blue rock, which is probably argillaceous and rather sandy, with layers from 2 to 6 inches thick and the average about 3 inches. This zone was thought by Dr. Foerste t The writer understands Dr. Foerste to now refer the Belfast bed to the Rich- mond, which is also the opinion of Dr. E. R. Cumings, based on a study of its Bryozoa. Dr. W. H. Shideler, however, has found certain Brachiopods in the Bel- fast which he thinks allies it with the Brassfield. THE NIAGARAN FORMATIONS OF WESTERN OHIO — 337 : TOTAL THICKNESS GRTIORNESS No. Feet Inches Feet Inches to represent the Belfast, since he wrote, “If they [the layers] represent the Belfast bed of more eastern sections, as is believed to be the case, they certainly have changed considerably from the typical form of the rock.”?......... 2 8 6 8 1. Richmond formation—Rather thin-bedded blue rock to shaly layers, perhaps with sandy to calcareous composition. This zone extends to water level and blue shale is washed out of a pit that has been dug still deeper by the water. About 5 feet of this zone are shown in the bank Peomechersouthermusider cry cet a2 ei) eye 4 ° 4 ° The foregoing section gives 26 feet 7 inches for the thickness of the Brassfield limestone. This agrees fairly well with the estimate based upon the thickness of the Brassfield on the northern bank at the falls and the section in the Big Four Railway cut west of the station on the southern side of the creek. Mr. W. Z. Miller, my assistant, made the top of the Brassfield limestone in the railway cut about 7 feet higher than the top of the ledge on the northern bank of the creek, which gave 284 feet for the total thickness of the Brassfield limestone on Ludlow Creek. The general section of Ludlow Creek is continued by the section of the western wall of the Colonel Samuel B. Smith quarry and the bank above it, below the house of Patrick Gallagher. SECTION OF WESTERN WALL OF THE SMITH QUARRY THICKNESS ee No. Feet. Inches Feet Inches 26. Laurel limestone-——Top of bank just below house of Mr. Patrick Gallagher. Light- to bluish-gray rock in fairly even layers varying from 2 to 5 inches in thickness. The upper weathered ones are rather buff and finely OOKOUS Pen Merwe rare Arline eaayd cua vote: om Reelerstots 3 5 PE So) 25. Partly covered interval. Light- to brownish- gray, rather thin-bedded, dolomite .......... 2 8 21 5 24. Osgood beds.—Partly covered zone; but at top bluish-gray shale to shaly limestone......... I 3 18 9 t Journal of the Cincinnati Society of Natural History, XVIII (1896), 182. 338 _ CHARLES S. PROSSER arcewess py ZORA No. Feet Inches Feet Inches 23. Brownish- to bluish-gray, thin layer, from 1 to 2 inches thick 4}: 3)) {22 3acee eee ee z= 17 6 22. Bluish-gray shaly limestone to shale, from 6 to 7 inches thick. A specimen of Leptaena rhom- boidalis (Wilckens) was noted.............0.. 6S ==) iy, 4 21. Dayton limestone—Top layer of western bank of the Colonel Samuel B. Smith quarry. Light gray, with rusty spots due to iron pyrite, from tolH 3 ‘to 4*inches' thick’... 7 ie aes eee tirt ok se 352% #16 n0 go. Layer 2== inches thick? 13 seer eer ese. a= | go 63 19. Layer from 3 to 4 inches thick with dark-gray blotches inatshipper partes eee. 33 16 Ae 18. Dark-gray irregular bands in the lighter colored rock. It contains iron pyrite and varies from 4, to5 inches in-thickmess eve ats. ashes ists 45 16 I 17. Light-gray rock without much dark color, which contains some iron pyrite and varies from 5 to.6 inchésam: thickness tae see ake 5 8 16. Three-inch layer at bottom of more or less massive zone (Nos. 16 to 21, inclusive), with average thickness of 1 foot 10 inches at top of tl quarry wall on the western and northern sides. . 3 15 3 15. Two rather compact layers, the upper one 4 inches and the lower, 6 inches thick......... ike) 15 ° 14. This layer will split up into thinner ones and it contains iron pyrite 2's fc. tee ae eh 9 14 2 73. ‘Compact layer =: 2 3: aes reese ee 4 13 5 12. This layer may split up to some extent on weathering sy iat aes ne eee ee oe 9 13 I 11. This layer on weathering splits into various layers from’ sto’ 5inches thick = ( tO 17 3 15. Light-gray to bluish-gray, compact layer... .. 5 15 5 14. Thin-bedded, light- to bluish-gray zone, with irregular bedding planes..’:.-5y. cas eerer = 10 rs 13. Compact, light- to bluish-gray layer......... 5 14 2 12. Rather /shaly layer iio... Sage seria A 5 13 9 11. Two compact, light-gray layers, the upper 9 and. the lower 8 inches thick< 2 act jcoesn esc I 5 13 10. Light-gray, shaly dolomite and blue shale... . 6 TE) eae g. Thin-bedded dolomite splitting into 3 or 4 laverss Wiis atelier eter tetors = I 4 II 5 THICKNESS on nN 20 9 Clave sa) Had fas Bicia ines Se See eee eter I 8 ie) I . Rather coarse, blocky blue shale which forms the lower part of the shale zone. Nos. 7 and 8 constitute the shale zone in the upper part of the Osgood beds with a thickness of 2 feet 2 INCHES so elave oc ok DNR RIN a tee ee 6 8 5 a | THE NIAGARAN FORMATIONS OF WESTERN OHIO — 343 8) THICKNESS Teacentse No. Feet Inches Feet Inches 6. Dayton limestone.—Massive zone at top of Day- ton limestone with a thickness of 2 feet 3 inches, which splits into 5 layers and the first three in descending order are from 4 to 5 inches each in thickness, while the fourth from the top is 8 inches and the fifth 4 inches thick. Light-gray with blotches and spots of dark-gray, compact rock with stylolites or hackletooth structur2 at thelbeddinepplameseiey. seein ces sce. et eiele 2 3 Te 5. Layer of shaly, greenish limestone........... 10 5 8 4. Compact layer of light-gray color with some ‘dark-gray blotches containing iron pyrite and Call CHC HEE y aU ete te oars Wig Mes I ° Li Sie) 3. Shale to shaly limestone parting............ 2 LO 2. Dark- to light-gray limestone which splits into Severdlulavenst asin iment aytiki ed we 3 2 3 8 1. To water level in old quarry pit............. 6 6 In the foregoing section, Nos. 7 and 8 correspond to the shale zone (No.7) of the Ehlers quarry and form the upper member of the Osgood beds. The Laurel limestone comprises Nos. 9 to 16, inclusive, with a total thickness of 7 feet 2 inches. The extreme upper part of this quarry wall shows nearly 3 feet of light-gray fossiliferous rock which is referred to the Springfield dolomite. Professor Bownocker has recently published a bulletin on the “Building Stones of Ohio’’ which contains a section of this quarry.* In this bulletin the zone called the ‘‘ West Union limestone” corre- sponds to the “‘ Mottled zone,”’ No. 17 of the foregoing section, and the upper limestone of the ‘“‘Osgood beds” is what the writer is correlating with the Laurel limestone of Indiana, and includes Nos. 9 to 16, inclusive, of the foregoing section. The blue shales overlying the Dayton limestone of Professor Bownocker’s section correspond to Nos. 7 and 8 of the writer’s section, the top of which he regards as corresponding to the top of the Osgood beds of Indiana. GENERAL SECTION ALONG LUDLOW CREEK A general section of the formations exposed at Ludlow Falls and in the series of quarries on the northern bank of the stream has been t Geological Survey of Ohio, 4th Ser., Bull. 18 (1915), Pp. 37- 344 CHARLES S. PROSSER compiled from the separate sections described above, which shows the formations from the upper part of the Richmond in the Ordo- vician system to the lower part of the Springfield dolomite in the Silurian system. Some of the zones or formations vary in thick- ness in different outcrops, in which case the variation in thickness has been given. This necessarily causes a variation in the entries in the column of total thickness. GENERAL SECTION ALONG LUDLOW, CREEK Thickness of Zone r : Feast Gopaet re ues or Puneet Names of Series and Formations Total Thickness sat toi6r” old ecee ee hy’ | Springfield dolomite (only lower part shown) Niagaran Series | Mottled zone nm tae ~ 7h Laurel limestone , | 42)’ to 45y%" | Top of Osgood beds THY’ to 3%’ Shale zone | Aad toast [ee ee ieee oe Dayion limestone II yy’ Base of Osgood beds Oswegan Series 26} to 28}’ Brassfield limestone a? RAS 8 ee ee | Cincinnatian Series 23’ Belfast bed at top of Richmond formation , Ol: 1. 2 pibexcee namie se A mere outline of the classification of the formations of the Niagaran series along the Ohio-Indiana state line was published by the writer on April 20, 1915, and the Brassfield limestone was given in the Oswegan series.” Correlation of the Brassfield limestone—The limestone in the foregoing sections, which is called the Brassfield, is the one which in Ohio has generally been called the Clinton and correlated with the well-known New York formation of that name, which forms the * Outlines of Field Trips in Geology for Central Ohio, The College Book Store, Columbus, Ohio, p. 18. 2 Ibid. THE NIAGARAN FORMATIONS OF WESTERN OHIO — 345 basal part of the Niagaran series of that state. As early as 18096 Dr. Foerste stated that— The identity between the Clinton faunae of the two states [Ohio and New York] on closer examination is not found to be sc close as at first supposed. Whether this is due to geographical causes, the Clinton of New York being more litoral, or whether it is due to moderate differences of horizon, can not be told until the Clinton of New York is much more closely studied. Although I have been accustomed to call the Ohio formation the Clinton, yet I should be willing to recognize the fact that the identity is not very marked, by giving it a name of its own, for instance, the Montgomery formation, on account of its typical development in Montgomery County, in Ohio.t In 1906 Dr. Foerste proposed the name Brassfield formation for this limestone from outcrops ‘“‘along the Louisville and Atlantic Railroad, between Brassfield and Panola, in Madison County,” Kentucky.? It was stated that “for the . . . . limestone section at the base of the Niagaran division of the Silurian, hitherto identi- fied with the Clinton of New York, the name Brassfield limestone is proposed.’ After listing the fauna of the Brassfield limestone in Kentucky, Ohio, and Indiana, and noting the absence in it of certain charac- teristic Brachiopods of the New York Clinton, Dr. Foerste wrote as follows: The identification of the Brassfield limestone of Kentucky, and of its northern extension in Ohio and Indiana, in former years, with the Clinton limestone of New York, rests rather upon a somewhat similar facies of the two faunas, and upon the general absence of the more typical species of the Rochester shale fauna of New York in these limestones at the base of the Silurian in Ohio, Indiana, and Kentucky, than upon the presence of any con- siderable number of species common to both areas. On closer inspection, the fauna of the Brassfield limestone of Ohio, Indiana, and Kentucky appears to differ sufficiently from the fauna of the Clinton limestone of New York to warrant the assumption of the presence of some sort of barrier between these two areas.4 Dr. Foerste has also stated in a later publication that “the Brassfield limestone is the southern continuation of the strata which were identified in Ohio, by Professor Orton, as Clinton.’’s t Journal of the Cincinnati Society of Natural History, XVIII, 189. 2 Kentucky Geological Survey, Bull. 7, p. 27. 3 [bid., p. 18. 4 [bid., p. 35. 5 Journal of the Cincinnati Society of Natural History, XXI (September, 1909), I. 340 CHARLES S. PROSSER e | At the 1912 meeting of the Geological Society of America, Pro- fessor Charles Schuchert proposed the Cataract formation: ‘‘a new formation at the base of the Siluric in Ontario and New York,” from a locality called the Cataract in the Credit River region of Ontario, 48 miles northwest of Toronto.' In August of the same year Professor William A. Parks in describing “The Palaeozoic section at Hamilton, Ontario,” stated that ‘“‘a new formation—the Cataract— . . . . represents an invasion from the north and west at the commencement of Silurian time. The upper limestones and shales of this formation are highly fossiliferous and present a fauna comparable with that of the Brassfield formation of Ohio and Kentucky.’” Dr. Merton Y. Williams described a series of sections in the Niagara escarpment of Ontario in a paper before the Geological Society of America in December, 1913, in which he reported that “the Medina sandstones of Niagara gorge (125 feet thick) are represented farther north by dolomite and shales (Cataract forma- tion).’’s An article by Dr. Kindle on *‘What Does the Medina Sandstone of the Niagara Section Include ?”’ contains the following sentence: “The examination by the writer of a number of sections holding this fauna [Cataract] in connection with a review of the Niagara section has convinced him that all of the terranes associated with the Cataract fauna are represented in the Medina of the Niagara section.’ In a later and exhaustive paper on the “‘ Medina and Cataract Formations of the Siluric of New York and Ontario,”’ Professor Schuchert shows the close relationship of the Brassfield fauna to that of the Cataract formation of Ontario and also “that the Cataract is a close correlate with the Medina” formation of New York. In another place is the statement that ‘in other words, the t Bulletin of the Geological Society of America, XXIV (March, 1913), 107. 2 Guide Book No. 4 (Twelfth International Congress of Geologists), “Excursions in Southwestern Ontario,” B;, p. 128. 3 Bulletin of the Geological Soctety of America, XXV (March, rgr4), 40. 4 Science, N.S., NXCXTX (June ro, ror4), 918. S Bulletin of the Geological Society of America, XXV, (September, 1914) 291. THE NIAGARAN FORMATIONS OF WESTERN OHIO = 347 typical Medina formation shades through lateral alteration into the typical Cataract.’ This apparently agrees with the idea expressed by Professor R. Zuber, of the University of Lemberg, on the escarpment at Hamil- ton, Ontario, in August, 1913,’when he said that the Medina and Cataract appeared to him to be different facies of the same forma- tion. Concerning the relation to the Brassfield, Professor Schuchert wrote: The Cataract may also be compared with the Brassfield formation of Ohio and Indiana, as the two are clearly related, and also both are of a limestone facies. The former has 76 species and the latter 140. Between the two there are 24 forms in common... . . When the two biotas are finally carefully compared with each other, there will undoubtedly be added more significant forms strengthening the view that the Cataract and Brassfield are fairly close correlates in time. However, as these two faunas are not of the same epicon- tinental basin, one cannot expect a large percentage of the forms to be common to both; the Brassfield element came in from the Gulf of Mexico region, while the Cataract migrated into Ontario through the Gulf cf St. Lawrence embay- ment across the Province of Quebec or came in from the Arctic A little later Dr. M. Y. Williams, in his article on the ‘‘Stra- tigraphy of the Niagara Escarpment of Southwestern Ontario,” has stated that ‘“‘Medina is used in the sense in which Grabau has redefined the term, that is, to include the beds above the (Queenstown shale and below the Clinton formation. It is extended, however, laterally to include the Cataract formation as defined by Scehuchert.:3 The Medina sandstone underlies the Clinton beds of New York and is not included in the Niagaran series, but is the upper forma- tion of the Oswegan series as classified by the New York Geological Survey. Therefore, if the correlation reviewed above be accepted, then the Brassfield limestone (formerly called Clinton) of Kentucky, Ohio, and Indiana is to be transferred from the Niagaran to the Oswegan series of the Silurian system. Furthermore, Professor T. E. Savage believes that in the Mis- sissippi Valley the Sexton Creek limestone “‘represents about the t [bid., p. 204. 2 Thid., p. 291. 3 Summary Report of the Geological Survey [Canada] for the Calendar Year 1913 (1914), p. 182. 348 CHARLES S. PROSSER same general period of deposition as the Brassfield limestone.”* The Sexton Creek limestone is the upper formation of the Alexan- drian series, named and described by Professor Savage,” a series that in Illinois and Missouri contains all the formations between the Richmond-Maquoketa formation, at the top of the Cincinnatian series, and the base of the Niagaran series. Since the above was written, advance pages of a work on Historical Geology by Professor Schuchert have been received in which the following correlation appears: Lower Silurian (Medina, Cataract, and Brassfield formations. or Oswegan \ Becsie limestone.3 OTHER SECTIONS OF WESTERN OHIO Sections farther up the Stillwater River toward Covington show the middle and upper parts of the section exposed along Ludlow Creek, while those in Covington carry the general section still higher. Sections farther west, near Lewisburg and New Paris, agree essentially with those of the Stillwater Valley. Sections in and near Covington——About two miles south of Covington is the Jackson Stone Co. quarry, near the Stillwater River, on the Charles H. Jackson farm. It is easily reached by the Piqua, Covington, and Dayton trolley line, leaving the car at stop 45, which has the name of Sugar Grove. The section given below is of the eastern wall of the quarry, the top of it near the engine house, a short distance southwest of the crusher. SECTION OF THE JACKSON STONE CoO. QUARRY THICKNESS eee No. Feet Inches Feet Inches 22. Cedarville dolomite-—Buff, mostly porous, crys- talline dolomite, with Pentamerus oblongus Sowerby common all through the outcrop. Zone at base with large number of specimens of Pentamerus oblongus Sowerby as well as Favosites niagarensis Hall. ...............-. 3 8 109 9 I Illinois State Geological Survey, Bull. 23 (1913), Pp. 33. 2 American Journal of Science, 4th Ser., XXV (1908), 434, 443; Illinois State Geological Survey, Bull. 23, pp. 14, 15. 3 A Text-book of Geology, Part II, Historical Geology (1915), p. 661. No. 21. 20. IQ. 18. 17e 16. 4. THE NIAGARAN FORMATIONS OF WESTERN OHIO Springfield dolomite.-—Buff-colored, rather thin, layers of even-bedded dolomite. The layers vary in thickness from 2 to 8 inches and the majority of the layers are probably from 4 to 6 inches thick. No one of the exposed layers is blue. Fossils, as Pentamerus oblongus Sowerby, and corals are common and certain layers con- tain numerous specimens of Pentamerus ob- longus. ‘This zone, with its thin even beds of buff color, has clearly the lithologic appearance of the Springheld dolomite: ) 5255.5... 05. 242 Bluish-gray, weathering to a buff color, rather compact, slightly porous, dolomite which gener- allyssplitswmto\twolayersiac:2c).)..). 2.00 Jeo) Bluish-gray, somewhat mottled, layer which Containssbut tewstossilstmens qc. .s seo Bluish-gray layer, fairly compact, in part sub- crystalline, with some small holes and contain- ing large numbers of Pentamerus oblongus Sowerby. On account of the large number of fossils this zone may be called a Pentamerus layer and clearly belongs in the Springfield GKENGICAMIYS 5c, Hide Gite ih ain GuSeNS Ieee Neer A Motiled zone.—Massive layer of bluish-gray dolomite marked with large, irregular-shaped spots of light-gray color, so that the entire sur- face has a coarsely mottled color. It has a porous structure with medium-sized cavities. It contains some fossils, as, for example, cup - corals and crinoids, with an occasional specimen of Peniamerus. At the base is a stylolites part- Laurel limestone.—Lithologically like lower layers, thickness varying from 9 to ro inches, more porous in upper 4 inches, with small and larger holes. Some fossils, as a cup coral and LEG HOTIGLUSG 9.5 Se tii es Sa Re MRE RE Eee . Bluish-gray, somewhat porous, limestone. A CUprcoralpwasimotediyy jr mera my. 5.0 rae Layer of light-gray mottled with dark-gray, fairly compact, limestone, varying in thickness HOM PA VLOMTGh IM CHES ewes vrene sre toe ciscccs eich THICKNESS Feet Inches Io 6 I 2 105 Io ii 2 1 oz= Io I z+ 349 TOTAL THICKNESS Feet Inches 106 I 95 Uf 94 63 Q3) aS 92 10 85 8 84 10% 84 2 350 CHARLES S. PROSSER " TorTraL THICKNESS THICKNESS ‘No. : Feet Inches Feet Inches 13. Light-gray limestone in 4 layers, with shaly partings, varying in thickness from 19 to 21 inches. Lithologically similar to subjacent layers . hc. iS bak cnievec Rae eee epee I 8+ 83 ° 12. Light-gray, with dark-gray spots and blotches, compact limestone, which is harder than lower layers and varying in thickness from 12} to 13 INCHES «y's lace bak We AI tee i fu: = Sr 32 11. Gray, weathering to a brownish color, gritty shale, about’ r.inch: thick ino. aot sateen os. x 1 = 80 3 ro. Layer lithologically about the same as the sub- jacent OME. < achcos ccc’ ee eRe et ede 491°" an Ww Io (os) fon > On Bate Light-gray, compact course.........-- Ga 12. Il. THE NIAGARAN FORMATIONS OF WESTERN OHIO base of the mottled zone and the top of the Osgood shale is about the same in all three sec- TSR TERED Wislentiste/aielelansnlc Laurel limestone-——The following courses all have a bluish-gray color when seen on fresh surface: Light-gray, compact course.......---- Aa Light-gray course, the upper part blotched with dark-gray spots, but not many in lower part........----+---- 8 Shaly limestone parting. ......------- <-1 Light-gray, compact layer.........--- 8 Light-gray course, with dark-gray spots and bands which tend to split into 2 AWA so nig bb babiooeeesosocdeaenMe avr ats) ight-oray, layenun ie. « aoa) 9-92 Very mottled light- and dark-gray, thin ANE 6 Soup des bavbansesnowooocueOn Ze Light-gray layer..........-----++-+:- ie) Similar to above layer with a broken parting at base.........--++----+--> 8 Bluish-gray layer with dark-gray streaks HAG NOUS, obeccseesaaecseneecocoous 6 The total thickness of the above layers varies from s feet 103 inches to 6 feet 2 inches. The three following layers are of light-gray color with dark-gray spots, blotches, and streaks and have about the same lithologic appearance: Feet Inches Idle JERE oo oo Kaboom ea don oO oUOK I 3 Secondilayerme aes] oe ie) Compact, massive layer containing calcite) crystalsy. 222.522 ---5-5---- 2 I Osgood beds.—Bluish-gray, soft shale which forms the floor of the quarry. Six inches or more are shown in the quarry. Mr. Robert Mollett, foreman of the quarry, stated that at the time of the March flood of 1913 the shale was shown by the side of the railroad track THICKNESS Feet Inches 4 I §=5 © 4 2 Syl TOTAL THICKNESS Feet Inches 33 9 29 8 23 8 Io. Q. wy Os below the crusher and varies in thickness from 2 feet 6 inches to ¢ feet. cosh eens tins Dayton limestone.—Mr. Mollett stated that the upper surface is somewhat uneven and that the limestone extends about 1 foot higher than the top of the exposed ledge on the south bank of Twin Creek below the Lewisburg Stone Co. crusher. Upper foot, according to Mr. Mollett, not ex- POSE. ck eee ee Light-gray layer, very rusty colored from weathered iron pyrite. Varies in thickness from ‘r ‘to:ak INChes ones eee ee Mostly light-gray, thin-bedded to shaly, lime- C1 KO) 0 RE a TE ies cho Choe wank ce eh AG een c . Thin layer of more or less crystalline structure, Which:containS 1OSSIIS wa eee eee eee . Bluish-gray, thin-bedded layers, weathering to a very light-gray or whitish color. Rarely thin, somewhat irregular, finely crystalline layers in which is an occasional fossil. ............... . This zone will split into 3 layers. ‘The upper one contains much pyrite and has weathered in spots to a very rusty color. The middle and lower parts of light-gray color with spots and irregular layers of dark-gray color from iron PYTICEN Fe CRs RS Pe ees . Light-gray to bluish-gray, rather thin-bedded layers on bank of creek, with slightly glistening surface. Upper part of zone contains imperfect BrachiGbods scis..5% Shacsa ao eee . Light-gray, compact layer which is harder than rock above. Base at creek level on Septem- BOP TOG Sh oon 5 x ical v4 an te . Two thin layers of compact, bluish-gray rock, the upper one 2 inches and the lower one 23 ches thiee s. 4 vou. tess eee Rte ew igs . Brassfield limestone —The upper surface rough. Mottled pink and gray crystalline limestone in bed of creek, just below the bank of Dayton limestone. In the bed of the creek 13 inches CHARLES S, PROSSER THICKNESS Feet to Inches Io ~] bol TOTraL THICKNESS Feet Inches 19 6 16 re) 15.) "Gee ee 14 867 14 oe wn” nm role ws bol THE NIAGARAN FORMATIONS OF WESTERN OHIO 359 THICKNESS cae No. Feet Inches Feet Inches of Brassfield limestone was measured; but it extends on down the creek and a quarryman stated that about 8 feet is shown along the CLEC Kite oe nn Lea eaR agen ess) oe 2 otmanal 8= oO 8 ° A view of the southern wall of the Lewisburg Stone Co. quarry is shown in Fig. 2. The lower part of the wall is the Laurel even- bedded limestone, the top of which Mr. Cottingham is indicating by the hammer, above which is the conspicuous ‘“‘mottled zone,” and above this zone is the Springfield dolomite: In the foregoing section the Laurel limestone is called the “blue building stone” by the quarrymen and comprises Zones 12 and 13 with a thickness of about 1o feet 2 inches. A section at a different part of the southern quarry wall gave a thickness of g feet 10 inches. It is well shown in the picture of the southern wall of this quarry (Fig. 2) where Mr. Cottingham is marking its top with the hammer. Samples of the Laurel limestone from the southern wall of this quarry were analyzed by Professor Demorest with the following result: Silicious Residue | Fe.0; and Al.0; CaCO; MgCo, 8.85 1.20 56.00 BO) Ot The massive ‘“‘mottled zone,” with a thickness of 4 feet 1 inch, immediately above the Laurel limestone, is also well shown in Fig. 2. Samples from this zone in the southern wall of the quarry were also analyzed by Professor Demorest with the following result: Silicious Residue | Fe.O; and Al.O3 CaCO; MgCO; ary 0.75 BA pL2 42.13 2.55 0.80 54.12 42.31 The following analysis by Professor Demorest of samples of the West Union limestone from Sproull Ravine, about 14 miles 360 CHARLES S. PROSSER northeast of Duncanville and 73 miles northeast of West Union, Adams County, is given for comparison with that of the ‘‘mottled zone:”’ SiO, Fe:O; and ALO, CaCO, MgCO, 20.24 | 4.05 44.67 290.14 It will be seen from these analyses that the West Union is a much more silicious rock than the “‘mottled .zone”’ and that the latter is a dolomite. It is to be noted that the chemical composi- tion and lithologic character of the ‘‘mottled zone” differ consider- ably from those of the West Union limestone in its typical region. The rock between the first and second cap rocks of the quarry- men is called by them the ‘buff building stone” and corresponds to Zones r5 and 16 of the last-given section, all of which evidently belongs in the Springfield dolomite. All the rock above the shale zone (No. 11) of the Osgood beds is quarried and crushed for concrete and road material. The fine rock, which the men call “‘sand,” binds well on the roads, and it was stated that the entire product of the quarry for 1914 was used on the Ohio roads by the State Highway Commissioner. Dr. Foerste some years ago published a brief description of the Weaver quarries, located on the northern side of Twin Creek, oppo- site the eastern part of the Lewisburg Stone Co. quarry. Recently Professor Bownocker has published a section of the Lewisburg quarry in which the upper limestone of the Osgood beds with a thickness of 9 feet 11 inches corresponds to the Laurel limestone of the last-given section.? The 3 feet of blue clay beneath is the Osgood shale and the subjacent 1o feet of ‘‘blue-gray limestone” the Dayton. The ‘‘West Union limestone,” 43 feet thick,> corre- sponds to the ‘‘mottled zone” of the last-given section, overlying which is the Springfield with a thickness of 8 feet and then the Cedarville which forms the highest part of the quarry. Apparently the line of division between the Springfield and Cedarville dolomites * Journal of the Cincinnati Society of Natural History, XVIII (18096), 183, 184. 2 Geological Survey of Ohio, 4th Ser., Bull. 18 (1915), p. 40. 3 Ibid., p. 39. 361 THE NIAGARAN FORMATIONS OF WESTERN OHIO snonoidsuod ay} St sUOJSOWT Janey] oy} 2 qied IMO] ay, “OIyO ‘SInqstMo'T Jo JSoMTIIOU oTTU ou AOqY ‘ayWMOTOP ppeysurds ay} st yTyM SutAj19A0 ‘IoWUIeYy 94} AQ poyYIeUr st YOIyM Jo do} ay} ‘ ,¢9U0Z poa]}jOUr,, QUO}SIUN] JOINe'T IY} SI ‘Arzenb °07) 9u01¢ SanqsimeT 9 O ]]/eM WINyIN0G—*z “OIF ) S qSIMa'T 9} Jo T {INOS A 362 CHARLES S. PROSSER is drawn at the same horizon as the top of Zone 16 in the writer’s section, which gives a thickness of 8 feet 7 inches for the Spring- field or 8 feet as measured by Professor Bownocker. SECTION NEAR LAUREL, INDIANA In the foregoing sections the correlation of the terranes referred to the Osgood beds and the Laurel limestone, both of which were named by Dr. Foerste,* was decided upon after visiting Laurel, Indiana, and studying some of his sections in that typical region. A section at one of these localities, in a somewhat condensed form, is given below. The section is on the bank of a stream at a locality known as Derbyshire Falls, on the C. J. Valkenburg farm, nearly 3 miles southwest of Laurel and some 47 miles southwest of the Lewisburg Stone Co. quarry. A section of the Laurel limestone, Osgood beds, and Clinton limestone measured at this locality and the Lower Derbyshire Falls was published by Dr. Foerste in 1898.? The measurements in the following section are those of the writer and his assistant, Mr. Kenneth Cottingham; but the classification is in accordance with that of Dr. Foerste, except where differences are noted: DERBYSHIRE FALLS SECTION THICKNESS Tee No. Feet Inches Feet Inches 18. Laurel limestone—This limestone is shown in the old quarry just across the quarry road to the south of Derbyshire Falls and this zone extends to the top of the quarry wall. It is light gray, as weathered, rather thin-bedded, the layers varying from 2 to 8 inches in thick- ness. There are also at least 3 chert layers ranging from 1 to 3 inches in thickness....... 4 7 52 ° * Osgood beds: Indiana Department of Geology and Natural Resources, 21st Annual Report (1897), pp. 217, 227-29. Laurel limestone: Journal of the Cincinnati Society of Natural History, XVIII (February, 1896), pp. 190, 191, and Indiana Department of Geology and Natural Resources, 21st Annual Report (1897), pp. 217, 230, 231. 2 Indiana Department of Geology and Natural Resources, 22d Annual Report (1808), pp. 244, 245. An illustration of Derbyshire Falls is given on Pl. XVI, which faces Pp. 244: No. i 16. is, 14. 13. 2. If. Io. THE NIAGARAN FORMATIONS OF WESTERN OHIO Thicker layers of compact limestone, light gray to buff, when weathered, with shaly partings. The majority of the layers are perhaps 3 to 5__ inches in thickness; but there are thicker ones which apparently range from 8 to 14 inches. No chert was noticed in this zone........... ?Top of Osgood beds.—Buff, compact, 8-inch layer at top of ledge on south side of falls, which is apparently the one across the quarry road in the base of the old quarry at the spring. The top of this layer is apparently the horizon where Dr. Foerste has drawn the line of separation be- tween the Osgood beds and Laurel limestone. Lithologic characters, however, in the vicinity of Laurel would apparently favor classing it withthe Iaurel limestones. ....2..-..--..< Blue, argillaceous, soft shale or clay. This is the blue-clay stratum of Dr. Foerste......... Shaly, light-gray limestone................ Light-gray, compact, even-bedded limestone; some of the bedding planes rather rough. The layers vary in thickness from 5 to g inches, and perhaps the majority of them average about 8 inches. This is the Lower Quarry or Osgood rock of Dr. Foerste, and is apparently the con- tinuation of the Dayton limestone in Indiana Brassfield limestone.—Light-gray, crystalline limestone. Apparently the upper foot and 3 inches of this zone was regarded by Dr. Foerste as a “‘doubtful horizon: White Clinton or base ofiNiagara rocks? Gin en haere Payee ees 2 Crystalline gray to pinkish limestone. It is very irregularly bedded and contains pyrite, so that it is frequently rusty colored on the weathered surface. Dr. Foerste’s section re- ports ‘‘Clinton; 7 feet 6 inches; reddish”’.... Richmond formation.—Light-gray, impure lime- stone with portions that are darker colored... . Gray, impure limestone, the upper layer a foot thick separated by a shaly parting from a lower layer of similar limestone, 1 foot 3 inches thick THICKNESS Feet Inches 5 3 8 I 6 4 Oa 2 3 Oly I 6 2 3 363 TOTAL THICKNESS Feet Inches 47 5 42 2 41 6 40 fo} fo ain, 33 4 Bia I 24 6 23 364 CHARLES S. PROSSER THICKNESS Tee No. Feet Inches Feet Inches 8. One layer of gray, massive, impure limestone. No fossils seen, 4 6... Ga Oe Sertraline cx 4 fo) 20 fe) 7. Layer similar to one above. .iie¢...60..55.- 2 2 16 9 6. Massive, light-gray layer with dark-gray spots. No fossils noted).<\/c. See ree eios tak ee 3 ° 14 7 5. Shale parting GO. > Jin meee mere eres 3. I II 7 4. Grayish, somewhat crystalline, limestone which tends to split into thinner layers............ ) Il 6 3. Bluish shales alternating with gray, fossilifer- GUS LIMeOSEOME :s:/c5-.icicrs ete eI ecdeos FR 2 3 ate) 9 2. Grayish, somewhat crystalline, limestone, hard and ‘very fossiliferotiss eee os ee. SS cals 5 4 8 6 1. Grayish to bluish shales which are not very fossiliferous.< (Foot or falls es yoatce es cae a 3 2 3 2 In the foregoing section Zones 11 and 12, with a thickness of 8 feet ro inches, have been classed together and considered the western continuation of the Brassfield limestone of Ohio. Zones 13 and 14 of light-gray limestone with a thickness of 6 feet 8 inches are considered the western continuation of the Dayton limestone of Ohio. Dr. Foerste has stated that ‘in Ohio Pentamerus oblongus occurs in the Dayton limestone, equivalent to the base of the Osgood bed.’’* The soft blue shale or clay of Zone 15 is believed to correspond to the blue shale of Zone 11 in the Lewisburg Stone Co. quarry and the shale at the same horizon in the various quarries along the Stillwater River. As stated above in the description of the section, the lithologic break occurs at the top of this shale, which appears to the writer from the sections which he has studied to be the horizon where he would draw the line of division between the Osgood beds and Laurel limestone. If the 8-inch layer of compact, buff limestone (No. 16) immediately above the soft blue shale zone be classed with the Laurel limestone, then ro} feet of it are shown in the wall of the old quarry on the bank above and south of the falls. It is believed to be the eastern continuation of this limestone which makes Zones 12 and 13 with a thickness of ro feet 2 inches in the Lewisburg Stone Co. quarry and the t American Journal of Science, 4th Ser., XVIII (1904), 341. THE NIAGARAN FORMATIONS OF WESTERN OHIO — 365 limestone which has been called the Laurel in the sections farther east along the Stillwater River. Samples of the Laurel limestone were collected at the quarry above Derbyshire Falls and analyzed Bx: Professor Demorest with the following result: SiO. FeO; and ALO; CaCO; MgCo; 17.84 I.00 47.89 31.54 This analysis shows that the Laurel limestone at Derbyshire Falls is a more silicious one than that at the Lewisburg Stone Co. quarry , which contains but 8-85 of silicious residue. On the other hand, the Lewisburg stone contains a larger percentage of CaCO,, where it amounts to 56 per cent of the rock; the other constituents at the two localities do not differ to any marked degree. ORIGIN OF THE LYMAN SCHISTS OF NEW HAMPSHIRE FREDERIC H. LAHEE Massachusetts Institute of Technology TABLE OF CONTENTS : PAGE INTRODUCTION. , 3 % « SEE Cs 4 os Ne SUMMARY oo o) 0) QpteRUee eens A 3. | ll TEE LISBON AND LYMANISERIES (7705 5 450. 6) 4s 4 2 GENERAL DISTRIBUTION OF THE LYMAN SERIES... . . . . 369 Frecp Retations AND MerGAscopic DESCRIPTIONS OF THE LYMAN SERIES The Young’s' Pond Locality." ek ws Le The Mormon: Bill Locality “ark Ge kk The Parker Hull Locality. "cage ehh) ss Microscopic DESCRIPTIONS Porphyry Schist of the Young’s Pond Locality . «4 Pebbles of the Conglomerate Schist of the Young’s Pond Locality . 374 oy Parker Hill White Schist . SAdss ey ek ioe NS Goya 5 Parker Hill Dark Schist #S) si ixie ees he ae) cs 8s a a GENERAL CONCLUSIONS ON THE ORIGIN OF THE LYMAN SCHISTS Porphyry Schist of the Young’s Pond Locality . . ... . . 377 Patker Fill “White Schist ~) “2 cite ee x Conglomerate Schist of the Young’s Pond Locality . . . . . 3790 STRUCTURAL RELATIONS OF THE LYMAN SERIES... . . . . 380 Acipic ErFrusIve Rocks EAst OF THE APPALACHIAN PROTAXIS . . . 380 INTRODUCTION During the summers of tort and 1912, through the generosity of Mr. R. W. Sayles, I made a geological investigation of part of the townships of Littleton, Lisbon, Lyman, Bethlehem, Bath, and Landati, in New Hampshire. In the fifteen weeks devoted to this work about 250 square miles were examined. The results that have been published' have reference to an area of only eight or nine square miles (Fig. 1), in the study of which three of the fifteen «F. H. Lahee, “Geology of the New Fossiliferous Horizon and the Underlying Rocks, in Littleton, New Hampshire,” Am. Jour. Sci. (4), XXXVI (1913), 231-50. 3 60 THE LYMAN SCHISTS OF NEW HAMPSHIRE 367 weeks were consumed. ‘The survey of the rest of the area amounted to a reconnaissance of intricately metamorphosed schists, regarding which final conclusions were deferred until more could be learned. Having had opportunity to revisit some parts of the field where the “Lyman schists’”’ are exposed and having examined a suite of thin sections of these schists, I submit the present paper as a second chapter on the geology of the ‘“‘Ammonoosuc District.” Fic. 1.—Index map of the localities mentioned in the text. The shaded rec- tangular area. is the strip of country described in the article on the Littleton fossiliferous horizon. 1, Blueberry Mountain. 2, Bald Hill. 3, Young’s Pond. 4, Partridge Lake. 5, Parker Hill. 6, the “Parker Hill locality.” 7, Lyman village. 8, Black Mountain. 9, the “Black Mountain locality.” 10, Mormon Hill. 11, Lit- tleton. 12, Lisbon. Only a few roads are drawn. I am happy to express my gratitude to Mr. W. L. Whitehead, a candidate for the doctorate of science in geology at the Institute, for making the accompanying microphotographs, and to Mr. Sayles for the use of Figs. 2 and 15. SUMMARY 1. The term ‘“‘Lyman schists’? was applied by Hitchcock to a group of schists many of which are characteristically whitish on their weathered surfaces. 368 FREDERIC H. LAHEE 2. These Lyman schists are broadly exposed in the area north- west of the Blueberry Mountain—Bald Hill range and its south- westward extension. 3. Hitherto the Lyman series has been regarded as a group of metamorphosed sedimentary rocks. 4. Field evidence, megascopic examination of hand specimens, and microscopic examination of thin sections indicate that the Lyman series contains interbedded members which appear to be of volcanic origin. 5. These metamorphosed volcanic rocks include, among others, species related to the quartz keratophyres and the keratophyres, and probably also tuffs and agglomerates of similar composition. 6. It seems more reasonable to attribute the origin of the coarse conglomerate schist near Young’s Pond, Lyman, to vulcanism than to glaciation. 7. The association of acidic effusives with the Paleozoic rocks east of the main Appalachian protaxis is not exceptional for New Hampshire. Such effusives are found in Maine, in the Maritime Provinces of Canada, and in the Piedmont Plateau of our middle and southern Atlantic border states. THE LISBON AND LYMAN SERIES On the northwestern side of the Blueberry Mountain formation" are highly metamorphosed greenish and whitish schists which out- crop over many square miles. The greenish varieties are often chloritic. They and certain other associated rocks were called by Hitchcock the “Lisbon group.”’ The whitish schists belong to his ‘Lyman group.” In his earlier reports he relegated these two formations to the Huronian,? the Lyman group being con- sidered the younger member; but subsequently he referred them to the Cambrian or to the Ordovician,’ and concluded that “the Lyman schists . . . . do not represent a stratigraphical terrane”’; the term “is a petrographical designation.’” *F. H. Lahee, op. cit. 2 C. H. Hitchcock, Geology of New Hampshire (1877), I, 277. 8 Ibid.; Geology of Littleton, New Hampshire, reprint from the History of Littleton (1905), pp. IX, 20. 4 Ibid., p. 31. THE LYMAN SCHISTS OF NEW HAMPSHIRE 369 GENERAL DISTRIBUTION OF THE LYMAN SERIES No attempt will be made here to mark out the exact distri- bution of this series. It is exposed over broad areas west of the Blueberry Mountain Siluro-Devonian belt and it may be related to schists on the east of this belt. The principal localities (Fig. 1) to which reference will be made in the sequel are Mormon Hill, the valley just northwest of Young’s Pond, and Parker Hill. Lyman schists are also exposed on the hill southwest of Young’s Pond, on the lower eastern slope of Black Mountain, near Partridge Lake, and elsewhere in the broad valley between Gardner Mountain and the Mormon Hill-Parker Hill range. In all these places the group has a breadth of outcrop of several hundred feet, the rocks are schistose, the strikes are roughly northeast-southwest, and the cleavage and the bedding (where visible) have steep dips. FIELD RELATIONS AND MEGASCOPIC DESCRIPTIONS OF THE LYMAN SERIES The Young’s Pond locality—Near a small schoolhouse about one-third of a mile northwest of Young’s Pond is a large outcrop of a conglomeratic schist which we used to speak of as the ‘“school- house conglomerate” (Fig. 2). It is the rock which Mr. Sayles has recently described as possibly being a tillite.". This conglom- erate schist appears to rest unconformably upon a whitish or light- grayish porphyritic sericite schist. The contact, which is very irregular, runs northeast-southwest and may be seen not only near the schoolhouse, but also, a few hundred feet southwest, near the brook that flows into Young’s Pond. ‘The conglomerate schist is on the northwest, the porphyry schist on the southeast, of this con- tact. The conglomerate schist is between 300 and 400 feet wide across the strike, and the porphyry schist is between 150 and 250 feet wide. This statement should not be understood to imply that these rocks have no variation across their outcrop belts. There is evidence that the porphyry is interrupted at least at one horizon by a bed of fine conglomerate schist. West of the “schoolhouse conglomerate”’ is a series of whitish rocks including fine, non-porphyritic schists, fine porphyritic «R. W. Sayles, ‘‘Tillite in New Hampshire,” Science, N.S., XLI (1915), 220. 370 FREDERIC H. LAHEE schists, tine conglomerate schists in which the pebbles and paste are of similar whitish materials and the pebbles are much sheared, and schists of the texture of medium sandstones. The last- mentioned material and some of the finer conglomerate schists looked so much lke altered pyroclastics that they were called tulls in the field. (See description of the Parker Hill white schist, below.) All these rocks are exposed more than once across the strike. The porphyry schist that underlies the -“‘schoolhouse con- glomerate”’ has an exceedingly fine groundmass, which is sericitic and in places rich in chlorite. The pheno- crysts, ranging up to nearly 3 inch in diam- eter, but averaging $-i5s inch, are of quartz and plagio- clase. They are uni- formly distributed and constitute about Lyman (‘schoolhouse’) conglomerate, half of the rock. The showing obscure outlines of pebbles. The dark masses, . cee a ee ocak tas at a eee “quartz has a very high edged with black enamel, are of the fine drab schist. ; Sc oe Photograph by Mr. Sayles. luster, is somewhat | bluish and very trans- parent, and has tendency to break with a rude cleavage. The plagioclase is so fresh that the striations are distinct and the frac- ture surfaces are brightly reflecting. The shearing of the rock seems to have been localized in the groundmass. Within its own mass this schist has no bedding. The conglomerate schist (“schoolhouse conglomerate’’), also, is quite devoid of bedding. It contains many large and small fragments of the porphyry schist, together with pieces of other whitish rocks of the Lyman series, and masses of rust-brown, smooth-looking schist, which are conspicuously darker than the Lyman varieties. All the pebbles have been more or less sheared and in such a way that their longest axes are parallel with the dip of the cleavage. Some seem to have been roundish, but more Fic. 2. THE LYMAN SCHISTS OF NEW HAMPSHIRE Dia often they were obviously angular. Naturally the shearing has increased the irregularity of their shapes. The quartz-plagioclase porphyry pebbles have been deformed less than the other varieties. Their average dimensions may be expressed by the ratio 2:1.25:1. They range up to two feet in length. The dark fragments, mentioned above, are often drab-colored on the weathered surface, but dark-greenish when fresh. As will be explained presently, they are chlorite and actinolite schists. They are of medium fine grain and of uniform texture, they are markedly sheared, and their shape is very irregular, even jagged. Some were found to be ro or 12 feet long and only 1 or 2 feet wide. Others are less elongate (Fig. 2). As compared with the pebbles, the paste of this conglomerate schist is in relatively small amount. Nor does it seem to have been derived from an argillitic substance. It looks rather as if it had been fine clastic débris, of texture ranging from fine sand- stone to fine conglomerate, derived from the same sources whence came the larger pebbles. Its appearance is that of metamorphosed tufaceous material. Partly on account of the metamorphism and partly because of similarity of character, the paste is not sharply defined from the pebbles. On fresh surfaces of the rock the pebbles are distinguished with difficulty, and even on weathered exposures, where the psephitic structure is best brought out, the details are blurred. The Mormon Hill locality—On a traverse across the strikes on Mormon Hill, one passes over a series of schists much like those near Young’s Pond. ‘They are porphyries, tuff-like clastic rocks, and fine and coarse psephites, all sheared. They dip steeply north- west. In one place obscure cross-bedding indicated that the stratigraphic sequence was younger westward; and in another locality the same fact was shown by an unconformable contact between tuff-like schist on the southeast and conglomerate schist on the northwest. The series was roughly measured as having a breadth of outcrop of 800 feet. The porphyry schist of this region differs from that near Young’s Pond in lacking quartz phenocrysts. Its phenocrysts are all of 372 FREDERIC H. LAHEE plagioclase. Otherwise, in respect to size, abundance, and arrange- ment of phenocrysts and characters of the groundmass, it is mega- scopically like the type above described. The coarse psephitic rocks here contain fragments of whitish and drab schists, as at Young’s Pond. The Parker Hill locality.—Por- phyry schists, tuff-like schists, and psephite schists are here exposed in a belt several hundred feet wide. The porphyry schists are finer than the Young’s Pond type. Along the road that crosses Parker Hill between Lisbon and Lyman villages exposures near the top of the hill exhibit curi- ous relations. A dark-gray schist Fic. 3.—Rounded plagioclase (called hereafter the Parker Hill dark phenocryst surrounded by ground- schist) is associated with a whitish Tae lice ee ain se rock of the Lyman series. The latter, imperfections on the negative. which will be called the Parker Hill white schist, in several outcrops, con- stitutes a matrix in which are scattered large and small irregular blocks and strips of the dark schist, many being greatly con- torted. The phenomenon resembles intra- formational pebbles in clastic material. The white schist is of the texture of a medium sandstone. Quartz and feldspar are both abundant and the interstices between the larger grains are filled with finer particles ot quartz and feldspar, together chenéect Thee with secondary mica. The dark schist is ines indicate the arrange- phyllitic. Some specimens of it have mi- ment of mica laths. En- nute phenocrysts in a still finer groundmass. @tged_ 16 diameters. Fic. 4.—Broken quartz MICROSCOPIC DESCRIPTIONS Porphyry schist of the Young’s Pond locality —Thin sections of the porphyry schist from near the schoolhouse northwest of Young’s Pond show a nearly uniform groundmass. The diameters of the THE LYMAN SCHISTS OF NEW HAMPSHIRE aya smaller phenocrysts are thirty or forty times those of the larger groundmass grains and there. is no gradation between the two (Fig. 3). Clear, but with a few vacuoles, the quartz phenocrysts display only slight crushing, and this is mostly peripheral (Figs. 4-7). Their corners have been rounded. Several individuals are invaded by long narrow bays of the groundmass (Figs. 5 and 8). It is worth while noting that, although the outer borders of these pheno- crysts are jagged on’ account of granulation and penetration by mica laths, the edges of the embayments are clean-cut and smoothly curving. \\ee7 ff Z Fic. 5 Fic. 6 Fic. 7 Fic. 8 Fic. 5.—Quartz phenocryst with edges serrate on account of marginal granu- lation and penetration by mica laths. Note smooth outline of the embayment (a). Enlarged 26 diameters. Fic. 6.—Detail of outer edge of the quartz phenocryst represented in Fig. 5. The shaded area is quartz (Q). The small laths are mica (M/). Fic. 7.—Quartz phenocryst showing terminal granulation at pole of minimum compression (¢). a-—b is the edge of the thin section. Enlarged 20 diameters. Fic. 8.—Part of a large quartz phenocryst. The border zone, a—b, is a granu- lated portion of the phenocryst. Outside this zone is the much finer, uniform ground- mass (not figured). Several embayments of the groundmass are shown, one cut longitudinally (¢), and the others intersected transversely at various angles. Note regular outlines of these embayments as compared with the jagged border of the grain as a whole. The figure is drawn from a microphotograph, but not with absolute precision in the finer details. Enlarged 16 diameters. Like the quartz, the plagioclase phenocrysts are not severely crushed (Fig.9). The angle of extinction on sections approximately perpendicular to the albite twinning varied between 13° and 17°. This and the fact that the index of refraction was always lower than that of balsam indicate that the mineral is albite. Some grains have very good crystal outlines (Fig. 10), but as a rule they have had their corners more or less rounded (Fig. 3). 374 FREDERIC H. LAHEE As for the groundmass, it consists of abundant quartz and chlorite, with some sericite and feldspar. It is so fine that micro- scopic discrimination between quartz and feldspar is almost impossible. Both chlorite and sericite show some tendency to parallel arrangement. The chlorite is likely to occur in larger flakes in the lee of the phenocrysts and in the quartz embayments—in other words, where there was some protec- tion from the shearing stress, such as it was. Neither quartz nor feld- spar phenocrysts contain scattered inclusions of the groundmass. The porphyry from the channel of the brook running into Young’s Pond is somewhat more sheared than Fic. 9.—A plagioclase pheno- that from the schoolhouse ledge. The cryst showing the bundles of seri- quartz and feldspar phenocrysts are cite laths that wrap round it at : : the poles of maximum compression. more strained and granulated, and in Enlarged 11 diameters. some places the broken fragments are separated by a strip having cata- clastic structure (Fig. 11). Sericite is more abundant and has better orientation. It is often thickly plastered against the pheno- crysts about which it wraps at the poles of maximum compression (Fig. 9). In other respects this schist is similar to the less meta- Fie, 16.2—-Plawioelaee morphosed specimens. phenocryst with zonal Pebbles of the conglomerate schist of the SE Young’s Pond locality—FPebbles of the por- ted ort eae Bs phyry schist are identical with the bedrock js somewhat decayed. of thesame. They need no further descrip- Enlarged 32 diameters. tion here. The dark masses which have been mentioned as greenish when fresh, but drab when weathered, are of peculiar interest, since they have been regarded as highly metamorphosed blocks of argil- lite. They have been compared with the slate blocks in the Squan- tum tillite.t. Thin sections were prepared from several fragments *R. W. Sayles, op. cit. THE LYMAN SCHISTS OF NEW HAMPSHIRE 375 taken from different outcrops of the ‘schoolhouse conglomerate.” Microscopic examination revealed two distinct varieties. One is composed largely of zoisite and bundles of parallel actinolite needles, with some chlorite and titanite and a little plagioclase. The other is rich in chlorite, plagioclase, and zoisite, and contains titanite and a little epidote and sericite. This schist is finely por- phyritic, the phenocrysts being of plagioclase (Fig. 12). Many have been bent, sliced, or granulated by the shearing. The white mica is fairly well oriented and, together Fic. 1r.—“Torn” plagioclase With the other minerals, is inclosed phenocryst. The large dark area jn an irregular background or net- in the middle of the photograph : L. Fy ne Ms eee work of chlorite. There are some crystal, the other two-thirds being chlorite aggregates which look as if below the picture. The irregular they had formed from a mineral that ee Pane ae SS as once present as phenocrysts. No quartz is recognizable in either rock. With reference to the origin of these rocks, it would seem as if their source was most probably igne- ous. Some of them might have been derived from sedimentary products of incomplete decomposition, but they can hardly have come from normal argillitic material. Parker Hill white schist—The whitish schist that contains the dark phyllitic fragments is composed prin- Fic. 12.—Section of a pebble of * j fine porphyry schist from the cipally of quartz and orthoclase in “schoolhouse conglomerate” at nearly equal proportions, somewhat the Black Mountain locality. less plagioclase, and some chlorite and Balad tO) tr: sericite. The particles of quartz and feldspar have every gradation in size from the largest to the smallest (Fig. 13). There is relatively little matrix, the larger grains being so numerous that they nearly, 76 FREDERIC H. LAHEE &» if they do not quite, touch one another. Many of the grains are an- gular; none is conspicuously rounded, and they do not seem to have been rounded before metamorphism. Of the two minerals, sericite and chlorite, the latter is most plentiful. The chlorite is least likely Fic. 13.—Section of a typical Fic. 14.—Section of a por- psammitic member of the Lyman phyritic pebble of the ‘Parker series. Enlarged r2 diameters. Hill dark schist.’’ Enlarged 15 diameters. to have parallel orientation. Another specimen from farther north- east near this horizon exhibited similar characters in thin section. Parker Hill dark schist.—The specimen from which the sec- tion was cut was ob- tained northeast of Parker Hill along the strike of the rock ex- posed in the Lisbon- Lyman road, and it is thought to belong to the same horizon Fig. 15.—Part of the “egg conglomerate’? where the majority of the pebbles are angular and subangular. Che phyllite is a small Photograph by R. W. Sayles. angular fragment ina whitish clastic matrix of the kind just described. It is porphyritic (Fig. 14), having small phenocrysts (o.13-1.0 mm.) of plagioclase, often with well-preserved crystal form, distributed in a very fine groundmass of plagioclase and chlorite. The phenocrysts do not THE LYMAN SCHISTS OF NEW HAMPSHIRE ea contain inclusions of the groundmass. This rock is remarkably like a pebble taken from a conglomerate which Hitchcock’ called the “egg conglomerate” (see Fig. 15), a pebble which would be called a felsitic rock without hesitation. GENERAL CONCLUSIONS ON THE ORIGIN OF THE LYMAN SCHISTS Among the characters described in the foregoing paragraphs many are of such a nature as to suggest that at least some members of the Lyman schist group are igneous rocks—probably volcanic— more or less modified by dynamic metamorphism. For the sake of clearness and emphasis, the significant facts which lead to this conclusion are reviewed below in summary form for three typical rocks. Porphyry schist of the Young’s Pond locality.—The quartz- plagioclase porphyry schist which has been described from the Young’s Pond locality is assumed to have been an effusive rock for the following reasons: t. It has a strong resemblance to the quartz porphyry type of igneous rock. ; 2. It is massive, without bedding. 3. Locally it has faint indications of flow structure. 4. The phenocrysts are rather uniformly distributed through the groundmass. 5. The quartz phenocrysts have high luster and high trans- parency, a faint bluish opalescence, and a tendency toward cleavage, all these being features not uncommon in true quartz porphyries. 6. Some of the feldspar phenocrysts have crystal form. 7. There is a great difference between the sizé of the smaller phenocrysts and that of the larger groundmass particles. 8. Many. quartz phenocrysts and some feldspar phenocrysts contain embayments of the groundmass, these embayments having 1 C. H. Hitchcock, Geology of New Hampshire (1877), I, 333. This “egg con glomerate” is exposed on the northwest slope of Blueberry Mountain and may grade laterally into the Fitch Hill arkose. Most of its pebbles are of quartz porphyry, quartz keratophyre, granophyre, devitrified rhyolite, and other felsitic types. The great preponderance of effusive rocks among the pebbles is to be noted for comparison with the Lyman series. 378 FREDERIC H. LAHEE regular curving edges evidently due to magmatic corrosion and not to subsequent localized granulation or recrystallization. g. Some feldspar phenocrysts have zonal structure parallel to the outlines of the corrosion inserts of the groundmass (Fig. 10). 1o. Both quartz and feldspar phenocrysts, and even those feldspar crystals which have zonal structure, are quite free from the type of inclusion of the groundmass which is so common in many metacrysts. 11. Although two or three feldspar crystals are sometimes attached as if they had grown so (Fig. 16), quartz and feldspar never occur thus together. If the materials of the schist were derived from the breaking up } of a granitoid rock, small pebbles composed of both quartz and feldspar would be expected. Fee cena 12. All, or nearly all, of the feldspar pheno- attached plagioclase crysts are albite. A clastic rock would be likely individuals which to contain several species of fragmental feldspar, a: Sane for there are plenty of rocks in this region which diameters, ~~~-—«zave orthoclase, microcline, microperthite, micro- pegmatite, etc., among their constituents. 13. The abundance of albite in this rock suggests that the content of Na may be abnormally high for an argillitic sediment. However, I have not analyzed the rock chemically, and the Rosiwal method would be unsatisfactory on account of the difficulty of determining the particles of the groundmass. 14. The preponderance of easily recognized pebbles of felsitic rocks in the “egg conglomerate”’ proves that acidic effusive rocks may be expected in the region. If, then, we grant that the quartz-plagioclase porphyry schist is effusive, its composition places it in the class of quartz kera- tophyres, and the similar quartzless rock of Mormon Hill belongs to the keratophyres. : Parker Hill white schist—This is taken as representative of the “tuff-like schists’? of the Lyman group. It is undoubtedly a metamorphosed clastic as is demonstrated by the following facts: t. Megascopically and microscopically it has a distinctly frag- mental aspect. os ne ve THE LYMAN SCHISTS OF NEW HAMPSHIRE 379 2. Locally it contains angular blocks and sometimes isolated pebbles. ; 3. It grades into the fine variety of Lyman conglomerate schist. 4. It has obscure bedding (layers differing in texture) in some places. : 5. There are all gradations in size between its largest and small- est particles. 6. There is a relatively small proportion of small grains as compared with the porphyry schist. This Parker Hill white schist certainly was never a normal clastic in the ordinary sense of the term. It might be called an arkose schist on account of its having abundant clastic feldspar. However, its more or less intimate association with effusive rocks, the presence in it of the constitutents of the porphyry schists, both as small particles (quartz and albite grains, etc.) and as larger fragments (some whitish like the Lyman schists and some dark like the Parker Hill dark schist), and the observed gradations between it and the fine Lyman conglomerate schist which is com- posed chiefly of sheared pebbles of felsitic nature—these facts induce me to classify the rock as a metamorphosed tuff. Conglomerate schist of the YVoung’s Pond locality.—There is no doubt that this “‘schoolhouse conglomerate”’ looks very much like a glacial deposit, as stated by Hitchcock’ and recently by Sayles; but if the accepted criteria for till ever did exist here, they have been entirely destroyed by metamorphism. For this reason it is futile to look for signs of glacial abrasion on an underlying rock pavement. ‘The weakness of the evidence for glacial origin was fully appreciated by Mr. Sayles, and I may add,that evidence against such glacial origin is almost, if not quite, as inconclusive. However, one should bear in mind that great variation in size of * constituent fragments and absence of bedding are characters shared by talus, landslide débris, pyroclastic materials, and not infre- quently even by river-laid alluvial cone deposits. Is it not more probable that the “‘schoolhouse conglomerate,” being closely asso- ciated with an effusive rock (quartz-plagioclase porphyry schist), « C. H. Hitchcock, ‘‘ New Studies in the Ammonoosuc District of New Hampshire,” Bull. Geol. Soc. Am., XV (1904), 472, and Geology of New Hampshire (1877), Il, 302. 380 FREDERIC H. LAHEE having a large proportion of its pebbles and bowlders consisting of effusive rocks, and having a paste which resembles the Parker Hill white schist (cf. above), is a coarse pyroclastic, an agglomer- ate, rather than a tillite ? It is important to note that both Hitchcock and Hawes! main- tained that rocks of the Lyman group were the outcome of the metamorphism of sediments. Neither of these geologists held the view that these schists were of volcanic origin, yet both were struck by the resemblance between some members of the series and ordinary felsite. STRUCTURAL RELATIONS OF THE LYMAN SERIES The structural relations, and therefore the age, of the Lyman schists still remain obscure. These rocks are surely not younger than Devonian, and they may be older, as has been stated by Hitchcock. In several places on the Parker Hill-Mormon Hill range stratigraphic structures point to an anticlinal axis eastward. If this is so, since the Blueberry Mountain—Bald Hill range is regarded as synclinal, the intervening valley is anticlinal. Again, if this is so, it becomes necessary to explain the lack of correlation between the Lyman schists on the western range and the marine argillites and sandstones of the eastern range. Unconformity or extensive faulting may be the cause. At present I do not feel justified in discussing this subject further. More time should be given to field investigation. The region offers ample opportunity for research in petrology and in structural geology. ACIDIC EFFUSIVE ROCKS EAST OF THE APPALACHIAN PROTANIS South of the latitude of New York City the Appalachian Mountains are flanked on the east by the Piedmont Plateau. North of the same latitude, the New England Province corresponds to the Plateau. Both regions are underlain by plutonic rocks and folded, sheared, sedimentary rocks, chiefly of Paleozoic and pre- Cambrian age. With extended study of these complex rocks two results stand out conspicuously: an increasing number of meta- *C. H. Hitchcock, Bull. Geol. Soc. Am., XV (1904), 468, 469; and G. W. Hawes, “Mineralogy and Lithology of New Hampshire,” p. 176, in Hitchcock’s Geology of New Hampshire, III (1878). THE LYMAN SCHISTS OF NEW HAMPSHIRE 381 morphosed rocks is being transferred from the pre-Cambrain to the Paleozoic, and the origin of the schists is found to be more diverse than was at first conceded. It is in line with these results that the variety and distribution of effusive rocks, both flows and pyroclastics, interbedded with the schistose Paleozoic sediments; are being expanded as investigation proceeds. In a paper written in. 1894, G. H. Williams called attention to this fact.1 He briefly described the following localities in eastern North America, where volcanic rocks were known at that time: Newfoundland, Nova Scotia, New Brunswick, that part of the province of Quebec lying west of Maine and north of Vermont and New Hampshire, Maine, Massachusetts, Pennsylvania, Maryland, Virginia, and the Caro- linas. To this list it is probable that Rhode Island and New Hampshire may now be added. Williams mentioned New Hamp- shire in his text, but did not show any volcanic rocks in this state on his map. He observed that ‘in New Hampshire felsites and quartz-porphyries abound. They were regarded as eruptive by Hitchcock and by Hawes when they occur in dykes, although the latter regarded many of them, especially when interstratified, as sediments fused in situ.’ This paper by Williams contains numer- ous references. A list of more recent articles on the subject is given by J. E. Pogue, Jr., in his “Geology and Structure of the Ancient Volcanic Rocks of Davidson County, North Carolina.’’s « “T)istribution of Ancient Volcanic Rocks along the Eastern Border of North America,” Jour. Geol., II (1894), 1-31. ZO Dp cits, Dp» 24: 3 Am. Jour. Sci. (4), XXVIII (1909), 218-38. NOTES ON THE DISINTEGRATION OF GRANITE IN EGYPT: DONALD C. BARTON Washington University, St. Louis, Missouri INTRODUCTION THREE PERIODS OF DISINTEGRATION IN THE ASWAN DISTRICT THe RATE OF THE DISINTEGRATION OF THE GRANITE At Aswan At Luxor and Gizeh THE CAUSES OF THE DISINTEGRATION The Conventional Explanation Inadequacy of This Explanation in the Case of the Disintegration in Egypt Moisture and the Disintegration Agreement of the Observations in Egypt with Those Made in Europe and in the Eastern United States INTRODUCTION The disintegration of the granite in Egypt has been treated in a general way by Walther in his discussions of disintegration in desert regions. It has been commented on by Ball and others. The disintegration of the New York obelisk, composed of the Syene red granite, has been reported upon by Julien. The follow- ing notes were made by the writer in a recent trip to Egypt in which especial attention was paid to the disintegration manifested by the granite in the ancient quarries and ledges of the Aswan district and in the temples and monuments of Upper and Lower Egypt. THREE PERIODS OF DISINTEGRATION IN THE ASWAN DISTRICT The disintegration in the Aswan district seems to belong to two periods other than the present. The products of what seem to be the earliest period of disintegration are found at the contact of « This paper is the result of work done as Sheldon Traveling Fellow, Harvard University. 382 THE DISINTEGRATION OF GRANITE IN EGYPT 383 coarse. Syene red granite with the base of the overlying Nubian sandstone (Cretaceous) (Fig. 1): They form a zone of what Ball designates as “‘broken-down granite, a kaolinic mass with quartz grains.” The zone is 1 to 13 m. in thickness, has a relatively sharp even contact with the overlying sandstone and conglomerate, but below grades through less and less disintegrated granite into comparatively unaltered rock. The upper part of the zone is composed of material that has suffered slight rearrangement, but the middle and lower portions consist of the untransported débris of disintegration. The feldspar of the upper portion is almost completely kaolinized.. In the middle portion, the kaolinization i! GH NA Fic. 1.—Diagrammatic section across the region of the Aswan Cataract. 1, Dis- integrated and decomposed zone at the base of the Nubian Sandstone. 2, The massive granular disintegration. 3, Disintegration of the present. is much less and in the lower portion it is not megascopically noticeable. Although the disintegration was seemingly not of the exfoliation type,’ it nevertheless took place roughly parallel to the very level upper surface of the granite. The surface is so level as to suggest a peneplain surface. The disintegration would seem to have taken place contemporaneously with or immediately preceding the deposition of the Nubian sandstone, and the kaolinization may t According to the present use of the term, it is possible to distinguish several types of disintegration. The term is applied in some cases to the breaking up of a rock-mass into blocks and in other cases to the breaking up of a rock-mass through the loss of cohesion between the constituent grains. The former process may be termed ‘‘block-disintegration” and the latter, “‘granular-disintegration.”’ Allied to the block disintegration is what may be termed the ‘‘exfoliation”’ type of dis- integration in which thin plates of rock rift off parallel to the surface of a ledge or block. The process seems to involve considerable loss of cohesion between the grains and readily goes over into granular disintegration. The term disintegration is also used in a loose sense to denote the chemical breaking up of the rock-mass. But as Merrill advises, it would seem better exclusively to use for that process the term ““decomposition”’ and to reserve the term “disintegration” for the process of mechan- ical breaking up. 384 DONALD C. BARTON have taken place under the estuarine or marine conditions then prevailing or at some later time under the action of ground-water. The disintegration and decomposition penetrate cracks to the depths of ro to 20 ft. and in one or two places may be seen to have taken place before the disintegration next to be mentioned. ~ The effects of the second period of disintegration are mani- fested in a tendency toward deep, granular disintegration massively affecting the coarse Syene red granite at a level which is approxi- mately that of the Aswan Reservoir when full. This disintegration is best seen on the island of El Hesa, in some recently excavated graves at the site of the former village of Garba. The graves are cut back into hill slopes of various inclination and have a hori- zontal depth of about 3 meters on the average and a height of about 1; m. The roof at the back commonly has a thickness of 1 to 2m. The greatest distance to which a grave was seen to extend hori- zontally backward from the surface of the hill slope was 4 m. The height of the grave was about 13 m. and the roof at the back was slightly over 2 m. in thickness. The graves are cut entirely in the disintegrated, or rather, partially disintegrated, granite and in no case were they seen to penetrate unaffected rock. The disintegration has therefore penetrated to a depth of at least 3 to 4 meters from the surface. The disintegration is of the granular type and, although affecting the rock uniformly, is not quite com- plete; blocks of the granite may be obtained which to the eye seem entirely sound but which crumble readily under the hammer. In the shallow sections that are afforded there is not any appreci- able decrease of intensity of the disintegration with depth. The disintegration has been accompanied by slight but megascopically noticeable kaolinization and in its products resembles closely the partial granular disintegration which is found at a depth of about 4m. in the Morvan and Plateau Central regions of Central France. This tendency toward massive granular disintegration is mani- fested also at several points along the old Nile Valley about 1 km. north-northeast of Shellal, at several points along the river trail from Shellal to the Aswan Dam, at several points immediately north of the village of Kuror along the river trail to Aswan, and about one-half mile northeast of Kuror in a small pass on the trail THE DISINTEGRATION OF GRANITE IN EGYPT 385 to Aswan. The disintegration in each of the cases is approximately at the elevation of that on El Hesa. It is a distinctly noticeable fact that in many of these cases, as for instance in the case of the grave on El Hesa cut 4 m. back into a northerly facing 40° slope lying at the foot of a cliff about 25 m. high, the disintegration has penetrated to a depth of at least 3 to 4 m. in spite of the fact that direct isolation is received only during the summer and then only at a low angle. Disintegration taking place under the present conditions is abundantly shown by most of the exposed ledges and loose blocks of the region and is manifested in three ways: (1) Surfaces which have been exposed for a relatively short while show a slight rough- ening. Individual grains and fragments of feldspar and of quartz become loosened and are removed. (2) Surfaces which have been exposed for a longer time show in addition exfoliation of thin superficial layers commonly of about two-thirds of a centimeter in thickness. Cross-sections afforded by broken blocks show that megascopically noticeable incipient exfoliation has penetrated to a depth of 10 to 15 cm. from the surface. (3) Disintegration takes place also by the spalling and splitting of large blocks and frag- ments, but the amount of disintegration taking place in this manner in the Aswan region is not very great. Of these three methods of disintegration that by exfoliation is by far the more important. In the excavation for the dam and navigation canal, concentric disintegration and decomposition were found to have penetrated to a depth of several meters below the high Nile level and are probably to be considered as going on at the present. The chief granite of the Aswan region is the famous Syene red granite, a coarse red porphyritic granite composed chiefly of large phenocrysts of orthoclase. Where the joints are comparatively far apart, its outcrops under the effect of the concentric exfoliation of the joint blocks resemble huge piles of bowlders. Where the jointing is more pronounced and the joint blocks of much smaller size, the concentric exfoliation is much less in evidence and the outcrops are composed of, and surrounded by, detrital masses of angular and subangular blocks and in appearance are very similar 386 DONALD C. BARTON to the outcrop and surrounding detrital slopes of similar types of rocks in New England. The fine-grained granite of the region, somewhat similar in composition to the coarse red granite, although with a lower content of colored silicates, is not so severely affected by the dis- integration. Exfoliation takes place very slowly, and although the edges and corners of exposed blocks have in most cases been rounded, the general form of the blocks is angular. The outcrops in general aspect are not unlike these of the -more jointed phases of the coarse red granite. The fine-grained granite is, however, itself much jointed. It was not seen massively disintegrated and small dikes cutting the massively disintegrated coarse granite showed merely slight exfoliation of the edges and corners of the joint blocks into which the dike is broken. Flaking and the loosen- ing of single grains on exposed surfaces do not seem severely to affect the fine-grained granite. THE RATE OF THE DISINTEGRATION OF THE GRANITE The rate of disintegration of exposed surfaces of granite at Aswan is not as rapid as at first might seem. Many of the numer- ous hieroglyphic inscriptions of this region show noticeable disin- tegration and on this account relatively rapid rates of disintegration have been postulated. These inscriptions almost without excep- tion are carved on bowlders of exfoliation, and in but few cases was there seemingly much effort on the part of the ancient Egyp- tians to remove more than the most readily detachable plates of exfoliation. The greater number of the inscriptions therefore were carved on surfaces that were already partially disintegrated. In the few cases in which the writer was able to satisfy himself that the inscriptions had been cut in surfaces dressed back into fresh rock, there was no disintegration noticeable and the inscriptions were entirely fresh and sharp. Such inscriptions can be seen on one of the two natural obelisks on the island of El] Hesa. The inscrip- tions date from the reigns of Mentuhotep I, about 2100 B.c.; Thutmoses IV, 1420-1411 B.c.; Amenhotep III, 1411-1395 B.c.; and Psammeticus II, 588-583 B.c. The inscriptions show no noticeable disintegration, and tapping with the finger or hammer THE DISINTEGRATION OF GRANITE IN EGYPT 387 does not reveal the presence of incipient exfoliation or flaking. The exposure is southerly and therefore one that affords the maxi- mum exposure to insolation. These rocks, as can be seen from Fig. 2, rise directly out of the Nile and there would seem to have been no chance of their having been buried and protected by accu- mulations of sand or débris. Other examples of inscriptions carved in fresh surfaces and not showing disintegration are those Fic. 2.—The Island of Konosso. View taken looking north-northeast. : The hieroglyphics mentioned are on the south face of the right hand of the two natural monoliths. along the trail from Aswan to Shellal numbered by Weigall 323, 320, 334, 343, and 350 and dating from the eleventh, twelfth, and thirteenth dynasties; an obelisk and a statue lying unfinished in the ancient quarries and referred by Wiegall to the reign of Amen- hotep III, 1411-1375 B.c., and numerous discarded quarry blocks in the ancient quarries and along the ancient quarry roads, dating probably from not later than the last century B.c. These blocks in many cases consist of a half, a quarter, or an eighth of a bowlder of exfoliation and in most cases it is readily possible to determine which were the originally fresh and which were the originally exfo- liating surfaces. The surfaces which were originally fresh are still 388 DONALD C. BARTON fresh and show merely an infinitesimally thin film of tarnish and alteration. Tapping revealed no incipient exfoliation. In micro- scopic thin sections taken at right angles to the surface of a block, the orthoclase is seen to be comparatively fresh; the oligoclase is much clouded by decomposition products, but the alteration is not sufficient to obscure the specific determination of the feldspar. The ferro-magnesian minerals show slight decomposition and in one of the sections there is considerable limonitic staining. The degree of decomposition is no greater than -that which is very commonly observed in sections of granite and is no greater toward the surface than deeper in. There is no tendency, as far as could be seen, toward incipient rifting parallel to the surface. The sections were taken at right angles to surfaces which had a southerly exposure and which were therefore exposed to the maximum heating effects of the insolation. Farther north in Egypt the rate of disintegration is more rapid. At Luxor, Thebes, Gizeh, and in the museum at Cairo, the granite (chiefly the coarse Syene granite) of statues, of obelisks, of por- tions of the temples, and of the facing of the pyramids, shows in the greater number of cases noticeable disintegration. That manifested by the statues is manifested chiefly as exfoliation of a thin film, o.5-o.7 cm. in thickness, from the pedestal, feet, and lower portion of the legs. Above the knees, the original high polish is commonly still intact, and tapping does not reveal incipient exfoliation or flaking. Examples of this type of disintegration can be seen on many, but not all, of the statues of Rameses II in the Forecourt of the Temple of Luxor and by the statue of Rameses II at the north entrance, by the colossal statue of Rameses II at the entrance to the great Hypostyle Hall, Karnak (Fig. 3), and by the medium-sized statue in the temple of Ptah, and by about half the statues of the coarse red Syene granite and also those of dark medium-grained rock possibly diorite in the museum at Cairo. The statue in the Temple of Ptah is situated in a small dark sanctuary and is not directly exposed to insolation. The other statues at Luxor and Karnak are less well protected, but neverthe- less are only very poorly exposed to the temperature changes con- sequent upon solar heating. In the Great Temple of Karnak, THE DISINTEGRATION OF GRANITE IN EGYPT 389 disintegration manifests itself as the spalling of the corners of the uprights; as the exfoliation to the depth of about 1 cm. of the walls in the Granite Sanctuary, erected in 313 B.c. by Phillip Arrhidaeus; and as spalling and exfoliation of the lower 6 to 8 ft. of the fluted columns in front of the Sanctuary, and also of the obelisk of Queen Hatshepsut, 1591-1447 B.c. The obelisk of Thotmes I, now lying in pieces on the ground, shows scattered, patchy flaking and under tapping much incipient exfoliation is revealed. At the Temple of Fic. 3.—Statue of Rameses II. Entrance to the Great Hypostyle Hall, Karnak, showing in a characteristic manner the exfoliation of the pedestal, feet, and lower legs. Medinet Habu, Thebes, disintegration is shown by the granite pillars of the doorway both on the sides which are exposed to the sun and on those which are not. In the Serapeum at Sakkara, on the other hand, the surfaces of the huge sarcophagi, which are hewn out of the coarse Syene red granite, still retain the high per- fection of their original polish and show not the faintest trace of incipient disintegration or exfoliation. The sarcophagi, however, are in dry underground chambers whose temperature, according to Baedecker, remains very constantly at about 80° F. At Gizeh, the granite blocks which formed a part of the facing of the second and third pyramids show for the most part on their 390 DONALD C. BARTON exposed surfaces a very marked exfoliation to the depth of 0.5 to o.8 cm. Minor exfoliation, in addition, is found along the joints between the blocks. Exposed surfaces not exfoliating commonly show marked granular flaking. The orientation of the surface, with ~ north, east, south, or west exposure, does not seem appreciably to affect the intensity of the disintegration and exfoliation. Disinte- gration and exfoliation are shown also by the granite facing that extends for 30 ft. down the shaft on the north side of the second pyramid, by the granite pavement of the temple at the east base of the second pyramid, and by the granite blocks immediately to the north of the east entrance to the temple. A striking feature Fic. 4.—Diagrammatic sketch showing the greater degree of disintegration below the old soil line than above. East entrance to the temple of the second pyramid, Gizeh. in this latter case, as can be seen from the accompanying sketch (Fig. 4), is that the disintegration is distinctly stronger below what seems to have been an old soil line than above it. A similar case was noted at one of the pyramids at Sakkara. The débris resulting from the disintegration and exfoliation in all these shows slight but megascopically noticeable decomposition. The degree of the alteration of the colored silicates is greater than that of the feld- spars, and that of the plagioclase is greater than that of the ortho- clase. The pyramids of Gizeh date from the Fourth Dynasty, about 2850-2700 B.c., the statues of Rameses II at Karnak and Luxor date from the Nineteenth Dynasty, 1292-1225 B.c., and the Granite Sanctuary, Karnak, dates from the reign of Phillip Arrhidaeus, 318 B.c. The average rate of disintegration and exfoliation would therefore seem to be about 1 cm. to 0.5 cm. in five thousand years. THE DISENTEGRATION OF GRANITE IN EGYPT 391 The maximum rate, shown by the Granite Sanctuary, would seem to be about 1 cm. in two thousand years, and the minimum rate _would seem to be so low that the effects are not apparent in three thousand years. In addition to this variation in the-rate of dis- integration apparently corresponding to a variation in the conditions to which the granite is exposed, there is apparently also a variation depending upon the orientation of the disintegrating surface in reference to certain directions within the rock, possibly the rift and the grain, or possibly a faint schistosity which is almost universally present in the Syene granite. THE CAUSES OF THE DISINTEGRATION The conventional explanation of disintegration in a region of desert climate like that of Egypt is that the disintegration results _ from the racking to pieces of the rock through the contraction and expansion consequent in the high-temperature ranges. In the case of Egypt, there would, however, seem to be serious objections to this explanation, although some disintegration undoubtedly does take place in that manner. The first objection is that, although the temperature range is of the same magnitude at both Aswan and at the pyramids of Gizeh, the rate of exfoliation is very much less at the former place than at the latter, and furthermore, that, although the statues in the temples are exposed in many cases only to very low temperature ranges, the rate of exfoliation in many of these cases is of the same magnitude as that at the pyramids of Gizeh. The second objection is that the massive granular dis- integration of the Aswan region penetrates to a greater depth than appreciable temperature changes can be expected to extend. The depth of the zone of warming at midday in desert regions is given by Walther as the result of many observations as only about 19 cm. The annual temperature variation is said by Sir William Thompson . to be reduced at a depth of 8 m. (25 ft.) to one-twentieth of its superficial amount. The mean annual temperature range in Egypt is less than 20° C and, at the depths to which disintegration can be seen to have penetrated at Aswan, 3 to 4 meters, must be reduced to amounts which are essentially negligible, especially since the period of the range is so long. The mean monthly range 392 DONALD C. BARTON is only 24°C. and at those depths must be even more seriously reduced in amount. The diurnal temperature range, furthermore, should be entirely absent at those depths, especially on slopes such as those in which many of the graves on El Hesa are cut, where direct insolation is received only during the summer and then at a low angle. Granite itself has a low coefficient of conductivity and that of dry granitic sand must be much lower; it would there- fore not seem surprising that a blanket of several feet of disinte- grated granite is found to be an effective insulating agent for the fresh rock beneath. At Aswan and at the pyramids of Gizeh, the only factor by which the conditions of exposure of the exfoliating rock differ is in the humidity. At Aswan there is no rainfall, there is only a light dewfall at night, and the relative humidity at 8:00-9:00 A.M. varies from 28 to 58, average 39; while at the pyramids of Gizeh there are several light showers,each year, there is a moderately heavy dewfall at night, and the relative humidity at 8:00-9:00 A.M. runs from 64 to 87, average 72. In the case of the exfoliating statues, their sheltered positions in the temples and the connection between the exfoliation and the lower portions of the statues would seem to indicate that the cause of the exfoliation lay not so much in the temperature changes as in some factor connected with the ground, as for instance, in the ground-water or moisture, and it is to such a cause that the exfoli- ation is ascribed by G. Daressy of the Department of Antiquities, Egypt, who says: “Les granites exposés continuellment a leau ou au soleil se conservent bien, mais ot ils se degradent, c’est lorsqu’ils ont été enfouis dans un sol humide. La formation de sels nitrate et autre fait alors decomposer le granite, sirtout lorsque le terrain est alternatement sec et humide.” The expansion con- sequent upon the kaolinization of the feldspar is emphasized by Merrill as the cause of the disintegration of the granite near Washington, D.C. Although kaolinization is megascopically very noticeable in these cases, it would scarcely seem to be of sufficient amount alone to account for the observed disintegration and exfoliation. THE DISINTEGRATION OF GRANITE IN EGYPT 393 The massive granular disintegration of the Aswan region possibly also may be attributed directly or indirectly to the effect of moisture. The disintegration is found at and for some few meters below the level at which the Nile must have flowed when in the old Nile Valley between Aswan and Shellal. At that time the granite at the level of this disintegration must have been alternately above and below the ground-water level, as the Nile rose and fell, and must consequently have been alternately wet and dry. At the present level of the Nile, the granite was found in the excavations for the navigation canal and for the dam founda- tions to be almost completely disintegrated and decomposed to a depth of several meters below the level of the high Nile. Decom- position in this case has, however, rather predominated over simple disintegration. These observations in the light which they throw on the cause of the disintegration of granite are in agreement with similar observations which the writer made in the Odenwald, in the Vosges © Mountains, in the Norvan and Auvergne districts of France, and in the eastern United States. In the many places in which the disintegration has reached the depth of 20, 30, or even 4o ft., it seems impossible to believe that the temperature changes are of sufficient amount to be of any appreciable effect. Diurnal, weekly, and monthly temperature changes must be completely eliminated at those depths, and according to Sir William Thompson the annual temperature range is reduced at a depth of 25 ft. to one- twentieth of its superficial amount. The disintegration in these places is accompanied in many cases by very much more and in other cases by only slightly more decomposition than is the dis- integration in Egypt. A RECORDING MICROMETER FOR GEOMETRICAL ROCK ANALYSIS S. J. SHAND Victoria College, Stellenbosch, South Africa The quantitative éstimation of minerals in rock sections is generally recommended as a valuable exercise, but in practice it is far too seldom performed. The reason is that the recognized methods of estimation are very tedious, while the results, when obtained, have not hitherto been put to any systematic use in the classification of rocks. The usual methods are of two kinds, viz., (1) separations, either gravitational or magnetic, the separated portions being weighed directly; and (2) geometrical methods, involving measurement either of areas or of diameters, and sub- sequent calculation of percentage volumes and percentage weights. Of estimations of the latter class, the following variants are known to me: t. The method of Delesse: The surface of the rock is polished and oiled, and the outlines of the grains are traced on transparent paper, the areas corresponding to different minerals being distinc- tively colored. The paper is then pasted upon tinfoil, and cut up along the boundaries of the grains. The fragments having been grouped according to color, the paper is removed and the tinfoil weighed. The weights so found are proportional to the areas traced upon the paper, hence also to the volumes occupied by the different kinds of grains, provided that the rock is uniform throughout. To get the proportions of the various minerals by weight, each volume must be multiplied by the specific gravity of the corresponding mineral. 2. The outlines of the grains may be traced upon squared paper, and the areas obtained by counting the number of squares occupied by each mineral, all broken squares being reckoned as half-squares. 394 MICROMETER FOR GEOMETRICAL ROCK ANALYSIS — 395 3. Methods 1 and 2 can be applied to microscopic sections by the aid of a camera lucida attached to the microscope. An ordinary photographic camera can also be used, the outlines of the grains being sketched from the enlarged image on the focusing screen. 4. Ifa dark room is at hand, it is sometimes preferable to photo- graph a rock section instead of sketching it. The print can be examined either by weighing or by means of squared paper. 5. A squared ocular micrometer, by means of which the areas of the grains in a section can be measured directly under the microscope, was tried by Rosiwal. It was found to be less advan- tageous than the following method, viz.: 6. Rosiwal’s linear traversing method:" If the work be executed with care and under all necessary precautions, this is the simplest and perhaps for that reason the most accurate of all geometrical methods of rock analysis. The measurement of areas is replaced by the measurement of: diameters along a selected line or lines. Either a microscopic section or the smooth face of a hand-specimen of the rock may be employed, according to whether the rock is of fine or coarse grain. In the latter case, a graduated rule or tape is required; in the former, an ocular micrometer. Any kind of micrometer will do, but the estimation is facilitated by the use of certain special types, such as the “‘planimeter ocular” of Hirsch- wald.? Subject to certain conditions, the number representing the sum of the diameters of all grains of one kind is proportional to the volume of the mineral concerned. . So far as tediousness is concerned, all these methods are more or less on the same level; the measurements are very wearisome and take a long time to perform. Generally speaking, one would expect weighing to be a more exact process than the use of squared paper, but then the weighing must be preceded by sketching and cutting out, and appreciable errors may creep in during these manipulations; furthermore, one cannot be sure that the material weighed, be it tinfoil, cardboard, or paper, is everywhere of the same thickness. On the other hand, the counting of innumerable « Rosiwal, ‘Uber geometrische Gesteinsanalysen,’ Verhandlungen der k.k. geolog. Reichsanstalt, Wien (18098), No. 5. 2 J. Hirschwald, Centralblatt fiir Min., Geol., Pal. (1904), No. 20. 306 S. J. SHAND tiny squares is a most aggravating business, and the fewer the squares, the greater the probable error in the result. On the whole, the advantage seems to be with the Rosiwal method, as being both simpler and more direct than the others. The practical disad- vantages of the method are two: First, the making of very many minute measurements by the aid of the scarcely visible scratches on the eyepiece micrometer puts a severe strain upon the eyesight, as well as upon the patience, of the observer. After an hour or two of such work I have sometimes been nearly blind. Secondly, the writing down and adding up of some hundreds or even thousands of measurements is itself a most tedious operation. To obviate these serious disadvantages of the Rosiwal method, I have devised a stage micrometer which both makes the measure- ments and performs the addition of them; it consequently effects a great reduction in the time needed for the estimation, and inci- dentally reduces the strain on the eyes to a minimum. The first instrument was made from my drawings by Mr. T. A. Linton, at the South African College, Cape Town, and I have pleasure in expressing my appreciation of his excellent workmanship. The design is very simple, and will easily be followed with the aid of the drawings (Figs. 1, 2, 3). The rock section, mounted as usual on a glass slide, fits into a rectangular brass sledge A, which is movable to right or left of the observer, within another sledge B, the movement being accomplished, and its amount recorded, by the micrometer screw L. Sledge B moves in the same manner and direction within sledge C, the movement being performed and recorded by the micrometer screw R. Sledge C has no transverse movement; it carries two runners on its under surface which travel in grooves on the sides of the rectangular stage of the micro- scope; the only movement of this sledge is to and from the observer and is effected simply by hand. . Suppose it is required to estimate the volume of augite in a dolerite. The section is put in place and adjusted till one edge of it appears against the point of intersection of the cross-wires in the eyepiece of the microscope. The readings of screws R and L are written down. Then screw & is turned continuously until a grain of augite is brought up to the cross (i.e., the point of inter- MICROMETER FOR GEOMETRICAL ROCK ANALYSIS 397 S J SHAND. | 1 mE LARAANUUIUUUTL Micro S)iide ! RG mimi mutans: ae Fig. J | | | ! } \ l t Working drawings of recording micrometer. FIGS. I, 2, 3. 308 S. J. SHAND section of the cross-wires); screw Z is now turned until the grain travels past and its other margin lies exactly beneath the cross; then screw X is turned till the next augite grain comes into position; then screw Z till the grain has passed, and so on. When the traverse has been completed, which with a section of ordinary size may take from one to three minutes, the readings of the microm- eter screws are again noted and written down below the former readings. It is obvious that the difference between the two read- ings of screw L gives the sum of the diameters of all the augite grains which were intersected during the traverse, while the differ- ence for screw RX gives the sum for all other minerals in the rock. Without stopping to make these subtractions, however, sledge C is pushed forward into a new position and a second traverse is made in the return direction; at its completion the micrometers are again read and the readings jotted down beneath the previous ones. Sledge C is again. pushed forward, and another traverse made, and so on until the number of traverses is considered sufhi- cient. One may now proceed to the subtraction of the successive readings, and the calculation of the percentage of augite, which is obviously sum of successive differences of L sum of successive differences of L + sum of successive differences of R 2 TG0: The most expeditious manner of recording the readings is to write them down in parallel, vertical columns in the middle of the page; then, when all the measurements have been made, the differences of ZL are quickly filled in to the left and the differences of R to the right, as in Table I, on p. 399, which is part of an actual estimation. It will be seen that by this method only two numbers have to be recorded for each traverse after the first, while by any other method some twenty or more may be necessary. It may be advisable at this stage to recall the conditions which, as Rosiwal has pointed out, must be observed if the linear traversing method is to give reliable results. 1. The length measured must be at least one hundred times the average grain of the rock. (With the additional facility afforded by the recording micrometer, it would involve little extra labor MICROMETER FOR GEOMETRICAL ROCK ANALYSIS 399 to increase this minimum to two hundred or even four hundred, and I strongly advocate the increase.) 2. Two measured lines should be at least the width of a grain apart. 3. When the constituents are fine-grained and uniformly distributed, measurement of a single section may be sufficient; if coarse-grained, several sections may be necessary in order to satisfy conditions 1 and 2. TABLE I Diff L R Diff | 3.82 8.05 6.43 16.60 I0.25 24.65 9.18 13.40 TOT EE GAG 5.86 17.45 6.93 28.70 6.93 16.25 0.00 12.45 7.62 15.23 7.62 27.68 36.02 78.93 oc 36.02 fie Percentage = |-5 o> +7803 X100= 31.3 Total distance traversed, 57.5 mm.; time taken for measurements and calculation, 12 minutes. 4. In the case of a rock with parallel structure it is necessary, and in most cases it is desirable, to take measurements both along and across the section. 5. The most accurate method is to measure all the minerals present at the same time, rather than one at a time, although the latter is the quicker way. (This recommendation applies to the use of an ordinary micrometer; it is of course inapplicable to the recording micrometer.) 6. In the case of coarse-grained rocks it is often quickest to measure a polished face macroscopically, using sections only for minor or microscopic constituents. | Rosiwal’s practice is to draw fine lines on the cover-glass with ink, and then measure along these lines. This is not necessary 400 S.J: SHAND if the stage of the microscope happens to be ruled with cross-lines, as is often the case. The method having been described, it remains to add some details about the construction of the instrument. In making the first model, we took the micrometer screws out of two small spherom- eters; these had just the right length (about 1} inches) and pitch Fic. 4.—Photograph of original micrometer (o.5 mm.) and were used almost without modification. The num- bers on the graduated disk run in the correct direction for screw L, but must be reversed for screw R. These screws proved, on trial, to be unsatisfactory, having been taken from a cheap type of spherom- eter (the only kind obtainable in South Africa at the time), and better ones have since been substituted for them by Messrs. Swift & Son, London. It is of course essential that the screws shall be machine-cut with the highest degree of accuracy; that the axis shall be perfectly straight; and the graduated disks truly plane and set exactly at right angles to the axis. The precise pitch of the MICROMETER FOR GEOMETRICAL ROCK ANALYSIS 401 screw does not matter at all, but a pitch of about one-half to one millimeter is convenient. The scales which record the movement of the screws are of course graduated to correspond to the pitch, and the divisions on both scales are numbered from right to left. Fic. 5.—The micrometer in position on the stage of an “ Allan Dick”’ petrological microscope (Messrs. Swift & Son, London). The ends of the screws may be rigidly attached to the sledges A and B, but can be more simply secured by means of backlash springs. The innermost sledge A carries two slots, one of them (abcd) made to take the standard English size of slide (75X25 mm.), the other 402 S. J. SHAND (efgh) to take German slides (48X28 mm.). Two small spring clips (not shown in the figure) serve to prevent any slight movement of the glass. In order to allow of measurements being made in two perpendicular directions across a rock section, four short pins are inserted in the sides of sledge A; these make it possible to fix a German-sized slide in cross position upon the micrometer, but the English slides are too long to permit this. . The attachment of the outermost sledge C to the stage of the microscope is effected by means of runners with beveled edges which fit into grooves in the sides of the stage. One runner is fixed directly to the under surface of the sledge, the other is attached by a simple wire spring to a fixed cross-bar on the under surface of the sledge. This simple arrangement could be replaced, if desired, by a mechanical movement; in any case it would be an advantage to have an additional screw (V, Figs. 1, 2) by means of which the instrument could be clamped to the stage in any desired position. My instrument was made to fit the stage of an ‘Allan Dick” microscope (Swift & Son), but it is obviously adaptable to any microscope with a rectangular stage. It could even be adapted to a circular stage by means of a rectangular plate clamped temporarily on top of the stage. The instrument, as actually made for me, differs in some minor points from the drawings. For instance, the micrometer screws are not rigidly connected to the sledges A and B, as the drawing suggests, but are attached by simple backlash springs which can be seen in the photographs (Figs. 4, 5). The dimensions of the side- and end-pieces of the sledges were slightly increased for the sake of greater strength. The figures relating to these parts are as follows: Length of sledge A, 85 mm. Width of sledge A (including bevel), 37 mm. Travel of sledge A, 16 mm. (this might be increased). Length of sledge B, 108 mm. (this might be increased). Width of sledge (including bevel), 45 mm. Travel of sledge B, 35 mm. Length of sledge C, 127 mm. Width of sledge C, 54 mm. (might be reduced to 52). Thickness of material throughout, 3.5 mm. MICROMETER FOR GEOMETRICAL ROCK ANALYSIS 403° It would be well to allow a greater travel to sledge A by slightly increasing the length of sledge B—say to 112 mm. A well-finished instrument of this kind could not be made for much less than £5, hence it is not likely to become part df every -petrologist’s equipment; but in view of the saving of time and eyesight which it effects the initial outlay is inconsiderable, and the gain to descriptive petrography, if it should succeed in popularizing the geometrical analysis of rocks, would be very great. The instrument has been examined by Messrs. J. Swift & Son, London, who have all the information necessary for executing copies of it. EXPRESSION OF THE “COLOR RATIO” OF A ROCK It has always appeared to me to be a matter of great importance in rock descriptions to state the proportion of light to dark minerals accurately. The terms “‘leucocratic” and “melanocratic’”’ have proved extremely useful in giving a rough indication of this pro- portion (which I am in the habit of calling the “color ratio” of the rock), but something more is urgently required. Of course the fundamental point of difference between the light and the dark minerals does not lie in their color, but in their specific gravity, to which, however, the color affords a convenient index, inasmuch as all minerals of gravity less than 2.8 are leucocratic (predomi- nantly light-colored) and those of higher gravity are melanocratic (predominantly dark-colored). It is becoming more and more apparent that differences of specific gravity must be reckoned among the chief causes of magmatic differentiation, and for this reason, as well as for its purely descriptive value, the color ratio must receive quantitative recognition in the future. With the micrometer described above, it is possible to measure the color ratio of a fine-grained rock with a high degree of accuracy in a period of ten to thirty minutes, according to circumstances. Having ascertained the ratio, the next question is how best to express it, and the way which does least violence to our accepted, illogical system of nomenclature is to resort to a system of prefixes. 404 S. J. SHAND The following ratio and the corresponding prefixes are simple and expressive : Light minerals more than er cent lo or L 8 “ “ “ “ go a“ 1 i “ “ “ “ oO a9 ; or l So ls { “ “ “ “ 70 a lL \ : ‘ oe or “ “ “ ‘ 60 le J “ “ “ “ i « « « oa « s | or lm Dark 50 m; if “ “ “ “ 60 a m6 \ « «“ «“ «“ 70 « m; f or m “ “ “ “ So ac ms \ “ “ “ “ 90 iT; m J Or 7 = 9 “ “ “ “ 07 “ Myo or M Examples: L=granite (alaskite); ls=syenite; Im=dolerite; m=syenite (shonkinite); myo=pyroxenite, etc. REVIEWS Origin of the Bighorn Dolomite of Wyoming. By Exior BLAacK- WELDER. Bull. Geol. Soc. Am., XXIV, 607-24, plates 8, . December 22, 1913. The Bighorn dolomite is widely distributed in northwestern Wyo- ming. Its fossils, mostly corals and crinoid stems, are rare and seldom well preserved, but indicate an Ordovician age, possibly including Silurian also. Chemically the formation is a very pure, normal dolomite, with very little terrigenous matter. Its weathered surface is character- istically coarsely pitted and fretted, owing, not to intermingling of sili- ceous with calcareous matter, but to compact fine-grained dolomite structures imbedded in a matrix of more coarsely crystalline and porous dolomite. The ill-defined branching patterns due to differential weather- ing are probably of organic origin, more likely representing banks of calcareous algae than plantlike animals. The obliteration of original organic structures is assigned to the process of crystallization of the dolomite, probably taking place almost simultaneously with deposi- tion on the sea floor. The deposits were doubtless made in an epicon- tinental sea less than 100-120 meters deep. Re CaaVie On Oceanic Deep-Sea Deposits of Central Borneo. By G. A. F. MOoLeNGRAAFF. Koninklijke Akademie van Wetenschappen te Amsterdam, Proceedings of the meeting Saturday, June 26, 1909. Pp. 7, map tr. . The Danau formation, which outcrops over an area of approxi- mately 40,000 sq. km. in central Borneo, consists of cherts and hornstones formed almost entirely from the tests of Radiolaria. The char- acter of the formation is very constant throughout the area. It consists of two types: the one, a true Radiolite, is semitransparent, hard, and brittle, with a color varying from milk-white to red or green, and is composed almost exclusively of the closely packed tests of Radiolaria; the other is an argillaceous chert, always red in color. The latter con- tains fewer Radiolaria and is analogous to modern deep-sea red clay deposits. The former corresponds to Radiolarian ooze. This large 405 400 REVIEWS area of deep-sea deposits clearly indicates a very deep submergence of this region, probably during the Jurassic period. The deposits probably occur in geosynclines developed at the edge of the permanent Australa- sian continental segment. R. Ca Mie Glaciology of the South Orkneys: Scottish National Antarctic Expedi- tion. By J. H. Harvey Pirie. Trans. Roy. Soc. Edin., XLIX, Part IV, pp. 831-61. Figs. 14, pls. 11, including one map. The South Orkneys have such a climate that the line of perpetual snow is practically at sea-level; the summer temperatures are rarely above freezing-point. The mean annual temperature is 22°7 F. The mean temperatures of the warmest and coldest months are 31°5 and 12°0 F., respectively. Foehn winds having passed over the central highlands sometimes produce as high a temperature in midwinter as in summertime. The islands are almost entirely snow-covered throughout the year. The resulting glaciers are characteristically antarctic in type. The surfaces are practically all covered with névé, there are no surface moraines, crevasses are rare except at escarpments, the whole mass of the glaciers shows stratification, and the glaciers terminate in sea-cliffs. The land: relief gives rise to various forms of glaciers: I. Ice-sheets, including: a) Inland ice. b) Ice caps of the Norwegian type. c) Much of the Spitzbergen type of ice caps. These are ice sheets which conform to the topography, overlying both valley and hill. II. Glaciers properly so called. a) Valley glaciers. b) Suspended cliff glaciers. III. Piedmont glaciers. These cover the low slopes between the mountain sides and the seashore. They end in cliffs from 15 to 20 meters high; their surfaces are uniform and snow-covered, having a gentle slope from the sea to the hills behind. They are fed by local precipitation and are not dependent upon snow-field reserves; they show well-marked horizontal stratification: IV. Glaciers of the coastal belt and shelf. a) Shelf-ice, such as the Great Ross Barrier. b) Ice-foot glaciers which lie in the zone between land and sea. They ’ are composed of layers of névé ice formed in place chiefly of drift snow supplied by wind action. REVIEWS 407 The surface of the ice is undulatory, conforming to the surface of the underlying ground. In one place the slope is opposite to the movement of the ice. The glacier as a whole may be traveling uphill for several hundred feet, but the total rise of the upper surface is thought not to exceed 20 feet. The glaciers are not notably advancing or retreating; in most places cliff terminations at the shoreline indicate advance, but occasional rounded ‘‘snout’’ endings bear witness to a slight retreat. The small number of ice falls from the cliffs also disproves any notable advance. Tn certain places the ice strata are slightly turned up at the glacier edges, but there is no sudden upturning at the end; this conforms to the other indications of but slight movement. Englacial material is rare and consists chiefly of wind-dropped rock; sand, pebbles, and bowlders are uncommon. In one place the ice grains were seen to be drawn out and arranged in sweeping curved lines which follow the direction of glacial flow. The pounding crystal faces are usually not plane but curved. From imperfect data collected on the speed of temperature waves through the ice, it appears that a wave of about 5° change in temperature _ penetrates 2 feet in about 2 days, and 4 feet in about 5 days. The islands seem to be the serrated tops of a mountain range which has been deeply dissected by glaciers while the islands stood at a mark- edly higher elevation above the sea. Hie 03), The Upper Devonian Delia of the Appalachian Geosyncline. By JosEPH BarRELL. In three parts. Am. Jour. Sci. [4th Ser.j, XXXVI (November, 1913), 429-72; XXXVII (January, 1914), 87-109; XX XVII (March, 1914), 225-53, Figs. 5. The Upper Devonian Oneonta and Catskill formations, that consist of alternating red shales and gray sandstones, of the Appalachian geosyn- cline in southeastern New York and northeastern Pennyslvania are believed to be “‘subaerial delta deposits [of westward-flowing streams} in a dry but not arid climate; a climate probably equable in temperature but subject to seasonal rainfall.” The Oneonta is 1,000 feet thick in the Catskill Mountains; the Catskill runs up into thousands of feet in thickness. The Portage and Chemung formations are the shallow-sea equivalents of the Oneonta and Catskill beds. The inland sea in which the former were deposited bordered the subaerial delta on the west and southwest. ‘The included map ‘“‘shows the shore line at the close of the 408 REVIEWS Devonian farther west than any previous map but the margin of the sediments farther east, except for the New Jersey strait of Schuchert which is here eliminated.”” The Upper Devonian sediments are believed to have extended northward beyond Lake Ontario and as far eastward as the margin of the present coastal plain; their removal over a great part of this area is referred to pre-Newark (Mid-Triassic), Jurassic, and post-Jurassic (Comanche and Cretaceous) erosion epochs. These Upper Devonian beds apparently formed a great piedmont plain that stretched westward from Appalachia; the Skunnemunk conglomerate (2,500 feet thick) is aremnant. This plain thickened from east to west and in so doing changed from coarse to fine sediments. The indications are that the drainage divide of Appalachia ‘“‘was at least as far east as the present 100-fathom line southeast of Long Island and New Jersey.” V- O28) Geology and Ore Deposits of the Monarch and Tomichi Districts, Colorado. By R. D. CRAwrorp. Bull. Colo. Geol. Surv. No. 4, 1913. Pp. 317, pls. 15 (including 4 maps), figs. 15. The Monarch district les in the southwestern part of Chaffee County, Colorado, on the east slope of the Sawatch Range. The Tomichi district, which is in Gunnison County, is on the west slope of the range and joins the Monarch district on the west. The sequence of formations is as follows: pre-Cambrian gneisses, schists, granites, pegmatite, and quartzite; the probably Upper Cam- brian Sawatch -quartzite (20+ feet); the mid-Ordovician Tomichi limestone (400 feet); the Upper Devonian and early Mississippian Ouray limestone (600-800 feet); the Pennsylvania Garfield formation (2,800 feet); the Permo-Pennsylvanian (?) Kangaroo formation (about 3,000 feet); post-Carboniferous quartz monzonite, granular rocks, porphyries, flow, and volcanic breccia; Pleistocene and later glacial and fluvio-glacial deposits; recent deposits. The Sawatch quartzite does not outcrop in the Monarch district. In the Tomichi district, the Garfield formation is only a few hundred feet thick, and the Kangaroo formation is wanting. Unconformities exist between the following formations: the pre- Cambrian and Sawatch, the Sawatch and Tomichi, the Ouray and Garfield, the Garfield and Kangaroo, the Kangaroo and _ volcanic breccia, the Pleistocene and older deposits. Regarding the interval between the Tomichi and Ouray, the author notes that “although one REVIEWS 409 or more stratigraphic breaks may be present, the beds show no angular unconformity.” An unconformity is probable between the Devonian and Mississippian portions of the Ouray. Pronounced folding and faulting occurred in post-Kangaroo (Permo- Pennsylvanian(?)) time. The primary ore minerals of the Monarch and Tomichi districts are believed to be genetically related to the quartz monzonite intrusion. _The principal ores produced in the Monarch district are of lead, silver, gold, copper, and zinc. They occur as replacements in limestone and dolomite, filling of fault fissures in limestone and quartzite, fissure veins in igneous rocks, contact deposits, and as deposits in pegmatite, gneisss, and schist. Further development is encouraged. The silver-lead ores now mined in the Tomichi district are chiefly sulphide ores. The iron ore bodies are of magnetite and limonite. The ores occur as replacements in limestone and dolomite, contact deposits, fissure veins, and as bog iron deposits. The future of the district lies in ‘the development of claims in groups.” Detailed descriptions of the mines of both districts are given. WeOe AR Reconnaissance of the Geology of the Rabbit Ears Region, Routt, Grand and Jackson Counties, Colorado. By F. F. Grout, P. G. WorcesTER, and Juntus HENDERSON. Bull. Colo. Geolj Surva No, 5, Part 1, 1913. Pp. 57, pl: 1: The Rabbit Ears region includes about 212 square miles in Routt, Grand, and Jackson counties, Colorado, ‘‘along and near the west end of the Rabbit Ears Range.”’ The geological sequence is as follows: Archean gneisses, schists, and granites; Permian or possibly Triassic ‘‘Red Beds”; the Upper Jurassic or Lower Cretaceous Morrison formation; the Cretaceous Dakota(?), Benton, Niobrara, and Pierre formations; early Eocene coal-bearing beds; post-Eocene volcanic breccia, dikes, and sheets; Pleistocene and Recent deposits. Unconformities occur between the Archean and the “Red Beds,” the ‘Red Beds” and the Morrison, probably between the Morrison and the Dakota(?), the Pierre and the early Eocene coal-bearing beds, and the coal-bearing beds and later deposits. Folding “began at or just after the close of Cretaceous time, probably 2) continuing for some time into the Tertiary. .... It is ‘‘probable 410 REVIEWS ' that igneous activity began in late Cretaceous or very early Eocene time and continued till very recently, and that there were several quite distinct periods.” VOU Permian of “ Permo-Carboniferous” of the Eastern Foothills of the Rocky Mountains in Colorado. By R. M. Burrers. Bull. Colo. Geol. Surv: No. 5, Part 2, pp. 6s-xor. Fig. 1- This report is concerned with the determination of the age of the Lykins formation, which was assigned by Fenneman to the upper part of the ‘‘Red Beds”’ in the Front Range of eastern Colorado. The Lykins formation, which varies considerably in thickness, con- sists of red shales and shaly sandstones, with a few sandy or shaly limestone beds. On the basis of the faunal evidence, the lower part of the Lykins is placed in the Pennsylvanian and an intermediate zone above is tentatively correlated with the Rico formation (Permian( ?)) of the San Juan region. ‘This leaves too-400 feet of shales to represent the Permian or the remainder of the Permian, the Triassic, and all the Jurassic up to the Morrison.” V. Oise The Geology of Central Ross-shire. By B. N. Peacu, L. W. HInx- MAN, E. M. ANDERSON, J. Horne, C. B. Crampton, R. G. CARRUTHERS. Petrological Notes by J. S. FLrerr. Memoirs of the Geological Survey of Scotland, No. 82, 1913. Pp. 114, pls. 8, figs. ro. The western quarter of County Ross is cut off by the great Moine thrust line. The Strathconan fault is the dominant one in the area considered. It is of the type described as a “wrench fault.”” The direc- tion of the fault line is a little east of north, and the lateral movement was to the northeast on the east side, and to the southwest on the western side. There are many Lewisian inliers thrust upon the younger Moine series. The Lewisian gneiss is a basement complex of various rocks of differ- ent ages and includes altered sedimentary rocks that were denuded and affected by contact metamorphism before the deposition of the Moine sediments. The Moine series comprises quartzose schists or quartz-biotite granulites and garnetiferous mica-schists or pilitic gneiss, representing REVIEWS 411 respectively metamorphosed silicious and argillaceous sediments. Locally the base of the series is conglomeratic. “a are divided into an upper and a lower silicious zone. Torridonian strata occupy most of the unmoved area east of the fault line. The beds are chiefly coarse, chocolate and red arkoses and pebbly grits which carry occasional layers of shale and flagstone. Unconformable upon the Torridon beds lie the Cambrian. The Cambrian is based with a gritty quartzite; the upper Fucoid limestones carry an Olenellus fauna. An apparent metamorphic transition of Torridon into Moine schists is reported, but no suggestion is made as to the age of the Moine schists relative to the Cambrian and Torridonian. The petrology of the district is marked by unusual lamprophyre dikes of minette and monchiquite relationships. The last twenty pages are given to a discussion of Pleistocene glacia- tion and glacial deposits. ©} The Archean Geology of Rainy Lake Re-studied. By ANDREW C. Lawson. Geol. Surv. Canada, Memoir No. 40, 1913. Pp. 115, pls. 9, map 1. Field study confirms the author’s earlier opinion (of 1887) that the Coutchiching sedimentary series is older than the Keewatin igneous rocks. He found that there were two widely separated periods of plu- tonic activity; to the earlier he proposes to confine the name Laurentian, and for the younger he introduces the term Algoman. Lawson’s classification of Archean formations from the top down- ward is as follows: 1. Eparchean interval—peneplanation. 2. Algoman. Vast batholiths of granite- and syenite-gneisses. 3. Seine series (Upper Huronian, Middle Huronian of some authors). Conglomerates, quartzites and slates. 4. Uplift, deformation and erosion, followed by depression. s. Steep rock series (Lower Huronian). Sediments and volcanics. Several hundred feet of fossiliferous limestones. Erosion which extensively exposed the granite batholiths. 7. Laurentian. Granites and granite-gneiss. 8. Keewatin. Chiefly volcanic rocks with intercalated sedimentary beds. Certain intrusive gabbros. g. Coutchiching. Sedimentary strata. Mica schist and paragneiss. 412 REVIEWS Only two divisions of the Huronian are admitted, and the Animikie and Keweenawan are not grouped with the Archean but with the Paleo- zoics. However, these two divisions are associated under the higher name Algonkian. He also proposes the name Ontarian to cover the closely associated Keewatin and Coutchiching. The major part of the report is given to a detailed discussion and description of the criteria whereby the Coutchiching is represented to be older than the Keewatin. The arguments presented are based upon structural relations, and actual contacts at which the Keewatin lies upon the Coutchiching. The conglomerate which Lawson formerly thought part of the Coutchiching, and which others used to show that the Coutchiching is younger than the Keewatin, Lawson now distinguishes as part of another, very much younger, group, the Seine series. The stratigraphical position of this series is not clearly established, and therefore the upper part of his Archean classification is not much more than tentative. TRE The Pre-Cambrian Geology of Southeastern Ontario. By WILLET G. MILLER and Cyrit W. Knicur. Report of the Bureau of Mines, Vol. XXII, Part II, t914. Pp. 151, illustrations 67, portraits 4, maps 13. The chief results of the work were to show that: (1) the sedimentary rocks have a basement of Keewatin green schists and ellipsoidal lavas; (2) the Grenville series were deposited upon the Keewatin lavas, but no erosional interval has been proved; (3) granites of two ages have been recognized; the older one is gneissoid and intrudes the Keewatin and Grenville rocks, the younger granite intrudes all the local pre-Cambrian rocks; (4) most of the metamorphosed blue limestones are classed with the Grenville series, but the conglomerates and some other sediments are younger and differentiated as the Hastings series; (5) post-Hastings igneous rocks are gabbro, basalt, and tuffs, and the Algoman granite which is later than the gabbro group. Because the great Grenville limestone series (94,000 feet thick— Adams) was pre-Laurentian, the authors think there is no special sig- nificance to be attached to the Laurentian as an epoch-marking time. They drop the terms Algonkian and Archean, and Proterozoic and Archeozoic. They do not reach definite conclusions about the correla- tion of the limestone conglomerate and other formations in the Madoc area. REVIEWS . 413 As an appendix the authors present their correlation of the pre- Cambrian rocks of Ontario, western Quebec, and southeastern Ontario. They follow Lawson in calling the older granites Laurentian and the younger ones Algoman. They drop the name Huronian because they think confusion of application has ended its usefulness. They group the sedimentary rocks of the classic Huronian district at Bruce as “Animikean,” and correlate with them the Cobalt and Whitewater series and the Ramsey Lake Conglomerate. All post-Algoman, pre- Keweenawan rocks are classed as Animikean. Pre-Algoman, post- Laurentian rocks are ““Temiskamian.” This name covers the Sudbury, Temiskaming, and Hastings series. For the group including the Keewatin and Grenville they propose the name ‘“Loganian.”” They think it unnecessary to retain the name Coutchiching, nor do they con- sider that the position of those beds has been proved to be below the Keewatin. Their classification is as follows: Keweenawan. Upper copper-bearing rocks of Lake Superior. Igneous rocks are both massive and in flows. Sedimentary rocks are little altered in horizontal positions. Unconformity. Animikian. Upper Huronian, Cobalt series, etc. Quartzite, arkose, con- glomerates in usually only gently folded positions. Great unconformity. Algoman. Lorrain, Moira, Killarney, Younger Laurentian granites. Generally massive, color pink. Temiskamian. Lower Huronian, Sudbury series. Quartzites, arkose, con- glomerate, Hastings limestone. Usually dips at high angles and is schistose. Unconformity. Laurentian. Granites and gneiss. Color typically gray. Loganian. Grenville and Keewatin. Highly metamorphosed. Lime- stones, iron formations, and igneous flows. at. ©) RECENT PUBLICATIONS —FIELDNER, A. C., Smiru, H.I., Fay, A‘H., AND SANFORD, SAMUEL. Analyses of Mine and Car Samples of Coal Collected in the Fiscal Years 1911 to 1913. [U.S. Bureau of Mines, Bulletin 84. Washington, 1914.] —Futron, C. H. Metallurgical Smoke. [U.S. Bureau of Mines, Bulletin 84. Washington, 1914.] —Ger, L. C. E. (Compiler). A Review of Mining Operations in the State of South Australia during the Half-Year Ended December 31, 1914. {South Australia Department of Mines, No. 21. Adelaide, ror5.] —Geological Survey of Canada. Summary Report of the Geological Survey, Department of Mines, for the Calendar Year, 1914. [Ottawa, 1915.] —GotptHwait, J. W. The Occurrence of Glacial Drift on the Magdalen Islands. [Canada Department of Mines, Museum Bulletin 14, Geological Survey, Geological Series, No. 25. Ottawa, 1915.] —GREEN, J. F. N. The Older Paleozoic Succession of the Duddon Estuary. [London: Dulau & Co., 1913.] ———. The Structure of the Eastern Part of the Lake District. [Reprinted from the Proceedings of the Geologists’ Association, Vol. XXVI, Part 3, 1915. London.] —Grover, N. C., Chief Hydraulic Engineer; and Hoyt, U. G., Horton, A.H., anp Covert, C. C., District Engineers. Surface Water Supply of the United States, 1913. Part IV. St. Lawrence River Basin. [U.S. Geological Survey, Water-Supply Paper 354. (Prepared in co-operation with the states of Minnesota, New York, and Vermont.) Washington, 1915.] —Hackman, Victor. Der gemischte Gang von Tuntijiérvi im Nérdlichen Finland. [Bulletin No. 39 de la Commission Géologique de Finlande. Helsingfors, 1914.] Ueber Comptonitginge im Mittleren Finnland. [Bulletin No. 42 de la Commission Géologique de Finlande. Helsingfors, 1914.] —Hancocx, E. T. The History of a Portion of Yampa River, Colorado River, and Its Possible Bearing on That of Green River. [U.S. Geological Survey, Professional Paper 90-K. Washington, r915.] —Havusen, H. Studier dfver de Sydfinska Ledblockens Spridning i Ryss- land, Jamte en Ofversikt af Is-Recessionens Férlopp i Ostbaltikum. Preliminairt Meddelande med Tvenne Kartor. Mit deutschem Referat. [Bulletin No. 30 de la Commission Géologique de Finlande. Helsingfors, Mars, 1912.] Undersékning af Porfyrblock fran Sydviastra Finlands Glaciala Aflagringar. Mit deutschen Referat. [Bulletin No. 31 de la Commission Géologique de Finlande. Helsingfors, Mars, 1912.] ; 414 RECENT PUBLICATIONS 415 —Hawaiian Volcano Observatory, Weekly Bulletins of the. Vol. II, Nos. 28, 29, 30. (Honolulu, 1914.] —Hay, O. P. Contributions to the Knowledge of the Mammals of the Pleistocene of North America. No. 2086. [From the Proceedings of the ‘U.S. National Museum, Vol. XLVIII, pp. 515-75, with Plates 30-37. Washington: Government Printing Office, 1915.] —HENNEN, R. V., AND REGER, D.B. Report on Logan and Mingo Counties. With Part IV, Paleontology, by U. Armstronc Pricr. Accompanied by maps showing topography and general and economic geology. [West Virginia Geological Survey, County Reports, 1914. Morgantown, 1or1s.| —Hicks, W.B. The Composition of Muds from Columbus Marsh, Nevada. [U.S. Geological Survey, Professional Paper 95-A. Washington, 1915.| —Hi1, J. H. Some Mining Districts in Northeastern California and North- western Nevada. [U.S. Geological Survey, Bulletin 594. Washington, 1915.| —Hinps, H., AND GREENE, F.C. The Stratigraphy of the Pennsylvanian Series in Missouri. With a Chapter on Invertebrate Paleontology, by G. H. Girty. [Missouri Bureau of Geology and Mines, Vol. XIII, Second Series. Rolla, 1o15.] —Hinxman, L. W., AnD ANDERSON, E. M. The Geology of Mid-Strathspey and Strathdearn, Including the Country between Kingussie and Gran- town. (Explanation of Sheet 74.) With contributions by J. Horne, R. G. CARRUTHERS, and C. B. CRAMPTON, and a Petrographical Chapter by J. S. Fretr. [Scotland Geological Survey, Memoirs of the, No. 74. Edinburgh: Morrison & Gibb, at Tanfield, ro15.] —Hosss, W.H. The Role of the Glacial Anticyclone in the Air Circulation of the Globe. [Proceedings of the American Philological Society, Vol. LIV, No. 218, August, 1915.] —Horton, A. H., Hatt, W. E., AND JAcKson, H. J. Surface Water Supply of the United States, 1913. Part III. Ohio River Basin. [U.S. Geo- logical Survey, Water-Supply Paper 3 53. (Prepared in co-operation with the State of West Virginia.) Washington, ro15.| —Horcnkiss, W. O., AND STEIDTMANN, E. Limestone Road ewe riale of Wisconsin. FW iceonsin Geological and Natural History Survey, Bulletin XXXIV. Madison, 1915.] —Hovyt, W. G., anp Ryan, H. J. Gazetteer of Surface Wiener of Iowa. [U.S. Goolonieal Survey, Water-Supply Paper 345-1. Washington, 1915.| —Hunrt ey, L. G. The Mexican Oil Fields. [Transactions of the American Institute of Mining Engineers. New York, 1o15.| —Ippincs, J. P. The Problem of Vulcanism. [New Haven: Yale Uni- versity Press, 1914.| —JIMENEz, CarRtos P. Estadistica Minera en 1913. (Boletin del Cuerpo de Ingenieros de Minas del Peru No. 81. Lima, 1o15.] 416 RECENT PUBLICATIONS —Jounson, R. H., AND Hunttey, L. G. The Influence of the Cushing Pool in the Oil Industry. [Proceedings of the Engineers’ Society of Western Pennsylvania, Vol. XX XI, pp. 460-87. Pittsburgh, 1915.] —Jounston, R. A. A. Gay Gulch and Skookum Meteorites. [Canada Department of Mines, No. 1533, Museum Bulletin 15, Geological Survey, Geological Series No. 26. Ottawa, 1o15.] —Jones, F. A. The Mineral Resources of New Mexico. [Mineral Resources Survey of New Mexico, Bulletin 1. Socorro, 1915.] —Kay, F. H. Coal Resources of District VII (Coal No. 6 West of Duquoin Anticline). [Bulletin 11, Illinois Mining Investigations. Co-operative Agreement. Illinois Geological Survey, Department of Mining Engi- neering, University of Illinois; U.S. Bureau of Mines. Urbana, 1915.] —KEELE, J. Preliminary Report on the Clay and Shale Deposits of the Province of Quebec. [Canada Department of Mines, Memoir 64, No. 1451, Geological Survey, Geological Series, No. 52. Ottawa, 1o15.] —Keves, C. Chart of the Geologic Terranes of Iowa. [Des Moines, 1914.] Iowa’s Great Period of Mountain Making. ———. Our Pre-Cambrian Rocks. Serial Subdivision of the Early Carbonic Succession in the Conti- nental Interior. Syllabus of a Course of Lectures on Geologic Processes and Geo- graphic Products. [Socorro: New Mexico School of Mines Press, 1914.] —Kwnoprr, A. A Gold-Platinum Palladium Lode in Southern Nevada. [U.S. Geological Survey, Bulletin 620-A. Washington, 1915.] —Kress, C. E., anp TEETS, D. D., Jk. West Virginia Geological Survey, County Reports, 1915. Boone County. Part IV. Paleontology, by W. ARMSTRONG PRICE. With Maps of Topography, and General and Economic Geology, 1914. [Morgantown, 1o915.] —KREISINGER, H. Hand-Firing Soft Coal under Power-Plant Boilers. [U.S. Bureau of Mines, Technical Paper 80. Washington, 1914.] —LamBE,L.M. On Eoceratops canadensis, gen. nov., with Remarks on Other Genera of Cretaceous Horned Dinosaurs. [Canada Department of Mines, Museum Bulletin 12, Geological Survey, Geological Series No. 24. Ottawa, 1915.] —LEE, W. T., STONE, R. W., GALE, H. S., AND OTHERS. Guidebook of the Western United States. Part B. The Overland Route, with a Side Trip to Yellowstone Park. [U.S. Geological Survey, Bulletin 612. Washington, ro915.| —Lewis, J. V. Determinative Mineralogy, with Tables. [New York: John Wiley & Sons, ro1t5.] —Locan, W. N. The Structural Minerals of Mississippi. A Preliminary Report. [Mississippi Geological Survey, Bulletin 9. Jackson, 1o11.] —Lowe, E. N. A Preliminary Study of Soils of Mississippi. [Mississippi Geological Survey, Bulletin 8. Jackson, 1911.] THE SUMMER QUARTER | 3 one OF THE UNIVERSITY OF CHICAGO A New P roposition Afford ot . oe A TYPEWRITER BY PARCEL POST ords opportunity for instruction on the same basis as during the other quarters of the academic year. ON TEN DAYS’ FREE EXAMINATION The undergraduate colleges, the graduate schools, and It is the ' the professional schools provide courses in Arts, Literature, Science, Commerce and Adminis- eration: ewe Medicine, eee eae } ‘ ivinity. Instruction is given by regular members p of the University staff which is augmented in the sum- Remington mer by appointment of professors and instructors J e Summer Quarter, 1916 unior ist Term June 19--July 26 gee 2d Term July 27--Sept. 1 the very latest addition to the great Remington fami iters. Detailed announcement will be sent upon application gto amily of ‘YB (ave s. Anew tothe and smaller Remington—built for the Dean of the Faculties needsof the professional man and thehome from other institutions. THE UNIVERSITY OF CHICAGO Mail this coupon today _ Chicago, Illinois é ie EE Se ES Remington Typewriter Company (Incorporated) 327 Broadway, New York | = _Send me a Remington Junior Typewriter, The Botanical Gazette price $50, on free examination. It is understood that I may return the machine, if 1 choose, within ten days. If I decide to purchase it, I Edited by John M, Coulter e agree to pay for it in ro monthly payments of Published monthly, with illustrations. Subscription $5 each. § price, $7.00 a year; single copies, 75 cents; foreign postage, 84 cents THE UNIVERSITY OF CHICAGO PRESS CUP < CHICAGO, ILLINOIS Daus Improved Tip Top Duplicator is an ideal assistant, always ready to reproduce clean, clear, perfect copies of letters, manuscripts, price lists, bills, in- voices, reports, drawings, menus, etc., etc. Don’t take our word, try it yourself for Ten days’ trial without deposit Tf not satisfactory, return and no questions asked. Makes rt50 handwritten and 50 typewritten copies in a few min- utes. Negatives require no washing out after using, and negative rolls can be used from 15 to 20 times. Complete outfit contains a continuous roll of Daus Non-Peeling Anti- Stick Negative Roll. Capsize,8$x13inches. Price $7.50 FELIX K. DAUS DUPLICATOR CO. Daus Building, 111 John Street, NEW YORK the American Library Association’s meeting to be held at Asbury Park, June 26th to July 1st and we should be very glad to have you visit our establishment while here. You will welcome an opportunity “ to see how the various functions of a large book concern operate, and to become acquainted with the facilities and methods of a business of such vital and economic importance to your profession. D7Ka0 doubt you will stop over at New York on your way to and from THE BAKER & TAYLOR CO. 354 Fourth Avenue © © At Twenty-sixth Street Wholesale Booksellers NEW YORK FINE INKS 4x» ADHESIVES For those who KNOW Drawing Inks Bternal Writing Ink Engrossing Ink e e 9 Taurine Mucilage Hig eins’ ¢ proto Mounter Paste Drawing Board Paste : Liguid Paste Office Paste 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 will bea revelation to you, they are so sweet, clean, well put up, and withal so efficient, At Dealers Generally CHAS. M. HIGGINS & CO., Mfrs. Branches: Chicago, London 271 Ninth Street Brooklyn, N.Y. Practical Projection Do You Want to Know What Your Handwriting : Reveals? WE have published a fascinating booklet of 32 pages on the sub- ject of graphology, by William Leslie French. Every reader of this publi- cation will be interested in comparing the examples of handwriting in judg- ing character and disposition of the writer. A copy of this booklet and 12 differ- ent patterns of The Spencerian Steel Pens will be sent postpaid on receipt of ten cents. If you can do good work with a poor pen you can do better work with a good pen. Silver plated Falcon, pens, smooth points, wear long, will not corrode. SPENCERIAN PEN CO. 349 Broadway Pe New York City for Schools For classroom work, lecture platform, or lyceum N use, in school or college—or for home entertain- ment and _ instruction—every requirement of a picture projection is met by the Bausch |omb BALOPTICON THE PERFECT STEREOPTICON Our recent application of the gas-filled Mazda lamp to projection gives the Balopticon an il- luminator far superior to the old-style arc lamp—and one using less current for equal power. The New Combined Balopticon projects either lantern slides or opaque objects, with instant interchange. It has the new lamp of 1,000 watts, using but 9 amperes of current, instead of 35 amperes as used by the arc lamp of equal illumination, and giving much better results on alternating current. Price $120.00. Other models for lantern slide or opaque projection from $20.00 up. Special terms to Educational Institutions. Write for illustrated catalog. Bausch €F lomb Optical GC. 554 ST. PAUL STREET ROCHESTER, N.Y. NEW YORK WASHINGTON CHICAGO SAN FRANCISCO Leading American Manufacturers of Microscopes, Projection Apparatus, Photographic Lenses, Engineering Insirumenis, Binoculars, and other high-grade Optical Products ~ VOLUME XXIV é NUMBER 5 THE A SEMI-QUARTERLY EDITED BY THOMAS C. CHAMBERLIN AND ROLLIN D. SALISBURY With the Active Collaboration of SAMUEL W. WILLISTON, Vertebrate Paleontology ALBERT, JOHANNSEN, Petrology STUART WELLER, Invertebrate Paleontology ROLLIN T. CHAMBERLIN, Dynamic Geology ALBERT D. BROKAW, Economic Geology ASSOCIATE EDITORS JOURNAL or GEOLOGY SIR ARCHIBALD GEIKIE, Great Britain JOSEPH P.IDDINGS, Washington, D.C. CHARLES BARROIS, France JOHN C, BRANNER, Leland Stanford Junior University ALBRECHT PENCK, Germany _ RICHARD A. F. PENROSE, JR., Philadelphia, Pa. HANS REUSCH, Norway WILLIAM B. CLARK, Johns Hopkins University GERARD DEGEER, Sweden WILLIAM H. HOBBS, University of Michigan T. W. EDGEWORTH DAVID, Australia FRANK D. ADAMS, McGill University BAILEY WILLIS, Leland Stanford Junior University CHARLES K. LEITH, University of Wisconsin — GROVE K. GILBERT, Washington, D.C. WALLACE W. ATWOOD, Harvard University CHARLES D. WALCOTT, Smithsonian Institution WILLIAM H. EMMONS, University of Minnesota HENRY S. WILLIAMS, Cornell University ARTHUR L. DAY, Carnegie Institution JULY-AUGUST 1916 THE GEOLOGICAL ‘SIGNIFICANCE AND. GENETIC CLASSIFICATION OF ARKOSE DEPOSITS - - - - - - - - - - - - Donatp C. BARTON AN UNUSUAL FORM OF VOLCANIC EJECTA - - - - WALLACE E. PRATT RIPPLE-MARKS IN OHIO LIMESTONES_ - - - - - Cartes S. Prosser THE RELATIONSHIPS OF THE OLENTANGY SHALE AND ASSOCIATED DEVONIAN DEPOSITS OF NORTHERN OHIO - - - - - - - C. R. STAUFFER . EVOLUTION OF THE BASAL PLATES IN MONOCYCLIC CRINOIDEA CAMERATA. I Herrick E. WILSON VARIATIONS OF GLACIERS. XX - : = bids tease ar - Harry Frerpivc Rep REVIEWS - - - - - - - Ea = < - = i ee z uh is RECENT PUBLICATIONS - - - - e = = A y es a 4 = THE TUNIVERSITY OF (\CHICAGO” PRESS CHICAGO, ILLINOIS, U.S.A. AGENTS THE CAMBRIDGE UNIVERSITY PRESS, Lonpon anp EDINBURGH KARL W, HIERSEMANN, Letpzic THE MARUZEN-KABUSHIKI-KAISHA, Toxyo, Osaka, Kyoto Pee THE MISSION BOOK COMPANY, SHANGHAI 4l7 450 456 476 488 THE JOURNAL OF GEOLOGY | EDITED By THOMAS C. CHAMBERLIN AND ROLLIN D. SALISBURY With the Active Collaboration of SAMUEL W. WILLISTON : ALBERT JOHANNSEN Vertebrate Paleontology Petrology STUART WELLER ROLLIN T. CHAMBERLIN Invertebrate Paleontology Dynamic Geology ALBERT D. BROKAW Economic Geology The Journal of Geology is published by the University of Chicago at the University Press semi- quarterly, on or about the following dates: February 1, 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 follows: For Canada, 30 cents on annual subscriptions (total $4.30), on single copies, 4 cents (total 69 cents); for all other countries in the Posta] Union, 53 cents on annual subscriptions (total $4.53), on single copies, 11-cents (total 76 cents). {| Patrons are requested to make all remittances payable to the University of Chicago Press in postal or express money orders or bank drafts. The following agents have been appointed and are authorized to quote the prices indicated: For the British Empire: The Cambridge University Press, Fetter Lane, London, E.C. Yearly subscriptions, including postage, 18s. 8d. each; single copies, including postage, 3s. 2d. each. For the continent of Europe: Karl W. Hiersemann, Konigstrasse 29, Leipzig, Germany. Yearly subscriptions, including postage, M. 22.65 each; single copies, including postage, M. 3.80 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, including postage, Yen 1.50 each. For the Chinese Empire: The Mission Book Company, 18 Peking Road, Shanghai, China. Yearly subscriptions, $4.00; single copies, 65 cents, or their equivalents in Chinese money, Postage extra, if mailed direct outside of Shanghai, on yearly subscriptions 53 cents, on single copies II cents, Claims for missing numbers should be made within the month following the regular month of publication. The publishers expect to supply missing numbers free only when they have been lost in transit. Business correspondence should be addressed to The University of Chicago Press, Chicago, III. Communications for the editors and manuscripts should be addressed to the Editors of THE JOURNAL OF GEOLOGY, the University of Chicago, Chicago, Ill. Contributors are requested to write scientific and proper names with particular care and 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 leading articles will be printed as ordered, of which 50 (without covers) will be furnished gva¢zs, the actual cost of the remainder (and covers, if desired) to be paid for by the author. The actual cost may vary, and will depend upon the amount of work in re-making the pages into forms, presswork, paper, binding, etc. Separates containing half-tones may be expected to cost somewhat more, the increase depending upon the number of cuts and the amount of work required upon them. Entered as second-class matter, March 20, 1893 at the Post-office at Chicago, Ill., under the Act of March 3, 1879. ‘hee m a ee a oe VOLUME XXIV NUMBER 5 THE hROURNAL OF GEOLOGY JULVAUGUSE 1016 THE GEOLOGICAL SIGNIFICANCE AND GENETIC CLASSIFICATION OF ARKOSE DEPOSITS? DONALD C. BARTON Arkose has been held by different geologists to be significant respectively of several different types of conditions at the time of its formation. By Walther,? for instance, it is considered to be so distinctive of desert formations as to be, next to salt deposits, the the most important index of. the desert origin of a formation. Mackie,’ although not mentioning arkose by name, in his discus- sion of the significance of fresh feldspar in sediments seems to con- sider the rock especially characteristic of deposits that have formed under rigorous climatic conditions. Von Hauer‘ believes that arkose is especially characteristic of coal-bearing formations. Shaler is of the opinion that it is formed when a granitic terrane, long under moist temperate climatic conditions, is exposed to more rigorous conditions or to marine or lacustrine transgression. Mans- field® believes, on the other hand, that the conditions for the formation t Portion of a thesis accepted in partial fulfilment of the requirements for the degree of Doctor of Philosophy at Harvard University. 2 J. Walther, Das Gesetz der Wiistenbildung, 2d ed., p. 174. 3 William Mackie, Trans. Edin. Geol. Soc., VII (1898), No. LV. 4 Franz von Hauer, Die Geologie, 1875. 5 W.S. Shaler, U.S.G.S. Monograph XX XIII, 1899, pp. 50-55. ®°G. R. Mansfield, Bul. Mus. Comp. Zoél. Harvard, XLIX (1906), 293-04. Vol. XXIV, No. 5 417 418 DONALD C. BARTON of arkose are intermediate between these extremes, and that a moderately cool and arid climate such as would prevail at mod- erately high altitudes in the lee of high mountain ranges or in con- tinental interiors would more probably be suitable. The present - paper is an attempt to delimit the significance of arkose. The fundamental conditions essential for the formation of arkose’ are: (a) a granitic terrane, (b) conditions favorable to the disintegration of the granite or gneiss with but slight accompanying decomposition, and (c) conditions favorable to the erosion and deposition of the débris of disintegration with merely slight loss of the feldspar. In the investigation of the regions which today can supply débris of disintegration for the formation of arkose (Figs. 1 and 2), it was found that disintegration is much more widespread than is perhaps usually realized, and that it takes place in marked amounts under practically all the conditions under which a granitic terrane is exposed (see Table I, a list of the occur- rences of disintegration which have been observed by the writer, or which he has been able to find described in the literature, together with a tabular view of the conditions under which the disintegration is taking place). The investigation of the conditions under which the disintegrated material could be eroded and deposited as arkose seems to show that in some cases the conditions favorable to the disintegration are likewise favorable to contemporaneous erosion and deposition of the disintegrated material as arkose, as, for instance, in desert regions, and that in other cases erosion can take place only after some change of conditions, as, for example, a change from the conditions of a moist temperate climate to those of a semi-arid climate. In yet other cases, erosion may take place contemporaneously with disintegration but be followed by decom- t Arkose by original definition and according to most general usage is a rock formed of the relatively undecomposed débris of granite or of rods of granitic mineral- ogical composition. It may be thought, however, that the original definition should be extended to cover feldspathic clastics derived from the disintegration of syenites, diorites, gabbros. Feldspathic clastics of this type, however, should be rare, since there are practically no purely syenitic, dioritic, or gabbroic terranes, and since the plagioclase feldspar of the diorites and gabbros is more or less readily decomposed. No specimen of this type of feldspathic clastic has been seen by the writer, and only one or two reputed occurrences are reported in the literature. Loca EK S.E. Ireland . lies Dartmoor, Engl Forez, France. | Plateau Central) | | | | | NTER TEMPERATURE Morvan, Francéut the same as above Hohkénigsburg Alsacee ice Sib Heidelberg Shee, . | Adlersberg, Thi, . Grimsel Pass, A. . Mer-de-Glace, A d Aswan, Egypt. r. Pyramids of Giz, Bushmanland, ¢. Tits th a) Himalayas....... Ceylone sos. | Lung-wang-shar, ie Around Inland § : Mt. Chocorua, 1 .. Jackson, N.H..... Rockport, Mass. Sykesville, Md.) . Washington, D.|. H Richmond, Va... Georgia...... 4. \t- ) Tron Mt., Mo... Mt. Stuart, Was. Wasatch, Utah ,. Butte, Mont.. .). Pike’s Peak, Col .. | Globe District, 2. California: si Sierra Madre, . Lower Califerni¢ . Valparaiso, Chil, _ Sao Francisco, E |. Min Mean sl) 2186 43 F. (Mean min. Jan.) 7F. 36 F. Lil Seaevaiten este 34 F. St a octets 34 F. 21k. 32F. 45 F. 62 F 30 F. 55F aie oe 80 F SEAR Io F. B aieichinecaye 32F —13F. 65 F. (For Boston) —8F 44F 45F 74 ¥F —29F 24 F eee ee: 6F. 22F 50F Re oeyete 62F a Vnchaparenete 82 F RAINFALL aL 4 + in Oct.—-Nov. REMARKS Medium-grained granite Coarse porphyritic granite; fine- grained granite not affected Coarse and porphyritic granites; fine-grained granite not affected Coarse, porphyritic granite Coarse, porphyritic granite Medium-grained granite Monthly Summer] Winter 4 in. 5 in. 4 in. 5 in. 4 in. 3 in. ) ° ° fo) 0.3in 12 in. 1.5to4|1.2to2 in. in. AAT 8\| (Seyret te in June and July oin. |r to 2 in. Medium-grained gneissic granite Medium-grained gneissic granite Coarse, porphyritic, biotite granite Coarse, porphyritic, biotite granite Medium-grained pyroxene granite Medium-grained, with a slight amount of biotite _ : Medium- to coarse-grained granite; hornblende granite Fine-grained biotite granite Moderately fine mica granite Medium-grained mica granite Fine- to medium-grained, porphy- ritic mica granite Medium-grained granodiorite Medium to coarse-grained mica, hornblende, quartz monzonite Coarse to coarse and porphyritic granite Granodiorite Granodiorite Fe-bearing minerals decomposed Fic. 1.—Distribution of the occurrences of marked granular disintegration as reported in the literature or observed by the writer. TABULAR VIEW OF THE OCCURRENCES OF DISINTEGRATION AND OF THE CONDITIONS UNDER W TABLE I HICH THE DISINTEGRATION Is Taxtnc PLACE RAINFALL 3 DEPTH To Warcu| AMOUNT OF DE- SuaoreR TEMPERATURE Winter Tew: Lat. | Lone. | Torocrapay ELEVATION BOLLUATION ore Depta to Warcy Dis- | tye coxposrtion |q COMPOSITION | VEGETATIVE CovER- s SERRE LocaLit¥ GRATIO INTEGRATION EXTENDS TSSaS SHowN By Distn- ING Monthly REMARK TEGRATED RocK | Yearly — . . la Max Min. Mean | Max Min. Mean | Summer} Winter windee 53 N. 7 W. | Mature to old WRONG |s6o565e58e8 cn00| endo cnnacecbosonocuee: | poe asecedunenccel be asudenonenonoL Grass and brush |........].......- Pe al Meee |S nea Pea can AW in gp. Ireland «---+" an F ‘ sses} Soin, ‘ sin. ogee stN.| 4W. 1,500 ft. General 6 to rs ft. 3 to8ft. Local-| Considerable Heather and moss | 68F. |........] 60F. |........| 3 3 i ; i Partmoor, England. «-« a ds of f bun (Mean max. into) i fen iS Seite Soi fcada. Sin. | Medium-grained granite reds of fee ; : min. 5 Jan.) ae 4s N. 4 E. Nod 1,500 ft. 1 Noticeable Grass and woods | ae automa fey WS | boats Sse 36F 9 ; . + Forez, France: i ncuaee 46 N. 3 EI 5 Genera To to 30 ft. to 4 ft. thoughiaoe 95 (Man Sine 7F. 36F. | 32in. | gin. Sin, | Coarse porphyritic granite; fine- Plateau Central, Fre . B ltora eK a ' : - F ee advanced grained granite not affected / nas i eneral 5 to 25 ft. 1 to 4 ft. oticeal ‘ASS 2 v ita ranceneeenerre?e 47 4 ae ae Grass and woods About the same About the same as above |........{e. ccc ce cde sec ee Coarse and Rerahyritic granites; very advanced | fine-grained granite not affected mi wes Mts., fohkanigsburg, Vosses 3.N. E. | Mature 1,500 ft. General 8 to 20 ft. plus 1 to 4 ft. Noticeabl Fores : eau Pon ~ Hahn Le iereke|| <4 7 r 04 aaa oe INTE = [logo nndllocoascus (YO |isspe3 fon) bonedscn 34 FP. UY. |andosue | Rratstercntete Coarse, porphyritic granite very advanced . j E. | Mature 800 ft. General to ro ft. 1 ft. Noticeabl F a4 FE wah A aie Heidelberg Shect..--- +++ 0+" 49 N 9 s thee et Forests) nny | sestetaieers | eitenietars CEOS | eons cq) soneoune 34 F. Gans | okereerers | hiertaenrahe Coarse, porphyritic granite . very advanced ey 3 E. | Late maturit AI, = Wesgandocacooars to 8 ft. 3 Noti oF 32 F , . ; Adlersherg, Thirringerwald . I SXEN 11 y sto 8 ft 1 to 2 ft “though not Forest = © ||Ratemecel| Soisrcis sta] tieeneccisn hoe eee 2F. RE) oh Vl ogoacian pucknond todirotesd Medium-grained granite very advanced 0 , glaciated 6 i f i J i ig! ei-K 9 . - = Grimsel Pass, AIDS. .-+-07+* 47 N. 8 E. Boung, slacia e 500 ft Only valley seen} Superficial 8 8 = |................ Very slight Light or none (Rigi-Kulm) RX) Ona Peettors Sas| eoodo tod Runoporal osostned ncacacd cera Medium-grained gneissic granite Merde-Glace AlpS.ceee cere 40.N. 7 E. surface of alpine 5,500 ft. Surface Sicsitiel = ane cansanoccand Very slight NOT CS 8 opeerond bsdodard sopaceed onde tcl hoo ona) saqdecad acto camo maancaurd ke Medium-grained gneissic granite etle-Grace, “Oi glacier eS 3 a 7 ; LE. | Shall oun 00 ft. i ia 2 | a 2F. R : nae Be Oh ‘ Avan, Egypte esvceeeeerers a4.N. | 32 alley 14 400 General Superficial, locally 10 ft.) None Torche mort part None 112 LN eee aS 1 oof 4sF 62 F. a) ° ° Coarse, porphyritic, biotite granite Pyramids of Gizch, Egypt.-..| 30N. | 31 E. paremids ona Less than 400 ft. Surtaceiof the ees. Sue) ileatosaBencoodoas Very slight None 104 F. 64 PF. 82 F. 75 BR. 30 F. ssF. | xr.sin ° 0.3 in, | Coarse, porphyritic, biotite granite Gindbssserececeenernenrerer SINE |) Ky BO) BOL be phaneanaaenpsced Valley sides artlyssuperticial aan | ettectciseretel erie Very slight 1\ Ci nn Soe poeas ssonasee| Poasomadl adoot 10 dllecsoacedlonqostsee MEP} Segdosn|lavoasnae CORSE al a than roin S, Africa....-| 30 S. fe} 105 |k. Anneetoadoo sons dlacsgcontonoooaisend sodmoaodada apie SUDETCia] Me | (essen Very slight None or very light ]........]........ sr F. EPI. llsnananas|jasoceoud Less Hushmanland, S ~ (Coldest (Warm- han month) est ro in month mean) Himalayas, .ss+ 20 GonGuGOO 34. N. | 76 E. | Young, moun- Above 44,000 ft. | Exposedsurfaces} Superficial 8 = |..............-. Very slight None se GB reaetoces | frases | Sosnreers AA chal loemnocod laaaqooenl Ketneacd Hatheeah hociicanc! tainous Ceylon. sss ccrcseesereeeers 8N.| 80 E. | Mountainous |.................. General Few feet GQianF nna ntencucessane Forests smn | sect tetcseys £2 eel esood onllondnooed 80 Fr Over |} in June-July so in. |} in Oct-Nov. Lung-wang-shan, China ..... 4x N. | 124 E. | Mountainous |................. Rolling alley “Tiefgreifend”” Slice h (inne eerste erst Nome et eteteyetstetite! | nverersrererax 75\Be: [ieee = Siete] cereinieiere 10 F. Op ockecerel hecannc in mountains Around Inland Sea, Japan ...] 38 N. | 138 I. | High coast hills ].................- @ayae flan canueooeooegs sagiceannn Slight Very slight egneyee Haag conelicononce FAPLOH: lence, ool ooumenae 32 FP. 85 in. 7in oin Mt. Chocorua, NH... 0... 44.N. | 71 W. | Mature, moun- 3,500 ft. Summit and Mostlyjsuperficial mes! Sete eetetletertetiele Very slight Mostlysnome;ssOme) frelateretersiel| tierra erorete|(evehsier eh tent |(cxeverercl@tatel | etetetareratere | (etetsteyar sists | eturstatertcts | ciateearencitte | Rretetetereraes Medium-grained pyroxene granite tainous flanks of forest : mountain Jackson, NW... AUCOURUOD 44.N. | 71 W. | Mature, moun- joo ft. Valley 5 ft. plus Slight Noticeable Rorestuand serassia| ererrelaere | eeetrerersters)| chetereteterars | steecate iS AllapBeearcl MEROOROA HoBobood loadanatd| actrorscac Medium-grained, with a slight tainous though slight ; i é amount of biotite Rockport, Mass......seee ee 43. N. | 71 W. | Glaciated pene- zoo ft. Chiefly shown in| Superficial = =... . 0.0 eee Noticeable Lichens 100 F 45 F 65 F 70 F. 45 in. |3 to 4 in./3 to 4 in.| Medium. to coarse-grained granite; plain glacial boul- though slight (Bor ) hornblende granite ders oston Sykesville, Md... ..cssseees 30 N. | 77 W. | Old 4oo ft. Surface generally} 30 ft. plus and minus 1.53 i tn | petetetor sdoaRenobeN Grass and forest See Washington Ant cojlooboanoo Inoooacool rar Sets | aersrane saislecsia's ....| Eine-grained biotite granite Washington, D.C, .s.sseeeee 39 N. | 77 W. | Old 50 ft. Surface generally] 20 to 80 ft. Considerable Combis at| Grass and forest 104 F. 36F. 75F. 78 I. jrin, |3toqgin| gin, | Moderately fine mica granite surface com- plete ; P i * . Richmond, Va... cece eserves 37 N. | 77 W. | Old 50 ft. Surface generally| 10 ft. plus 5 ft. plus Considerable; at} Grass and forest SeeWashington Jue fiuleieecewe|eeercees|aeesnese|isinest ss) sssiern es Medium-grained mica granite " surisce com- ; plete : A . 0 t , ' UTA snungonnoasnn Dnacen 34.N. | 84 W. | Old 1,200 ft. Surface generally| Incipient decay to 350 ft.| 95 ft. Considerable; at| Grass and forest 100 F 5s F. 75T. 73K. | -8F. | 44F. | goin. [3 togin.) sin, | Fine. to medium-grained, porphy- (around Atlanta) saree com- (tos - ritic mica granite plete lanta ‘ ‘ Won Mt, MO csseseece ees 37 N. | 01 W. | Old 1,200 ft. Surface generally} 20-80 ft. 20 to 8o ft. Consider ble; upeobonaanonsocosdn 105 F 45 7F (I 76 P. 457 74F. | goin 4in |2to2a.s5 surface com- in. Mt$ Y plete : i i i Stuart, Washi.....seeee 47 N. | 121 W. | Ruggedly moun- 9,470 ft. Summit and ex-| Superficial 8s [www see eee Slight Wight) ~ © © |! steyeyerer CLASSIFICATION OF ARKOSE DEPOSITS 4I9Q ae | Set ATU ISB hy eeieers eerie! ruse Fic. 1.—Distribution of the occurrences of marked granular disintegration as reported in the literature or observed by the writer. DONALD C. BARTON 420 Ss ! “IOWIM IY} Aq paaresqo IO 9INYVI19}I] VY} UT poz1odar se s}isodap asoyIv Jo 9dUEIINIQ—"z “OIT . or or oe co oF 09 0; cy oor ‘on roy co oF ont on oir oar ou oy Ost oer coy or ou oo g = Ra oe ——— oo = | ve vy — wes se uted’ = ! os 2 ' "'S : i. , 4 > a 0 [ =! P 4 ; fs I a MM ‘| o vie 2 fais i PR & 7 os oF "oe oz ot oF Be “op. Os °° CLASSIFICATION OF ARKOSE DEPOSITS 421 position of the feldspar, with the consequent loss of the material as a possible source of arkose. It was therefore found impossible to limit the significance of arkose to significance of any one or two special sets of conditions. The arkose deposits forming under the different conditions should expectedly be of certain characteristic types, which as a matter of fact agree with the types of arkose deposits as they are found (see Table II). The genetic classification of arkose deposits which appears in the following pages is therefore intended to embrace these various types of arkose deposits. GENETIC CLASSIFICATION OF ARKOSE DEPOSITS Arkose deposits may be divided broadly into two classes: (a) those formed directly through the effects of rigorous climatic conditions; and (b) those formed, at least indirectly, through the effects of moist and more temperate climatic conditions. The latter conditions allow much decomposition, which, however, commonly takes place at a slower rate than the disintegration. The arkose formed directly or indirectly under these conditions therefore has feldspars showing considerable decomposition, has in many cases a matrix of argillaceous material derived from the more easily decomposed grains of feldspar and other silicates, and is associated with beds of argillaceous material derived from the totally decom- posed portions of the rock. The former conditions are unfavorable to decomposition, and the arkose deposits forming under them have comparatively unaltered feldspars, have little or no argillaceous matrix, and are not associated to any great extent with argillaceous beds. The distinction, however, is not absolute. At Aswan, Egypt, in a region in which no rain is recorded over a period of many years, the disintegrated granite in some places shows marked decomposition, and the feldspar of modern arkose deposits in the beds of the wet-weather streams of the region is deeply decomposed. In regions of moist, temperate climate it is not uncommon, on the other hand, to find below the zone of complete disintegration a zone of rock which to the eye seems fresh but which crumbles readily under blows from the hammer, and from this rock, within a region of moist, temperate climate, it would be possible for arkose with relatively fresh feldspar to form. DONALD C. BARTON 422 Ssurppoq -ssoia. ‘syuuid = dorpures ‘syovIo-pnyy ‘ossoOyIe ov Spoq AY-AJ1OJ JO JNO DIT, 97IS zjrenb oy} Jo asvyd ouo ssoyry yy ynq ssoyre you ATqissog a}ULIZ 9Y} OVUL SapeIs ssoyly IPURIS IY} OUT pue ayizyrenb our YoOq uoTIsuRIy, uolyPoy -1}81}S OU IO 9]}}I] YIM osoyTy peyesys [Pa speMAII4) paevadn sur -peis ‘aseq ye osiv0d ‘osoyIy poqysosse Ay100d ‘ensure surery xLI}eul ouy ut redspyay poAvoop pure zyienb svjnsuy X1I}eU OUT pue vor ‘siedspyey ‘zj1en() spueq AAvoy UT I}IIIsS pue asoz}1eN¢) siedspyoy puv zyrenb ‘sures3 SULYPOIO}UL «= “SUL YOOT-OTFIS]O YT asSeIO -1svid pue zy1enb ssepoyj10 JO SUIRIS Iv[NZuUR sso] 10 dIO[T redspyay pue zjienb jo sayqqod pure SuIvIs PopuUNOI sso] IO IIOP, asoyie pur ‘9}vI10 -wo[su0D pur 9uU0}js -purs ‘sys ‘aeys poy APULIS PURBES asOyLY asOYIV 9}1Z}12NQ) aye190 -Wwosse ‘a}TWIOTOp ‘931Z -yarnb ‘oysemAoIs ITM ayUPIy dSOYIV O}IUVIL) asOyl peed aecs dSOYLV ayoeMmAdIry 9}1Z}1"NC) ayIeM -A913 pur ‘9}eI1IWO[SUO0D ‘oyizjienb ‘oyWOTOp YIM asOYIY “VJ oor dyIVMADIS PUL ISOYIY *}J CO MpeMAIIS pure 941] -[Is1e pure ou0yspues pay "yy OFZ NVIVENVO-dad VOIMANV HLYON SyILUIy, asOyIY JO 1o,vILYD u0rq9ag “SIM ‘TOA 4S “UOIJRULIOLT UO STUWLYy sInoT 95110 ALO Jowoedns ayv'y ‘yo1y4stcy soyousg ‘9}R[S zyienb ‘asoyly “WUO “pen? JAY Youd] “WUO “WLASIC SuIssIdINN ‘asoyie SUTUeOSTUIO T, ‘quo ‘WeqoD ‘osoyie oureI0'T ‘yug ‘Ainqpng ‘asoyie Pyo soddog IX ‘X ‘TIT sesuey ‘yoy 4G ‘OAR APMLIION :daqGan() (qsvoo ysva) Avg uospnyT UOIZdY PUP UOT}RWIO J NOILVAYASHQ IVNOSYTG HONOYH], YO AWALVAALI'T AHL HONOUHT, UAL, AHL OL NMONY Siisodaq ASOMAY AHL AO SNOILVIOOSSY GNV ‘AALOVAVHD ‘AONTANIOIG AHL AO MAIA AVINAV, Il WlaViL 423 CLASSIFICATION OF ARKOSE DEPOSITS 931z}1enb ay} Jo osvyd v asoyIY oq1zy1enb dq} JO [eseq 94} SI dsOxIe oY, poy10suyy aqzyrenb ay} Jo aseyd yeseq ayeUIpPIOgns v A][vII SI BSOyIY SUIPpoq-sso19 ‘syurid doipurei ‘syor19-pnyy oTYydr0ureya JL XLIVeUL OU UT Ivdspyoy pue zyienb = papunor AyeyeIIpoyy X11] VUL 9} 119S-Z}1eND poureis -ouy =e ul siedspjay pue zyienb papunoil Ajaye1opoy[ 9}VIDWIO]SUOD IISOYLY auojspues Avis pure poy auo}SOUN'T SsIous puv oTURID 9} eIOWO[SUOD VSOYIY a}IZJAeNC) UOJSOWNT dUO}SOUN] puv ozTURIT) aSOYIV PUL 9}1Z}.1eNC) dU0}SOWNT ssious puv o}1URI4) d}PIOWO[SU0D =| pue asOYIe YIM 9}1Z}1eNC) { sUO}SOWN] pur ouoyspurs d1y}edspyoj pue ‘oyxoemAois ‘9327S -OFIV YIM 9} eIIWIO[SUOD asoyie pue }"zy 1&2 "yy O00'S —oS¢ “yy OOS‘T —006 “yy 00S‘z —oof JUOWLIIA 0} VISIOI) DISOYIe uvlquiey jo oseq oy], NVIYaNVO poesoydioweyz yy SNOIETIIBIV 9S0zZ}1eN() IIOP Z11eNC) OqT[Is1v asoyre =) pue 9}e19WIO[SUOD ssOyTV { 9U0}SOWTT aque aISOyly d}IZJILNC) d[Is1e puv 9uo}soUTT asoye 0jUT SuIpess ‘ouojspues ayyM pue ‘por ‘uMmoIg auo}spues yurd pue o71yM ‘ouo0}sUMOIG ‘QUO}SpURS Pot PUL 9}IT W9°0 “VJ COg'I *qy OOS‘T 91ST [A JO ses JAOPDAQDT JUOUIIN A pur s}yosnyo -eSSeT, ‘UOTVEUIO JUOUTIOA Aasiof MON ‘9}1Z}JIVND a{[IAUO}STIpav yy euI[OIR) ‘UOT}BUWIOY prIqaous SIDS pay) Udaysvy YON oJOD “pend ouNUIYS ‘d}eIOWIO[SU0D ONeOFT ‘Od ‘uoAuR) Jloq;y ‘UoTRWIO, 39g UISUODSI\ ‘QUO}SPURS LIUITIO DONALD C. BARTON 424 saoRJINs poyeq -uns pu sjurid dorpures sey yoryM ‘ayeys pot oy} YA pappeqiejut — pappaq-ssoi9 yonut 10 dAISSeUI—o}TURIS OJUL Sapeis ssoyIe AT[RIO'T auojspurs 94} jo oseyd yeseq & SI 9sOyIV ay, UOI}VIYIVLIYS JULIET ayUeis 24} OUI sapvis ssoyIe jeseg SYILUIIY Jojoo ut Avis yIoq *XLIJeUL OUTF YIM 9S0z}1eNb poureis-ouy ssoyie soddq ISIVOI—XU}VU UT T oye) asOyly J oor > a[eys pay yjoor > 9} P1IWIO[SUOD pue |eoo pur 9ssoyie ‘gjeys pue ouo}spurg “yj OOl < asozyienb uazjo a10ur ‘o171 “URIS SOWIT}IWIOS ‘asOyIR [eseg SNOUAIINOAAVO aqui) asoyle ystuse13s pue YsIppory “3y Ob—of souvivodde ur dy1URI4 QUO}SPULS NVINOADA SSI9US-o} TURIN) ayizjrenb pure ssoyly “jf Og—z doy dUO0}SOUI’T |, j Jah ((olora ye o1V1ZJ1eNb saurosaq ‘WLIO; a}¥IS pur dUOJSOUT’T { -Tun a10ul ‘19uy ssoyire raddyQ asoyly “J Oof xiyew ouy ur zj1enb —007 pue siedspjoy ‘aouvired aT[Ls1e “AI “44 COS -de ul o1jtueis ‘asoyie yeseg —OST NVIaNTIS sslous pur oye) saseyd pur x11}euUl SOI YIM 94}1Z}IeNC) ArAydiod pur oyu ayizjrenb pur ssoyxly “yy Of Arp pur lay oie asizoo ‘ssoyry |‘pues useIS YIM pues ‘IID [ panuyuoj—NVIAANVO asOYIY JO JojovieyD u01q99S UISeg JOSUvdvIIT Ny purjusei4y ‘auo}spurs O¥[eBT "HN “voy AWT ‘Zl1y ‘vaqsig ‘a}Izj1eNQ) esjog PULOYLLYQ ‘VUOJspuvS URdvIY uolsoY pue uoTzeUWIO,T Pero A Lay di 425 CLASSIFICATION OF ARKOSE DEPOSITS SOSU9] IO spoq YIy} Ur ssoyIy gsoyle Jeopie~sng se owes ‘Oyo ‘syoRIy 9] -das ‘syurid dorpures ‘syaei9 “pout yy 3eyS ~PePPpeq -SSOID YONUE v[RYs Jnq spoq ][V UOWUIOD SUIPpeq-sso1g a WERE PMloy SME BO REG UOWUIOD SUIPpeq-sso1g as SIPEKEL PERI SMR BO ENE S[ISSO} -quejd §=SurArzed = suorjeyeo -I0JUI 9[eYS pol YM ooyxIy P9IPPeq-sso1o ssoyly SnOIaJIUOG -ID 94} Jo asvyd [eseq osoyIy ISOYIV JVOPIVSNG Sv sues [WHoJeUI SNOsdeTIISIV JO XIIARIY S}IIS JISOYIV 0} pur ISOYIV IIIUVIS 0} 9} eIIWIOTS -U0D ISOYAV WOT] SOLIVA ISOYIV Ystppot SyxoOr TLV ysippor ‘ysnor ‘asivog YSIpped pue 9s0z}1eNQ) AvIS SNOIdBOTUL ‘gouvivedde ut o1jtuRIs ‘pouieis-ouy pue wIniIpoyy _ Avi8-ysippor *A}}118 ‘asrv0y ARIS ‘9SIvOD aseq 94} }¥ 9710 -WO]sU0D pue adIsOyIV pure o9]eys pot suwos Y}IM ouUO}SpuRS AVIS 0} *zj COI‘E YSIMo]aA pue Ysippoxy —oo*z “JJ 009'E aTS1y — —Oog*T SSIoUS Pue o}TURIN asOyIV aseq—ovys YUM 9} RIOUIO[SUOD pue souojspuvs dISOWLV SIEuS auo}spurs OISSVIUL per ][z—ouo}spurs pur ‘ayeys ‘o}e19WI0;SU0D = *7y. COST 0]UI SUIPLIS VsOyIV YSNoyYy, —oo$ auoYS (aSOYIP) -OUIT pur ‘ayeys ‘soye1a = -yy. «Sz -wojsuo0d ‘s}Is ‘asoyry (e104) SoUO}SpULS YSIPpoy “IJ CoS‘1 NVINaudd SJUJUIIpEs puv 9}1URID) ISOYIV ISOYIV SYA [PEEL “ak C19) spoq [eoo uly} YIM oseys pue ‘auojspues ‘asoyry “3 OFS "PIN “EN “ACN ‘9S0¥1V 107319015 “UUOD PUL ‘sseP “IOATYT yNoauuO, ‘ssoyIy Jroy avsng ‘o]OD ‘UOTRULIO UTe}UNO,T ‘O[OD ‘UOT}VULIO J9]IND BI}JOIG VAON ‘SITIIG W0IOFT ‘eATM “STN youvig Moprayy ‘ayeyS sIoATY "PAM “SIA Tourrg MOpray, ‘UOVUIOT aM xoY asoyie ay} pur oyeys oy} Ur SsyxIeUL -gjddiu pue surewer jug soyeys oy} ul pue dSOYIV oY} Ul SUTeWIOI JUr[ yuosoid sjueyd pissoy z 5 Re ISOYIV OY} UT STISsOy OUTIL] =i isa) S) Q eS) = = Q UOUIUOD SUIppoq-sso1d "suOZLIOY [[@ Ayavou ye speq ur ynq aseq 7e ATUO JOU SIND9O BSOYIV SyILUIDY Aqye9 0] SnOddLUOGILI ISOYIe puv a[vys pesodurosa( Spoq YOry} ut ‘aAIs “seul pure privy “‘ouy 0} asIv0d OATS seul ‘Ud015-y1lep ‘pouless-ul yy AVIS ISOYIV ULL] ysippor osoyie roddy X11} PUL poureis-ouy = =out = sivdspypoy pue zjienb sureis airpnsuy souvivedde ur oIV1UeI4) asoyIY Jo JayoRIeYyD Surposoid oy} 0} reps Aa asoyIe pure gyeys ‘auo0}satuty puv ou0}spurs SUTYeUIOITY asoyie pue ‘auoyspurs ‘QUOYSIUNTT [JIM sayeys snosseuoqied pue Yxory{q SNOUOVLAYO ayluris ‘Avp pue ‘asoyie ‘pues yy ost NV@HONVWNOO 9}VIIWIO[SUOD PUL ISOYIV ISOYIL-9}LIOIp pur ‘saqRIs ‘souojsouy oanduty asoyie pure ‘soqeys ‘syn 7, oissvaal oyUPI dsOyy a[eys ‘souo}spurg “jj COf—-o souoyspurs styyed | -ppey ‘soyeys yoriq Geog 1 00S 9uoSpuvs Avis YM aeys Oe “3 C00'z (oryyedspyay ud}JO) oUOJSpuRS os1V0D “zy COS panuyuoj—OISSVIXL W01499S BYSLTY ‘SOLIS VIIQ ‘SoIag UNTO BYSLLY ‘SOG YNUYAOF] BYSLLY ‘SOIG O[[Ips0 J, COE dt TEASE) ‘UOI}BULIO, = JUaXnyeg suey BYSeTY ‘Saag youyeN BYSLLY ‘SIIIG ¥}}09-V1I9 T, VYSELY ‘SOLIIG VJUIMYS “RA SUIS [vod puowyory ‘dno soyryor J, UOISIY Pu UOTJLUIOT 426 panuy “od—II UTEV.L 427 CLASSIFICATION OF ARKOSE DEPOSITS Surppeq -SSOID. yonu «yy = =pue peppeq Ajury} SI UWoOneUWIO.T SNOIsdBUOGIeS Spaq 9} [TV suozii0y OM} ye sjissoy jurjqg ‘worry -17P1S 9]}}1] pue 19}9eIeY UT UOI}LIIVA 9]}}I]T SMOYS VsOyIV souvived -de ur o1jtuvis yred ut ssoyry ut Avis-ystnjq IOTOO ‘QUO]SpuUvS 0JUI SuIpeis owos ‘souvIved -dv Ul o1y1UvIS 9UIOS—SIDAP] quUsIoyIp Ut I]qeueVaA IsoyIy 9S0Z}1BN0) dPIOIP Z31eNC) ISOYIY survas jeoo yWM ‘97eIIWO]S -u0d pur ‘ayeys ‘au0}Spues spoq [evo pur oyeys YJIM soy puv sUoJspursg aseq oy} 32 d}VIIWIO[SUOD IO SOE YjIM ouojspues ‘soyeys snosovuoqivs ‘asoyay IsOyIV pue ‘soyeys snosveu0q -Iea ‘guojspuvs oul puv 9siv09 ‘9}e1OWIO;sU0D ANHOOSTIO GNV ANHOO asoyIe ySOUL UL UeY} JoYsory reds -P]ey fx1yeu ou YIM. z}1enb pue siedspyej jo sures repn3 -ue ‘souvivedde ur o1y1uRI38 ‘poureis-wmnipoul ‘ABI3-VYSVT a}1OIpoursy d}PIOWIO[SS ITULITOA, ISOYIV IAISS IA 9u0}s -purs puv oq]]Is1e poy auo}spues dy} edspjo,T 9} PIOUIO[SUOD auojspurs oryyedspyay use13 0} Avid) 9VIDULO[SUOD 9SIVOD au0js -puvs o1yyedspjey use14 oy d}LIOWIO[BUOD VSIvOD SOSUZ] O}T[Id1e rq YM suo}Spurs o1yyedspyey ude18 07 AvIn a}IT[Is1eV Yyoejq 07 AeriH "yy O0S'S *7J C00'S —o00S‘f "JJ OOOO Ay C29) “YJ OO1‘T “YJ OO "yy O0S‘T "yy 002 “yy 0oS‘E "JJ COf "J OOb'T “yy oor‘ "}¥ OOO'f VYSLTY ‘UOLVULIO LeU sy UO}SUTYSE ‘punog josng ‘uonjeuUio,, josng UO}SUTYSe A ‘AQT[VA CUTYVA “UOIZVULIO YURMS Oa ‘xTUs0Yg ‘UOTIVULIO IOART 971993T 10) tel ‘UdIUIOZOFT “I ‘osoye uoqyesheg DONALD C. BARTON 428 ISOYIV UPTUOPIIIOT, 0} ILTLWUTS qnoysnoiy} wiaojytun AIZA SI 9SOYIe IY, quasaid Ayjeuorse990 squid doipurel puv = sajue yer Ppeppeq-sso1a sARMR JSOUITe pure poappeq A]PAIsseU VsOyIY uornedar snouojouout Ul d1B SYDOI yUdIOyIp IY, zyienb pur zedspyay |YselJ JO SUIvIS PpapuUNoOI-]jaM |Ajoye1opour §=0}—s AR[NSueqns XTI}VU OU YIM zyrenb pue aedspjey ysory Airey jo suleis popunol-jaMm 0} /pepunos Aja}e19pour Jo pasod -woo ‘gouvivedde ut o1y1uPIs (ssorsajddy) osoyie soddq SouTT[eySAID Uva xPIY aseq 7® ssoyIe O}JUL SUIpeis 9UO0}SpUues NVIwaWvo ayueiry ISOYLY d}LIOULO[SUOD YIM pue oyeys uMOIq YIM ay1zyVNb MoTPA 10 JT 9}eIOWIO][SUO0I pur ‘aqyizqyrenb ‘e1s901q YIM JUOJSPURS PUL ISOYIV poy SSIOUS UPISIMIT asoyie yoq pue ‘Avis ‘aAys uy aseq Je BIII0Iq ayeys yep YIM ou0js -pnul puv 9uoyspurs poy sayeys poet d121 YIM VSOIe pdr soyeys yep ‘ssvy pur auoyspurs Avis puR pdx NVIadnvo-dad adOdna a} eIOWIO[S -uod pue ‘asoyie ‘[Roo YIM 9UOSpues puP aeYS panuyuo) —ANHOIOOITO AGNV ANAOOW "yy C00'S < “yy 007‘ "yy 008 "yy 000'R —000'S “yy CooL —000'£ "yy 0OS‘z PIAPUIPURIS puryjurunosuy ‘ueruy0f ARMION ‘oyusvieds purp}oos ‘ouospuvs uUPIUOpLO y, uoise1 Avg Jo][O1]U0_ SYIPWIIY asOylY JO JojovIeyy) wo1y295 UOIsayY PUP UOTJRWIO TF panuyuoj—{y ATAVL 429 CLASSIFICATION OF ARKOSE DEPOSITS ISOYIL UI SUIeUel yuR{d poztuoqgie) UOWWOI SUIPPIq-sso1d Vs1vOD QSOFIL 9G 0} Wes Pjnom suoyspuesS pay PIO IaMOT 943 JO soseyd [eIaAIS S[ISSO] QUIIVU YIM Spoq po}eIossy ‘yspIy} “yy S-€ spoq ul ssoyIy uaA9 SUIPpeg SNOIIEUOILS ‘IAISSPUL ‘QUIf SOWITJIUIOS ‘9STVOD U9}JO ‘gouvivedde ul dIyIURIS BsOyIYy ssoyxIv puv 9}v][SsS Apues SoInsvoul [VOD JO oseq [Zevw pur ajeys pur JUO}SPUYS YSIPpat ‘9soylV OTTTSTV ISOyIY OUTST YR adUOJSpuvs ARID ISOYLY Avo Apues o}IYM pue poy SSIOUL) ISOYI ISIVOD au0ys -pures snosovoIu UuMOIg auo}spurs pue jroo YM oss0xI1y Surpedeid oy} 0} APIS YONUI UISeq [[eUIS ssfous pur oyIUPRIN d}PIOUIO[SUOD pur ‘ayeys ‘jeoo ‘asoyle pappoqiszUy SNOWAATINOGAVO ysory redspja,q pei 0} yurd ‘10jo9 IIJLIOWIO[SUOI IUIOS ‘gsouvivodde ul o1ytUeIs UIOS pepunod Ayjensed sureis zz1eN0) ISOYIV-BYIU 9}IOIIG 9} VIOUIO[SUOD UTYY ‘guojspues pue dssoxIy Souo}SpuRS PUL SoTeYS NVINOAGG "yy 06 “Vy Of "yy OOf "JJ CO£-O1 UISEG [VOD J9ZUOYLY-OUpryy aSIIGISUISOTY SUOT}VULIOT VUlRY pur neuneig a“ UISeg [VOD Uss[axy uopeg ‘xpospporosusyor, uopeg ‘neusdg uIseg YORqs[yo1yUTFY uepeg ‘uapeg-uopeg uopeg ‘uoydneysiog UIeYIg yaIy ‘aUO}spuRS Poy PIO SOUISIP [aya pur souusply ‘UOT}VUIO,, UeIUUIpey DONALD C. BARTON 430 UOWIUOD SUIPpaq -SSOID ‘peppeq = ATYysnoxy SUIppoq-sso19 yonut ‘fpappoq Ayrepnsory Aasiaf MaN pue ‘x10 XK MAIN ‘s}jJasnyoesseypy JO oIs SULIT, JO oSOYIV 0} ALTIUS A19 A, uoWOD ~SuIppoq -sso1o ‘yporyy “4p © ATYysnor speq ‘ysnor uoNPoyeys ysodap v}jap B aq 0} JYySnoy Ty, asoyle oy} UL suTeUOI yuRTd paztuoqieg 9u0}s ou] puv gfeys Avis dIPI YIM 9} R.1OUIO[SUOD pasoduioo [pu ‘ayeys ‘asoxre ‘au0}s -9p Mou sedspyey ‘asozjien() |-purs o1yzedspyey Yystppory STAONAOAITLON souvived -dv ur o1jmueis f10;09 ut Ain ystXvis 10 ystnyq tojoo ‘faouvivedde ur oy1uURIy quad iad oz ynoqe xtyeu ‘rejnsueqns sureis ‘asozjz1vN() SoINsvou [VOD UT a[vys puv 9UO}SpuURS PdY ayeys Avan auojspues o1y}edspyey puv ‘aeys ‘asoyre YSIPpIy jee Ra Be a[eys pue suojspurs AeIny 9} LIIWOTSUOI 9S1BOD YIM 9UOspuRSs pue s[vYS 9UO}SpULS PIUTeIS-dUy ‘MOT[PA pu aeYys Ywrypq auoyspurs o1y}ed -Sppy pure o}RIBWIO[SUOD SUOT}VIII PUL IOJOVIvVYO TLITWIS JO ISOYIV xL}eu su ut reds -PppPy} pezruyory yonut pue zyienb «sures = re_nSueqns soouvivodde ur o1y1uvs13 Javed uy sstoury ASOYI YSIAvID eee Ope eS 5.32 9} LIIWIO][SUOD VSIvOD "1 9 9} eIDULOTSUOD ou YIM ssoyre YSstAviy “VJ O1Z panunuoy—SNOUATINOPAVO Auruliory “Jouysiq, ayeN pure Ievg ‘spog JaAajoy, J, pue sJopasng purysug “ploy pure urginy FIP pursuy ‘pp [vog s11ysy10 x QOUTAOI DUTY ‘IOTI9M7}O ‘Snossjruoqiey Joddy pue sopuasajoy JoMO'T “STA sossoA “AITIOM, pue Yorquoply “S}IN SosSOA “YOeR'T SyIRUII x IsOYIY JO 1ojvIvyO u01q99g UOLsay pPue UOT]UIO.T Pannyyod— Th w Tav.L 431 CLASSIFICATION OF ARKOSE DEPOSITS quojspurs ay} jo osvyd yeooy] B AyUO SI gSsOyIV oy, yuosoaid s]issoy quLTg qyuosoid Ajqissod sojqqed = peaqjaoe “QAISSVUL ‘UdAd JOYJVI = WOTPBOYT}LI}S QTPYs oY} Ul suorssoidurt jury Agsiof MON pur ‘YOK MON ‘qnoTJDIUUOD ‘s}Josnyoesse JO OSISSVIIT, 94} JO soy 0} IepTUMIs ATOA AT[VIISOT[OYIWT Ssurppeq -SSOID IVI YUM ‘sIoAVT UdAD ‘QAISSVUL UL SI ISOYIV 9S0Z}.1UNQ) souvived -de ut omiueis ArvA yied uy pozrurjory Aja}e[du109 mou iavdsppoy} por ‘asozj1enQ) g0ue -Ivoddv ul o1y1UeIs pue ss1v0D pesodwuiosep Ajojo,d -ul0d Mou Iedspfay ‘fesozjreN) BIIIIIG pure ‘o}e10UL0;sU0D ‘soyeys ‘osoyre YSIpparyy speq yorq Ud] 10 Ystyos ‘oyuUvIs 9} VIIULO[SUOD YTM ISOYIV greys YIM osoyre Avi SUOT}L[VIIOJUL asOYIV YIM g[eys oding 9uo}Sspues pur ‘osoyre ‘9}e1VWO[SUOD souo}s YSIPp9ad oq TUL -O]Op pur osoyIe YIM guoyspues s1yyedsppaT gjeys pur ssoyie Avis SuryeUIoy 9d} V.IOULO[S -uod pur ‘auojyspues ‘so[evys YIVp YIM ssoyIy oyURID) ayeys poyesouvA pure 9s9soyIy a[BYSs poyeSoIva pue Yyorlq_ ISOYIV YIM 9} LIOWIO[S -U0d aTvYS ‘VUO}SpURS PIY oyIURID syn} pue osoyiy -pniu ‘sayeus 1 syny [zew pur ‘au0js -pnur ‘auoyspuvs ‘oyeys pol YUM ss0yxIe por “YF OOO -obS “VW C9 34 09 “yy 06 “JJ COI ‘UISeg “S}JAL SOBSOA ‘Spoq JoyDeqTYyosy SJ], SossOA ‘spoq Yyoequorty, solo] Joryg ‘wreyydgyos }S010,J yprpg ‘sioquivry9s ‘spaq yorqoy] ysolo,y yorpg_ ‘neusddg Auvulloy ‘uspeg-uopeg Auvulioy ‘s1oqjeploy AUBULIO) ZUR, ‘spaq JaAoOUL, DONALD C. BARTON 432 XLIJUI SNOddETIBIV Sutppaq-sso1o auros = |ssay «IO 910W YM AT[ensn St o19y} Ajarer {spaq oats |‘sourreodde ut = orjURIZ -SPUI SSd] JO 91OU UI SI asOyIY [AIBA Jed ul ‘AUG 0} asIROD a}UPIDy (Ayuo urseq 24} Jo aed wroyznos ul 9soyIe jeseq pur auojsouy ~‘sAep> Apues) s}113 ‘spew pue soauojspurs ‘sosoyly JUOJSOUT] PUL SLIP] ASOMAV AUVILAAL a syisodap a10ys aq 0} yoRIny Aq _ pesoddns oie ‘asoyie xl}eW YON yNoYyyIM ay} JO Ayeiadsa ‘sjisodap ayy, |10 AIM 9sozjenb Aaa ATISOPL asoyle pur ‘jrew ‘Arp ‘au0jspues yew pue ‘Avo ysippel ‘au0js -suT] ‘asOyIe dIITULO[O oryyeds -PpPy Ayyeooy = ‘auoyspues “qj 09 ose pay yy OS1 AYTUIIA pur Ule}SyxIeY 1B a7eurye[eg 9} UI asoyIe OJUL Suipeis ‘auojspues “4 009 asoyie Surpadeid ay JO yuayeAtnba ayy Ayqissod ‘asoyre IepIUIS JEYMIWIOG syoei} atdor YIM do} 94} pieMo} speq afeys |poureis-ouy ‘surpunor yonut UIYy} YIM ‘aseq oy} 3e Speq BZuLmoys sureiZs yuanzsuOos UIAI ‘JAISSBU UI VSOYIV IOMOT ‘payflopIs A[IAvaY pu 9SOZ}IENC) XEJeUL_PayLoryis ayUeIL) SuOI}e[eO -Iajyut AVARPD YIM VsOyIy “J 06 pew pur ssoyxiy 3 Of uinsdAS pue sU0}SoWII'T OISSVIAL ayueis Sursjiopun 94} WOIJ psatiep jOU SI yng ‘aseyd yeseq & SI asoyIe ayy, [xLyew ouy YIM ‘AvIZ ‘asivO_ | aque BID991q pue ‘asoyie ‘au0jspues ‘gjeys ‘a}eIIWIO;ZU0d panuyuo)j—_SAGNADaITLOa syIeUlaYy aSOxIY JO JoyovIeYyD w01999S panuyuwoy—iIl TTaVL aouRiy ‘ouseuT spoq uPIstIouuRs speg uvidure}s eIUOIUPL spoq rodnay jung SRee eet ct suoa’yT Ieou ‘Assay ayspeply ‘sreuaqny uoli] re9U ‘AT]TUUD aouely ‘OLMISIP UPAIOSY yso10,q BIsuLINy J, uolgay pur uor}eUIIO.J 433 CLASSIFICATION OF ARKOSE DEPOSITS o}ULIS 94} OUI Sopeis ssOyIe IY} AT[RIO'T asOyIe quosoider 0} Ajqissod wey “PIO Aq paararaq st sstous oy, souvivodde ur dIy1UPID ays pur Z}1eNb jo x1yVeU ‘souvivodde ut d1y1uPIs ‘osre0* ayuPIS SulAjIopun oy} 0} oduervedde ul Ir] -IWIS AI9A puv OT}TURIS syed UT . pesoydiouejyour yonzy URI 9} LIDULO[SUOD S}Is o1qyedspjoy pure osoyiy UO}SOWIN'T VOTSAV ISOYIV IWOS YIM ATQIS -sod ‘so}v1aWOp, suo. pue SauO}SpURS SNODUASO19}9F] aque SOIL PUL 9}1Z}1eN() sayeys sno -OVUOGIVS PUL JUOISOWTT ouIZ -yzenb pur ‘ystyos ‘sstour) ISOWIL YM —-941z} RNC) SSIoUS ULYS-IP, J, aSOYLy SoPIUIWUeS J Be lec ayuedyy d}eIOULO[SUOD pue a ayeIOULO[S +4 OST -U09 PUP 9UO}SpURS as1vOD | yeseq (aeEqureD) OTL VISV wy} 0} JuaTeAinbs ATqeqord pue Ssurpaseid oy} 07 reprams sjisodep ssoy1y AUOTOD adey ‘yoox] your ‘sattag osuo) Auojog odey ‘ATY salyepy ‘soag osuo0d snpuy toddyQ ‘spoq sus007%y VIPU] ‘UOT}VUIOY IeMeg eipuy ‘dnorry oArry PIPUT ‘UOTJVULIO IVMTY WI9}ShG 12, I-NA\ DUTAOLY IZ [Suv A OURI Z910.J JO ulseg UOIquIO'T JO Use DONALD C. BARTON 434 IOJUM 9Y} JO 1jOU dy} 0} JWOD SPY BOLOUY YINOG UL JO VITPAYSNY UL oyueid ayy 07UL soprid ssoyie ayy ‘asoyie ayy jo yard szoddn ut yuososd Ajqissod = uoreoyryesys jure y SHIVUIY VOIMAWNVY HLNOS GNV VITIVALSNV pies) ISOLY dVRIDULO[JU0D pur souravodde ut o1ytueid AIDA aUOJSpuRS snNoddeT[1y xXUyeuL SISIYIS DUI[eISAI auy YIM ssozjienb ‘ysippory ISOYI PUP OYOBMNIL) S}SIYOS OUI][eISAI) ISOALY souojspurs snoaieo[v9 pur sareys ysippet pure ysrudasry S19AR] DITWO[Op pur aUOTSPURS YSIIYM ‘osoyry sures spaq Ayqqad iepndue yy ys Ysippay [pur Addvpo YIM ssoyLy OUP) AVPIDULO[SJUOD puv asOyLy Aes pur oyIZ41eNC) Ssstaury DUOSpURS PUP oyeYsS YIM 9VBIDULO[SUO0D IISOYLY JUOJSPURS PUR 9[PYs YLVC] panuyuoy—VOTAAV AIYWO IDOYAV JO JUNOIOV ON JUOPSPULS UPICN NT wOLYY Sey YSN Gopny oye] OAUOD) YOUIL ST ‘URIUOAX(T ~=pur = SNoOsayruOqan, ysVoy ploy Auopodg adey ‘satag jaoy yori Auojog adeyg ‘sarsiag ysnsaMmnoryy aBOYLY JO 1ayIvILY u01499g panuyuoj—Il ATAVL uoWsay pue UOIyeULIO CLASSIFICATION OF ARKOSE DEPOSITS 435 A. ARKOSE DEPOSITS FORMED ENTIRELY UNDER RIGOROUS CLIMATIC CONDITIONS Feldspar showing merely slight decomposition.’ Argillaceous material absent or present in minor amounts.’ 1. Deposits formed in desert regions.—The deposits are massive, homogeneous, and in some cases of very large size. In desert regions, the disintegration takes place chiefly on outcrops directly exposed and not protected, as in moist temperate regions, by a mantle of vegetation. The débris of disintegration is easily eroded and the processes of erosion and disintegration and deposi- tion of the eroded, disintegrated material as arkose can therefore take place contemporaneously, and can continue as long as the desert conditions persist and as long as the granitic terrane remains unburied. The size of an arkose deposit formed under such con- ditions will therefore depend chiefly on the size of the terrane of disintegration, the size of the basin of deposition, and the length of time the desert conditions prevail. ‘The constancy of the condi- tions during the period of formation should be marked by a massive- ness and homogeneity of the deposits. a) Terrestrial: In deserts, wind action prevails the greater part of the time, but rare storms do occur and in the short space of their existence do an immense amount of work. Deposits of arkose formed in desert regions therefore are likely to be in part of eolian and in part of aqueous origin. The arkose shows in part the eolian characteristics of well-rounded sand grains, faceted pebbles, local lag-gravels, dune stratification, etc., and in part the ordinary characteristics of water action. In deposits of arkose forming in arid mountain regions, the greater part of the transportation of the disintegrated material may be by water action, during the rare cloudbursts, and by the streaming of the débris down the hill slopes under gravity. The constituent grains in this case are subangular, and quartz and feldspar grains should be present in about the same proportions as in the original granite or gneiss, since the amount of t At the time of formation of the arkose. Subsequent exposure may produce complete decomposition. 2 The possibility of the presence of exotic material brought in by a river whose headwaters lie in a temperate or tropical region should not be forgotten. 436 DONALD C. BARTON transportation undergone should not be sufficient to cause marked comminution and loss of the feldspar. The deposits as a whole should be rather massive, but with cut-and-fill stratification rather common, and with considerable intercalation of coarse block débris toward the sides of the valleys. An example of this type of deposit is found in the Applecross group of the Torridonian sandstone, a pre-Cambrian formation extensively developed in the Northwest Highlands of Scotland. The ideal section of the Torridonian sandstone is given by the Geological Survey* as shown in Table III. The Applecross group TABLE III Groups Thickness Composition Chief Occurrence 3. Aultbea....| 3,000-4,000 ft.| Sandstone, flags, dark and black; Loch Ewe and shales, and calcareous bands) Loch Broom passing down into chocolate and red sandstones, and gray micace- ous flags with parting of gray and green shale 2. Applecross..} 5,000-8,000 ft.| Chocolate and red arkose with peb-| Cape Wrath to bles of quartzite, quartz-schist| Skye (felsite, jasper, etc.), and occa- sional red and chocolate shales 2. Diabaig....} 500 ft. in Gair-| At top, fine red sandstone with red} Assynt to Skye loch; 7,000/ mudstone and gray shale; at ft. in Skye base,’ coarse breccia of Lewisian gneiss. In Skye, gray and buff arkose in great thickness is a formation composed of massive arkose of very uniform character (Fig. 3), marked off by thin intercalations of fine quartzitic and shaly sandstone into persistent layers of rather uniform thickness—in each case three, six, or ten feet or so. Although extremely massive in texture, the arkose shows most irregular stratification and almost invariably is strongly cross-bedded. Walther reports cross-bedding indicative of dune formation, and faceted pebbles. The lithologic character of the arkose may be seen from the accompanying photomi- crograph. In the finer-grained phases, quartz composes a slightly greater proportion of the whole than in this medium-grained phase (quartz about 60 per cent and feldspar 4o per cent), and the grains «B. N. Peach, J. Horne, and others, ‘‘The Geologic Structure of the Northwest Highlands of Scotland,” Mem. Geol. Surv. Great Britain, 1907. CLASSIFICATION OF ARKOSE DEPOSITS 437 show a slightly greater degree of rounding. It is notable that the predominant feldspar of the arkose, microcline, does not occur in the underlying terrane. Where the arkose rests directly on the mountainous topography of the underlying terrane, the arkose basally becomes a coarse breccia. One of the most striking features of this formation, with the exception of this basal portion, is its uniformity throughout its great thickness. To this type of arkose deposit should be referred the following: The Torridonian Arkose, pre-Cambrian of Scotland. Arkose of the Sparagmite formation, pre-Cambrian of Scandinavia Lower Old Red Sand- stone, Scotland (in part) Paysaten Arkose, Creta- ceous, British Columbia-— Washington b) Marine: If marine conditions prevail adja- Fic. 3.—Photomicrograph of Terridonian cent to a granitic terrane arkose, Applecross, Scotland, showing the lack of in a desert region, marine matrix in a desert arkose. The rounding of the grains is rather obscured by secondary growth of the grains. Magnification, 15 diameters. arkose may form, having in part the characteristics of an eolian arkose. Some of the constituent grains in this case should show the rounded outlines of eolian sand grains. The deposits as a whole, however, should show the structure and stratification of marine sediments. To this type of deposit should be referred the arkose that is now forming along the east shore of the Gulf of California. : 2. High-altitude deposits—Local deposits, of small size and extent. The conditions of high altitude, according to Oldham and others, are peculiarly favorable to disintegration. Erosion of the disintegrated material takes place rapidly, with rapid deposition of it in many cases as arkose in local catchment basins of the intra-mountain valleys. As such a region is subject to general 438 -» DONALD C. BARTON degradation in the course of time, the deposits must be temporary in character, and usually of recent geologic age. ‘They may be wholly or in part lacustrine, fluviatile, alluvial (cone or fan), landslide, or fluvioglacial. The stratification should be rather irregular, and the constituent grains should be angular to subangular. To this type of deposit should be referred possibly some of the deposits of the Upper Indus Valley, although from the descriptions of the deposits it is not quite clear whether they are really arkosic or not. 3. Deposits of cold (high-latitude, subglacial) climate.—In the high lati- tudes, the effects of dis- integration are not pro- nounced, or at least they are, not noted in the lit- erature. ‘‘Disintegra- tion’ is reported many times, but in most cases 1G. 4.—Photomicrograph of Pondville (Mass.) it is clearly’ block disin- arkose, an arkose formed under moist temperate tegration that is meant, conditions, showing the quartz and feldspar grains and in no case has the in a fine-grained matrix of quartz and argillaceous : material. Magnification, 15 diameters. writer been able to make it out clearly to be granu- lar disintegration. ‘That the effects of the latter are not notice- able may be due in large part to the relatively recent glacial erosion of the products of the preglacial disintegration, or, in regions of considerable relief, it may be due in part to excessive block disintegration and erosion. As the temperature range is often great, and the lower part includes the critical point of freezing, and as, furthermore, hydration can take place at the surface during the summer and, in regions not too far north, at all times below the level of freezing, there would seem to be no theoretical reason why granular disintegration should not take place. Granitic and gneissic blocks exposed on the surface of glaciers in many cases show noticeable disintegration, although CLASSIFICATION OF ARKOSE DEPOSITS 439 the amount that takes place in this manner is slight. If disintegra- tion takes place, the conditions would seem favorable to the erosion of the disintegrated material and its deposition in arkose deposits of small or moderate size, probably in association with glacial or fluvioglacial beds. ‘To this type of deposit may possibly be referred some of the pre-Cambrian arkose of Canada. B. DEPOSITS FORMED DIRECTLY OR INDIRECTLY THROUGH THE EFFECTS OF MOIST AND USUALLY TEMPERATURE CONDITIONS Deposits of small or moderate size; the arkose commonly with a matrix of fine-grained argillaceous material and usually associated with argillaceous beds; feldspars commonly showing a moderate amount of decomposition (Fig. 4). In the present regions of moist temperate climate, especially where the topography is in a mature or old-age stage of develop- ment, there is almost universally present a very considerable accumulation of disintegrated material which is available as a source of material for the formation of arkose. The following section, from the vicinity of Autun, France, in its essential features is characteristic of such regions as the granite terranes of Morvan, the Plateau Central, and Forez, France; the Vosges Mountains, the Odenwald, and the Thiiringerwald, Germany; Dartmoor, England; the Piedmont belt and the Pike’s Peak region, United States. iLO PR anon sae Mantle of vegetation; surface soil and subsoil of gritty brown clay with quartz and feldspar grains 6ft.........Granitic sand and gravel, stained with limonite; feld- spar showing considerable decomposition toward the surface, the amount decreasing with depth Dy on Granite more or less fresh on superficial examination, but crumbling under light blows of the hammer; depth difficult to estimate; fresh granite The relative and absolute proportions of these zones vary greatly. The maximum depth to which disintegration was observed by the writer to have extended was 4o feet, at Royat (Puy-de-Déme), France, and at Hohkénigsburg, Vosges Mountains. On Dartmoor and in the Piedmont belt decomposition is more in evidence than in France and, in the Piedmont belt especially, the zone of soil and 440 DONALD C. BARTON subsoil is much larger in proportion to the zone of disintegrated material. The rate of disintegration under the conditions of a moist temperate climate seems to be rather slow—in New England there has been since the Glacial Period disintegration sufficient barely to efface the glacial striae and polish on granitic and gneissic ledges—and the very considerable amounts of disintegrated material generally found in those regions are the result of slow accumula- tion under the protection of the mantle of vegetation. General erosion of this disintegrated material and its subsequent deposition as arkose can take place only when the mantle of vegetation is critically weakened or destroyed. When this has once happened and the mantle of disintegrated material has been swept away, a long time must elapse before considerable amounts of the disin- tegrated material can again accumulate. The arkose deposits formed from the accumulated débris of disintegration in a moist temperate climate will therefore be of small or moderate size. As the mantle of soil and completely decomposed rock is eroded at the same time as the mantle of disintegrated material, the arkose is commonly associated with mudstones and shales, and, as the disintegration is accompanied by considerable decomposition, the arkose itself is likely to contain much argillaceous material and to have feldspars showing noticeable decomposition. Since stream transportation of débris results in the rather rapid elimina- tion of the feldspars, the arkose is likely to grade into quartzite. The causes which might critically weaken or overcome the mantle of vegetation and result in the erosion of the accumulated products of disintegration are: introduction of arid or semi-arid conditions, introduction of subglacial or glacial conditions, a marked increase of rainfall, a marine transgression, deforestation by forest fire, and marked upwarping. A marked change of climatic con- ditions toward aridity in a region previously of moist temperate climate would necessarily result in a marked diminution of the vegetation and in the exposure of the underlying disintegrated material to erosion during the occasional storms. Glacial condi- tions might result either in the erosion of the disintegrated material by the ice itself or in the exposure of the disintegrated material to erosion through the destruction of the vegetation of the temperate CLASSIFICATION OF ARKOSE DEPOSITS 441 conditions without the introduction of an arctic flora sufficiently luxuriant to form anew the protective mantle of vegetation. A marked increase of the rainfall, it was suggested by Shaler, might be such that the streams would be competent to waste generally the land surface. A marine transgression would necessarily result in the working over of the materials of the regolith, irrespective of the luxuriance of the mantle of vegetation, and might easily result in the deposition of arkose. Forest fires are not uncommonly due to lightning and often are effective agents of deforestation. It would seem possible that a period of heavy rains following a severe forest fire might result in the géneral erosion of the mantle of disintegration. Upwarping of considerable amount would result in an increase of the stream gradients, in an increase or decrease of the rainfall, and in the lowering of the mean temperature. The total effect might possibly be conditions favorable to the general erosion of the mantle of disintegration. A very special cause is to be found in volcanic activity of the explosive type, which not uncommonly results in deforestation and desolation in limited local area. Of this type of deposit, in which the arkose should be asso- ciated with tuffs, there is at least one example, the Rotliegendes arkose north of Heidelberg, Germany. In regions of youthful topography and considerable relief in a moist temperate climate, there would seem to be no reason why disintegration should not take place. That it is not seriously in evidence is probably due to the fact that it is masked by block dis- integration and by rapid erosion. If it does take place, the débris that can be eroded at one time is of small amount and is lost through decomposition of the feldspars or through intermixture with the heterogeneous stream-borne sediment. In tropical regions, decomposition commonly prevails over disintegration, but in two localities disintegration is reported as occurring with but slight accompanying decomposition. The débris in these cases, if eroded under normal conditions, would probably be lost through decomposition, but if eroded under the conditions of a marine transgression, or under the conditions of aridity, there would seem to be a strong possibility that a deposit of arkose might be formed. Except by means of a contained fauna or flora, such 442 DONALD C. BARTON deposits would probably be indistinguishable from the correspond- ing types of deposits of the temperate zone. No deposits have been recognized to be of this type. . Terrestrial—(a) Deposits laid damn under semi-arid condi- es Arkose reddish, composed of subangular iron-stained grains of quartz and partially decomposed. feldspar deeply in an iron- stained matrix of fine-grained quartz and of argillaceous material. When the moist temperate conditions give way to those of aridity, the mantle of vegetation, weakened by the change, is no longer able to protect the accumulated products of decomposi- tion and disintegration, and during the occasional violent storms they are quickly eroded, to be deposited with rapidity usually in the near-by valleys and catchment basins. Owing to deposition from torrential streams, the materials of the mantle of disintegrated material are laid down in coarsely stratified banks and lenses of arkose, showing much foreset and cut-and-fill cross-bedding. The soil and mantle of completely decomposed rock are deposited, partially sorted, as cross-bedded, argillaceous sandstones and as more finely and evenly stratified gritty mudstones. As the temporary lakes dry up, these mud beds become sun-baked and glazed and cracked and may receive raindrop prints. Under the conditions of alternate wetness and dryness, there should be almost complete decomposition of organic matter and oxidation of the iron. Deposits of this type are not rare, and a good example may be found in the Sugarloaf arkose of the Connecticut River Triassic. The formation occurs in what was possibly a Triassic basin, and consists essentially of an unordered alternation and repetition of gritty, argillaceous sandstones, conglomerates, arkose, and sandy and calcareous mudstones. There is a coarse, general stratification whose dip initially was apparently low. In the beds of mudstone, even, fine stratification is the rule, but cut-and-fill bedding is found in a few places. The coarser strata are strongly cross-bedded, mostly with the foreset type of bedding. Cut-and-fill bedding, however, is common. The mudstones show mud-cracks, raindrop prints, glazed surfaces, and reptile footprints. The arkose is found in banks and lenses, chiefly at or near the base, but also at CLASSIFICATION OF ARKOSE DEPOSITS 443 numerous higher horizons. It is.composed of subangular grains of quartz and subangular grains and pebbles of feldspar in a fine- grained matrix of argillaceous material. The color of the whole formation is deep red, due to a heavy stain of ferriciron. Fossils are rare in the formation. The following deposits are apparently of this type: Arkose of the Amnicon formation, pre-Cambrian, Wisconsin Sugarloaf arkose (Triassic), Connecticut River area, Massachusetts and Connecticut Stockton arkose (Triassic), New York, New Jersey, Pennsylvania Arkose of the Upper Carboniferous, Ottweiler, Rhine Province, Germany Arkose of the Lower Rotliegendes, Rhine Province, Germany Arkose of the Rotliegendes, Mainz Basin, Vosges Mountains and Black Forest, Germany Arkose of the Old Red Sandstone, England Arkose of the Cutler formation, Permian, Colorado( ?)* Arkose of the Fountain and Lower Wyoming formations (Permian), Colorado(?)! (b) Deposits laid down under moist, chiefly temperate, condi- tions of climate: Arkose grayish, composed of subangular grains of quartz and of considerably decomposed feldspar in a matrix of fine-grained quartz and argillaceous material, in most cases carbonaceous, and in some cases carrying plant fossils; the arkose commonly associated with coal deposits. The causes of a general erosion of the regolith in a region of moist temperate climate are not completely evident. The sug- gestion of the introduction of subglacial conditions as a possible cause seems not well founded, since the several glacial epochs of the Pleistocene do not seem to have caused a general erosion of the regolith of the Piedmont belt to the south of the glaciated area. It would seem reasonable to expect, furthermore, that the effect of the change on the mantle of vegetation would be a replacement of the temperate by arctic flora. The suggestion that a marked increase in the amount of the rainfall might be a sufficient cause would likewise seem not well founded, as an increase in the rainfall characteristically results in more luxuriant vegetation, with a con- sequent increase in the protective power of the mantle of vegetation. * Commonly considered marine, but apparently very like the Newark beds of the Connecticut River and the New York-New Jersey Triassi ‘»reas. 444 DONALD C. BARTON Forest fires are another possible cause. They are very commonly due to natural causes and often are effective agents of deforestation. It would seem possible that a period of heavy rains following a severe forest fire might easily result in the general erosion of the regolith. Upwarping of considerable amount, with the consequent increase of stream gradients and lowering of the mean temperature, if associated with decrease in the amount of the rainfall, might possibly result in the general erosion of the regolith. While these causes are thus in doubt, the fact of the formation of deposits under these general conditions seems to be indubitable. There is a characteristic type of arkose deposit which is usually associated with carboniferous beds or coal, which is itself carbona- ceous or may even carry carbonized plant remains, and which there- fore must have formed under moist climatic conditions. As the feldspar shows much decomposition, as there is an argillaceous matrix, and as the quartz and feldspar grains are distinctly angular to subangular, the constituent materials of the arkose would seem to have been derived from the débris of disintegration under moist temperate conditions. The arkose, commonly in part, is coarse and granitic in appearance and seems not to have been transported far from the point of origin of its constituent material, and in part usually is finer and less feldspathic, and seems to have been trans- ported for a greater distance. Besides being associated with coal beds, the arkose is associated with conglomerates, impure sand- stones, and silty mudstones. Cross-bedding is common, and many of the beds seem to be the result of rather rapid deposition. The color of this type of arkose is gray. As an example of this type of deposit, there may be taken the arkose of the Richmond (Triassic) Coal Basin in Virginia. The lower portion of the section in the basin is as follows: Productive Coal Measures.....500 ft. Interstratified beds of bituminous coal, black shale, feldspathic and micaceous sandstones Lower Barren Beds.........0-300 ft. Sandstones and shales under the coal beds, often with arkose Boscabel Bowlder Beds....... o-so ft. Local deposits of bowlders of gneiss and granite tN. S. Shaler and J. B. Woodworth, U.S.G.S. Nineteenth Ann. Rept., Part II (1897-98), pp. 423-26. CLASSIFICATION OF ARKOSE DEPOSITS 445 The arkosic beds are best developed about the granitic masses of the eastern margin, but reappear from horizon to horizon with increasing marks of waterwear. ‘The arkose of the basal horizons is granitic in appearance, and by the inexpert eye might not be distinguished from the granite. The arkose is gray in color and is composed of subangular grains of quartz and much decomposed feldspar in an argillaceous matrix of small amount. The arkose of the higher horizons is not so granitic in appearance, there is a somewhat lower content of feldspar, and the quartz and feldspar grains are slightly more rounded. Some of the associated shale beds are carbonaceous, and locally there are small intercalations in the arkose of carbonaceous silty material. . To this type of deposit there should probably be referred: The Carboniferous arkose of the Narragansett Basin, Rhode Island and Massachusetts ‘ The arkose of the Rockwell formation (Mississippian), Meadow Branch Mountains, West Virginia The arkose of the Vosges Mountains, the Black Forest, and adjacent parts of Bavaria The arkose of the Coal Measures of the Yorkshire Coal Field, England The arkose of the Coal Measures of the Flint, Rhutin, and Mold districts, England The arkose of the Richmond (Triassic) Coal Basin, Virginia The arkose of the Corwin formation (Jurassic), Alaska Comanchean arkose at the base of the Coastal Plain series, Maryland, Virginia, North Carolina The arkose of the Swauk formation (Tertiary), Washington The arkose of the Puget formation (Tertiary), Washington. The arkose (Early Tertiary) of the Matanuska and Controller Bay regions, Alaska c) Deposits formed under glacial conditions: If an ice sheet advances over a granitic terrane in which there is a mantle of dis- integration, it would seem possible for small amounts of this débris locally to be preserved as arkose among the various glacial deposits. In New England, there is in several localities deep preglacial dis- integration, showing that the mantle of disintegration of the Piedmont district probably extended in former times northward over this area. Arkose is apparently lacking however, in the New England glacial deposits. No example of this type of deposit is known to the writer. 446 DONALD C. BARTON 2. Marine and lacustrine—(a) The basal member of a new, transgressive marine series is commonly composed of the materials of the former regolith. If the shore forces are not too violent in their working over of the débris, the basal deposit in regions of granitic rocks may be arkosic. The constituent grains show more or less rounding. There may be present a small amount of argil- laceous matrix. Through the elimination of the feldspar the arkose may grade into quartzite. Arkose deposits of this type may grade into deposits of the type discussed in (0). To this type of deposit (a) there should be apparently referred: Arkose of the Hotauto formation (pre-Cambrian), Shinumo Quad., Arizona. The Cambrian arkose of Eastern United States (in large part) The arkose (Silurian) of Littleton, New Hampshire (in part) The arkose of the Igaliko formation (Devonian), Greenland The arkose (Triassic) of the Morvan (in part), France The arkose (Tertiary) of the Limagne (in part), France (b) When erosion of the mantle of disintegration in a granitic terrane adjacent to the sea or to a lake occurs, deposition of the disintegrated material is likely to take place in the sea or lake with, as a consequence, the formation of arkose. Near the shore the arkose is in banks and lenses and is interstratified with beds com- posed of the material from the soil and zone of decomposition and of argillaceous material eliminated from the débris of disintegration. The constituent grains of the arkose are subangular to poorly rounded, the degree of rounding being greater in the more quartzose beds. There may in some cases be a slight amount of argillaceous matrix. Unless the feldspar is itself reddish, the arkose is grayish in color. Although not necessarily basal, the arkose is more likely to be near the base of the formation than not, since the change of conditions which causes the erosion of the mantle of disintegration is likely to mark the inauguration of a new period of sedimentation. The arkose formed far from shore is less granitic in appearance than that formed immediately at the shore, there is considerable rounding and sizing of the constituent grains, and there is elimination of much feldspar and argillaceous material. The arkose in many cases grades into quartzite. An example of this type of deposit is the Tertiary arkose of the Limagne, France. During the Oligocene times the Limagne was CLASSIFICATION OF ARKOSE DEPOSITS 447 first a brackish-water basin some thirty km. in width and later a fresh-water lake lying then as now in the granite plateau of the Plateau Central. The arkose is found chiefly near the base and is found in banks alternating with greenish marl and in some cases extending out a considerable distance from the edge of the basin. Some of the arkose, especially that at Royat, is massive, coarse, composed largely of good sized fragments of the coarse phenocrysts of the underlying granite, and is extremely granitic in appearance. The greater part of the arkose, however, is much finer, is more quartzose, is composed of more-rounded grains, and grades into quartzite. There is in some cases an argillaceous matrix, in some cases a sericitic matrix, and, in some of the more quartzose phases, there is very little matrix. There is a general even horizontal stratification. Where cross-bedding is present in individual strata it is usually of the simple foreset type. To this type of deposit are probably to be referred: Much of the pre-Cambrian arkose of Ontario Arkose of the Congo Series, French Hoeck, Cape Colony The Cambrian arkose, North Carolina-Tennessee Fitch Hill arkose, Silurian, Littleton, New Hampshire Haybes arkose and Weismes arkose (Devonian) Ardennes—Eifel District Arkose of the Grés bigarrés and Grés vosgien (Triassic) of the Morvan region, France Dolomitic arkose (Keuper), Franconia and Thuringia Arkose of the Blasensandstein and Coburgerbausandstein (Keuper), eastern Palatinate Lower Stampian-Sannoisian arkose (Oligocene), Limagne, Forez, and Roannais basins, France Much of the Jura—Cretaceous arkose of southwestern Alaska Arkose of the Cutler formation (Permian), Colorado( ?) Arkose of the Fountain and Lower Wyoming formations (Permian), Colorado( ?) C. UNTRANSPORTED OR SEDENTARY ARKOSE Basal, unstratified deposits grading into the underlying granite. When deposition begins in a district, the original regolith locally may be buried before it has been eroded to any considerable extent. It is thus possible for arkose to be formed without the usual inter- mediate steps of erosion, transportation and deposition of the 448 DONALD C. BARTON disintegrated material. The arkose is composed of the constituent minerals of the granite or gneiss in essentially their original propor- tions. Some of the silicates, especially the biotite, hornblende, and plagioclase, are in many cases highly decomposed. Theiconstituent grains are angular. The upper part of the arkose may show a rude stratification and may grade upward into a well-stratified deposit. The lower portion is massive and grades downward into the granite, and may show the unaltered cores of bowlders of exfolia- tion. A good example of this type of arkose is to be found in the lower arkose in the Silurian at Littleton, New Hampshire. Between the Niagaran Limestone and the granite there is from two to eighty feet of arkose which is coarse and granitic in appearance. The quartz and feldspar are the same as those of the underlying granite and are present in practically their original proportions. Thereis a slight amount of a fine-grained dark matrix. In its upper portion, the formation shows faint traces of stratification and in its lower portion it grades into the underlying granite-gneiss. Locally the original spheroidal weathering and the unaltered cores and shells of concentric weathering are distinguishable. To this type of deposit are to be referred: The arkose (in part) of the Vermont formation (Cambrian), Massachusetts and Vermont The basal arkose (Silurian) at Littleton, New Hampshire Pre-Cambrian arkose (in part) of the Cobalt District, Ontario Basal arkose, Narragansett Basin, Massachusetts and Rhode Island Nubian arkose, Aswan, Egypt D. SUMMARY The geological significance of arkose in brief, then, varies from case to case and cannot be limited in the general statement to significance of a special set of conditions. Each deposit is sig- nificant of some special set of conditions and these in many cases can be determined from the individual deposit or its associations. In the preceding discussions an attempt has been made to show a grouping of these in conformity with a genetic classification of arkose, each type of which is significant of some special type of CLASSIFICATION OF ARKOSE DEPOSITS 449 conditions. But even if the premises of this classification should be seriously disputed, it still remains a fact that most formations lying unconformably on a former granitic terrane have arkose at or near the base, and there seems to be a more or less general rule that, whenever a period of deposition is inaugurated over a granitic terrane, arkose is the first or one of the first deposits to be laid down, whatever the prevailing conditions. ‘This basal arkose commonly shows but slight effects of wear and is apparently near the point of origin of its constituent material. At higher horizons, there often is yet other arkose, in most cases showing more signs of wear and apparently having been transported for a greater distance; and in still other cases, as has been noted, arkose composes the whole of a formation, thousands of feet in thickness. The deposits are of such differing types and have such different associations with coal measures, with mud-cracked red beds, with beds containing faceted pebbles, and with beds carrying marine fossils, that the old conception of the limited significance of arkose is manifestly incorrect, and arkose must be significant of several types of conditions. AN UNUSUAL FORM OF VOLCANIC EJECTA? WALLACE E. PRATT Chief Division of Mines, Bureau of Science, Manila, Philippine Islands In the course of a study of the eruption of Taal Volcano, in southwestern Luzon, Philippine Islands, during the month of February, 1911, I noted the presence of small concretion-like bodies in the finest-grained portion of the blanket of fragmental ejecta which the eruption spread over the surrounding country. It will be recalled that the eruption in question was characterized, by the expulsion of great volumes of water-vapor, charged with ash or sand, together with a small proportion of coarser fragmental material. The eruption destroyed completely a dozen small villages, with attendant damage to crops and live stock, and killed 1,335 people. A thin layer of mud and dust was spread over an area of about 1,000 square kilometers, extending principally to the north and west of the crater. I commented upon the presence of the spherical bodies in the ash-fall at the time as follows: An interesting feature of the fall of the ejecta is the formation of drops or balls of mud. These were observed most abundantly on the island itseif, but were seen at Talisay and Banadero also. They range in size from large shot to hazelnuts, and when broken sometimes show concentric markings. Apparently they fell late during the activity, being found just below the sur- face of the deposit. These mud balls cannot be classed as lapilli in the strict sense of that term, since they are built up, probably through the condensa- tion of steam into drops of water. The accompanying vertical section of the fall of mud or ash [text Fig. 2] was taken on the southwest slope of the volcano? Text Fig. 2, referred to in the quotation, is reproduced here- with as text Fig. 1. Taal Volcano forms an island near the center of a lake from 15 to 20 kilometers in diameter. Thus the mud balls, which were found both on the slopes of the volcano and at the villages of Talisay « Published by permission of the Director, Bureau of Science, Manila, Philippine Islands. 2 Wallace E..Pratt, Phil. Jour. Sci., Sec. A (t911), VI, 71. 450 AN UNUSUAL FORM OF VOLCANIC EJECTA 451 and Bafiadero on the margin of the lake, from 6 to 8 kilometers distant from the crater, must have been widely distributed; never- theless, at the time I was inclined to attribute their formation to accidental, rather than to common, conditions of explosive volcanism. The literature accessible to me revealed little evi- dence that ejecta of this character had been observed generally, although the following description by Edward Otis Hovey of “drops of mud,’ which he encountered after the eruptions on ‘Martinique in 1902, shows that simi- lar phenomena have been noted: Present Surface ane Very fine ash In addition to the showers of dry dust | and ashes, there fell during the eruption an immense amount of liquid mud which had been formed within the eruption cloud through the condensation of its moisture. This mud formed a tenacious coating over everything with which it came in contact. That drops of mud, too, formed in the air and fell as a feature of the eruption is proved by the condition of the walls the of houses in Precheur, on which I found flat- tened spheroids of dried mud which could [/ermer Surface have formed only in the manner indicated. These flecks of mud were two, four, and Gill om ine soniimedion loss of even six inches across, where two or more Jaaj Volcano in January, rgrr, had coalesced. They occurred mostly on showing balls of dried mud near top the northern and eastern walls of the of layer. houses. The testimony of the people as to the occurrence of rain during the great eruption is conflicting, but the evidence of the coating and these drops of mud proves that much aerial condensation of steam accompanied these outbursts.t Mud balls (ash) | | | | § | | | | Fic. 1.—Section of ash which More recently I have come upon evidence which leads me to the belief that the formation of mud balls has been rather character- istic of that type of volcanic activity which results in the explosive eruption of great clouds of dust-laden steam, at least where atmos- pheric conditions similar to those on the island of Luzon prevail. In the examination of samples of strata pierced in drilling for artesian water at the towns of Bauan and Taal, distant 25 and 15 kilo- raeters, respectively, from the crater of Taal Volcano, abundant 1 Edward Otis Hovey, Am. Jour. Sci., XIV (1902), 343. { 452 WALLACE E. PRATT spheroidal and ellipsoidal inclusions were found which are indis- tinguishable from the mud balls of the last eruption of Taal. These ejecta may have come from Taal itself, or from some other of the numerous small craters which are known to have existed in southwestern Luzon formerly. The wells were drilled by the Bureau of Public Works with standard drilling rigs, and the samples Pe | | = Fic. 2.—Photograph of balls of dried mud which fell with the fine tuff portion of the ejecta of Taal Volcano, in the eruption of rort. studied were submitted by the drillers. The balls of dried mud came from depths of from too to 150 meters in very loosely con- solidated, silt-like volcanic tuff, fragments of which had evidently caved into the well and had been brought to the surface by the sand pump or bailer. Some of the balls were broken, but many were intact in spite of the disintegrating effect which the rushing action of the water into the bailer must have caused. The size and appearance of the balls are well shown in the accompanying photograph (Fig. 2). One specimen still imbedded AN UNUSUAL FORM OF VOLCANIC EJECTA 453 in the tuff appears near the center of the photograph. The broken surfaces display clearly the concentric structure which is charac- teristic of these bodies. The balls can be disintegrated between the fingers when wetted, and the individual particles prove to be like dust in size. That these aggregates have not resulted from solution processes nor from dynamism is evidenced by the facts that they do not contain calcium carbonate nor any other extra- neous cementing agent, and that the beds in which they occur have certainly not experienced metamorphism. The theory which Dr. Hovey advanced to explain the presence of ‘‘drops of mud”’ in the ejecta from Mont Pelée accounts satisfactorily for the similar, although apparently smaller, balls of dried mud in the loose tuffs of southwestern Luzon. W. H. Brown, botanist, Bureau of Science, has submitted to me several hundred balls of dried mud which he found included “in the upper part of a thick bed of volcanic tuff”? on the slopes of Mount Maquiling, an extinct volcano about 20 kilometers north- west of Taal. He had been engaged in a study of the flora of Mount Maquiling and had encountered these balls in the course of a soil survey. They are precisely like those already described in shape and structure, but many of them are larger and they have a brownish-yellow color, whereas the Taal products are light gray in color. They consist of the same material as the inclosing bed—clayey, fine-grained tuff. The balls from Maquiling attain a diameter in rare specimens of as much as 4 centimeters, thus being comparable in size with the drops of mud observed by Dr. Hovey, and are so hard that they can be broken only with diffi- culty between the fingers. The appearance of a face in the tuff bed containing these balls is shown in Fig. 3. The concentric structure of the balls is again revealed in this photograph. Recently, also, I have encountered well-preserved balls inclosed in clayey tuff on Bondoc Peninsula, Tayabas Province, and near the Santa Lutgarda iron mine at Angat, Bulacan Province, widely separated parts of Luzon. The tuff beds in these localities are of greater age than the recent tuffs in the Taal volcanic region, dating back, probably, to the late Miocene. The tuff is slightly indurated, but the balls have retained their form and display 454 WALLACE E. PRATT clearly the characteristics already recorded in describing the ejecta from Taal. Iam confident that they originated in the same manner in each case. The suggestion arises, in view of the foregoing observations, that the condensation of mud into drops or balls must be a rather common feature of volcanic eruptions which throw out great clouds of water-vapor and fine sand or dust. The product may be Fic. 3.—Photograph showing close view of a face in a bed of clayey tuff con- taining “‘mud balls’’; slopes of Mount Maquiling, southwestern Luzon. About one- third natural size. described, perhaps, as a volcanic hailstone. Undoubtedly, the contour of such bodies is often destroyed by the impact of the fall to the ground surface. Probably only where the drops have had opportunity to dry out somewhat before reaching the earth and where they strike in soft, unconsolidated beds of recently fallen tuff, is their form preserved under subaerial conditions. It would appear to be equally remarkable that they should retain their form upon falling into water. Yet it is beyond question that the tuff series into which the wells at Bauan and Taal penetrated is in great part water-laid, and it is to be presumed that the mud AN UNUSUAL FORM OF VOLCANIC EJECTA 455 balls encountered in the wells at these towns fell into the sea originally. Unless conditions peculiar to the tropics, such as high temper- ature and, perhaps, excessive humidity, are essential factors in the phenomena which have been described, it would appear that mud balls should have been formed in the eruption cloud from Katmai Volcano in Alaska and in the recent eruptions of Mount Lassen in California. So far as I have observed, none of the published accounts of the eruptions of these volcanoes have mentioned ejecta of this character. RIPPLE-MARKS IN OHIO LIMESTONES* CHARLES S. PROSSER Ohio State University INTRODUCTION ORDOVICIAN RIPPLE-MARKS Elk Run Cherry Fork Trebers Run Review of Previous Work SILURIAN RIppLE-MARKS Beasleys Fork Lick Fork Sproull Glen Lawshe Quarry Elk Run Sharpsville Near Peebles Near Locust Grove Leesburg Monroe Formation DEVONIAN RIPPLE-MARKS Sandusky INTRODUCTION Dr. Edward M. Kindle, in a “Note on a Ripple-Marked Limestone” in the Devonian of northern Manitoba, published in 1912, stated that: ‘‘the occurrence of ripple-marks on sandstone is a common phenomenon to every geologist... . . The literature on ripple-marks relates almost entirely to these familiar sand or sandstone ripples. The occurrence of ripple-marks on limestone seems to be a phenomenon of such relative infrequency that it appears desirable to record an example which has come under the writer’s notice.” In a later paper Dr. Kindle has described ripple- t Published by permission of the State Geologist of Ohio. 2 Ottawa Naturalist, XX VI (December, 1912), 1 (reprint). 450 RIPPLE-MARKS IN OHIO LIMESTONES 457 marks in the Trenton limestone near Hull, Quebec,’ and given a summary of previously described ripple-marks in American lime- stones.? At an earlier date Dr. August Foerste had noted the occurrence of wave-marks (ripple-marks) in the Ordovician and Silurian limestones at a number of localities in Kentucky, Ohio, and Indiana. Recently Professor J. A. Udden has described ripple-marks in the Burlington limestone of Iowa and in limestones of Comanchean age in Texas.4 On account of the comparative infrequency of described observa- tions of ripple-marks in limestone the writer has concluded to record the most conspicuous of those which he has seen while engaged in field work in Ohio. These will be grouped in the several geologic systems in which they were observed, arranged in ascending order. ORDOVICIAN RIPPLE-MARKS Elk Run.—The best ripple-marks seen in the Ordovician are in the upper part of the Richmond formation on Elk Run in the northwestern part of Adams County. This locality is on the Marion Dunlap farm, about 14 miles east of Winchester and 3 miles west of Seaman, where the ripple-marked layer of limestone forms the floor of the run for a considerable part of the distance between the Norfolk & Western Railway trestle and the highway bridge. An excellent view of these ripple-marks may be had from the Norfolk & Western passenger trains while crossing the trestle if one looks downstream to the north. The first series of ripple-marks is on a layer forming the bed of the run a short distance below the trestle and continuing up a branch from the west for about two rods. The direction of the ripple-marks is about due north and south. The more gradual slope (stoss) is to the east, and the steeper (lee) to the west. The distance apart (amplitude) of the crests varies from 28 to 32 inches t Jour. Geol., XXII (1914), 707-0. 2 [bid., pp. 709-11. 3 Jour. Geol., III (1895), 50-60 and 169-97; 1-40 (reprint); Jour. Cincinnati Soc. Nat. Hist., XVIII (1896), 167; Am. Geologist, XX XI (1903), 333-61. 4 Jour. Geol., XXIV (March, 1916), 125, 126; illustrated by Fig. 3, p, 126, Fig. 4, Pp. 127, and Fig. 5, p. 128. 458 CHARLES S. PROSSER in the more normal ones, with, in some instances, shorter and more irregular ones between these crests. The depth of the troughs, from the crest to the bottom, varies from 2 to 3 inches. The crests undulate or curve slightly in crossing the surface of the lime- stone, and this undulation is conspicuous in the bed of Elk Run a little below the branch. Fig. 1 shows the ripple-marks in the Fic. 1.—View of ripple-marks in bed of Elk Run, just below the Norfolk & Western Railway trestle, 1} miles east of Winchester, Ohio. Photograph by C. S. Prosser. bed of Elk Run, just below the railway trestle, and also on the same layer of limestone in the bed of the small western branch. There is a dip of at least 1° to the east as measured on the crests of the npple-marks in the lower part of the branch, and farther up this stream it increases to 2°. The layer of limestone contains large numbers of shells which Dr. Foerste states are covered by sand.t It is rather difficult to make out the sand, although there is granular material to some extent. This horizon t Jour. Geol., III (1895), 50. RIPPLE-MARKS IN OHIO LIMESTONES 459 is given by Dr. Foerste as 60 feet below the top of the Richmond formation.* The dip downstream carries this ripple-marked layer down Elk Run to where it covers the entire bed of the stream for some dis- tance above the highway bridge. At this locality the bed of the stream which is covered by the ripple-marked layer is some 60 feet wide and extends 156 feet along the bed of the stream. This is a beautiful example on a large scale of a ripple-marked limestone. The trend of the crests of some of these ripple-marks is N. 3° W., and they undulate to a considerable extent in crossing the stream. The crests of the normal ripple-marks are from 27 to 32 inches apart. The crests of about three out of five are 29 inches apart, and the average over different parts of the surface of this layer is 29, 30, and 31, inches apart. The more gradual and longer slope (stoss) is to the east, the steeper and shorter slope to the west, and the ripple-marks are clearly asymmetrical. A view of the ripple- marked bed of Elk Run at this locality, looking downstream toward the highway bridge, is shown in Fig. 2. A view of the bed of Elk Run from the highway bridge looking upstream, with the railway trestle in the distance, is shown in Fig. 3. The layers on the western side of Elk Run below the highway bridge are dipping from 3° to 425 N. 10 E. The majority of readings on the different layers, however, gave 3° for the amount of dip. The barometer gave a dip of 5 feet to the east for the surface of the ripple-marked layer from the branch to the bed of Elk Run under the highway bridge, a horizontal distance of 500 feet. Just below the highway bridge on the western bank is a ripple- marked layer, between 2 and 3 feet higher than the fine one in the bed of Elk Run which has just been described. The ripple-marks of this higher layer run N. 30° W., are not so conspicuous as in the lower layer, and do not show much difference in the slope of the two sides. On the western side of Elk Run, not far below the highway bridge, is the house of Mr. Charles L. Bailey, and near water-level above the house is a set that runs about northwest and southeast. The eastern slope of these ripple-marks is more gradual than the western slope. Just below the Bailey house a Ibid., p: 58. 460 CHARLES S. PROSSER fall is formed by a ripple-marked layer in which the ripples run N. 3-5° W., and the steeper slope is to the west. The crests are from 27 to 28 inches apart, and the ripples are more irregular than those in the bed of Elk Run above the highway bridge. It is not certain that this is the same layer as the one above the highway bridge, and Mr. Bailey, who is interested in geology, states that ~ ‘5 ~w Fic. 2.—View of ripple-marked bed of Elk Run, looking downstream toward the highway bridge. Photograph by C. S. Prosser. it is lower. Below the ripple-marked layer in the fall is another one with the ripples running N. 10° W., and the steeper slope to the east, with a more gradual one to the west. Farther down Elk Run, shown for some rods in the bed of the stream and making a lower fall, is a conspicuous ripple-marked layer, which is 5+ feet below the one forming the Bailey fall. These ripple-marks run N. 1o-16° W., the crests are undulating, the slopes steeper to the west than to the east, and the crests from 21, 26, to 27 inches apart. Farther down the run on the eastern side, opposite the E. E. Jamison house, where the pike comes down into RIPPLE-MARKS IN OHIO LIMESTONES 461 the valley of the stream, is a ripple-marked layer. The ripple- marks are not so clearly defined as in the layers farther up the stream; but they apparently run about north and south. Loose blocks of limestone containing pebbles were noted at this locality; but the layer was not located in place. Probably this is the locality described by Dr. Foerste when he says, “Within half a Fic. 3.—View of ripple-marked bed of Elk Run from the highway bridge looking upstream, with the railway trestle in the distance. Photograph by C. S. Prosser. mile of the bridge, farther down, opposite a house on the east bank, plenty of pebbles occur in the rock.”* Farther down the stream, below the next house on its western side and a ford, is a still lower layer, with not very clear ripple-marks. As noted above, the ripple-marks in the limestones along this stream were first described by Dr. Foerste as wave-marks on Elk Horn Creek.? Cherry Fork.—Ripple-marks in the Upper Richmond were also seen in the bed of Cherry Fork, below the highway bridge at 1 Op. cit., Pp. 59. 2 Ibid., pp. 58-60. 462 CHARLES S. PROSSER Harshaville, 4 miles southeast of Seaman, Adams County. The ripple-marks in the highest limestone layer below the bridge run about N. 60° W. and have the more gradual slope ‘to the east of south, and the more abrupt to the west of north. ‘The crests vary from 20 to 28 inches apart. In another limestone layer about 6 inches lower than the first one the ripple-marks run north and south, with the steeper slope to the west and the more gradual one to the east. The crests are from 20 to 30 inches apart. Over part of the floor of the creek below the bridge is a less distinctly ripple- marked layer between the two which have just been described. These run almost directly northwest and southeast, the crests are from 22 to 28 inches apart and are rather flat, and the slope is about the same on each side, making them nearly symmetrical. The northeast slope on a few of them is a little steeper and these do not have such flat crests. Perhaps the tops of the crests of the others have been worn away, which gives them the present some- what flattened form, Trebers Run.—This stream is a western tributary of Lick Fork (called Lick Creek on the Highway Map of Adams County), about 5 miles northeast of West Union, 2} miles below Young’s Chapel, about a mile above Dunkinville, and 9 miles southwest of Peebles. On the southern bank of Trebers Run, about 150 yards above the covered bridge on the West Union and Jacksonville Pike, is a ripple-marked limestone layer exposed for 90 feet or more along the bank. This limestone layer is in the upper part of the Richmond formation. At the time this locality was visited the streams were high and the layer was partly covered by water, so that the con- ditions were not especially favorable for study. Three adjacent ripple-marks run as follows: N. 12° E., N. about 12° E., and the third one about N. and S. The distance between the crests varies from 2 feet 2 inches to 2 feet 7 inches, while the depths of the furrows (troughs) is from 23 to 3 inches. The eastern slope of the ripple-marks is perhaps a little steeper; but there is comparatively little difference in the slope of the two sides. There is a heavy dip for this region downstream to the east, and certain layers on the northern bank some rods farther up the run dip from 4°5 to 5° N. 100 E. RIPPLE-MARKS IN OHIO LIMESTONES 463 This stratum is known locally as the ‘‘washboard” layer, and it is apparently the one described by Dr. Locke in 1838. He stated that ‘‘the waved stratum at Treber’s is exposed in the bed of the fork, about 400 feet in length, and 50 feet in width.”* Mr. William Treber, now eighty-nine years old (July, 1915), who lives on the Treber farm just south of the run, remembers when Dr. Locke studied this locality, and his daughter, Lizzie Treber, stated that the layer described by Locke is believed to have been exposed in the field a few rods northeast of the lower part of Trebers Run, on the eastern side of the Pike. Lick Fork has shifted its bed some- what to the east and the locality is now covered by soil. The strong easterly dip would probably carry the layer now exposed on Trebers Run down to the locality where it is stated that the ripple- marks described by Dr. Locke were exposed. The barometer gave a difference in altitude of 110 feet from Lick Fork at Trebers to the top of the Richmond formation on Lick Fork above Young’s Chapel, 25 miles above Trebers. Miss Treber also stated that formerly the ripple-marked layer was exposed in Lick Fork, about opposite their house, as well as above it; but the high water at the time this locality was studied prevented determining whether any of the layer is now shown when the water in the stream Is at its normal height. Review of previous work.—Ripple-marks in the Upper Richmond in Ohio, so far as known to the writer, were first described by Professor John Locke from outcrops on Lick Fork,? about 5 miles northeast of West Union, Adams County. Locke called it the ““waved stratum”’ and located its horizon as 55 feet below the top of the blue limestone, No. V of his section, and he stated that near a house known as Trebers it was exposed ‘“‘in the bed of the fork, about 400 feet in length, and 50 feet in width.’ Professor Locke, however, stated that ‘‘these waves are not local, but may be traced in the same stratum over tracts of many miles. They have been called ‘ripple marks’; but all geologists will agree that the blue limestone has been formed far below the reach of ‘ripples.’’”4 * Second Ann. Rept. Geol. Survey Ohio, 1838, p. 247 and bottom of Pl. 6, opposite Pp. 242. 2 Second Ann. Rept. Geol. Survey Ohio, 1838, pp. 246, 247, and Pl. 6, opposite p. 242. 3 [bid., p. 247. 4 [bid., p. 246. 464 CHARLES S. PROSSER Messrs. Joseph Moore and Allen D. Hole have described “ripple-marks in Hudson River limestone, in Wayne County, Indiana, 5 miles southwest of Richmond,” which are illustrated by 3 plates... These are probably in the Richmond formation and it is stated that ‘‘the mean distance from crest to crest is approxi- mately uniform for the series, and the average for twenty such distances is found to be 2.63 feet. The average depth of lowest part of troughs below crests is 14 to 1;%y inches.’ At an earlier date W. P. Shannon had described ‘“‘wave-marks on Cincinnati limestone”’ in the bed of Salt Creek, 3 miles west of Oldenburg, in the southwestern part of Franklin County, Indiana. Dr. Orton called attention to the observations of Professor Locke and stated that ‘it is an even more striking characteristic of the rock in its lower beds [Cincinnati group], as shown in the river quarries of Cincinnati, or in the lowermost too feet that are there exposed. . . . . The interval between the ridges varies, but in many instances it is about 4 feet. The greatest thickness of the ridge is 6 or 7 inches, while the stone is reduced to 1 or 2 inches at the bottom of the furrow, and sometimes it entirely disappears.’ Dr. Foerste also noted wave-marks and ripple-marks in the “ Lower Hudson, or Utica’ opposite Cincinnati, at West Covington, Kentucky,’ which in general “run slightly east of north.’ Recently Dr. Kindle has reported Dr. Foerste as stating that “wave-marks (ripple-marks) occur in Ohio, Indiana, and Kentucky, in abundance in the Lower Eden, Upper Richmond, and Upper Brasstield limestones. They occur in great numbers, but not so abundant, also in the Middle Eden. In Kentucky they are common also locally in the Mount Hope bed, at the base of the Maysville. They occur often near the middle of the Arnheim and at various intervals in the Lower and Middle Richmond in the three states mentioned.’” * Proc. Indiana Acad. Sci., t901 (1902), pp. 216-20. 2 Tbid., p..217. 3 Ibid., 1894 (1895), pp. 53, 54. 4 Rept. Geol. Survey Ohio, I (1873), 377- 5 Jour. Geol., III (1895), 56-58. ® Tbid., p. 58. ? Ibid., XXII (1914), 709, 710. RIPPLE-MARKS IN OHIO LIMESTONES 465 SILURIAN RIPPLE-MARKS Beasleys Fork.—Ripple-marks in the upper part of the Brass- field limestone (formerly called Clinton) were noted at several localities in Adams County. One of these localities is in the bed of the upper part of Beasleys Fork, some distance above the house of Walter D. Grooms, which is about 1} miles south of West Union on the Wrightsville Pike. This stream is crossed by three layers of limestone in which ripple-marks are conspicuously shown. The lowest layer is a very crinoidal limestone, from 3 to 10 inches thick, which forms a small fall, and its top is about 8 feet 3 inches below the top of the Brassfield limestone. The ripple- marks run about east and west, with the more gradual slope to the north and the steeper toward the south. ‘The top of the second or middle ripple-marked layer is 4 feet 11 inches below the top of the Brassfield formation, and the layer itself is 5+ inches thick, but, the ripple-marks are not so conspicuous as in the layer below and the one above. These ripple-marks run north and south with the steeper slope on the east side and the more gentle slope on the west. Finally, there is the third or highest ripple-marked layer, the top of which is 4 feet 5 inches below the top of the Brassfield, and opposite the small house on the bank of the creek on the Joe Morrison farm. The layer is a grayish, somewhat greenish- spotted, crystalline limestone, 8+ inches thick. The ripple- marks run about north and south, with the steeper slope to the east and the more gradual to the west. ‘These ripple-marks are heavy and the crests are 26, 31, 34, 35, and 38 inches apart. The distance from the bottom of the trough to the top of the ridge varies from 3 to 9 inches. These ripple-marks were noted by Dr. Foerste in his description of the section ‘‘along the road to Beasley Fork.” The following is that part of this section in which the ripple-marks occur, as described by Dr. Foerste:* Ft. In WESTON eC BWAVC-INAE KEG Meme ce til heres cece aie gies ¢ aedldie ni coetargiehens mie eters a CERT dio ao MRIS Oop CHAE aE ol errene cathe INIA” Searls IMR ERR Bein ESTE eRe ANre vere SLA 8 TGA ESEOMLC MN ONT soe ret eric AS cavalrou vlc Ye drs carats By a areal ceapha see 3 CEST @ Bach aight PMN SINS 0) 2 Dad a a a anaes ee airs Mad Sa 3 t Kentucky Geol. Survey, Bull. No. 7, 1907, Pp. 42. 466 CHARLES S. PROSSER Ft. In Limestone, with large waveamarksys a. 2.0). ss. 0.5245 aes ee ate 6 LAMIESCONENS gi) ice As jae Fekete Mem RS ETA aS ed sesisLxeh es ocek gtd eke oy 6 Clay, witha little thimilitnest@ner cack. s cits ds. oe cs aie ee ee Ee O Limestone with large wave-marks and containing large crinoid beads. . 6 Lick Fork.—At least two ripple-marked layers in the upper part of the Brassfield limestone occur on Lick Fork (called Lick Creek on the Highway Map of Adams County) above and below the highway bridge on the West Union and Jacksonville Pike, about 2} miles northeast of West Union and about opposite the house of J. Frank Young. Stratigraphically, what is apparently the lowest ripple-marked layer outcrops a few rods above the bridge, where there is a strong dip downstream. ‘The rock of this layer is a very crystalline and crinoidal limestone containing large cup corals and numerous fragments of other fossils in its upper surface. The ripple-marks are large and one of them runs N. 70° E., although some of them run perhaps more nearly east and west. The crests of two of them are 30 inches apart, and of another set 31 inches apart. The trough is 33 inches deep and the slope much steeper on the southern than on the northern side. A few rods below the bridge is a layer with ripple-marks which run N. 20° W. and S. 20° E. The crests of these ripple-marks are about 31 inches apart and the western slope is steeper than the eastern one. This ripple-marked layer is about 19} feet higher than the base of the Brassfield limestone. Not far above the highway bridge are ripple-marks running N. 18° W. and S. 18° E., which apparently occur in the same layer as those below the bridge, which ‘have just been described. Farther upstream the direction of the ripple- marks, apparently on this same layer, has changed to N. 74° W. Not much farther upstream than the ripple-marked layer first described one is shown in the bed of the stream, which may be higher than the others; but its stratigraphic position was not certainly determined. The ripple-marks of this layer run north and south, the crests are 30-35 inches apart, and there does not appear to be a marked difference in the angles of the slopes. This is probably the locality where Professor John Locke noted two waved layers in the Flinty limestone (Brassfield), No. III of his RIPPLE-MARKS IN OHIO LIMESTONES 467 section, on Lick Fork.t He gave the top of the upper one (No. 13) as 4 feet 93 inches below the top of a flinty layer (No. 1) (appar- ently one of the layers of the Dayton limestone), which he seemed to consider the top of the flinty limestone. This upper “waved stratum” is given as 3 inches thick, with 20 inches mostly of marl between it and the lower “‘ waved layer” (No. 17) which is reported as 7 inches thick. Sproull Glen.—This glen is on the R. C. Sproull farm, now owned by Mrs. Jennie Black and Dr. O. T. Sproull, not far from Sproull Bridge over Ohio Brush Creek, 6 miles southwest of Peebles. The heavy rains of July, 1915, had deepened and cleared out the bed of this stream to such an extent that three layers of ripple-marked limestone were exposed which on a visit to the same glen in Sep- tember, 1914, were not seen. The lowest ripple-marked layer was shown on the northern side of the stream with ripple-marks running N. 10° W. to N. 30° W. One foot 5 inches higher is another limestone layer 7 inches thick with ripple-marks imperfectly shown on its upper surface. Also 3 feet higher ripple-marks occur on thin layered limestones; but the last two layers were so poorly shown that not many data could be obtained concerning the ripple-marks. The top of the third or highest ripple-marked layer is 9 feet 9 inches below the base of the to-inch zone of Whitfieldella quadrangularis Foerste} in the Brass- field limestone, and 9g feet 3 inches below the top of this formation as exposed in the third fall, or 13 feet 6 inches below the top of the very hard Dayton limestone as exposed in the stream above this fall. Lawshe quarry.—This old quarry is located on the Vincent Robbins farm, north of Lawshe, Adams County, on the Cincinnati Division of the Norfolk & Western Railway. Ripple-marks were noted near the western end of the quarry, which are not well exposed but run N. 16° E. The ripple-marked layer occurs from 1 foot 4 inches to 2 feet below the base of a rather conspicuous 2 foot 2 inch Second Ann. Rept. Geol. Survey Ohio, 1838, pp. 244, 246, and Pl. 6, opposite p. 242, on which only one waved layer is indicated in the flinty limestone. 2 [bid., p. 244, where he states that ‘‘the upper layer of the flinty stratum is peculiarly marked. It is about one foot thick, and contains so much silex that it has the sharp conchoidal or flinty fracture, and gives fire with steel.” 3 Kentucky Geol. Survey, Bull. No. 7, 1906, p. 41. 468 CHARLES S. PROSSER layer, which is 12 feet 10 inches above the top of the ‘‘chert zone”’ in the basal part of the Brassfield limestone and 16 feet 10 inches higher than the lowest outcrop in this quarry. The 2 foot 2 inch layer is variously colored crystalline limestone, which on the weathered faces is apparently cross-bedded and contains a good many pebbles, more or less flat, and some of them of considerable size. The top of the massive layer just described is about 22 feet below the top of the Brassfield limestone and nearly 243 feet below the top of the hard Dayton limestone. Elk Run.—On Elk Run (called Elm Run on the Highway Map of Highland County), 2} miles northeast of Belfast, High- land County, are ripple-marked layers in the Brassfield limestone, which are well shown in the stream a few rods below the iron high- way bridge on the upper road from Belfast to Elmville. This layer is a crystalline limestone, from } to 3 inches thick, which contains fossils and some pebbles. The ripple-marks run N. 80° E., the steeper slope is to the north and the more gradual one to the south. The crests range from 26 to 36 inches apart, 32 inches being the most frequent distance. The deepest trough noted is about 2} inches lower than the crest. A ferruginous limestone 1 foot 4 inches thick occurs just above the crystalline, ripple-marked layer, and another 4-inch ferruginous limestone layer just above this, the upper surface of which is apparently ripple-marked. Below the ripple-marked, crystalline limestone is a layer 1 foot thick containing fossils and numerous pebbles of Brassfield limestone. The majority of the pebbles are rather flat and fairly well rounded on the margins. The size of some of the larger pebbles is indicated by the following figures: one 8} inches long, and three rectangular ones respectively 9X6, 9X8, and 9X83 inches. The pebbles asia rule lie flat (horizontal) or at least nearly so in the rock; but there are some that are imbedded at more or less of an angle. Attention was first called to this locality by Dr. Foerste, who has written as follows concerning it: By far the most interesting feature of the locality, however, was the presence of great wave-marks, wonderfully distinct and well exposed for a distance of a hundred feet down the creek. The line of strike of these wave- marks was magnetically about north 65° east. The crests of the wave-marks RIPPLE-MARKS IN OHIO LIMESTONES 469 were about two inches above their greatest depressions, and the distance from one crest to the next was on the average about 28 inches. They sloped northwards a little more steeply than southwards. This wave-marked layer is only from one to two inches in thickness; and immediately overlies a great mass of pebbles, imbedded in the Clinton just beneath. These pebbles sometimes project strongly into the sandy layer above, which shows the wave-marks. The pebbles are on the average larger than at any place where pebbles have so far been seen in the Clinton. Plenty of them are 12 inches in Fic. 4.—View of ripple-marks as formerly shown in the old Schoepfle quarry, Sandusky, looking southeast. Photograph by C. W. Platt. diameter, and many of them range between four and eight inches. As usual, the pebbles are only an inch to an inch and a half in thickness. Lithologically they are similar to the sandy stratified layers of the Clinton limestone, found characteristically in the lower half of the Clinton in this part of the state, and occurring also at higher levels. If there had been any doubt hitherto about the Clinton age of these pebbles, it was dispelled by the fossils found in some of the pebbles at this locality.t Shar psville-—On Turtle Creek, above Sharpsville, in the western part of Highland County, ripple-marks were noted in the Brass- field limestone; but there was not an opportunity to measure t Jour. Geol. III (1895), 184. 470 CHARLES S. PROSSER them. Opposite Sharpsville and farther down Turtle Creek than the place above mentioned Dr. Foerste reported “‘a thin, sandy layer, very undulated, like ripple-marks where waves have crossed from various directions. Their importance was not appreciated, when observed, and their direction was not carefully noted. Judging from the memory alone the larger ripples had a general northeast course and indicated currents transverse to this direc- tion.””? Near Peebles ——In the Peebles Stone Company quarry, on the northern side of the Norfolk & Western Railway, one-half mile west of Peebles, is a layer 7 feet 10 inches above the base of the West Union limestone, the surface of which is conspicuously ripple- marked. The rock is bluish gray in color, massive, contains Brachiopods, and the upper surface is very crinoidal. On the crests of the ripple-marks are furrows which are apparently trails, The majority of the ripple-marks run in a regular direction, which is N. 6° W.; but an occasional one runs in an irregular direction. The distance between two parallel crests is 27 inches and the trough is 17 inches deep. The distance between two other conspicuous crests is 3 feet ro inches, with a much smaller ripple-mark about half-way between them. In general the eastern slopes appear to be the steeper, although part of them do not show any particular difference, and one is apparently steeper on the western side. About one-fourth mile west of Peebles in the bank of a small stream on the northern side of the Norfolk & Western Railway ripple-marks occur in Niagaran limestone. Ripple-marks are clearly shown in two layers of crinoidal limestone at this locality. In the lower layer one of the ripple-marks runs N. 6° W., and another one, N. 2? W. The crests are 22-23 inches apart and one trough is 4 inches deep. In general the more gradual slope is to the east and the steeper to the west, although in one of them the eastern slope appeared to be the steeper. In the upper layer the ripple- marks run N. 5° E., the crests are 44-45 inches apart, a trough is 43 inches deep, and the steeper slope is to the east. This locality was described as follows by Dr. Foerste: Along the railroad about a quarter of a mile west of Peebles, where the railroad crosses a creek, there are very good wave-marks in the rock on the t Jour. Geol., III (1895), 178. , RIPPLE-MARKS IN OHIO LIMESTONES A471 north side of the railroad, in a sort of quarry. The rock is of a bluish tint, and is some distance above the Niagara shales. It is presumably of the Springfield horizon. The crests of the waves run here north 3° east; they are about 3% inches above the troughs of the waves, and are about 42 inches apart, showing therefore approximately the same characteristics as the waves of Clinton age in Elk Run. They descend more rapidly eastward than westward. The wave-marks are seen at several levels through a thickness of 23 feet oe Fic. 5.—Nearer view of part of layer shown in Fig. 4. Photograph by C. W. Platt. s ; Ole LOCK oc a. Where the railroad crosses the creek, 50 feet towards the southeast, the wave-marks are shown over a larger area. ‘The crests here run north 5° west.t s Dr. Kindle has also made reference to this locality. Another locality of Niagaran ripple-marks is southwest of Peebles on the dirt road for Lawshe, about one-fourth mile west of the West Union and Locust Grove Pike. The ripple-marks are shown in the highway to the west of a run, and the first house to the west is that on the farm of James Graham. The ripple- marks run north and south, the crests are 20, 21, and three of them 24 inches apart. The more gentle slope appears to be to the west t Jour. Geol., III (1895), 190. 2 [bid., XXII (1914), 711. 472 CHARLES S. PROSSER ; and the steeper to the east. These ripple-marks occur in somewhat thin-bedded dolomites, like those to the west of Peebles, called Springfield by Dr. Foerste, and apparently are in the southern continuation of those beds. Near Locust Grove.—On the western side of Locust Grove branch of Crooked Creek, 1 mile west of Locust Grove and nearly opposite the house of James Ogden, in the West Union limestone are ripple- marks running N. 12° W. and S. 12° E. Some rods farther down- stream and 30 feet higher on the bank in the West Union limestone are rather large ripple-marks running N. 10° W. and S. 10° E. Leesburg.—Ripple-marks occur in Lee’s Creek in the gorge below the Davis Mill and north of Leesburg in the northern part of Highland County. The ripple-marks occur in the Blue Cliff dolomite, which Dr. Orton correlated with the Springfield dolomite, and just below a massive layer 7} feet thick, which is well shown on the eastern side of the creek where the rock has been quarried to a small extent. The ripple-marks are rather irregular; but one runs N. 30° E. and they appear to be steeper on the western slope than the eastern. Monroe formation.—Ripple-marks were noted at three horizons in the Monroe formation, 23 miles east of Peebles; but the area exposed in each case was too small and the ripple-marks too imper- fect to make any definite measurements. The lowest layer is exposed in the bed of Morrison Run some distance below the spring | and house of John K. Morrison, on the road to Bacon Flat school- house, and it is about 15 feet higher than the base of the formation. The highest one is shown in the highway leading up the hill to the west of the John K. Morrison house, and is stratigraphically some 86 feet higher than the base of the Monroe. The direction of ‘these ripple-marks is somewhat irregular; but they run about east and west. The third and intermediate horizon is on a layer in the Virginia Product quarry, just south of the Norfolk & Western Railway, about 25 miles east of Peebles, and it is more than 25 feet above the base of the Monroe dolomite. DEVONIAN RIPPLE-MARKS Sandusky.—The best example of ripple-marks in the Devonian limestone that the writer has seen is in Plant No. 2 of the Wagner RIPPLE-MARKS IN OHIO LIMESTONES 473 Stone Company (formerly known as the Schoepfle quarry), on Hancock Street in the southern part of Sandusky. Formerly there was a considerable area of the floor of this quarry on which ripple- marks were beautifully shown, the three views of which repro- duced in this article were furnished by Mr. DeLos C. Ransom, of Fic. 6.—View of ripple-marks as formerly ‘shown in the old Schoepfle quarry, Sandusky, looking northeast and at right angles to Fig. 4. Photograph by C. W. Platt. Sandusky, Ohio, who first called the writer’s attention to them in January, 1902. Most of this area, however, has been destroyed by the work of the Wagner Stone Company, and when studied on September 19, 1914, only a small portion of it remained, which was located about east of the crusher and on both the east and west sides of the quarry track running from the crusher around to the new part of the quarry to the northwest. This ripple-marked 474 CHARLES S, PROSSER layer is in the upper part of the Columbus limestone, the western continuation of the Onondaga limestone of New York. In general, the crests of the ripple-marks which remained in September, 1914, run N. about 20° W., but those farthest to the west run N. about 25° W. Occasionally~two will run together and, in one case at least, then diverge again. The crests of those studied vary from 22, 24, 25 to 26 inches apart, the greater number of them being about 2 feet apart. Mr. Ransom reported those shown in the half-tones as ‘‘being 3 to 4 feet from crest to crest.’ In the somewhat worn surface of the layer it is dificult to distin- guish any particular difference in the angle of slope of the two sides; although, perhaps, some of the ripple-marks west of the quarry track are a little steeper on the southwest side, with a little longer slope on the northeast side. The troughs of these ripple-marks are rather shallow, although the exact depth was not determined. Figs. 4, 5, and 6 are from photographs, furnished by Mr. DeLos C. Ransom, of the floor of the old (Schoepfle) quarry, showing the tipple-marks as they formerly appeared before this layer was mostly destroyed by the extensive work of the Wagner Stone Company. Mr. Ransom has stated that in Figs. 4 and 5 the camera was pointed toward the southeast and parallel to the direc- tion of the ripple-marks. In Fig. 6 the direction across the ripple- marks is toward the northeast, at right angles to Fig. 4, and the crests are three or four feet apart.2 Mr. Ransom has written several times concerning these: ripple-marks, and the following quotation is from one of his letters: Now such ripple-marks 3 or 4 feet wide as are in limestone must have been constructed in water 50 to roo feet deep and waves of immense size in the ancient comparatively shallow and wide sea when our limestones were laid down. “Ripple-marks” hardly expresses or describes these large parallel stone waves. They are perpendicular to the direction of the wind, hence aeons ago winds were as now largely S.W. and N.E. and hence the poles of the earth have not changed since. t Letters of January 25, 1902, and October 28, 1903. ? Letter of September 25, rors. 3 Letter of October 28, 1903. RIPPLE-MARKS IN OHIO LIMESTONES 475 On September 19, 1914, a considerable area in the new (north- west) part of the quarry had been stripped of soil and this upper surface of the limestone was beautifully glaciated. It was all worn smooth and marked with striae of various degrees of strength; but part of them are rather strong. ‘The direction of the striae was determined at different places on this surface and in all cases they ran N. 100° W. (W. 10° S.) and N. 80° E. (E. 10° N.). THE RELATIONSHIPS OF THE OLENTANGY SHALE AND ASSOCIATED DEVONIAN DEPOSITS OF NORTHERN OHIO' C. R. STAUFFER The University of Minnesota In the flat glaciated region to the south and east of Sandusky there are few outcrops of the older rocks. The drainage is mostly by small, weak streams which have not yet had time to erode exten- sive valleys, and railroads have not been compelled to cut deeply in order to establish their grades. About seven miles south of Sandusky, however, where the land is a little higher than in the city and the mantle rock is exceedingly thin, some of the creeks have exposed small sections of bedrock which are somewhat exaggerated by a considerable local dip. Two of the more impor- tant of these are to be found along Plum and Pipe creeks. These sections have been discussed elsewhere,’ but a recent study of the region has added some valuable facts to those formerly given and has made it possible to correlate this Ohio Hamilton with the Devonian deposits of the same age to the north of Lake Erie. Plum Creek heads about nine miles southeast of Sandusky and flows, in a general northerly direction, to the lake. At a point about two miles east-northeast of Prout station, on the Baltimore & Ohio Railroad, it cuts into Huron shale, and a little farther north into the Hamilton beds, exposing the following section: SECTION OF THE HAMILTON ROCKS AND HURON SHALE ALONG PLUM CREEK Thickness Huron Shale Ft. In. PRM SHAE, BICHDIIGHS DUAR Senin GMC Toile, sia shaw cca nase ante 4 ° Widder Beds 11. Prout or Encrinal limestone. A very hard siliceous blue lime- stone containing a little chert and much pyrite. Silicified corals and crinoid stems are abundant, the latter especially in the WICH Gs FAVORS suey ttre 6,5 0 cee eee aie oleae Fares Ke 8 10 * Published with the permission of the Deputy Minister of Mines, Ottawa, Canada. 2 Geological Survey of Ohio, Bulletin No. 10, 4th Series, 1909, pp. 119-22, Pls. VIII and IX. 476 OLENTANGY SHALE AND DEVONIAN DEPOSITS 477 FAUNA Favosites billingst, Zaphrentis prolifica, Fenestella sp., Atrypa reticularis, Spirifer sp., Stropheodonta per plana Olentangy Shale On SOVELe dsl Lenya oy pyeennpamiastemarit cee s asec sods ode 9 yo vans Geen 6 fo) FAUNA Autodetus lindstroemi, Spirorbis angulatus, Streblotrypa hamilto- nensis, Fistulipora corrugatus ?, Hederella canadensis, Tremato- spira sp., Chonetes deflectus, Crania hamiltonensis, Leiorhynchus kelloggt, Leiorhynchus laura, Rhipidomella cyclas, Sprifer mucro- natus, Stropheodonta demissa, Stropheodonta concava, A ctinopteria boydi, Aviculopecten fasciculatus, Glosseletina triquetra, Leiopteri rafinesquit, Muicrodon bellistriatus, Modiomorpha subalata, Mytalarca oviforme, Pterinea flabellum, Pterinopecten vertumnus, Schizodus appressus, Tellinopsis subemarginata, Bembexia sulcomarginata, Cyrtonella mitella, Pleurotomaria capillaria, Or- thoceras sp., Bairdia devonica, Bollia sp., Bythocypris indianensis, Primitiopsis punctulifera, Phacops rana, Fish plates Seolalewargillaceous sort yblucsa ane iens ake c\sciusls sie eter eae 3 6 FAUNA Chonetes deflectus, Chonetes setigerus, Crania crenistriata, Leiorhynchus kelloggi, Spirifer mucronatus, Styliolina fissurella, Bythocypris indianensis FAmimestone ys Gquite Inara DIMEN tania. so eer: we ce Eee fo) 6 FAUNA Cystodictya hamilionensis, Trematopora sp., Chonetes deflectus, Leiorhynchus kelloggi, Leiorhynchus laura, Spirifer mucronatus, Actinopteria boydi, Platyceras erectum, Phacops rana GeO ewSOlemOlI Chee a rms cat eiltre Na oe Us aleccteie) om atroeniare ae cmet 5 fe) FAUNA Chonetes deflectus, Bythocypris indianensis 5. Limestone, blue, lower part shale, very fossiliferous.......... ° 6 FAUNA Ambocoelia umbonata, Chonetes deflectus, Leiorhynchus kelloggi, Stropheodonta demissa, Cypricardinia indenta, Glyptocardia speciosa, Grammysia arcuata, Grammysia bellatula, Grammysia bisulcata, Grammysia constricta, Modiomorphia subalata, Nucula corbuliformis, Nuculites oblongatus, Nuculites triqueter, Nyassa recta, Pholadella radiata, Schizodus appressus, Bellerophon lyra, 478 C. R. STAUFFER Pleurotomaria planodorsalis, Pleurotomaria rotalia, Orthoceras sp., Bairdia devonica, Barchilina sp., Bollia sp., Primitiopsis punctulifera, Bythocypris punctatus, Phacops rana, Fish plate 4. Shale, argillaceous, soft, blue. This shale contains numerous small pyrite concretions, some of which are beautifully twinned crystals. The fossils are probably rare in most of it and appear to be in streaks or layers:,; Mostly covered .5.... 5. oc o.Neee Io oO FAUNA Orbignyella monticula, Leiorhynchus laura, Spirifer mucronatus, Athyris spiriferoides 3. Shale, blue; with pyritized fossils’) ).0.4.. car oo wate soeuee o. 66+ FAUNA (PYRITIZED) Leiorhynchus sp., Leda rostellata, Nuculites triqueter, Bactrites arkonense, Tornoceras uniangulare 2. Shale, blue... Fossilsirather abundant. < .c5 ssi .cdn yc ec ae 3+ o FAUNA Alveolites monroei, Aulopora cornuta, Aulopora serpens, Cerato- pora rugosa, Zaphrentis prolifica, Spirorbis omphalodes, -Acan- thoclema hamiltonense, Ascodictyon stellatum, Batostomella obliqua, Fistulipora involvens, Fistulipora spinulifera, Hederella canadensis, Hederella filiformis, Heteronema monroei, Orbignyella monticula, Orbignyella tenera, Athyris spiriferoides, Chonetes coro- natus, Chonetes deflectus, Crania hamiltoniae, Cryptonella plani- rostra, Cyrtina hamiltonensis, Leiorhynchus kelloggi, Phili- dostrophia iowaensis, Spirifer mucronatus, Stropheodonta demissa Tropidoleptus carinatus, Styliolina fissurella, Bairdia devonica, Primitiopsis punctulifera 1. Shale, soft blue, with some flat calcareous concretions. Fossils TARO Sie cp ss ais 8 Wale cree Ree RIN RRR reece Seat re 5+ o The other important Ohio outcrop is along the south branch of Pipe Creek, one-fourth mile east of Bloomville and about three miles west of the one just given on Plum Creek. At that point the Prout limestone has been quarried to a limited extent, and is therefore well exposed, although the beds of shale below are pretty well sodded over. SECTION OF THE HAMILTON BEDS ALONG SOUTH BRANCH OF PIPE CREEK Widder Thickness Beds Ft. In. 4. Prout or Encrinal limestone. An impure blue limestone some layers of which are very crinoidal and the upper one containing TALE MAUMIGLOUS CORAIS saci otto iete cer Aes wer< Gisle ex cle ner 9 ° OLENTANGY SHALE AND DEVONIAN DEPOSITS 479 FAUNA Baryphyllum verneuilanum, Cladopora canadensis, Cladapora roemert, Cystiphyllum vesiculosum, Favosites alpenaensis, Favo- sites billingsi, Favosites placenta, Favosites radiciformis, Fav- osites turbinatus, Heliophyllum halli, Syringopora intermedia, Zaphrentis prolifica, Polypora; sp., Atrypa reticularis, Athyris vittata, Chonetes mucronatus, Chonetes scitulus, Rhipidomella vanuxemt, Schizophoria striatula, Spirifer audaculus macronotus, Spirifer mucronatus, Stropheodonta demissa, Platyceras erectum, Phacops rana Olentangy Shale 3. Shale, blue, alternating with blue argillaceous limestone; very poorly exposed but weathered blocks of the limestone lie on the steep bank and numerous fossils have weathered out of the SOLLCTMMANVICES ep intesrs ema airs HueinMerumtareeriscern tM uN 2 MaMa I5= o FAUNA Ceratopora rugosa, Zaphrentis prolifica, Spirorbis, angulatus, Spirorbis omphalodes, Acanthoclema hamiltonensis, Batostomella obliqua, Botryllopora socialis, Cystodictya hamiltonesnsis, Fis- tulipora spinulifera, Hederella filiformis, Orbignyella monticula, Polypora sp., Ambocoelia umbonata, Atrypa reticularis, Chonetes deflectus, Chonetes scitulus, Crania hamuiltonensis, Cryptonella planirostra, Leiorhynchus laura, Rhipodomella cyclas, Schizo- phoria striatula, Spirifer mucronatus, Stropheodonta demissa, Fish plate 2. Shale, marly, blue, badly weathered and soil-covered........ Io oOo 1. Shale, blue, with disk-like blue limestone concretions......... 5+ 0 The lower part of the shales in these sections is apparently not very fossiliferous. ‘The same is usually true of the Lower Hamilton of Ontario and in many places of the similar deposits in New York. This is probably even more characteristic of the Olentangy shale in central Ohio. However, at Delaware, Winchell’s type section of the Olentangy, a few poorly preserved fossils' were found which, although only given’ generic identification, are believed to be identical with others found in the shales of the Sandusky region. It was at Delaware also that the crinoid bed was found in the Olentangy shale—a lens corresponding in every way with the thin lenses of crinoidal limestone common in the lower shales of t Geological Survey of Ohio, Bulletin No. 10, 4th series, 1900, p. 89. 480 C. R. STAUFFER the Hamilton beds exposed along Aux Sable River in Ontario. The general make-up, appearance, and physical properties of the shale below the Prout limetsone and the Olentangy shale are the same (Figs. 1, 2, and 3). Moreover, the Delaware limestone, which underlies this deposit at Delaware and at Sandusky, carries the Fic. 1.—A weathered bank of fossiliferous Olentangy shale showing one of the common calcareous layers in this shale along Plum Creek, near Prout station. Fic. 2.—A bank of Olentangy shale along the Olentangy River at Delaware, Ohio. The limestone disks in this bank contain an occasional fragmentary fossil. The lens of crinoidal limestone was found at this locality. This is Winchell’s type locality for the Olentangy shale and shows its marked contrast to the Ohio or even to the blue bands occurring in the middle portion of the latter. same fauna at both places and extends northward into Ontario. Whitfield found the lower part of the Delaware to be the western extension of the Marcellus shale, to which he considered it to be in part equivalent. Wherever the Delaware limestone becomes ™R. P. Whitfield, Proceedings of the American Association for the Advancement of Science, XXVIII (1879), 208. OLENTANGY SHALE AND DEVONIAN DEPOSITS 481 especially shaly, as is often the case, the fauna tends to revert to that of the more typical Marcellus, so that these forms are not limited to the basal portion. But the occurrence in it of certain fossil forms more characteristic of the true Hamilton beds of New York than of the Marcellus of that region has led to the use of Marcellus-Hamilton or Lower Hamilton for the Delaware. In this use of the Hamilton, it is the older and broader sense of that term, rather than the restricted present usage, that was intended. It would more properly be called the lower Erian. The Olentangy Fic. 3.—A bank of Olentangy shale downstream a short distance from the one illustrated in Fig. 2. shale is overlaid by the Ohio shale in the central part of the state and by the Huron or lower Ohio shale at Sandusky. ‘The strati- graphic position of the blue shale in question, therefore, suggests the same correlation that has been made on the meager fauna and the lithological similarity. When it is recalled that the regions in question lie within the same Devonian basin and that the deposits are a continuation of the same general line of Devonian outcrops, traceable by well-records in the covered interval between, this relationship seems worthy of consideration. The relation of the Sandusky deposit to the Hamilton beds of Ontario is much more easily determined. In a memoir on the Devonian of southwestern Ontario, which was recently published by the Geological Survey of Canada, the correlation of the shale below the Prout limestone with the Olentangy has been adopted, and the 482 C. R. STAUFFER Hamilton beds have been divided, in descending order, into the Ipperwash limestone, the Petrola shale, the Widder beds, and the Olentangy shale. In the sections at Arkona, Lambton County, the two lower stages are well exposed. At no place in the province is the bottom of the Olentangy shale exposed, although well- records indicate that it rests directly upon the Delaware limestone. This latter formation is well shown in the excellent outcrops at St. Mary’s and at several places in the vicinity of Brussels and Goderich. In the following section of the Hamilton beds in Rock Glen at Arkona, only the fauna of the Encrinal limestone and of the beds below are given, since the subsequent comparison is made with those portions of the section only." SECTION OF THE HAMILTON Rocks AT ARKONA, ONTARIO Widder Thickness Beds Ft. In. ri. Soil and drifts os5 (eo ie eee res Ne eee Cee 15 ° 1o. Limestone, argillaceous, massive, blue, partly cystalline, alter- nating with thin layers of shale. These beds form the falls at the ‘old:mill’.. oko sek se ree eee ee ene Secret ree 10 8 g. Shale, soft, blue, with calcareous nodules or concretions...... 8 4 8. Limestone, argillaceous; blues.5,.a. 13Scas000 525 260s. 0 seek I 6 7. Shale, usually soft, blue, but some layers harder and more THASSEVE 2.55 Sh ORS Ses retort rea eI eck ee ah eae 17 5 6: ‘Shale.and shaly ‘blue limestones 2-3 jr se 52 sae Seles sk wees 7 ° 5. Coral zone. A decomposed blue or gray shale, often an impure shaly limestone, filled with corals’). ..00i. 2.02. s oeee eects 3 6 4. Encrinal limestone. A hard, pyritiferous, bluish-gray limestone which is full of crinoid segments, coral fragments, and other fossils. It includes some brownish shale near the base....... 2 4 FAUNA Aulopora serpens, Ceratopora dichotoma, Cladopora canadensis, Cladopora roemeri, Craspedophyllum archiaci, Cystiphyllum vesiculosum, Favosites alpenaensis, Favosites placentus, Favosites turbinatus, Heliophyllum halli, Syringopora intermedia, Syringo- pora nobilis, Trachypora elegantula, Zaphrentis prolifica, Dola- tocrinus liratus, Hederella filiformis, Streblotrypa hamiltonensis, Taeniopora exigua, Ambocoelia umbonata, Atrypa reticularis, Athyris vittata, Chonetes coronatus, Chonetes lepidus, Delthyris « The fauna of the other beds may be found listed in the Geological Survey of Canada Memoir (No. 34, p. 164), on the Devonian of southwestern Ontario. OLENTANGY SHALE AND DEVONIAN DEPOSITS 483 | sculptilis, Leiorhynchus laura, Pholidops hamiltoniae, Pholido- strophia iowaensis, Productella productoides, Rhipidomella penelope, Rhipidomella vanuxemi, Schuchertella perversus, Spirifer divaricatus, Spirifer mucronatus, Stropheodonta concava, Stropheodonta demissa, Stropheodonta perplana, Pterinea flabel- lum, Platyceras erectum, Tentaculites bellulus, Phacops rana Olentangy Shale 3. Shale, soft, gritless, blue, containing few fossils except in certain SELEAKSHOLMAV EES ero nameatia | Cormtara ge acs ed Mic veh airine, aie 19 ° * FAUNA Microcyclas discus, Arthracantha punctobranchiata, Chonetes lepidus, Pholidostrophia iowaensis, Platyceras erectum 2. Shale, soft, gritless, blue, with a few thin lenses of very fossilif- erous crinoidal limestone. These beds include a thin zone in which the fossils are small and always pyritized............. 6 ° FAUNA (PYRITIZED) Leda rostellata, Nucula lirata, Nuculites triqueter, Paracyclas lirata, Bactrites arkonensis, Tornoceras uniangulare FAUNA (NON-PYRITIZED) Arthracantha punctobranchiata, Gennaeocrinus arkonensis Palaeaster eucharis, Chonetes deflectus, Chonetes lepidus, Chonetes scitulus, Cyrtina hamiltonensis, Parazyga hirsuta, Productella spinulicosta, Schuchertella perversus, Spirifer mucronatus, Actinopteria boydi, Glypodesma erectum, Bellerophon triliratus, Platyceras erectum, Platyceras rarispinosum, Platyceras sub- spinosum, Pleurotomaria delicatula, Styliolina fissurella, Tentaculites attenuatus, Tentaculites bellulus, Primitiopsis punctulifera, Phacops rana 1. Covered interval to the level of the Ausable River.......... sie) ° The most important point of similarity between the Ohio and Ontario sections is to be found in the fauna of the thin layer about 25 feet below the base of the Prout limestone, and at a similar dis- tance below the Encrinal limestone of Ontario. In both cases the fossils are pyritized and of small size. Although the fauna of this layer is somewhat more limited in Ohio, the four species that have been found in it are identical with those of the similarly located layer of the Arkona section. At no other horizon has this fauna been found in Ohio, and three of the species have not been 484 C. R. STAUFFER found outside of it in Ontario, and even the fourth but sparingly. It seems certain, therefore, that this is the same horizon in both cases. From the prominence of Bactrites arkonensis in this layer at Arkona and Sandusky, it may be termed the Bactrites horizon. The question next arises as to the relationship of the beds above the Bactrites horizon. In the Sandusky region the fossils of this portion of the formation seem to be more abundant in certain streaks or beds. To a limited extent the same is true of the shales below the Encrinal limestone in Ontario, but there the great body Fic. 4.—The Prout or Encrinal limestone overlain by the Huron shale at Slate Cut along the Lake Shore and Michigan Southern Railway three miles east of San- dusky, Ohio. of the deposit between the Bactrites layer and the Encrinal lime- stone is very sparingly fossiliferous. The Encrinal limestone may be described as several layers of a hard, pyritiferous, bluish-gray limestone which is often full of crinoid fragments—a description which fits equally well the Prout limestone (Fig. 4) in the sections here under consideration and especially the middle layers at Bloomingville. Along Eighteen Mile Creek, New York, one of the important fossils of the Encrinal limestone is Delthyris sculptilis. Grabau says: ‘‘This species is entirely restricted in this region to the Encrinal limestone, and may be regarded as the typical fossil of the fauna.’’' This is also the case in Ontario and probably led Shimer and Grabau to correlate the limestone in Ontario with t Bulletin of the Buffalo Society of Natural Science, Vol. VI (1898), 32. OLENTANGY SHALE AND DEVONIAN DEPOSITS 485 the similar one in western New York,’ although many other forms are also common to the Encrinal limestone of the two regions. The fossils in the beds immediately below the Prout limestone are more abundant than in the shale just below the Encrinal lime- stone of Ontario. In this respect the northern Ohio deposit shows more decided relationship to the western New York section. In fact, the upper part of it includes a portion of the fauna of the “Demissa bed,” although it lacks Spirifer granulosus and some of the other prominent forms. However, this suggested relationship with the New York section is not fully substantiated. In addition to the marked lithological similarity and strati- graphic relation of the Prout and Encrinal limestones, over 75 per cent of the fauna of the Prout limestone also appears in the Encrinal limestone of Ontario, and the upper layers contain many of the corals of the coral zone at Arkona, Ontario. It seems reasonably certain, therefore, that the Prout limestone is the Ohio representative of the Encrinal limestone to the north and perhaps to the east as well. At Kettle Point, Ontario, the Devonian black shale rests upon a limestone of the Hamilton which lies about 150 feet above the Encrinal limestone, and well-records? show that this is the usual succession of beds in Lambton County, Ontario. In Middlesex and Kent counties, which lie to the south of Lambton, this limestone is sometimes present, but at other places is wanting,’ as might be the case where erosion has taken place prior to the deposition of the black shale. The Huron, or basal portion of the Ohio, lies directly upon the Prout limestone at Sandusky. It therefore either represents the upper Hamilton shale and limestone of Ontario, or these deposits are wanting in northern Ohio and the Huron shale rests unconformably on the Encrinal limestone. On the basis of the fossils and the occurrence of spheroidal concretions in both deposits, Kindle has correlated the black shale at Kettle Point, Ontario, with the Huron shale of northern Ohio.’ If this correlation t Bulletin of the Geological Society of America, Vol. XIII (1902), 164, 166. 2H. P. H. Brumell, Geological Survey of Canada, Ann. Rept., V, Part Q (1892), 61-70. 3 Ibid., pp. 52, 73, 74- 4 Geological Survey of Canada, Summary Report for 1912 (1914), pp. 287, 288. 486 C. R. STAUFFER is correct, as seems probable, the Huron shale does not repre- sent the Upper Hamilton, but rests unconformably on the Prout or Encrinal limestone. Mr. Allen R. Stuckey, who has drilled numerous wells in Crawford and adjoining counties, reports that at Bucyrus’ the drift ranges from 55 to 80 feet in thickness. Under this is 35 to 200 feet of black shale, which is usually succeeded below by about 10 feet of gray shale, so tough and sticky that it is difficult to drill. This gray shale immediately overlies the limestone, but in a few wells it has been found to be absent where the black shale rests Fic. 5.—The Olentangy-Ohio shale contact at ‘Dripping Rock,” near Liberty Church, Delaware County, Ohio. Here again there is an undulating contact. directly upon the limestone. In the eastern part of this county, 30 feet of the gray shale is found at many places. It is evident, therefore, that in Crawford County the Olentangy shale is even more variable in thickness than it is in central Ohio and that the Prout limestone of the Sandusky region has disappeared. At “Dripping Rock” (Fig. 5), in Delaware County, where the Prout limestone is wanting and the Olentangy shale is only about 31 feet in thickness, the contact between it and the overlying Ohio shale is most marked and slightly undulating. The contact is equally marked at High Banks, near the Franklin-Delaware county line, and again in the town of Delaware (Fig. 6). At this latter place the basal Ohio shale is somewhat arenaceous. Near the Ohio River, at Kinkead Springs, Pike County, the Ohio extends down to the Silurian limestone and is firmly welded to it. OLENTANGY SHALE AND DEVONIAN DEPOSITS 487 Southward from Kettle Point, Ontario, therefore, the Huron or lower portion of the Ohio shale rests on older and older beds to which its relationship must be that of unconformity (disconformity). This relation is not strikingly perceptible at any one place, but in southern Ohio the time interval between the Silurian and Devonian strata, which are in contact, is enormous. When it is recalled that the first effect of running water on a newly uplifted land surface is to roughen it, and that continued erosion tends to produce planeness, it is clear that, where little or no folding or tilting of the stratified rocks has taken place, slight (apparent) unconformities are likely to Fic. 6.—The sharp and slightly undulating contact between the Ohio and the Olentangy shales in the clay pit at Delaware, Ohio. represent great intervals of time, while conspicuous ones may stand for shorter intervals. Or, in other words, the greater the time inter- val whichis represented by an erosional unconformity (disconformity) in undisturbed strata, the more evasive it is likely to be. This is probably one of the chief reasons for the marked differences in the interpretation of sections where such gaps in sedimentation occur. With the Hamilton beds at Sandusky and in central Ohio resting on the Delaware limestone (Lower Erian) and overlaid uncontormably by the Ohio shale in both places, the advisability of calling the soft marly beds to the south of Sandusky, Olentangy shale seems to be justified, even though the faunal evidence may not be as conclusive as could be desired. EVOLUTION OF THE BASAL PLATES IN MONOCYCLIC CRINOIDEA CAMERATA. I + HERRICK E. WILSON United States National Museum, Washington, D.C. CONTENTS PAR EE INTRODUCTION ACKNOWLEDGMENTS REVIEW OF WACHSMUTH AND SPRINGER’S THEORY OF BASAL PLATE EVOLUTION REVIEW OF CERTAIN PHYLOGENETIC CHARACTERISTICS IN MONOCYCLIC CAMERATA : DETAILED STUDY OF PROCESSES ACTIVE IN PLATE EVOLUTION 1. Changes Not Primarily Modifying the Relations of Plate Contact and Position a) Symmetrical Growth b) Reduction of Parts by Absorption c) Anchylosis PARESTE 2. Changes Primarily Modifying the Relations of Plate Contact and Position a) Reduction and Compensating Growth b) Enlargement and Compensating Reduction c) Plate Division d) Plate Migration e) Plate Interpolation tf) Anchylosis ANCHYLOsSIS: ITs ANTECEDENTS AND CONSEQUENCES 1. Anchylosis and Reduction 2. Delayed Anchylosis 3. Anchylosis and the Phylogenetic Reappearance of Sutures DEVELOPMENT OF THE INTESTINE AND THE CONSEQUENT ZONES OF POTENTIAL WIDENING ORIGIN OF THE ANAL PLATE 488 BASAL PLATES IN CRINOIDEA CAMERATA 489 PART III EVOLUTION OF THE BASE IN MoNocycLic CAMERATA 1. Evolution of the Pentagonal Base 2. Theories for the Evolution of the Base in Hexagonal Camerata a) Enlargement of the Posterior Basal b) Development of the Quadripartite Base c) Development of the Tripartite Base 3. Evolution of the Base in the Hexacrinidae a) Evolution of the Tripartite Base b) Evolution of the Bipartite Base 4. The Succession of Basal Changes in the Platycrinidae and the Hexa- crinidae SUMMARY OF CONCLUSIONS BIBLIOGRAPHY PART I INTRODUCTION Most students of fossil crinoids have been interested in the morphological results of evolution rather than in the morpho- logical processes of evolution and their results, and the description of extinct, fossil crinoids as well as of most living ones has been in general a tabulation of those results. In many cases the tabulation is incomplete, not only because of the rarity and incompleteness of the specimens described, their poor preservation and sometimes poorer preparation, but also because the obscure morphological processes which have given rise to their characteristics are either overlooked or misinterpreted. The greatest difficulty in estab- lishing a natural classification lies in the fact that processes giving rise to morphological characteristics must be known before a cor- rect interpretation and classification can be made. From necessity paleontologists are familiar with the obscurity of processes. The ontogeny of the fossil crinoid is only partially revealed, and that mostly in the phylogenetic succession, for the delicately constructed embryos are ill-adapted for preservation, and comparatively few immature individuals have been preserved. The complete em- bryonic and larval development of modern crinoids is known only in one highly specialized genus, Antedon; the larval stages, however, are partially known in four other genera, Promacocrinus, Thaumato- crinus, Comactinia, and Hathromeira. 490 HERRICK E. WILSON Various phases of the evolutional changes in the basal plates of crinoids have been considered by both paleontologists and zodlogists. The relation of these plates to the column, their modification by the anal plate, and their characteristic decrease in number and size with the passage of geologic time attracted attention long before the importance of tegminal structure was realized," and some interesting though unsuccessful attempts have been made to use basals as a foundation for classification.? Such artificial classi- fications have been swept away, but the fact remains that the changes of each plate and system of plates, as they have passed through the varying stages of evolution, must have classificatory value. When, where, and from what the monocyclic Crinoidea originated are questions of great importance, yet not of special import for the subject as herein discussed. The fact that mono- cyclic crinoids have existed, that their basal plates exhibit a series of remarkable changes, and that certain features in their phylogeny throw much light upon the character of the changes and their succession, is sufficient for present purposes. The process by which these changes came about is the problem to be considered here. In this paper no attempt at reclassification will be made, but the changes exhibited by the basal plates will be reviewed, the nature of these changes studied, and suggestions arising from these studies will be applied to certain theories of descent that have arisen from similar studies made by others. ACKNOWLEDGMENTS The writer desires to express his sincere appreciation to Professor Stuart Weller, of the University of Chicago; Mr. Frank Springer, Mr. Austin Hobart Clark, Dr. E..O. Ulrich, and Dr. R. S. Bassler, of the United States National Museum; and to Professor R. A. Budington, Dr. Charles G. Rogers, and Dr. Maynard M. Metcalf, of Oberlin College, for the material assistance given by them in the use of specimens and literature, and for a broad view of the * Ref. 38. (The reference figure is the number assigned to the work cited in the Bibliography.) 2 Ref. 2; ref. 22. BASAL PLATES IN CRINOIDEA CAMERATA 491 processes operative in the growth and modification of many groups of invertebrates. To Professor Weller, Mr. Springer, and Mr. Clark the writer is especially indebted for their many kindnesses, their helpful advice, and the privilege of freely studying their collections of fossil and recent crinoids. REVIEW OF WACHSMUTH AND SPRINGER’S THEORY OF BASAL PLATE EVOLUTION While various phases of basal plate evolution have been touched upon by numerous writers, the discussion by Wachsmuth and Springer’ is the only one in which a general treatment of the sub- ject has been undertaken, and, in order that the reader may have their theory clearly in mind before going farther, it will be quoted in substance. The base of a monocyclic crinoid is composed of a single cycle of plates, termed the basal plates, lying at the proximal end of the cup, between the stem and the first plates of the radial series. This plate cycle was primarily composed of five separate plates, but, by anchylosis or the union of two or more members, they were reduced from five plates to four, three, two, or one. The first monocyclic crinoids had five basals. Before the close of the Lower Silurian (Ordovician) there appeared two monocyclic genera with four basals, both having a special anal plate inter- posed between the radial. The quadripartite base reached its culmination in the Upper Silurian (Silurian), and disappeared before the close of the Devonian. The earliest genera with a tripartite base occur in the Upper Silurian; some of them have an anal plate, and others not. When that plate is represented, the basals are of equal size; when absent, two of the basals are equal, and the third about half smaller. The two forms continue to exist side by side to the end of the St. Louis group of the Carboniferous (Missis- sippian) when both became extinct. The bipartite base is restricted to the Carboniferous (Mississippian and Pennsylvanian). It occurs from the Kinder- hook group up to the Coal Measures, but is found only among genera with a large? anal plate. t Ref. 39, pp. 52-68. 2 The exact meaning of the statement that the bipartite base is found only among genera with a large anal plate is doubtful, as Pterotocrinus, a genus with a hexagonal bipartite base, usually has, in proportion to the size of its basals and radials, the 492 HERRICK E. WILSON It is evident from these observations that the number of basals was gradu- ally reduced in Paleozoic times, and that in the Camerata the anal plate was introduced after the quadripartite base had made its appearance. It will now be shown that this diminution of number was the result of fusion of two or more of the five original plates, and that by the introduction of the anal plate the base underwent further modifications. The manner in which the modifications in the number of basals were effected may be best understood by reference to the diagrams on Table A. [This table is photographically reproduced as Fig. 1 of this article.] Looking at these diagrams, the transmutation in the Camerata from five basals to a less number is readily understood among genera in which the anal plate is wanting. When the base is quadripartite, it is invariably the two anterior plates of the elementary five which are consolidated (2). In the tripartite base there is a fusion of the posterior with the left postero-lateral basal, and another between the right posterior and the adjoining antero- lateral plate (3). The figure shows that a bisection of the two larger plates will reproduce the original five pieces, interradially disposed. The case is not so simple in genera with an anal plate, where the form of the basal disk is changed from pentagonal to hexagonal (4), as a bisection of smallest anal plate.now known among Camerata in which the anal plate is in apposition with the base. See Pl. IIT, No. 11, and Ref. 30, Pl. 70, Figs. 2-0. When preparing figures for this paper from specimens in Mr. Springer’s collection, with his permission, I overlooked the fact that the dorsal cup of this Species had never been described or figured. As to this, Mr. Springer has furnished me the following note: “When proposing the species Pierolocrinus coronartus (Geology of Kentucky, III, 476, Pl. I, Figs re, 6) Lyon described and figured only the tegmen with the ponderous wing plates, as did Wachsmuth and Springer after him (V.H. Crin. Cam., p. 795). While this work was in press I discovered in the Museum of Comparative Zodlogy at Harvard a lead cast of what was apparently the same specimen, with the dorsal cup attached, which fact I mentioned in a footnote on the page cited. When I acquired in 1903 the collection of the late Col. Sidney S. Lyon, I found associated with the tegmen constituting the published type the dorsal cup reproduced in the cast; the two parts were separated, but I have again united them. It is probable that they belong to the same specimen, and the fact that they pertain to the same species is proved beyond question by another specimen from the same locality having the same dorsal cup with one of the wing plates attached. The species is remarkable, not only for its extraordinary wing plates, but also for the construction of the dorsal cup, in which the basal plates are very small and flat, while the radials are of enormous size, larger than all the other plates of the cup combined; this being the reverse of the structures in Pt. capifalis and all other known species of the genus. It occurs in the Birdsville formation of the Kaskaskia group, in Crittenden County, Kentucky, where it is extremely rare; and also in the Renault formation in Monroe County, Illinois, from which region a fragmentary specimen was described by Hall (Geol. Towa, IT, 689, Pl, XXV, Fig. 7) under the name Dichocrinus protuberans.” BASAL PLATES IN CRINOIDEA CAMERATA 493 the larger plates would produce six plates instead of five. This difficulty, however, is overcome if we consider that the introduction of the anal plate into the ring of radials necessitated corresponding modifications among the basals, as otherwise these plates would lose their interradial position. It 3 4 Ichthyocrinidg. Tehthyocrinidas 18 Rhodocrini a 6 ocrinide. Thysanocrinide Apiocrinides (yonng). Apiocrihidw (adult) Fig. 1.—Diagrams illustrating the evolution of the basals and infrabasals: all figures represent the anal side at the top; @=posterior basal; b and e=postero-lateral basals; ¢ and d=anterior basals; f, g, h, 7, k=infrabasals. required either the introduction of a basi-anal plate, or an increase in the size of the original pieces. That the latter occurred among the Camerata is clearly shown by the diagrams, and the evidence leaves no doubt at what part of the base the extra width was inserted. 404 HERRICK E. WILSON Taking first the quadripartite base, and comparing 2 of the diagrams with 8—one pentangular, the other hexangular—we find that in the latter the posterior basal has doubled in size (7), without materially changing the orien- tation of the plates, or disturbing their general arrangement. ... . In the tripartite base the change was accomplished in a different way. There x is added to plate c (9 and 10), and plates ab and ed have coalesced, and hold relatively the same position as in 3. The bipartite base is probably derived from the tripartite (4), which pre- ceded it in time, and x, which in the latter constituted a part of c, is united with de, and ab with c (11 and 12). Now taking up 7 and eliminating x, so that the side of plate a rests against the plate e, we obtain 2, and by a similar procedure we are enabled to trans- form g into 3. The hexagonal base is thus restored to its primitive pentagonal form without disturbing the orientation of any plate, compound or simple. This theory was thought by Wachsmuth and Springer to have been confirmed by an abnormal example of Teleiocrinus umbrosus n which the anal plate was wanting. Teleiocrinus umbrosus has normally three equal basals, but in this specimen the basal plate to the left of the anterior ray is reduced to one-half its normal size, leaving the basal disk exactly like that of forms which are normally without the anal plate. It is very remarkable that while in all crinoids with an unequally tripartite, monocyclic base, the smaller plate is located to the /eft of the anterior radial, this plate in the base of the blastoids lies invariably to the right (6). In the discussion of the changes from the pentagonal, five-basal form to the hexagonal, four-basal form, it is stated that the enlarge- ment of the posterior basal took place upon the right side of that plate, but no evidence is submitted for this statement. Why could not the enlargement have taken place as well upon the left side, or by symmetrical development upon both sides of the pos- terior basal? The statement that ‘“‘the evidence leaves no doubt at what part of the base the extra width was inserted” is not sufh- cient, and it in itself creates a doubt. Again we are told that “the introduction of the anal into the ring of radials necessitated cor- responding modifications among the basals, as otherwise these plates would lose their inter-radial position.’ However, in the discussion of the basals in dicyclic crinoids the statement is made that ‘“‘the introduction of the anal plate into the ring of TRef. 39, p: 50: BASAL PLATES IN CRINOIDEA CAMERATA 495 9 radials did not affect the basals of dicyclic crinoids” in the same manner as in the monocyclic. While in the latter, when the plate is represented, the orientation of the basals is slightly disturbed, in the dicyclic forms it remains unaltered. The anal plate of the latter rests invariably upon the trunkated upper face of the pos- terior basal; while in monocyclic crinoids it is supported by the basals a and e (Nos. 1o and 12), or occasionally by a and x (No. 8). This statement leads one to believe that no widening of the posterior basal took place upon the introduction of the anal plate in the dicyclic form, but Nos. 15-18 in Fig. 1 show a decided widen- ing of that plate. Overlooking this inconsistency and the fact that trunkation of the posterior basal is in itself an alteration, we are still unenlightened as to why alteration is demanded in the one case and not in the other, as no explanation of an alternative phenomenon is given. Neither is any reason given, other than the position of sutures, for the markedly different positions of enlarge- ment in basal plates of the four-, three-, and two-basal hexagonal forms. By reading between the lines one is able to supply various explanations, yet they are not the explanations of the writers, nor what is needed and demanded by the conditions of the problem. Furthermore, no reasons nor illustrations are given showing why the abnormal specimen of Teleiocrinus umbrosus, by which the theory was apparently confirmed, was oriented with the smaller basal in the left.anterior interray. The questions in the writer’s mind are: Why should the stimulus of enlarging the anal area in the quadripartite form cause enlargement of the right side of the posterior basal and not of the left or of both sides as well? Why should the same stimulus in a dicyclic crinoid have no effect upon the adjacent basal plates? Why should the same stimulus cause the enlargement of the left anterior basal in the equibasaled, tri- partite form, and of the left side of the right anterior basal in the equibasaled, bipartite form, derived therefrom? What was the nature of the change in shifting factor x from basal c to basal d? How could such changes take place without disturbing the orien- tation of any plate, compound or simple? What were the reasons for the orientation assigned to the abnormal specimen of Teleio- crinus umbrosus ? These questions have led the writer to investigate 496 . HERRICK E. WILSON the evolution of the basal plates in the monocyclic Crinoidea, and the result of this investigation will be stated in the latter part of this study. . REVIEW OF CERTAIN FHYLOGENETIC CHARACTERISTICS IN MONOCYCLIC CAMERATA In this review the phylogenetic characteristics of the various groups of monocyclic Camerata will be considered, for it is in this order that the succession is better understood and that the evolution of the basal plates is the most complex. There exist in this order two great groups, separated structurally but not phylogenetically according to the outline of the basal cup, and therefore according to the presence or absence of an anal plate in contact with and trunkating the posterior basal. The first of these groups possesses an anal plate in apposition with the posterior basal, and has in consequence a hexagonal basal. The second group, in which plates of the anal series are either present or absent but not in contact with the posterior basal, has a pentagonal base. In considering these facts, questions of descent and relationship necessarily arise. In the ontogenetic development of the external skeleton in the living Antedon' so many of the phylogenetic steps found in fossil crinoids are illustrated, that it may well be looked upon as a key to their methods of development and perhaps to their interrelation- ships. Leaving out of consideration the earlier embryo-logical stages, let us follow in detail .the methods of plate intercalation and development after the formation and contact of the basal and oral plates (Fig. 2). About the time of the attachment of the pentacrinoid larva, the basal plates assume a regular, trapezoidal outline, the lower part of each being an acute-angled triangle with its apex distally directed (Fig. 2, No. 1). The sides of the lower triangle are bor- dered by a somewhat thickened edge of solid, transparent stereom, the presence of which indicates that the plate has received its full proportionate increase in that direction.?, Even after the plate t Antedon, while more highly specialized in its larval development than some of the other modern crinoids, is chosen for comparison with the Camerata because it is the best known of any of the genera in which the larval development has been studied. 2 Ref. rr, p. 720- BASAL PLATES IN CRINOIDEA CAMERATA 497 edges are thus defined, the plates steadily increase in size, appar- ently by interstitial growth.t The adjacent borders of the plates, however, do not come into absolute contact, as a thin lamina of sarcode is interposed between them until the sutures are closed by anchylosis. The upper margins of the basals have at this time no distinct border,” but are still growing by the process of branching and anastomosis (see p. 501, plate growth). Fic. 2.—Formation of the dorsal cup and migration of the anal plate in Antedon rosaceus: 1, 2, after Thomson; 3-5, after Carpenter; 6, original from specimen in Oberlin College Museum; o=orals; 6=basals; r=radials; a=anal. Shortly after the fixing of the pentacrinoid and the opening of the cup, a third series of plates, the radials, make their appearance in the space left by the beveling off (absorption) of the adjacent lateral angles of the basals and orals (Fig. 2, No. 2),3 the beveling being caused apparently by the encroachment of the radials.4 About the period of the development of the second radials (costals) a forked spicule makes its appearance between the upper parts of the posterior radials. This plate gradually increases in the regular t Interstitial growth: ref. 35, p. 538. 3 Ref. 35, p. 530. 2 Ref. 11, p. 720, Pl. 41, Fig. 1, 0. 4 Ref. 11, p. 720. 408 HERRICK E. WILSON way (see p. 502), until it develops into a round, cubiform plate, the anal plate (Fig. 2, No. 3).1_ The radials, with the anal plate between, now form nearly a complete circle, resting on the basals and sepa- rating them completely from the orals.?,- Although greatly enlarged, the radials are still subquadrangular in outline, the proximal angle occupying the enlarged portion of the interbasal suture, and the distal angle, now trunkated, supporting the narrow costals.3 Con- siderable space still exists between the adjacent radials, except where they are in apposition with the anal plate (Fig. 2, No. 4), and these spaces are filled only with sarcodic substance. The anal plate from proximal growth now comes into apposition with the posterior basal, and the two are mutually trunkated. Upon fur- ther development the radials meet, and their margins assume the finished appearance previously noted in the proximal portion of the basal plates. The posterior radials, especially the right, how- ever, show marked asymmetry, owing to the non-development of the sides adjacent to the anal plate (Fig. 2, No. 5). Growth in the radials does not cease upon their meeting, and as the basals do not now further enlarge, the radials and the anal are forced by contact with each other to extend themselves in an oblique direc- — tion, thus enlarging their distal perimeter, and increasing the diameter of the tegmen. The anal plate by this time has reached its full development and, being more firmly attached to the visceral mass than to the adjacent radials,5 is gradually lifted out of the cup by the extension of the anal tube. The space left by the withdrawal of the anal is gradually filled by lateral growth from the adjacent radials, which, however, do not immediately come into contact. Before the anal is completely withdrawn from the radial cycle, however, the pos- terior radials meet below it, and, as withdrawal continues, cor- responding and continuous enlargement of the radials fills the re-entrant angle, and gives to the plates their bilaterally sym- metrical outline (Fig. 2, No. 6). About this time a very remarkable change takes place in the tegmen. The oral cycle, like the basal one, does not partake of * Ref. 35, p. 529 (anal). 3Ref. 11, p. 720. 2 Ref. 15, p. 314 (radianal). 4 Ref. 11, p. 729. SRef. a ipayse BASAL PLATES IN CRINOIDEA CAMERATA 499 the pronounced enlargement noted in the radial cycle, its diameter being neither increased by growth of its component parts nor augmented by their separation from one another; but, as the ventral disk expands, the orals become separated from the radials upon which they were previously superimposed and are carried upward and relatively inward and the costals and lower distichals are incorporated into the cup. The space now existing between the radials and orals generally remains as a simple, membranous perisome, traversed by the five ambulacral canals; but, in some specimens of Antedon rosaceus,’ and in other modern crinoids, well-defined groups of interradial plates develop in the angles between the brachials. When these plates appear in the cup they are known as interbrachials, and in the tegmen as interambulacrals. Further tracing of the development of the basals and the orals and consideration of stem formation are not here necessary, although they may at times be referred to in the following dis- cussion. ; In tracing the phylogeny of the Batocrinidae and Actinocri- nidae, the anal plate is found as a constant characteristic and the base throughout the series is hexagonal. Complete incorpora- tion of the ambulacral grooves has taken place in the tegmen, and the arms are incorporated in the cup up to and often beyond the second distichals. In Tanocrinus, a genus probably closely related to the ancestors of the Batocrinidae,? five basals are present; the anal plate separates the posterior radials, and is in apposition with and trunkates the posterior basal. In Xenocrinus, Comprocrinus, and Abacocrinus only four basals are present, the anterior pair being obviously united and somewhat reduced in width. In the other genera of the Batocrinidae and Actinocrinidae there are three equal basals, the basal sutures meet- - ing the antero-lateral radials and the anal plate. Why the third basal suture meets the anal plate will be considered later (Plate III). The Melocrinidae, while not showing as complex a basal evolu- tion as that shown in the hexagonal Camerata, are interesting in showing the absence of an anal plate in contact with and trunkating the posterior basal, and in some genera an absence of all plates of t Ref. 35, p. 540. Rete On ps LOA: 500 HERRICK E. WILSON the anal series. The base throughout this family is pentagonal, and the basals number either five, four, three, or one. When five basals are present the basal sutures meet the radials in the normal manner. When four basals are present either the anterior or left anterior suture is missing, and when only three basals are present the sutures meet the anterior, left-anterior, and right-posterior radials. The Calyptocrinidae have throughout a pentagonal base and only four basal plates, and the anal plates are entirely missing. In considering these families the question arises as to whether the base has been pentagonal throughout its whole phylogenetic history or whether there has been a hexagonal stage, as in the Batocrinidae and Actinocrinidae. The Platycrinidae and Hexacrinidae are characterized by having the ambulacral grooves and lower brachial plates but slightly incorporated in the calyx. The orals in the simpler forms are well developed, and the base is either pentagonal or hexagonal. In the Platycrinidae ‘no anals have been positively determined, and the base is pentagonal. Throughout this group there are ordinarily but three basals, five in youth and sometimes but one in age. Two of the basals are large, the third smaller. In the Hapalo- crinus,* the smaller basal is the right-anterior one, and the basal sutures meet the anterior, right-anterior, and the left-posterior radials, as in Stephanocrinus. In the other genera of Platycrinidae the left-anterior basal? is the smaller, and the basal sutures meet the anterior, left-anterior, and right-posterior radials. In the Hexacrinidae there are usually three, or two, equal basals, some- times only one. The base is hexagonal and the anal plate well developed. The basal sutures in the three-basal forms meet the antero-lateral radials and the anal plate; in the two-basal forms the sutures meet the anterior radial and the anal plate. A discussion of the monocyclic Inadunata as a whole cannot now be undertaken, but certain species of Larviformia which in their development illustrate very clearly, even diagrammatically, t Ref. 21, pp. 94-110. ? Exceptions are noted on p. 507 in the description of Fig. 5, No. 6. 3 Ref. 6, p. 158. BASAL PLATES IN CRINOIDEA CAMERATA 501 some of the slighter morphological changes found in the Camerata will be considered in the discussion of morphological principles. Having noted the more conspicuous changes through which the basal cup has passed, a detailed study of the processes involved in these may be undertaken. DETAILED STUDY OF PROCESSES ACTIVE IN PLATE EVOLUTION The processes active in the evolution of crinoid plates, especially the basals and radials, may be divided into two broadly separated though often co-operating groups: (1) those which do not necessarily modify the relation of contact and position of the plates; and (2) those which do modify these relations. The discussion of the first group of processes includes: (a) symmetrical growth; (0) sym- metrical reduction; and (c) anchylosis. The second group includes: (a) reduction and compensating growth; (b) enlargement and compensating reduction; (c) plate division; (d) plate migration; (e) plate interpolation; and (/) anchylosis. I. CHANGES NOT PRIMARILY MODIFYING THE RELATIONS OF PLATE CONTACT AND POSITION a) Symmetrical growth—Plate growth in the Echinoidea and Crinoidea’ is due to the deposition, by amebod cells, of crystalline calcium carbonate, or calcium and magnesium carbonate’ in re- ticulate pattern in the mesenchyme. Three, or perhaps more, of the ameboid calciferous cells fuse by means of pseudopodia into a plasmodium or reticulate tissue. There the pseudopodia meet, the protoplasm forms a small calcareous nodule (intracellular secre- tion, according to Theel; extracellular, according to Semon), which gradually increases along the pseudopodia, forming a triradiate spicule. By further branching and anastomosis, the branches of the spicule meet and fuse at the tips of their processes (Fig. 3, No. 3), thus building up a hard tissue (stereom), showing a strongly t Although no observations have been made upon the early details of plate deposi- tion (stereom formation) in the Crinoidea, the growth of the plate from the primary spicule on so closely parallels that in the Echinoidea that there can be no doubt con- cerning the method of formation. 2 Composition of crinoid skeletons: ref. 25, p. 31; ref. 14, p. 488; ref. 17. - 3Stereom formation: ref. 34. 502 HERRICK E. WILSON reticulated structure (Fig. 3, No. 4) when sectioned in any direc- tion. The co-ordination of deposition is such that each plate acts optically and mineralogically as a single crystal of calcite, without other planes of weakness than the cleavage planes developed by crystallization. The margins of the plates are at first rough with sprouting spicule branches (Fig. 2, No. 2, R), but later, upon coming into mutual contact, become smooth (Fig. 2, No. 5). Growth by branching and anastomosis gives way to interstitial growth,’ and the increase in size is more gradual. Each plate, as Fic. 3.—Stereom formation: 1, formation of the triradiate spicule by the fusion of seven calciferous cells; 2, basal from a larva of Antedon on the sixth day; 3, basal on the tenth day; 4, ideal representation of regular, reticulate stereom. (1, after Theel; 2, 3, after Bury; 4, after P. H. Carpenter.) it now enlarges, is carried relatively outward and away from-the adjacent plates, and in the basal cycle not only away from the adjacent plates, but also away from the axial canal, as is shown by growth lines whenever present. When growth of the plates is symmetrical, each plate in a cycle is the equivalent in size and shape of every other plate in that cycle, and has the same angles with reference to the central axis of the cup as the other plates in the cycle. b) Reduction of parts by absorption —This also is a function of ameboid cells, which are similar in appearance to the calciferous PURGE IS, Db beter BASAL PLATES IN CRINOIDEA CAMERATA 503 cells, but take the calcareous salts into solution and transmit them to the deposition cells... When reduction by proximal or distal, or proximal and distal, absorption is symmetrical throughout a cycle of plates, the relation of parts is not disturbed, unless the reduction Fic. 4.—Stereom absorption; absorptive cells operating upon the posterior end of a calcareous rod in a mature pluteus: 1, one portion of rod separated, the second partially cut; 2-4, advancing stages up to the nearly completed absorption of portion one, the separation of portion two (after Theel). of the cycle is complete, thus bringing previously separated cycles into apposition with each other. c) Anchylosis—Anchylosis is the uniting of apposed plates by an unbroken deposit of stereom in the sutures, and is, as far as we know, an ontogenetically repetitive process, taking place only *Ref. 34, pp. 349-351. 504 HERRICK E. WILSON between plates and not between their formative cell groups. The intrasutural deposit is formed through the activity of the ameboid, calciferous cells, and the firmness with which the plates are united depends upon the amount of stereom deposited. In youth the deposit is slight, the plates are easily separated, and the sutures are usually discernible as external or sometimes as internal grooves. In age, however, much variation exists, for the extent of deposition depends upon the stage of development reached by the group and upon the vitality of the individual. In some cases immature (in- complete) anchylosis is apparently an adult characteristic and the plates are easily separated.t In other cases the deposit is as strong as the stereom of the plates, and fracturing results as readily in the plates as in the old suture plane. Again, in cases where firm union is the rule, as in the Camerata, lowered vitality, or other physiological disturbance, sometimes results in the partial or total inhibition of anchylosis. Such abnormalities, or reversions, are of great value in determining the position of sutures otherwise un- traceable, and will be more fully considered under the topic of delayed anchylosis. From the foregoing definition of anchylosis the conclusion is drawn that any suture or group of sutures appearing in the primitive basal cup may be lost through anchylosis, and the following dis- cussion will show the actual and possible combinations due to simple anchylosis alone. In order to facilitate this description and the tabulation of the variations to be cosidered, the convention of lettering the basal plates from the posterior to the nght-posterior, in an anticlockwise direction, has been adopted, the letters separated by dashes making up the basal formula. The letters from @ to e denote the plates and the dashes the intervening sutures; a dash over a letter shows trunkation of that plate, while the absence of a dash between the letters denotes the absence of the suture between those plates. Thus, a—b—c—d—e— is the formula for the simple, pentagonal base while a—b—cd—e— shows a hexagonal base with the posterior basal trunkated and the anterior pair of basals united. Ref. 1, p. 20; ref, 10;'pp- 36, 37- BASAL PLATES IN CRINOIDEA CAMERATA 505 From a study of Figs. r and 9g it will be seen that eight types of basal modification, by the reduction in number of the basal plates, occur in the Camerata during the Paleozoic era. The primitive base (a—b—c—d—e—), Fig. 1, No. 1, appears in the Ordovician, but it probably originated long before that time. The four-basal type (a—b—cd—e—), Fig. 1, No. 2, appears in the Ordovician, becomes abundant in the Silurian, and disappears before the close of the Devonian. The three- inequi-basal type (ab—c—de—), Fig. 1, No. 3, makes its appearance in the Ordovician, increases in the Silurian, reaches its climax in the Devonian, and disappears in the Mississippian. The three- inequi-basal type (ea—bc—d—), Fig. 1, No. 6, is present during the Silurian, Devonian, and Missis- sippian, but never becomes very prominent. The one-basal type (abcde), Fig. 1, No. 5, occurs at various times during the Paleozoic period, but not as a generic or specific characteristic. The hexag- onal, five-basal type (a—b—c—d—e—), Fig. 9, No. 6, is present in the Ordovician. The hexagonal, four-basal type (a—b—cd—e—), Fig. 1, No. 8, appears first in the Silurian (Richmond) and dis- appears during the Silurian. The three- equi-basal type (al—ca— de—)', Fig. 1, No. 10, appears in the Silurian, increases in the Devonian, reaches its climax in the lower part of the Mississippian, and then disappears. The two- equi-basal type (abc—a«de—),' Fig. 1, No. 12, 1s introduced in the Kinderhook and becomes extinct in the lower part of the Pennsylvanian. Possible combinations of the primitive five-basals: Since Wachsmuth and Springer have assumed that all crinoids having an unequally tripartite base have the smaller basal in the left- anterior interray, it may be well to consider what combinations may be expected from anchylosis of the primitive five-basals. These combinations are explained below and illustrated in Fig: 5. Fig 5, No. 1, illustrates the primitive basal cup of pentagonal out- line, represented by the formula a—b—c—d—e—, and found in Glyptocrinus, Schizocrinus, Stelidiocrinus, and young individuals of Platycrinus. Fig. 5, No. 2, shows a common Ordovician and Silurian type of reduction (a—b—cd—e—), in which the two anterior basals are t Formula based upon Wachsmuth and Springer’s theory. 506 HERRICK E. WILSON anchylosed. In this group there are five possible combinations: ab—c—d—e—, a—bc—d—e—, a—b—cd—e—, a—b—c—de—, and’ ea—b—c—d—. Of these combinations, a—b—cd—e— has been the only one described in the four-basal Melocrinidae; however, the writer has found the a—bc—d—e— (No. 2a) combination in the following specimens in the Springer collection: of eight speci- mens of Melocrinus calvini from Calloway County, Missouri, four ©oe@ 2ee Fic. 5.—Diagrams illustrating types of the possible combinations due to anchylosis of two or more of the primitive five-basals. could be properly oriented’, and these showed the bc anchylosis (Pl. II, No. 4), as did two specimens of M. obconicus ? Hall, and one of the type specimens of M. roemeri. The other four-basal forms of the Melocrinidae(?) and the Eucalyptocrinidae? have the a—b—cd—e— type of base. Fig. 5, No. 3, illustrates a type in which three adjacent basals have been anchylosed (abc—d—e—). Here again five combinations ‘In orienting these specimens the anal tube was used as the reference point. ? This orientation of Eucalyptocrinus is strictly arbitrary, as no indices for proper orientation have as yet been discovered. BASAL PLATES IN CRINOIDEA CAMERATA 7 507 are possible: abc—d—e—, a—bcd—e—, a—b—cde—, b—c—dea—, and c—d—eab—. Of these combinations one, the abc—d—e— combination, is found in Zophocrinus, as figured. This orientation does not agree with Bather’s interpretation of the genus,’ but is based upon the discovery of the anus in a large number of speci- mens studied by the writer in the Springer and Walker Museum collections. ; Fig. 5, No. 4, shows the anchylosis of four adjacent plates (abcd—e—), leaving but one free plate. Here, however, any plate of the five might have been the free plate, and five combinations are possible: abcd—e—, a—bcde—, b—cdea—, c—deab—, and d—eabc—. Of these combinations none have been discovered. Fig. 5, No. 5, illustrates one of five possible combinations in which only one suture exists (abcde—). Combinations 4 and 5 seem too asymmetrical from a structural viewpoint to occur as either generic or specific characters, but might appear in cases of delayed anchylosis, in which complete anchylosis is the normal result. Fig. 5, No. 6, illustrates the simple tripartite combination (ab—c—de—), in which two pairs of basals are anchylosed. Of the five possible combinations of this type four are now known: a—bc—de—, in Allagecrinus americanus;? b—cd—ea—, in Stor- thingocrinus’ and Hyocrinus (No.6,a)4; ab—c—de—, in S ybathocrinus and the Platycrinidae (No. 6, 6); bc—d—ea—, in Hapalocrinus’ and Stephanocrinus® (No. 6, c). The fifth combination (al—cd—e—) has not yet been discovered. Fig. 5, No. 7, illustrates one of the five possible combinations in which three- and two-basals are anchylosed, as abc—de—, etc. Of these only one is known, abc—de— in Mycocrinus,' as figured. Fig. 5, No. 8, shows the complete type of anchylosis (abcde). While this formula recognizes but one type of complete anchylosis, = IRGi5 Oy 10/05 WSO), isi 2 This combination was found recently, by the writer, in two specimens of Allage- crinus americanus in a large collection of that species made by Professor Weller at Louisiana, Missouri. 3 Ref. 3, p. 426. 4 Ref. 6, p. 153. 6 Ref. 19, pp. 212, 351. 5 Ref. 21, pp. 95-105, Pls. 9, to. 7 Ref. 28, p. 110, Pl. 7, Fig. 4. 508 HERRICK E. WILSON the type may have been derived by the simultaneous anchylosis of the five primary plates, or from the closure of the remaining sutures in any of the thirty possible combinations given above. Dealing here with results alone, we see that thirty-one possible modifications of the five primary basals are obtainable through simple anchylosis. [To be continued] EXPLANATION OF PLATES PLATE I Teleiocrinus umbrosus Hall. Abnormal specimen cited by Wachsmuth and Springer as confirmation of their theory No. 1.—Oblique view of posterior interray, showing absence of anal plate and one of the first interbrachials, and the reduction of the right-posterior basal. Basal formula, ab—cd—e—; formula of posterior interray, o—1—3—4—2-. No. 2.—View of right-posterior interray, showing normal arrangement of interbrachials r—2—2—2-—, and reduction of right-posterior basal. No. 4.—Posterior view. No. 6.—Tegmen, showing position of anal tube. Teleiocrinus umbrosus Hall. Springer collection; normal specimens Fig. 3.—Posterior view, showing normal posterior interray and normal base. Basal formula, a)—cd—ex—; formula of posterior interray, A —2—3— 4—1—. To be compared with Nos. 1 and 4. No. 5.—Tegmen, showing position of anal tube in another specimen; to be compared with No. 6. PLATE II Glyptocrinus decadactylus Hall. Springer collection, specimens having a pentagonal base No. 1.—Basal view, showing normal pentagonal base with five basals. Basal formula, a—b—c—d—e-. Chicagocrinus inornatus Weller (type). University of Chicago Paleontological Collection, No. 10787 No. 2.—Basal view, showing anchylosis and reduction of antero-lateral basals and compensating enlargement of postero-lateral basals. Basal formula, a—b—cd—e—. Melocrinus calvint Wachsmuth and Springer. Springer collection No. 3.—Posterior view. No. 4.—Basal view of same specimen, showing anchylosis and reduction of sinstro-lateral basals and compensating enlargement of posterior and right- anterior basals. Basal formula, a—bc—d—e—. BASAL PLATES IN CRINOIDEA CAMERATA 509 Stephanocrinus angulatus Conrad. University of Chicago paleontological Collection, No. 10787 No. 5.—Basal view, showing anchylosis and reduction of right-posterior and posterior and sinstro-lateral basals, and compensating enlargement of right-anterior basal. Basal formula, ea—bc—d—. Note reduction of right- posterior and left-anterior radials. Platycrinus subspinosus Hall. Springer collection No. 6.—Basal view, showing normal Platycrinus type of anchylosis. Basal formula, a—c—de—. To be compared with No. 5. Abacocrinus tesselatus Angelin. Springer collection, specimens having a hexagonal base No. 7.—Basal view, showing asymmetry of posterior radials, enlarge- ment and trunkation of posterior basal, and anchylosis and reduction of antero-lateral basals, coupled with compensating enlargement of postero- lateral basals. Basal formula, a—b—cd—e-—. Melocrinus sampsoni M. and G. (type). University of Chicago paleontological collection, No. 6958; probably Actinocrinus chouteauensis S.A.M. No. 8.—Basal view showing reappearance of anterior basal suture. Basal formula, ab—c—d—ex—. _Batocrinus. University of Chicago paleontological collection No. 9082 No. 10.—Basal view of normal specimen. Basal formula, ab—cd—ex—. No. 11.—Basal view of abnormal specimen, showing reappearance of anterior basal suture. Basal formula, ab—c—d—ex—. Actinocrinus multiradiatus. University of Chicago paleontological collection No. 8959 No. 9.—Basal ‘view with basé removed, showing asymmetry and reduc- tion of posterior radials. Compare with Nos. 7-14. Steganocrinus pentagonus Hall. Springer collection No. 12.—Basal view of normal specimen, showing asymmetry and reduc- tion of posterior radials. Basal formula, ab—cd—ex—. No. 13.—Basal view of abnormal specimen, showing loss of the antero- lateral and posterior basal sutures and the reappearance of the anterior and left-posterior sutures. Basal formula bc—dea—; factor x may or may not be present. No. 14.—Basal view of another abnormal specimen, showing loss of right- anterior basal suture and reappearance of anterior suture. Basal formula, WG On. No. 15.—Basal view of abnormal specimen, University of Chicago pale- ontological collection, No. 8979, showing loss of anal plate. Basal formula, b—cd—ea—. See Plate III. 510 HERRICK E. WILSON PLATE III Steganocrinus pentagonus Hall No. 1.—Posterior view of abnormal form of PI. II, No. 15. No. 2.—Tegmen of same specimen, showing position of anal tube. No. 3.—Posterior view of normal specimen in Springer collection. To be compared with No. 1. Hexacrinus elongatus Goldiuss. Springer collection No. 4.—Basal view of normal form. Basal formula ab—cd—ex—. Note reduction of posterior radials. Hexacrinus anglypticus Goldfuss. Springer collection No. 5.—Basal view of abnormal form, showing loss of posterior basal suture and reappearance of left-posterior suture. Anchylosis of anterior basals normal. Basal formula, b—cd—ea—; factor x may or may not be present. No. 6.—Basal view of another abnormal specimen, showing loss of right- anterior basal suture and reappearance of anterior suture. Basal formula, ab—c—dex—. Talarocrinus patei M. & G, Springer collection No.7.—Basal view of normal specimen. X2. Basal formula, abec—dex—. No. 8.—Basal view of abnormal specimen, showing reappearance of left- posterior basal suture. X2. Basal formula, a—bc—dex—. No. 9.—Basal view of another abnormal specimen, showing reappearance of left-anterior basal suture. X2. Basal formula, ab—c—dex—. No. 1o.—Basal view of another abnormal specimen, showing loss of posterior basal suture and reappearance of left-posterior and right-anterior sutures. X2. Basal formula, bc—d—ea—. Factor «x may or may not be present. Pierotocrinus cirnarius Lyon, Springer collection; Tegmen figured by Wachsmuth and Springer, ref. 39, Pl. LX XIX, Figs. 7a, 76 No. 11.—Basal view, showing great reduction of anal plate, and asym- metry of posterior radials. Basal formula, abc—dex—. Dichocrinus inornatus W. and Sp. Springer collection No. 12.—Lateral view of normal specimen. Xr}. No. 15.—Tegmen of normal specimen showing flexible condition of anal tube and interambulacral areas. X 2. Platycrinus symmetricus W. and Sp. Springer collection No. 13.—Lateral view, showing similarity to No. 12. 1} No. 14.—Tegmen of young specimen, figured by ref. 39, Pl. LXIX, Fig. 1c. Note flexible condition of anal tube and interambulacral areas, and general similarity to No. 15. X2. PLATE I JournaL oF Grotocy, Vou. XXIV, No. 5 JourNaL or Georocy, Vor. XXIV, No. 5 PrAny 1t Journat oF Grotocy, VoL. XXIV, No. 5 ~ Prats TI VARIATIONS OF GLACIERS. XxX? HARRY FIELDING REID Johns Hopkins University, Baltimore, Maryland The following is a summary of the Nineteenth Annual Report of the International Committee on Glaciers.” THE REPORT OF GLACIERS FOR I9g13 Swiss Alps.—Sixty-one glaciers were measured in 1913; a larger proportion were retreating this year than in 1912. The Rhone Glacier, however, has grown in thickness, throughout, with an increase of velocity and an advance of the tongue. Eastern Alps.—The summer of 1913, like its predecessor, was very wet, especially on the north side of the Alps; the conditions during both summers must have influenced the glaciers in 1913. Of the 37 glaciers observed 8 were advancing, 4 were stationary, and only 25 continued their retreat. The increase in the number of advancing glaciers is certain, and the retreat of the glaciers of the Eastern Alps has diminished; it is doubtful if the retreat can now be said to be the prevailing condition. Italian Alps.—In the Piedmont Alps the snowfall has been too heavy to permit of good observations, but the glaciers are appar- ently retreating. Careful photographic surveys have been made of three of the large glaciers on the south side of the Mount Blanc massif; these glaciers are retreating; but the snowfall in the higher regions has increased, so much so in places as to, cause a marked change in the appearance of the mountains. In the Monte Rosa group the observed glaciers were retreating. In the Lombard Alps some glaciers were making slight advances, some slight retreats. On the whole, the large glaciers of the Italian Alps were retreating, but a number of the smaller ones were slightly advancing or were in a doubtful condition. t Earlier reports appeared in the Journal of Geology, Vols. III-X XIII. 2 Zeitschrift fiir Gletscherkunde, [X (1914), 42-65. Gara HARRY FIELDING REID on m bo Swedish Alps —But one glacier, the Mikka, was observed, and it showed no change. Norwegian Alps.—All the 16 glaciers examined in the Folgefon and the Jostedalsbrae were retreating with one exception. On the other hand, in the Swartisen, the Okstinderne, and the Frostisen, 7 glaciers were advancing, one retreating, and two were stationary. Russia.— Observations in the Caucasus and in Turkestan have laid the basis for the determination of future changes. One glacier in the Caucasus was retreating and two were stationary. Canada.—No observations have been recorded in the last two years;' but during the few years before r1o1o the Illecillewaet, the Asulkan, the Victoria, and the Yoho glaciers were all re- treating. Shortly before 1909, however, the Asulkan made an advance. Himalaya.—The greater part of the information collected refers to variations which occurred some years ago. The positions of the ends of many glaciers were determined in 1906, but later observa- tions are not available. At that date there was evidence that the glaciers were generally retreating. New Zealand Alps——Here also observations are scanty. For about ten years aiter the middle of the nineties, several of the larger glaciers advanced. Later conditions have not been reported. REPORT ON THE GLACIERS OF THE UNITED STATES FOR IQI4 Mr. F. E. Matthes sends me the following information: The snowfall during the winter of 1913-14 in the Sierras was so heavy that the glaciers were still completely covered at the end of September; the snow extended even beyond some moraines which encircle the glaciers at a short distance. These moraines are recent; the youngest is comparable to the moraines which marked the advance of the Alpine glaciers at the end of the eighteenth and the beginning of the nineteenth centuries. Historical evidence is not available to determine the actual time when these moraines were formed; but the presence of big trees (Sequoia Washigtoniana) near the glaciers may supply the information; for their rings of growth contain a trustworthy record of the climatic fluctuations of the last three thousand years. «The last report on the Canadian glaciers in this series was in Variations of Glaciers, XIII, report for 1906. See Journal of Geology, XVI (1908), p. 665. VARIATIONS OF GLACIERS 513 Professor Lawrence Martin sends me the following information regarding the Alaskan glaciers: College Fiord.—Miss Keen visited Prince William Sound during the sum- mer of 1914 for the exploration of the Harvard Glacier, and made careful observations of the variations of a number of other glaciers.t | She found that the Harvard Glacier is 18 miles long, or about 28 miles if the Brunonian Glacier tributary is included. It rises at an elevation of about 7,500 feet. The eastern side of the end of Harvard Glacier seems to have retreated slightly between 1910 and 1914, but the western edge near Radcliffe Glacier was still advancing. There was no observable change in Downer, Baltimore, and Smith glaciers. Bryn Mawr may have retreated slightly; the barren zone at its northern border was widest, but the evidence was conflicting, for a shrub north of the glacier was being overturned at the time of Miss Keen’s visit. Vassar Glacier was more crevassed Jn t914 than in 1910, but had not advanced appre- ciably; Wellesley Glacier had retreated slightly; Yale Glacier seems to have advanced a little; Barnard had a slight forward movement. Harriman Fiord.—The-recession of Barry Glacier, observed in 1913, con- tinued in 1914. The total recession of different parts of the ice front, from 1910 to September 25, 1914, was 3,000-7,000 feet. Cascade Glacier was nearly independent of the Barry in 1914. Of the other ice tongues in Harriman Fiord the Baker Glacier advanced at least 1,000 feet between 1910 and 1914, and spread considerably at both borders. The Harriman and Roaring glaciers seem to be still advancing. A small unnamed ice mass on the slopes of Mt. Muir moved forward slightly. The Surprise, Cataract, Serpentine, Toboggan, and Dirty glaciers were unchanged. Eastern Prince William Sound.—The Valdez Glacier, continuing its long- maintained recession, melted back about 200 feet from 1909 to August to, 1914. Shoup Glacier advanced very slightly. Columbia Glacier, the largest ice tongue in Prince William Sound, is also the most interesting, for it has continued the slow forward movement that has been in progress since before 1908. Miss Keen found that the eastern border advanced 1,500 feet between 1910 and September 30, 1914, and spread laterally; in other parts the advance was less, being perhaps 1,300 feet on Heather Island. Photographs of Childs Glacier, by Robert Sewall, show that in July, 1914, its northern border was retreating. Southeastern Alaska.—It was reported in the Juneau papers that Norris Glacier, in Taku Inlet, had made a considerable advance. A photograph of the Taku Glacier, taken about 1907, shows a distinct advance since 18go. Dr. Martin has tabulated the snowfall and temperature as recorded at several Alaskan stations. He finds not only great differences in different years, but the years of maximum snowfall t Bulletin of the American Geographical Society, XLVII (1915), 117-19. 514 HARRY FIELDING REID at the various stations are different, so that it is not safe to draw any detailed conclusions about the snowfall in the mountains from the records at the stations; though the general trends are the same. The annual snowfall at Killisnoo, somewhat more than one hundred miles southeast of Muir Glacier, between 1891 and 1896 was about twice as great as it has been since then; but the glaciers do not show corresponding variations. Temperature records have been kept at Sitka, with two short intermissions, since 1828. The average temperature for the five months from May to September in the years 1828-77 and in the years 1906-13 differs by only one-tenth of a degree Fahrenheit. The average temperature at Juneau for the same months during the years 1906-13 is about 2° F. higher than during the years 1883-96. It does not seem possible to infer any definite relations between temperature and glacier variations from these records. The United States Geological Survey has published a map of a portion of the Chugach Mountains, northeast of Prince William Sound, on a scale of about one inch to the mile (Port Valdez District, Alaska; sheet 602 B). It shows a large area of glaciers and snow- fields. Parts of the Columbia, Shoup, and Valdez glaciers appear onit. Unfortunately the contours are not carried over the glaciers, but the altitudes of a number of points are indicated so that marked future changes in the thickness of the ice will be determinable. Other new maps of Alaskan glaciers cover the Haganita~-Bremner region, northeast of the Copper River Cafion,' the Bering Glacier, and the western border of the Malaspina Glacier at Icy Bay,? and part of the Kenai Peninsula.s An excellent topographic map, on the one-inch scale, showing all the glaciers and their relations to the mountains and rivers, accompanies M. C. Campbell’s Popular Guide to the Glacier National Park.*| Pamphlets containing popular accounts of the glaciers of Mount Rainier and of Glacier National Park have been issued by the Department of the Interior.$ t Bull. U.S. Geol. Surv. No. 576, Plate I. 3 Tbid., Plate VIII. 2 Bull. U.S. Geol. Surv. No. 592, Plate IV. 4 Bull. U.S. Geol. Surv. No. 600. 5 Mount Rainier and Its Glaciers, by F. E. Matthes; Glaciers of Glacier National Park, by W. C. Alden. REVIEWS Climates of Geologic Time. By CHARLES SCHUCHERT. Carnegie Institution of Washington, Publ. No. 192, pp. 263-98, Figs. 87-90. There has been a progressive advance, in late years a most rapid one, from the conception of a former hot, dense, vaporous earth atmos- phere, the natural corollary of the Laplacian hypothesis. Knowledge of glacial climates, which had its beginning in studies in the Alps early in the eighteenth century, has grown until not only has there been demonstrated a world-wide lowering of temperature with glaciation of much of the Northern Hemisphere in recent geologic time, but there has been proved as well a number of such glacial periods in earlier times. The cold climates which have periodically affected the earth more or less widely since the beginning of geologic history have been of short geologic duration. The data at hand indicate at least four well- marked glaciations: (1) earliest Proterozoic, shown by the widespread “slate conglomerates”’ at the base of the Lower ‘Huronian in Canada; (2) latest Proterozoic, marked by thick, widespread tillites beneath the Lower Cambrian of Southern Australia, and by the Gaisa formation of Northern Norway, both now thought to be latest Proterozoic instead of Lower Cambrian; (3) Permian, abundantly proven by tillites in many parts of the world, mostly between latitudes 20° N. and 40°S.; and (4) Pleistocene, the deposits of which mantle much of the Northern Hemisphere. Less well-marked cold periods seem to have occurred (5) at another part of the Proterozoic, for the glacial materials of this age in South Africa represent neither the earlier nor the later part of the era; (6) in the Lower Devonian of.South Africa, shown by the Table Mountain series, and (7) in the early Eocene, indicated by deposits in the San Juan Mountains of Colorado. The greatest reductions of temperatures, so far as known, varied between the hemispheres. Guided by the postulate that the living things of sea and land always have been affected by climatic conditions much as now, climate varia- tions are to be observed in the succession of plants and animals recorded as fossils. In addition, the color and general character of the sedimen- tary deposits afford light on climatic conditions at the time of their 515 516 REVIEWS deposition. In spite of widespread glaciation at certain periods, the Proterozoic era had, in the main, a rather warm, equable climate. This is shown by the enormously thick limestone deposits (50,000 feet in Canada), abundance of large Archaeocyathinae, widely distributed graphites, and presence of coal. The Cambrian, with an abundance of shallow-water life, had a uniformly warm temperature which continued into the Ordovician and Silurian. The red shales, gypsum deposits, salt beds, and scant, depauperate fauna of the late Silurian indicate aridity and possible coolness, the latter expressed perhaps by local glacia- tion (South Africa). The deposits of Northern Europe in the Devonian probably marked a cool, somewhat arid climate, and the great change in the life-forms in the Middle Devonian may be further evidence of the same thing. The climate of the middle and later part of the Devonian was warm; that of the Carboniferous, warm-temperate to subtropic. The great variety of marine life, abundance of reef corals in high lati- tudes, extensive coal deposits, subtropical flora, and large-sized insects, all suggest this. The adverse climate of the Permian is clearly shown in the glacial tillites, red shales, salt and gypsum deposits (to thickness of 3,300 feet), and depauperate, scanty fauna. The sweeping change in the types of life seen in the Triassic is most convincing proof of climatic severities at this time. Large trees (to 8 feet diameter), and their absence of rings, luxuriant ferns, and thick deposits of limestones in high lati- tudes, all suggest warmth. The late Triassic-Lias probably saw a reduction of temperature, for of the Triassic ammonites (1,000 species) none passed into the Jurassic, the insects were uniformly dwarfed, and the corals, both numerically and geographically, were very much re- stricted. The Jurassic was a period of remarkably warm, equable climate. The wide distribution and variety of ammonites (15,000 species), their presence with corals and marine saurians in very high latitudes, and the very cosmopolitan, luxuriant floras are to be noted. The Comanchean- Cretaceous marks the introduction of hardwood forests and may indi- cate a cooler climate than the Jurassic; but the presence of magnolias in Greenland and Alaska shows at least warm-temperate conditions there. The Cretaceous is distinguished by a remarkable deployment of the immense land reptiles and very thick limestone deposits. Climatic conditions in the Tertiary are not sharply different from those of the Cretaceous. Middle and late Eocene floras show many tropical marks, Oligocene faunas are varied and large sized, especially the foraminifers (nummulites), the Miocene shows a distribution of warm-temperature plants in Spitzbergen and Grinnell Land, but the late Miocene was, at REVIEWS 517 least in many places, cooler. The Pliocene was rather warm but un- doubtedly became colder toward the beginning of the Pleistocene when glacial conditions reached full expression. The author concludes that the marked climatic variations of the past are primarily due to periodic changes in the topography of the land surface, modified by the variations in the amount of heat stored in the oceans, and the change in the composition of the atmosphere which conditions the storage of solar radiation. Supplementary notes with quotations from original descriptions of pre-Permian tillites and a bibliog- raphy of the subject are appended. ee = Cae Oceania. By P. Marswatyt. Handbuch der regionalen Geologie, 5. Heft, Band VII, Abteilung 2. Heidelberg, 1912. Pp. 36, figs. 10. Oceania, limited on the west by the Marianne, Pelew, and Caroline islands, on the east by the Sandwich Islands, is a region measuring about 8,400 miles east and west, and 4,200 miles north and south. Most of the islands are small, aggregated in rather well-defined groups or lines, and within the limits of each group the geological and physical structures are somewhat uniform. With the exception of the largest only, the islands are volcanic or composed of coralline limestone, and almost every island is fringed by coral reefs. The basin of the Pacific is of great, and nearly uniform, depth (2,500-3,000 fathoms), but in the west part of Oceania the ocean depths are far less regular. Very deep troughs are found subparallel to some of the island chains and their connecting submarine ridges, and the location of the shallows and basins is suggestive of important structural relations. The island chains seem to define at least four mountain ranges which are seen to converge toward North- ern New Zealand, a region therefore of great structural importance. The true border of the Pacific basin segment of the earth’s crust is marked by a fairly definite line, indicated by the submarine elevations, areas of raised coral rock, and the distribution of andesitic rocks. This line passes through the Kermadec, Tonga, Fiji, New Hebrides, and Solomon islands, and is noteworthy as the belt of present volcanic activity. Triassic fossils in New Zealand and New Caledonia indicate their coastal connection in the past, and the present faunal and floral distribution is strongly suggestive of the former existence of a continental area limited by the island line mentioned. Coral growths of the Pacific 518 REVIEWS generally imply considerable subsidence, though in places this has been superseded by elevation. A brief description of the physical character and geology of each of the island groups, so far as known, comprises the central part of the paper. A bibliography of the subject is appended. RCo Geology of the Gold Belt in the James River Basin, Virginia. By STEPHEN TABER. Virginia Geol. Surv., Bull. No. VII, 1913. Pp. 271, figs. 23, maps 2, pls. 8. The gold mines are localized mainly in Goochland and Fluvanna counties. Free gold occurs in quartz veins which cut pre-Cambrian quartzites, schists, and gneisses. The gold seems to be associated with granite intrusions, possibly of Cambrian age. The author suggests that this district illustrates the formation of quartz veins by the force of crystallization. The value of the gold produced in this region amounts to about $6,000.00 per annum. T Te Pre-Cambrian Algonkian Algal Flora. By CHARLES D. WALCOTT. Smithsonian Misc. Coll.; LXIV, No. 2, 1914. Pp. 153, pls. 19. Fossil algal flora, produced by blue-green algae, are found in the Algonkian formations of the Cordilleran region. Walcott describes and figures 8 new genera and 12 new species of algae from the Belt series. Before the discussion of the algal remains, there is a discussion of the continental conditions and sedimentation of Algonkian times. From Robson Peak, British Columbia, to Arizona and southern California, a> distance of over a thousand miles, there is a marked Algonkian-Cambrian unconformity. Preceding this advance of the Cambrian sea, the Algonkian was a time of continental elevation and of largely terrigenous sedimentation in non-marine bodies of water; also there was some sub- aerial deposition. Marine sediments accumulated along the shores of the continents, but they are now far buried, and everywhere lost to our knowledge. This unknown marine life, preceding the Cambrian invasion, belongs to what the author calls “Lipalian” time. Red sandstones and shales in the west suggest an arid and, possibly, a cold climate. The thick limestones in the western interior are explained as having been deposited from non-marine waters by algae. Ts Tate RAGE OBLIGATIONS —LuL1, R. S. Triassic Life of the Connecticut Valley. [Connecticut Geo- logical and Natural History Survey, Bulletin 24. Hartford, 1915.] —Martitanp, A. G. Annual Progress Report of the Geological Survey of Western Australia, for the Year 1913. [Perth, 1914.| Annual Progress Report of the Geological Survey of Western Aus- tralia for r914. [Western Australia Geological Survey. Perth, 1915.] MAKINEN, Erro. Die Granitpegmatite von Tammela in Finnland und ihre Minerale. [Bulletin No. 35 de la Commission Géologique de Finlande. Helsingfors, January, 1913.| —MarsHatt, R. B. Profile Surveys in Bear River Basin, Idaho. [U.S. Geological Survey, Water-Supply Paper 350. Washington, 1914.| Profile Surveys in Willamette River Basin, Oregon. [U.S. Geo- logical Survey, Water-Supply Paper 349. (Prepared in co-operation with the State of Oregon.) Washington, 1914.] Results of Spirit Leveling in Idaho, 1896 to 1914, inclusive. [U-S. _ Geological Survey, Bulletin 567. Washington, 1915.] Results of Spirit Leveling in Minnesota, 1897 to 1914, inclusive. [U.S. Geological Survey, Bulletin 560. (Work done in co-operation with the Sate of Minnesota from 1909 to 1914, inclusive; Geo. A. Ralph, Chief Engineer of State Drainage Commission.) Washington, 1915.| —McLetsu, Joun. Annual Report on the Mineral Production of Canada during the Calendar Year 1913. [Canada Department of Mines, Mines Branch No. 320. Ottawa, 1914.] —Metnzer, O. E., anp Hare, R. F. Geology and Water Resources of Tularosa Basin, New Mexico. [U.S. Geological Survey, Water-Supply Paper 343. (Prepared in co-operation with the New Mexico Agricultural Experiment Station.) Washington, 1915.] —Merrii1, G. P. On the Monticellite-like Mineral in Meteorites, and on Oldhamite as a Meteoric Constituent. [Proceedings of the National Academy of Sciences, Vol. I, p. 302. Washington, 1915.] The Fisher, Polk County, Minnesota, Meteorite. [No. 2084. From the Proceedings of the U.S. National Museum, Vol. XLVIII, pp. 503-6. Washington: Government Printing Office, May 3, 1915.] —Michigan College of Mines, Year Book of the, 1914-1915. Announcement of Courses for 1915-1916. [Houghton, 1915.] —Mwpteton, J. The Production of Sand-Lime Brick in 1914. [From Mineral Resources of the United States, 1914, Part IJ. Washington, 1915.| 51¢ 520 RECENT PUBLICATIONS —MINERALCHEMIE, Handbuch der. Bd. II 8 (Bog. 21-30). [Dresden und Leipzig: Verlag von Theodor Steinkopff, 1915.] —Mines and Metallurgy, School of, University of Missouri. Catalogue, 1914- to1s. Bulletin, March, 1915. Vol. VII, No. 2. [Rolla, ro15.] Bulletin, June, 1915. Vol. VII. No. 3. ([Rolla, r915.] —Mining Congress Journal, The. Vol. I, Nos. 1 and 2. [Washington, February, 1915.] —Mississippi Geological Survey Commission, Third Biennial Report of, June 30, 1909—June 30, r911. [Jackson, ro1t.] —Missouri Bureau of Geology and Mines. Base Map of Missouri. Com- piled in co-operation with the U.S. Geological Survey. [Rolla, 1914.] —Perkins, G. H. Report of the State Geologist on the Mineral Industries and Geology of Vermont, 1913-1914. [Burlington, 1914.] —PocuE, J. E. The Turquoise, A Study of Its History, Mineralogy, Geology, Ethnology, Archaeology, Mythology, Folklore, and Technology. [Mem- oirs of the National Academy of Sciences. Vol. XII. Third Memoir. Washington, 1915.] : —REINECKE, L. Physiography of the Beaverdell Map-Area and the Southern Part of the Interior Plateaus of British Columbia. [Canada Department of Mines, Museum Bulletin No. 11, Geological Survey, Geological Series No. 23. Ottawa, ror5.] —Resources of Tennessee, The. Vol. V. [Nashville: Tennessee Geological Survey, I915.] —Rice, G. S. What a Miner Can Do to Prevent Explosions of Gas and Coal Dust. [U.S. Bureau of Mines, Miners’ Circular 21. Washington, IQI5.| —Royal Geographical Society, Year-Book and Record, tg914. [London: Kensington Gore, S.W., 1914.] —ScHRADER, F. C. Mineral Deposits of the Santa Rita and Patagonia Mountains, Arizona. With Contributions by James H. Hitz. [U-S. Geological Survey, Bulletin 582. Washington, rors.] —SEDERHOLM, J. J. Weitere Mitteilungen iiber Bruchspalten mit beson- derer Beziehung zur Geomorphologie von Fennoskandia. [Bulletin No. 37 de la Commission Géologique de Finlande. Helsingfors, June, 1913.] —Seismological Society of America, Bulletin of the. Vol. V, No. 1. [Stan- ford University, California, rot5.] | Animal Communities in Temperate America A STUDY IN ANIMAL ECOLOGY Designed to serve as a reference work and textbook in the New Science of Animal Ecology By VICTOR ERNEST SHELFORD Assistant Professor of Zodlogy in the University of Illinois Illustrated with 300 figures, maps, and diagrams HIS volume by Professor Shelford pre- ; T sents the principles of field ecology, _ illustrated by the more widely distrib- uted animal habitats of the eastern half of temperate North America, and the aquatic habitats of a much larger territory. The material has been accumulated during ten _ years of field study, from the point of view of _ modern ecology, in various parts of the United States, though most of the material is drawn from the Chicago Region. The book is de- signed to serve as a reference work and a pexthook ct in es new Science of field ecology. — - 380 pages, 8v0, cloth; price $3.00, postage exira “(weight 2 lbs. 12 02: y The rately, of EE Press Chicago - Seas - Illinois Do You Want to Know What Your Handwriting ~ Reveals? WE have published a fascinating booklet of 32 pages on the sub- ject of graphology, by William Leslie French. Every reader of this publi- cation will be interested in comparing the examples of handwriting in judg- ing character and disposition of the writer. A copy of this booklet and 12 differ- _ ent patterns of The Spencerian Steel Pens will be sent postpaid on receipt of ten cents. If you can do good work with a poor pen _ you can do better work with a good pen. Silver plated Falcon pens, smooth points, wear long, will not corrode. SPENCERIAN PEN CO. 349 Broadway - - - New York City TEACHERS ne dae ee OF BIOLOGY AND GEOLOGY Will find the Bulletins of the Geo- graphic Society of Chicago of special value for field trips, lectures, etc. Write for descriptive circulars. . The University of Chicago Press Chicago Illinois WE WANT TO BUY Journal of Geology, Vols. I to XXI. Quote shorter runs also, and other geological periodicals. INTERNATIONAL MAGAZINE CO. Elizabeth, N. J. THE ASTROPHYSICAL JOURNAL Issued monthly except February and August. $5.00a year; single copies, 65 cents. Foreign postage, 62 cents. Whe Unibersity of Chicago }oress CHICAGO, ILLINOIS 966 Duss Zip 1 {FRE DUPLICATE COPIESARENECDED Daus Improved Tip Top Duplicator is an ideal assistant, always ready to reproduce clean, clear, perfect copies of letters, manuscripts, price lists, bills, i in- voices, reports, drawings, menus, ete., etc. Don’t take our word, try it yourself for Ten days” trial without deposit If not satisfactory, return and no questions asked. Makes 150 handwritten and 50 typewritten copies in a few min- utes. Negatives require no washing out after using, and negative rolls can be used from 15 to 20 times. Complete outfit contains a continuous roll of Daus Non-Peeling Anti- Stick Negative Roll. Capsize,8$x13inches. Price $7.50 _ FELIX K. DAUS DUPLICATOR CO. Daus Building, 111 John Street, NEW YORK and ee FINE INKS 482 ADHESIVES For those who KNOW A New Proposition A TYPEWRITER BY PARCEL POST ON TEN DAYS’ FREE EXAMINATION It is the Remington Junior the very latest addition to the great Remington family of typewriters. A new and smaller Remington—built for the needs of the professional man and the home Me SD Drawing Inks Eternal Writing Ink Engrossing Ink s ° 9 Taurine Mucilage Hi GQims? | Photo Mounter baste 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 ill-smelling inks and adhesives and adopt the Higgins Inks and Adhesives. They will bea revelation to you, they are so sweet, clean, well put up, and withal so efficient. At Dealers Generally Mail this coupon today Remington Typewriter Company (Incorporated) 327 Broadway, New York Send me a Remington Junior Typewriter, price $50, on free examination. It is understood that I may return the machine, if I choose, within ten days. If I decide to purchase it, I agree to pay for it in ro monthly payments of $5 each. CHAS. M. HIGGINS & CO., Mfrs. Branches: Chicago, London 271 Ninth Street Brooklyn, N.Y. The whole: school will welcome the Balopticon The school authorities, the parents, the teachers, and the pupils will all proht by the ee. of this notably superior lantern—the Bausch [omb BALOPTICON THE PERFECT STEREOPTICON The images projected by the Balopticon are wonderfully clear and vivid over the entire field; the lantern itself is simple to operate, and built for lang and steady use. Its newly developed, gas-filled Mazda lamp gives more brilliant illumination than the alternating-current arc, with much less current expense, and is noiseless and automatic. Models for lantern slides and opaque objects (maps, photos, pictures, specimens, etc.). Also combined models for both forms, with instant interchange. Model C (illustrated) for slides, price from $30.00 up. Other models from $20.00 up. Write for illustrated price list and booklets containing suggestions for the use of the Balopticon Bausch £3 jomb Optical G. 554 ST. PAUL STREET ROCHESTER, N.Y. NEW YORK WASHINGTON CHICAGO SAN FRANCISCO Leading American Makers of Photographic Lenses, Microscopes, Projection Lanterns (Balopticons), Stereo Prism Binoculars and other High-Grade Optical Products arenas hein Ee ao ee ) JOURNAL or GEOLOGY A -SEMI-QUARTERLY ae a gai EDITED BY o « ‘ RGN THOMAS C, CHAMBERLIN AND ROLLIN D. SALISBURY With the Active Collaboration of SAMUEL W. WILLISTON, Vertebrate Paleontology ALBERT JOHANNSEN, Petrology 2 : STUART WELLER, Invertebrate Paleontology ROLLIN T. CHAMBERLIN, Dynamic Geology _ ALBERT D. BROKAW, Economic Geology ASSOCIATE EDITORS SIR. ARCHIBALD GEIKIE, Great Britain JOSEPH P.IDDINGS, Washington, D.C. - CHARLES BARROIS, France JOHN C. BRANNER, Leland Stanford Junior University ALBRECHT PENCK, Germany RICHARD A. F. PENROSE, Jr., Philadelphia, Pa. HANS REUSCH, Norway : WILLIAM B. CLARK, Johns Hopkins University GERARD “DEGEER, Sweden ; 4 WILLIAM H. HOBBS, University of Michigan T. W. EDGEWORTH DAVID, Australia FRANK D. ADAMS, McGill University BAILEY WILLIS, Leland Stanford Junior University CHARLES K, LEITH, University of Wisconsin ‘GROVE K. GILBERT, Washington, D.C. WALLACE W. ATWOOD, Harvard University CHARLES D. WALCOTT, Smithsonian Institution WILLIAM H, EMMONS, University of Minnesota HENRY’ ‘Ss. WILLIAMS, Cornell University ARTHUR L. DAY, Carnegie Institution SEPTEMBER-OCTOBER 1916 THE PRE-WISCONSIN DRIFT OF NORTH DAKOTA shies. Wise ix Ce cel ONIN etc 2 EVOLUTION OF THE BASAL PLATES IN Bie A ates CRINOIDEA CAMERATA, II é HErrRIcK-E. WILSON = 533 BIRRERENTIATION IN INTERCRUSTAL MAGMA BASINS - - ALFRED HaRKER 554 STRATIGRAPHY OF THE SKYKOMISH BASIN, WASHINGTON "Warren S.. SMITH. 550 “PUFF” CONES ON MOUNT USU ele ee Mella ees, se! adam uate a a aS OINOURIE Seg ORIGIN OF FOLIATION IN THE PRE-CAMBRIAN ROCKS OF NORTHERN NEW SAGAS - - - - - - - - - - - WILLIAM J. MILLER 587 THE COMPOSITION OF THE AVERAGE IGNEOUS ROCK -~ - ApotpH KNopr 620 BORER Reese a Goria = ee cSt tes nO THE UNIVERSITY OF CHICAGO PRESS CHICAGO, ILLINOIS, U.S.A. AGENTS THE CAMBRIDGE UNIVERSITY PRESS, Lonpon AND EDINBURGH : KARL W. HIERSEMANN, LErpzic haan THE MARUZEN-KABUSHIKI-KAISHA, Toxyo, Osaka, Kyoto, Fukuoka, SENDAI tims THE MISSION BOOK COMPANY, SHANGHAI THE JOURNAL OF GEOLOGY EDITED BY veal THOMAS C. CHAMBERLIN AND ROLLIN D. SALISBURY With the Active Collaboration of SAMUEL W. WILLISTON ALBERT JOHANNSEN Vertebrate Paleontology Petrology STUART WELLER ROLLIN T. CHAMBERLIN Invertebrate Paleontology Dynamic Geology ALBERT D. BROKAW Economic Geology e_——=— —— as The Journal of Geology is published by the University of Chicago at the University Press semi- | quarterly, on or about the following dates: February 1, March 15, May I, 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 Cana] — Zone, Republic of Panama, Hawaiian Islands, Philippine Islands, Guam, Samoan Islands, Shanghai, {] Postage is charged extra as follows: 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 single copies, II cents (total 76 cents). [Patrons are requested to make all remittances payable to the University of Chicago Press in postal or express money orders or bank drafts. F The following agents have been appointed and are authorized to quote the prices indicated: _ For the British Empire: The Cambridge University Press, Fetter Lane, London, E.C. Yearly subscriptions, including postage, 18s. 8¢. each; single copies, including postage, 3s. 2¢. each. _ For the continent of Europe: Karl W. Hiersemann, Kénigstrasse 29, Leipzig, Germany. Yearly subscriptions, including postage, M. 22.65 each; single copies, including postage, M. 3.80 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, including postage, Yen 1,50 each. { : . For the Chinese Empire: The Mission Book Company, 18 Peking Road, Shanghai, China. Yearly subscriptions, $4.00; single copies, 65 cents, or their equivalents in Chinese money. Postage extra, if mailed direct outside of Shanghai, on yearly subscriptions 53 cents, on single copies II cents. Claims for missing numbers should be made within the month following the regular month of publication. The publishers expect to supply missing numbers free only when they have been lost in transit. Business correspondence should be addressed to The University of Chicago Press, Chicago, Ill. Communications for the editors and manuscripts should be addressed to the Editors of /THE JOURNAL OF GEOLOGY, the University of Chicago, Chicago, Ill. Contributors are requested to write scientific and proper names with particular care and 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 leading articles will be printed as ordered, of which 50 (without covers) will be furnished grafts, the actual cost of the remainder — (and covers, if desired) to be paid for by the author. The actual cost may vary, and will depend upon the amount of work in re-making the pages into forms, presswork, paper, binding, etc. Separates containing half-tones may be expected to cost somewhat more, the increase depending upon the number of cuts and the amount of work required upon them. Entered as second-class matter, March 20, 1893. at the Post-office at Chicago, Ill., under Ya Act of March 3, 1879. VOLUME XXIV NUMBER 6 THE WOURNAE OF GEOLOGY SALT EMBERAOCROBETR: 1070 THE PRE-WISCONSIN DRIFT OF NORTH DAKOTA _ A. G. LEONARD North Dakota Geological Survey Considerably over one-half of North Dakota is covered by Wisconsin drift, which is found east of the Missouri plateau and also occupies a belt of country along the eastern margin of the plateau. The western border of the Wisconsin drift is marked by the wide, massive Altamont Moraine which crosses the state diagonally from northwest to southeast and has a width in places of 20 miles. West of this moraine there is an older drift sheet which extends from 70 to 130 miles beyond the Wisconsin drift and in North Dakota covers an area at the surface of approxi- mately 19,000 square miles. The border of this older drift crosses the Montana boundary about 30 miles north of the Northern Pacific Railroad. This older drift undoubtedly underlies the Wis- consin drift of eastern North Dakota, but with possibly one excep- tion it has not been observed in outcrops or recognized in wells. East of the Missouri River the pre-Wisconsin drift, while covered in many places by outwash silt from the Altamont Moraine, is present in Emmons, Burleigh, eastern McLean and Mountrail, and perhaps in Williams and Divide counties. West of the Mis- sour! River the older drift covers most of Morton, Dunn, and McKenzie counties, a corner of Stark, and all of Mercer and Oliver Vol. XXIV, No. 6 521 A. G. LEONARD U1 to to counties. In this area west of the river outcrops are not uncommon, and thus the region is the most favorable in the state for the study of the older drift sheet. The following account of some of the features of the pre- Wisconsin drift is based on observations made during the course of tield work for the North Dakota Geological Survey, a portion of the time in co-operation with the United States Geological Survey, in Morton, Dunn, McKenzie, Burleigh, and other counties during the years 1909 to rgt4 inclusive. The older drift west of the Missouri is in most places thin and has undergone great erosion. The deposit perhaps never had any considerable thickness in this region except locally, where it forms moraines, and much of the glacial material which was formerly present has been swept away by streams. The drift throughout much of the area is thus represented by bowlders and gravel, the coarser materials left behind when the finer débris, such as clay and sand, was carried off. There are extensive tracts where little or no glacial material is present and where only an occasional bowlder or a patch of gravel indicates that the ice sheet once covered this region. The western margin of the older drift is therefore poorly defined, and the mapping of it is based largely on the distribution of glacial bowlders and gravel. As would be expected from the foregoing characters, the pre- Wisconsin drift has not, except in certain restricted areas, affected the topography to any large extent. The region is one of many streams and mature drainage, in striking contrast with the area of the Wisconsin drift, with its few streams, numerous lakes, and youthful topography. PRE-WISCONSIN DRIFT OF BURLEIGH COUNTY East of the Missouri River in Burleigh County, between the river and the Altamont Moraine, the older drift is present, but the occasional outcrops appear to indicate that it forms only a thin veneer over the underlying rocks, seldom exceeding 8 or 1o feet in thickness. It outcrops in the bluffs of the Missouri River 3 miles northwest of Bismarck, where ro feet of till is found, and in the cut at the east end of the Northern Pacific bridge it attains a thickness THE PRE-WISCONSIN DRIFT OF NORTH DAKOTA 523 of 15 to 20 feet. The bowlder clay appears in a number of the cuts along the Minneapolis, St. Paul & Sault Ste. Marie Railroad north of Bismarck, here generally associated with water-laid drift. It also outcrops at several points on Apple Creek, where it shows a thickness of 10-12 feet. DRIFT IN MORTON COUNTY West of the Missouri River in Morton County the glacial drift is represented almost wholly by gravel and bowlders, the latter occurring in great numbers. These bowlders, which are mostly granite, thickly cover the surface in many localities, resting directly on the bedrock, and except in rare instances no drift clay is associated with them. In some places they are scattered loosely over the ground, but in many others they form a bed or pavement in which the individual bowlders are in contact with each other. These bowlder deposits or bowlder beds are especially noticeable on the tops of divides and on upland areas. The bowl- ders vary in size from 6 or 8 inches to several feet in diameter, large ones measuring 8 and to feet being seen occasionally. MORAINES OF LITTLE HEART RIVER BASIN The most interesting and notable occurrence of glacial till in Morton County is found in the basin of the Little Heart River where the drift has been heaped up into morainic hills. During the Glacial Period this basin was _ probably occupied by an ice lobe of the continental glacier, and this lobe formed the belt of Fic. 1.—Bowlder-covered moraine hill of the pre-Wisconsin drift, Little Heart Basin, northeastern Morton County. morainic hills which nearly encircles the broad valley plain and deposited more or less drift in the pre- glacial valleys of the Little Heart and its tributaries. As the ice 524 A. G. LEONARD melted, the waters flowing from it deposited much outwash silt in the form of valley trains sloping away from the moraines. A number of the morainic hills have been partially buried by the outwash silt and rise like islands from the ; level plain of the valley ; train. The morainic i belt of bowlder-covered ; hills and ridges is found j near the base of the slopes on either side of the two valleys tribu- 4 tary to that of the Little 3 Heart River (Fig. 1). Moraines cross the Fic. 2.—Pre-Wisconsin till overlying Fort Union beds on Tobacco Garden Creek several miles : above its mouth, McKenzie County. valley in three places and one belt of ridges and hills continues unbroken along the south side of the valley of the South Branch for a distance of 12 miles. The cultivated fields extend up to the moraine and end there where the soil becomes too rocky and the slopes too steep for cultivation. DRIFT IN MC KENZIE COUNTY Near the western boundary of North Dakota the pre- Wisconsin ice sheet reached 35-40 miles Fic. 3.—Pre-Wisconsin till resting on stratified sand and gravel on Tobacco Garden Creek, near its south of the Missouri River and thus covered the greater part of McKenzie County which lies between the Yellowstone, Missouri, and Little Missouri rivers. The older drift left by this ice sheet is well shown in many places in this mouth. THE PRE-WISCONSIN DRIFT OF NORTH DAKOTA 525 region and the glacier also caused important drainage changes. The best exposures are found on Sand Creek, a short tributary of the Missouri River several miles east of Tobacco Ganden Creek, on Tobacco Garden Creek, and on Clear Creek, a tributary of the latter. Near the mouth of Tobacco Garden Creek appears 58 feet of till ineraecut, banks Tt) 1s Fic. 4.—Pre-Wisconsin drift resting on Fort yellowish gray or drab Union sandstone, valley of Clear Creek, a tributary of Tobacco Garden Creek, northeastern McKenzie County. in color, contains many small bowlders, and rests on Fort Union beds (Fig. 2). Not far from here is seen 18 feet of bowlder clay over- lying 15 feet of well- stratified sand and gravel (Fig. 3). The valley of Clear Creek was partly filled with drift and there are many good outcrops of bowlder clay in the fre- quent cut banks along the stream where from 30 to 40 feet and over of till is exposed (Fig. 4). The greatest thickness Fic. 5.—Pre-Wisconsin drift exposed on the foynd along this creek tributary of Sand Creek, northeastern McKenzie County. was in Sec. 36, T.152N.., R. 97 W., where in a high bluff too feet of dark gray till overlies 100 feet of soft Fort Union sandstone. This outcrop lies within the morainic area to 526 A. G. LEONARD be described later, which probably accounts for the exceptional thickness of the pre-Wisconsin drift at this point. Another excellent drift section occurs in this same morainic area 3 miles west and one-half mile south of Charlson, or 5 miles south of the Missouri River on a tributary of Sand Creek (Fig. 5). Here the following section appears in the steep bluff which rises abruptly from creek-level: Bt: In. Sandy claysand ‘Soultewe ereemeee eet con cra i wk coe i eee 2 Sand and. clay incaltermating bands®< ©... W046. 4.5. :@ 3 sn ose ee 8 Gravel... 3/055 oo eee ert te eters rere | f al nae 8 Sand and. clay invaltermatmyp layers oS. sce ac ee 3 Gravel CORPSES. le) ee aes cater es Rte ect ees er ios mes RTE 8 Sando!" 295 RUA ere Rar BC pee ce pet tS gL ee 2 4 Gravel Saeco 3 gp ee ee Ree ite ok eee 7 8 Till or bowlder clay, dark gray in color, and containing large numbers of pebbles and bowlders imbedded in the tough, hard clay. A large proportion of the bowlders and pebbles are composed of compact limestone and the till contains considerable blue shale (Pierre ?). Thickness of till exposed above the creek......... 61 SS ae The sand and gravel forming the stratified drift of the foregoing section are light yellow in color. The drift hills just back of the bluff rise 50-60 feet above the top of section, so that the total thickness of the drift above creek-level is from 140 to 150 feet. BOWLDER BED ON THE MISSOURI RIVER An interesting bowlder deposit belonging to the pre-Wisconsin drift is found along the Missouri River just above water-level, 2 miles below the Nesson Ferry and one-half mile below the mouth of Tobacco Garden Creek (Fig. 6). This bed of bowlders shows a thickness, above the normal stage of the river, of 12-14 feet and its depth below water-level was not determined. The deposit is well exposed along the water’s edge for a distance of nearly 100 yards, and scattered bowlders and ferruginous gravel occur at intervals for another 200 yards. Overlying the bowlders is 15 feet of gravel, which is overlain in turn by silt and fine sand extend- ing to the top of the terrace, 100 feet above the river. The bowlder THE PRE-WISCONSIN DRIFT OF NORTH DAKOTA 527 bed is composed of bowlders of all sizes, those from 2 to 3 feet in diameter being quite common, though those less than 1 foot in diameter are most abundant. One sandstone bowlder measured 14 feet in length, another was g feet long, and another 4 feet. Granite, limestone, petrified wood, and sandstone bowlders are found, together with other kinds of rock. The interstices between the coarser materials are filled with gravel and sand, and the whole deposit is cemented into a rather firm, indurated mass. It is very ferruginous and brown from the limonite forming the cement- ing material, and in many places the bowlders are firmly held by the iron cement and sand, which serve as a matrix in which the bowlders are imbedded; when the latter weather out their shape is pre- served in the matrix. While some of the bowlders of this de- posit may have been brought here by float- ing ice, it is probable that most of the deposit was left here by the pre-Wisconsin ice sheet when it advanced south near mouth of Tobacco Garden Creek, McKenzie of the river. The finer County. materials of the drift, if they were ever present, have been carried away, leaving the gravel and bowlders, which were subsequently cemented by the iron of the surface waters. Fic. 6.—Bowlder bed on the Missouri River MORAINE OF THE PRE-WISCONSIN DRIFT Mention has been made on a previous page of the moraines of the Little Heart River Basin, and in McKenzie County a much more extensive moraine of the pre-Wisconsin drift forms a promi- nent topographic feature of the region. It lies in the northeastern part of the county, east of Tobacco Garden Creek, where it extends on the upland from the edge of the Missouri Valley bluffs north 528 A. G. LEONARD of Charlson in a southwesterly direction to the valley of Timber Prong Creek in Secs. 14 and 15, T. 151 N., R. 97 W. It has a length of 16 miles, an average width of 2 miles, and its area is about 30 square miles. This moraine shows well from Charlson, where its ridges and hills are seen rising from too to 150 feet above the flat plain in the foreground (Fig. 7). Within the morainic area the surface is rough and hilly and the ground is thickly strewn with bowlders. The topography is typically morainic, with numerous irregular hills and ridges, while scattered among the hills are great numbers of hollows or kettle-holes, some containing water and others dry. Many of the hills rise 50-125 feet above the bottom of the kettle-holes. Where this moraine crosses the valley of Timber Prong Creek it forms a dam, which holds back the waters of the upper valley and forms a lake known locally as Dimick Lake. This moraine lake is very irregular in shape and has an area of between two and three square miles. Fic. 7—Small lake in the moraine of the pre-Wisconsin drift near Charlson, northeastern McKenzie County. North of the Missouri River another morainic belt occurs which may be part of the pre-Wisconsin moraine just described as extending southwest from the vicinity of Charlson, though it is perhaps more likely to have been formed by the Earlier Wiscon- sin ice sheet referred to on another page. It lies from 2-3 miles west of White Earth Creek and with a width of 1-2 miles extends north and south a distance of at least 6 or 8 miles between the Missouri River and the Great Northern Railroad. Fifteen or 20 miles or more west of the Altamont Moraine the railroad crosses a well-developed moraine extending 4 or 5 miles and perhaps more north and south of Temple, and has a width of several miles. This hilly belt lies about 9 miles west of the above-mentioned moraine near White Earth Creek. THE PRE-WISCONSIN DRIFT OF NORTH DAKOTA 529 CHARACTERISTICS OF THE PRE-WISCONSIN DRIFT The topography of the older drift and the large amount of ero- sion it has suffered compared with the Wisconsin till have been mentioned on a previous page, where it was shown that in many places only the coarser materials of the drift—the pebbles and bowl- ders—remain as evidence that the ice sheet once covered the region. But the color of the pre-Wisconsin till is generally unlike that of the Wisconsin drift. The latter is commonly light yellow to light gray in color as exposed in railroad cuts or along stream valleys, but where the deeper till appears in fresh excavations it is seen to have the blue color of the unoxidized clay. The pre-Wisconsin till, on the other hand, where best exposed on Sand and Clear creeks in northeastern McKenzie County (Fig. 8), is dark gray in color throughout the maximum observed Fic. 8.—The valley of Clear Creek where it cuts through the moraine of the pre-Wisconsin drift, northeastern McKenzie County. = Eee _ Sannin thickness of over 100 feet. Except in the morainic areas the thickness of the pre-Wisconsin drift is not great. West of the Missouri River it is seldom as much as 8 or 10 feet and generally the thickness is not over 2 or 3 feet or less. The thinness of the older drift is due partly to erosion which has removed much of the glacial material and over a large part of the area left only bowlders or a thin veneer of gravel, but partly perhaps also to the fact that the drift may never have been very thick in this region. BOUNDARY OF THE PRE-WISCONSIN DRIFT Chamberlin and Salisbury many years ago noted the presence of an older drift beyond the Altamont Moraine and its approximate boundary was shown on their map.t. The more detailed work of «Terminal Moraine of the Second Glacial Epoch,” Third Ann. Report U.S. Geol. Surv., pp. 291-402; also Plate XXXV. 530 A. G. LEONARD recent years has shifted the margin somewhat farther west and south and it is now provisionally located as shown on the accom- panying map (Fig. 9). Glacial drift is found 50 miles south of the mouth of the Yellow- stone River, or within less than 15 miles of Glendive, Montana. For a distance of 50 miles east of the western boundary of North Dakota the drift margin extends approximately east and west and lies 30-40 miles south of the Missouri River. North of the Kill- deer Mountains the boundary swings sharply to the south, and crossing the Knife River near the western edge of Dunn County it takes a general southeasterly course across the state. The margin of the drift is believed to cross the Northern Pacific Railroad and the Heart River between 2 and 3 miles west of Gladstone. That a lobe of the ice sheet crossed the Heart River at Gladstone is shown by the presence of thick deposits of drift gravels on the upland 1-2 miles south of the Heart and at an elevation of between too and 200 feet above river-level. In places the gravel and sand have a thickness of at least 90 feet, and the deposit contains a number of good-sized granite bowlders. A well-defined gravel ridge marks the edge of the drift for 3 or 4 miles in this area south of the Heart River at Gladstone. This ridge rises 30-40 feet above the surface on either side and falls away rather abruptly on the south, while on the north the slope is more gradual. There is no evidence that the ice sheet extended more than 2 or 3 miles south of the railroad between Gladstone and Richard- ton, but in this vicinity glacial gravel and a few small bowlders occur that far south. Between the Cannon Ball and Heart rivers glacial bowlders are found as far west as Elgin, or within 12 miles of the western border of Morton County. In general the drift margin between the Killdeer Mountains and the South Dakota line lies from 40-60 miles west of the Missouri River. Thickness of the ice sheet —In the vicinity of Berg in northeastern McKenzie County there are twelve or fifteen high buttes, known as the Blue Buttes, which are irregularly distributed over an area of 15 or more square miles. Many glacial bowlders occur on top of these buttes at an elevation of over 2,700 feet above sea-level, or 1,000 feet above the Missouri River only 6 miles to the east. Son THE PRE-WISCONSIN DRIFT OF NORTH DAKOTA Since the buttes rise nearly 500 feet above the surrounding upland the ice sheet probably had at least this thickness in order to over- Another ride them and deposit bowlders on their summits. vIOYLG YON Ur ourvesopy JUoMIeI]Y pur iylIp uIsuoosTAA-oAd Jo uIsIeUI surmoys deyy—6 “ong NDILD {Anoss! esc ee Teaopeo + Id $ |= QUIOAO'!W tdlug ulsuodsi\r- Satw gO WW JUOWDLY 40 Apes nog \ ihislyadanik \ 532 A. G. LEONARD explanation for the presence of the bowlders on top of the high buttes is that the ice upon encountering these obstructions was buckled up as it passed over them. But it seems more probable that the ice sheet which was able to push across the deep valley of the Missouri River and advance 40-60 miles beyond was thick enough to submerge the Blue Buttes and pass on over them. The terminus of the continental glacier was not far from 15 miles south of the Blue Buttes, since the ice advanced only as far as the Kill- deer Mountains. Back about 15 miles from the edge the ice sheet therefore doubtless had a thickness over the upland plain of con- siderably over 500 feet. The ice which filled the Missouri River Valley must have had locally a thickness of 1,000 feet or over. Age of the older drift—That the drift west of the Missouri River is much older than the Wisconsin drift is evident from the great amount of erosion it has undergone. Over much of the region the finer materials of the till have been swept away, leaving only bowlders and gravel. The older drift also differs in color, being considerably darker than the Wisconsin. ‘This pre-Wisconsin drift has commonly been regarded as Kansan and there is perhaps more reason for referring -it to the invasion of the Kansan ice sheet than any of the other early ice invasions. The drift north of the Missouri River and west of the Altamont Moraine appears to be younger than the drift south and west of the river. It has undergone less erosion and resembles the typical Wisconsin till in color. Calhoun believes that the drift of north- eastern Montana is of Wisconsin age, and on his map this drift is shown as extending to within about 40 miles of the North Dakota line.’ The extra-morainal drift north of the Missouri River in Williams County is undoubtedly continuous with the drift sheet west of here in Montana, in which case it is probably of Wisconsin age, and belongs to the Early Wisconsin stage. li this view is correct, there are three drift sheets in North Dakota—the Late Wisconsin east of the Altamont moraine, the Early Wisconsin west of the moraine and north of the Missouri River, and the Kansan drift west and south of the river. * Fred. H. H. Calhoun, “‘The Montana Lobe of the Keewatin Ice Sheet,” Prof. Paper No. 50, U.S. Geol. Survey, 1906, pp. 52-57. EVOLUTION OF THE BASAL PLATES IN MONOCYCLIC CRINOIDEA CAMERATA. II HERRICK E. WILSON United: States National Museum, Washington, D.C. PAK 2. CHANGES PRIMARILY MODIFYING THE RELATIONS OF PLATE CONTACT AND POSITION The second series of changes, those which modify the primary position and relation of the basals and radials, will now be con- sidered. These are: (a) reduction and compensating growth; (0) enlargement and compensating reduction; (c) plate division; (d) plate migration; (e) plate interpolation; and (f) anchylosis. a) Reduction and compensating growth—Reduction, or the diminution in size of a plate, may be either a function of the absorp- tive ameboid cells (see p. 502), or due to inhibited growth (atrophy). That is, the absorption of a fully outlined plate may take place, as in the absorption of the anal and oral plates in Aniedon, or a continuous diminution in development to a former standard of size may result in the atrophy and final disappearance of a plate, as in the great reduction of the basals in Pisocrinus quinquelobus* and the disappearance of the first costal in some specimens of Eucalyptocrinus rosaceus? and Alloprosallocrinus conicus. Atrophy in plate growth may be due either to plate contact, which inhibits the free branching and anastomosing type of development, or to some deep-seated morphological change. The simplest form of inhibition in plate development is that shown in the normal growth of plates after coming into mutual contact. It is the process which = IRGC, Bo 0s Lye 2 Ref. 28, p. go, Pl. XI, Figs. 6, 7. 3In Alloprosallocrinus conicus the writer has found that the apparent anchylosis of the costal plates (see ref. 39, p. 407) is due to the complete reduction of the first costal. 533 534 HERRICK E. WILSON gives to the plates their polygonal outline and must not be consid- ered a form of atrophy. Plate contact does, however, produce atrophy when the accelerated growth of one plate causes the reduction in size of some adjacent plate. This type of inhibition may be termed superficial atrophy, and is apparently the type just illustrated in the absorption of the first costal in Alloprosallo- crinus. Atrophy of the other type is apparently the result of marked internal changes which appear on the exterior in the reduc- tion of skeletal parts. This form of inhibition may be termed deep-seated atrophy, and is the type illustrated in the drawing together of the posterior radials in Pferotocrinus upon the reduction of the anal plate (PI. III, No. 11.) SARE yeaa) ce: Fig. 6.—Figures showing the reduction of the first costal in AlJloprosallocrinus conicus: 1 and 2, from specimen No. 9350; 2 and 3, from specimen No. 9357, in the University of Chicago collection. With the decrease in diameter of a plate in a closed cycle there must be (1) a compensating increase in the diameter of some plate or plates in the same cycle, or (2) a decrease in diameter of some plate or plates of the apposed cycle; otherwise the symmetry of the cup will be distorted. The first principle is clearly demon- strated by the increase in diameter of the first interbrachial plates in Amphoracrinus' upon the gradual reduction of the proximal portion of the second anal plate. The second principle is clearly demonstrated in the reduction of the apposed compound basal and radial of Zophocrinus. This form of change might be confused with vertical plate- splitting followed by anchylosis of the parts to the adjacent plates if the change were a sudden mutation and no knowledge of the ontogenetic development obtainable; otherwise the phylogenetic succession would show the factors involved. If the reduction of a TARet 32 ps Loy 2URet Oneal sit. BASAL PLATES IN CRINOIDEA CAMERATA 535 plate in a closed cycle is asymmetrical, the growth of one of the laterally adjacent plates will be greater than that of the other, and will, on completion of the reduction, occupy the entire area of the missing plate. This process is shown nearing completion in the reduction of the antero-lateral radials in Catzllocrinus,' and the consequent enlargement of the left posterior and anterior radials, and in Mycocrinus? in the reduction of the left anterior and dextro- lateral radials. In Pzsocrinus the principle is diagrammatically shown in both basal and radial cycles, where, by the reduction of two basals and three radials, three basals, two radials, and the radianal are greatly enlarged (Fig. 7, Nos. 1-4). = © Ww Fig. 7.—Diagrams illustrating reduction and compensating growth in Pisocrinus: 1, hypothetical, ancestral stage; 2-4, based upon specimens in the University of Chi- cago collection; x=position of first anal plate. b) Enlargement and compensating reduction.—Plate enlarge- ment, as we have seen, is due to the activity of the ameboid calcif- erous cells. In normal, symmetrical development the growth of the young plates is for a short time more rapid than that of the body wall, but upon plate contact the increased enlargement of both plates and body wall is theoretically balanced. If, however, a plate increases more rapidly than the adjacent plates, and is not controlled by the inhibiting influence of symmetrical development, and its accelerated growth is not compensated for by growth of the body wall, this growth must be compensated for in some other manner. Accelerated lateral increase of this type in a cycle plate demands then either (1) the decrease in diameter of some adjacent plate or plates in the same cycle or (2) the increase in diameter of some adjacent plate or plates in the apposed cycle and (3) distor- tion of the horizontal outline of the calyx or various combinations of the first two of these secondary developments may occur. t Ref. 6, p. 149, Fig. LXII. 2) Ret. 28) p. 110; Pl. 7, Bigs 4: 536 HERRICK E. WILSON Accelerated increase of an interpolated plate between cycles or plate groups demands: (1) the decrease in size of some adjacent plate or plates, as in the decrease in size of the dextro-lateral radials in Pisocrinus (Fig. 7), upon enlargement of the radianal, or (2) distortion of the cup. ) c) Plate division——This process is the splitting of a plate into two (perhaps more) parts, either during or after the formation of the primary, formative cell group. Division of a cell group is due to cell separation, and may or may not be accompanied by cell division. Division of the plates when they are once formed is due to the action of the absorptive, ameboid cells. Division differs essentially from intercalation, in that a fundamentally distinct cell group is demanded for the interpolated plate. But no matter how division may take place, apparent evidences of division in the plates of fossil crinoids must be very carefully investigated before too much significance is attached to the opinion that division and not interpolation has taken place. Division due to absorption is only known to occur during the absorption of the supporting rods in echinoid larvae,’ and in the reduction of the radianal in Antedon.? Division or duplication is assumed by Bather? in the formation of the paired, proximal inter- brachials in -Actinocrinidae, yet all evidence from the work on crinoid larvae shows duplication and not division to be the process involved. Horizontal bisection is assumed by Bather* in ten genera of monocyclic Inadunata, not including those in which bisection of the right-posterior radial only occurs. However, when it is noted how closely the development and migration of the radianal in the larvae of modern crinoids parallels the development and migration of the radianal in the Flexibilia,’ there is good reason to believe that bisection of the right-posterior radial has not occurred, but that the radianal and subradials are primary or interpolated (secondary) plates. Vertical splitting seemed beyond question in Ref. 34, Pp. 340-54- 2 Ref. 27, pp. 52, 53, Pl. 5, Fig. 11. 3 Ref. 3, p. 34, fifth notice. 4 Ref. 6, pp. 112, 144. 5 Ref. 16, p. 332, 333: BASAL PLATES IN CRINOIDEA CAMERATA 537 the formation of the compound, left-posterior radial in Anomalo- crinus,’ yet Springer? has shown it to be an abnormality, due per- haps to plate fracture There are, it is true, certain conditions surrounding apparent cases of plate division which lead us to believe that division and not interpolation has occurred. If, for example, two plates of the same cycle occupy the approximate area of one plate in that cycle, if they mutually fulfil the require- ments of but one plate in that cycle, and if in the obliteration of the intervening suture a plate would be formed indistinguishable from the other four undivided plates in that cycle or from the morphologically undivided equivalent of that plate in a closely related ancestral genus, division would seem the only logical con- clusion. But this conclusion is by no means proved. In the light of phylogenetic and ontogenetic development as ascertained from fossil crinoids division is very uncertain, for we cannot see the process taking place. Furthermore, if division had occurred anchylosis of the parts to the adjacent plates might take place and either complete absorption with compensating enlargement or migration would have to be called upon to explain the appearance of the new suture. Thus, no matter how carefully we attempt to ascertain the fact that division has occurred, the factor of inter- polation will usually appear as an alternative. Only through careful observation, in modern larval development, of plates not destined to obliteration in the adult stage can this process be satisfactorily determined. d) Plate migration—Any shifting which brings a plate, as a unit, into a new relation of contact and position with plates of the adjacent cycles, or of adjacent plates in the same series, may be termed a migration. For simplification in the discussion, the dif- ferent types of migration may be broadly separated into two divisions: simple migrations, or those unattended by movements of the sarcode; and complex migrations, or those dependent upon movements of the sarcode. Of the simpler forms, three types occur: portional migration, cell-group migration, and simple plate migration. 2 IRGE, Bip lewd IUD, jo, Qs TBI, MO), Os WEB 2 IRI, BA. De Divs. 3 The italics are the writer’s. 538 HERRICK E. WILSON If the absorption or atrophy of one side of a plate and the growth on the opposite side are approximately equal, the plate would appear to be shifting as a whole, although actually stationary in part. This type of migration may be called portional migration, and is the type illustrated in the shifting forward of the postero- lateral basals in Xenocrinus and sometimes in Eucalyptocrinus and Callicrinus. One form of cell-group migration involving the approximation and fusion of two groups into one would occur, if in the previous development of phylitic compression of characters anchylosis were to be carried back into the embryonic period. Thus, anchylosis which appeared as an adult character in early times might appear as an embryonic character in a later stage of development, and cause fusion of the formative cell groups. It is suggested that this type of fusion might be responsible for the interradial development of the two larger infrabasals in Antedon.‘ That cell groups as such may migrate in response to physiological stimuli from changed environment without any such evolutional change is also possible, and experimental evidence has been obtained to substantiate this hypothesis.” | Simple migration, after plate formation, without tortion or other movement of the sarcode, seems from the very nature of plate growth (see p. 501) to be impossible. This opinion is appar- ently substantiated by the system of migration of the anal plate in Antedon, and of the radianal in Promachocrinus* and Hathrometra,* and by the development of the posterior radials in Antedon upon the introduction of the anal. Equal spacing of the radials and the anal in the hexagonal stage of Antedon is not apparent; on the contrary, the space separating the postero-lateral and antero-lateral radials is much greater than that separating the postero-lateral radials and the anal. As the plates increase by branching and anastomosis, the adjacent margins of the anal and the postero- lateral radials meet and assume a finished appearance (Fig. 10) before the postero-lateral and antero-lateral radials meet. For- ward shoving of the postero-lateral radials into this unoccupied t Ref. 8, pp. 288, 280. 3 Ref. 16, p. 332- 2 Ref. 26, p. go. 4 Ref. 24, pls. 8-12. BASAL PLATES IN CRINOIDEA CAMERATA 539 area by growth pressure would be expected, but it does not occur. On the contrary, the posterior radials either assume an asym- metrical outlfne as in Antedon, most of the hexagonal Camerata (Pls. II-III), and many of the Fistulata, or the continued widening must be compensated for by accelerated growth of the sarcode in the posterior region. Since no better opportunity for simple plate migration could be conceived, and since it does not in this case occur, there is little reason for believing in its existence. Complex migrations consist of plate shiftings induced by accel- erated growth, tortion of the body wall, in either local or broad areas, which distorts the normal space and contact relation of plates. This form of migration is diagrammatically shown in the carrying up by elongation of the anal tube of the anal in Antedon, and of the radianal in Promacocrinus and Hathrometra. Where the anal and radianal, being more firmly attached to the viscera than to the adjacent plates, are bodily lifted out of the cup into the tegmen by the accelerated growth of the hind-gut, this process is the one which undoubtedly explains the migration of the radianal in all fossil crinoids. A more common form of complex migration, and one shown in many groups of crinoids, is that which results in the formation of biserial from uniserial ossicles in the arms. e) Plate interpolation—This process may be defined as the interpolation of some plate or plates, of primary or secondary derivation, between any plates forming the primitive crinoid cup and its appendages. It is one of the most common forms of evolu- tion found in the crinoidea, and may be broadly separated into two groups: primary interpolation, or the development of primary or secondary plates zz situ from a primary or secondary formative cell group; and secondary interpolation or migration. Only the first type need here be considered, as plate migration has already been discussed. Primary interpolation is the only known method by which additional stem ossicles appear; the new ossicles devel- oping either between the base and the adjacent stem ossicles, as in the Inadunata and Camerata, or between other stem ossicles, as in the Flexibilia. In the development of cirri, primary inter- polation is the rule, the interpolation taking place at the proximal end of the appendage. Here too belong the development of the 540° HERRICK E. WILSON perisomic" interbrachials, of the ambulacrals between the orals and interambulacrals, and of the peculiar plates appearing between the basal and radial cycles of Acrocrinus.? Interpolation in the basal cycle is known in but two genera. In Sagenocrinus and Homalocrinus the radianal is incorporated in the basal cycle, and in Sagenocrinus especially it assumes the appearance of a basal plate. Interpolation in the basal cycle, being well established in at least two instances, seems very rare, yet this may be due entirely to our lack of knowledge, for the evolution of the radianal plate in the Flexibilia leads one to believe that it appeared in its primitive state in the basal cycle A discussion of interpolation in the radial cycle has been purposely omitted in the preceding citations, as it belongs more properly in a discussion of the origin of the anal plate, where it will be fully considered. It appears then that every cycle of plates excepting the infra- basals, and every series excepting the brachials, are affected by interpolation, and even in the brachial series Clark* has evidence which points strongly to interpolation. There is then a possibility that every plate cycle and series may be subject to interpolation, but this point is not of immediate consequence and need receive no further attention. Interpolation in the calyx may demand (1) a reduction in some adjacent plate or plates of the same cycle or group, with or without oblique development of the plates,> as in the reduction of the posterior radials and oblique development of the radial cycle in Antedon upon interpolation of the anal plate; (2) an increase in diameter of some adjacent plate or plates in an apposed cycle or group, with or without trunkation, depending upon the alternating or superimposed position of the interpolated plate, as in the enlarge- ment and trunkation of the posterior basal in some of the Camerata t Ref. 16, p. 330. 2 Ref. 39, pp. 805-10, Pl. 53, Figs. 1-3, 4-9, and 10a, b. 3 Ref. 31, p. 493, Pl. 5, Fig. 9. 4Refs 13) purr: 5 Oblique development is here used in reference to the displacement of either the proximal or distal ends of a plate, from the vertical axis of the cup and not from the planes of pentamerous symmetry. BASAL PLATES IN CRINOIDEA CAMERATA 541 and Fistulata upon interpolation of the anal plate, and broadening of the radials without trunkation upon interpolation of interbra- chials, as in many of the Camerata; (3) increase in the body wall, as is shown in the lengthening of the calyx of Acrocrinus upon interpolation of the extra plates between the basals and radials; (4) interpolation of an extra plate in an adjacent cycle or group to satisfy the demands of plate alternation; (5) deformation of the cup; (6) various combinations of the first four effects named. f) Anchylosis —The process of anchylosis is provisionally placed under this group of processes, because of its intimate association with certain modifications of plate contact and position which cast some doubt upon the propriety of assuming it to be a result and not the cause of those modifications. This fact is shown in the following section. ANCHYLOSIS: ITS ANTECEDENTS AND CONSEQUENCES I. ANCHYLOSIS AND REDUCTION Anchylosis is the most potent factor operative in the obliteration of sutures, and while it has been discussed as a simple ontogenetic process, its antecedents and results have not been considered. The expression ‘‘reduction and anchylosis,”’ so commonly used in the description of brachials, means anchylosis following and depend- ing upon reduction, but whether or not this usage is morphologically correct is not clear to the writer. Anchylosis may take place without reduction, or reduction without anchylosis, although the former is not common. Anchylosis may perhaps be either preceded or followed by reduction, but the writer is inclined to believe that when anchylosis is preceded by reduction the reduction is phylo- genetic, and that in ontogenetic development anchylosis is followed by reduction. That is, plates which will appear in the adult as a reduced anchylosed unit are in ontogenetic development up to the time of anchylosis the equivalent of the other plates in the same cycle or series, and with anchylosis inhibition of growth causes the reduction of the compound plate. This inhibition of growth may be due either to deep-seated atrophy in local areas or to local superficial atrophy. 542 HERRICK E. WILSON The first change of interest in anchylosis of basals is the change of plate outline, from the pentagons through a heptagon to a hexa- gon. With lateral anchylosis of two basal plates, the pentagons of the basals are merged into a heptagon, with a re-entrant angle where the supported radial interlocks with the supporting compound plate. Upon further development, absorption decreases the angle of the radial plate, causing it to assume, first a lower angle, then a proximally convex outline, and finally, by complete absorption, a straight angle. At the same time, by increased deposition along the adjacent basal margin, the re-entrant angle of the heptagon is gradually filled, the filling comforming throughout with the reduc- tion of the inserted angle. The final step is the change of the heptagon to a hexagon. At some time either before, during, or after anchylosis, a remarkably persistent reduction of the compound plate or its component parts occurs: This reduction is parallel to the line of sutural closure, and is sometimes accompanied by reduction in the proximal diameter of the directly supported radial. From the principles set forth in the discussion of reduction and compensating enlargement, either compensating enlargement of the adjacent plates must take place or distortion follow the reduction. When only one pair of basals is anchylosed, as in the Xenocrinus, etc., and the Calyptocrinidae, the reduction is bilaterally symmetrical; the basals adjacent to the reduced plates are equally enlarged, and the reduction is apparently not due to deep-seated causes and affects only the basal plates and the proximal margin of the apposed radials (Pl. II, Nos. 2,7). When two pairs of plates are anchylosed, as in the Stephanocrinidae, Pentremitidae, and Platycrinidae, the problem is not so simple, for on one side the compound basals are mutually apposed, and the intervening suture meets the center of the proximal margin of the radial. If the reduction of the com- pound basals is asymmetrical and occurs only on the sides opposed to the simple basal, no distortion in symmetry is necessitated. If, however, the reduction is symmetrical, there must be a distortion in symmetry of the cup, for there is no basal to enlarge where the compound plates are mutually opposed. In the Platycrinidae the reduction is asymmetrical, and is perhaps due to superficial atrophy. BASAL PLATES IN CRINOIDEA CAMERATA 543 The simple basal is in general symmetrically enlarged, and the base is occasionally a regular pentagon. In the Stephanocrinidae and Pentremitidae the reduction of the compound basals is bilaterally symmetrical, and is usually accompanied by reduction of the radials directly supported by the compound plates. This reduction is compensated for on the anterior side by lateral growth of the simple basals and the radials obliquely — supported by it. On the posterior side, however, distortion has taken place. If absorption had caused the reduction, either plate shoving or shrinkage of the sarcode would be necessary to keep the plates in contact; but plate shoving is apparently impossible, and shrinkage of the sarcode improbable. Upon comparison of the ornamentation in the reduced and unreduced radials in Ste- phanocrinus, another change seems to have taken place. The neural ridges of the reduced radials are fused, farming a single broad ridge, which apparently indicates that the underlying nerves are in closer relation than in the other radials. The reduction seems, then, to have been caused by the inhibition of lateral expan- sion in the sarcode in the reduced areas, and not to absorption, and isa very good example of deep-seated atrophy. In the three- basaled, hexagonal Camerata, anchylosis and reduction of the basals are complicated by the appearance of the anal plate, and cannot now be considered. 2. DELAYED ANCHYLOSIS In many genera, especially of the hexagonal Camerata, anchy- losis takes place at such an early period in development that no trace of the immature forms with unanchylosed basals is preserved. The only hope, then, of locating the missing suture is by delayed anchylosis, or characteristics of ornamentation. Ornamentation has, however, received so little study, and the subject is so broad, that the writer cannot at present give it adequate consideration. Examples of delayed anchylosis, however, are not unknown, and a number of cases will be cited in the latter part of this paper. Delayed anchylosis is simply a nascent stage of anchylosis, due to the inhibition of activity in the ameboid calciferous cells. It may appear in the form of internal or external grooves, or in its 544 HERRICK E. WILSON completeness as an unchanged suture. In the latter case, it is in the Camerata always accompanied by the inserted angle of the apposed radial. In the pentagonal Camerata the reappearance of a suture or group of sutures would unhesitatingly be described as cases of delayed anchylosis. In the hexagonal forms, however, doubts might arise concerning the reappearance of the anterior suture in the three-basaled genera and of the right-anterior suture in the two-basaled genera, since such reappearance would, according to Wachsmuth and Springer, indicate the presence of six basal plates (pp. 492-93). There might also be some question as to whether these reappearances are due to delayed anchylosis or resorption; but until more light can be thrown upon the problem of sutural reappearance by reabsorption of the intrasutural deposit, these abnormalities may well be ascribed to delayed anchylosis. 3. ANCHYLOSIS AND THE PHYLOGENETIC REAPPEARANCE OF SUTURES Anchylosis of the basals, as far as we now know, is an onto- genetically repetitive process, confined to plates and not taking» place as a result of cell-group fusion. The basal sutures are always present in ontogenetic development, and constitute phylogene- tically a plane of weakness in the compound plate. This is appar- ently not true of the infrabasals, at least in modern forms, as is shown in the interradial development of two of the infrabasals in the embryo of Antedon. Atavistic reappearance of sutures, by delayed anchylosis, is a possibility, but the cenogenic or phylogenic reappearance of a suture lost through anchylosis is another question. The skeleton of the Echinoderm is deposited in the midst of living tissue and remains under the full control of the ordinary processes of growth, reabsorption, and modification by living tissue.t The partial or total absorption of plates, the shifting of sutures, and the reabsorption and modification of the basals in the formation of the centro-dorsal are sufficient evidence of this statement. It seems possible, then, that under the conditions of physiological disturbance (loss of vitality) common in the paracme of develop- ment,’ failure of anchylosis or reabsorption of the intrasutural deposit might take place, and the sutures reappear as phylogenetic t Ref. 7, p- 350- 2) [bid., Pp. 350. BASAL PLATES IN CRINOIDEA CAMERATA 545 characters; but we have no evidence of such a reversion, and, until such evidence is brought forward, theories of descent demanding the reappearance of lost sutures should be carefully scrutinized. DEVELOPMENT OF THE INTESTINE AND THE CONSEQUENT ZONES OF POTENTIAL WIDENING “Tn all the Echinoderm classes it is the digestive tube that controls any departure from pentamerous radial symmetry.”’ This statement by Clark" may perhaps be too sweeping in extent, especially when we consider the potency of atrophy and compen- sating hypertrophy (p. 533) in distorting symmetry; but the fact that the digestive tube is one of the most powerful factors in distorting symmetry cannot be too strongly emphasized. The most striking and therefore the most widely known effect of this power is that shown in the various distortions of the mouth and ambulacral grooves by excessive growth of the hind-gut. These tegminal distortions have been so frequently described that there is no necessity of reviewing them here; distortions produced by the intestine in the basal and radial cycles have, however, received too little attention. In the development of the digestive system in Antedon the gastric sack is elongated horizontally into a form somewhat resem- bling the human stomach, having a large end into which the funnel- shaped oesophagus opens, and a small end with a caecal termination, which is the potential intestine. Upon further development the intestine is also horizontally prolonged, and coils to the right, around the stomach; in the space left for it by the enlargement of the calyx;? before the coil is completed, however, the anal appears and the intestine directs itself toward that plate. The pressure _exerted by the intestine upon the anal tends to keep separated the posterior radials and prevents the right-posterior radial from encroaching upon the anal. Soon, however, the intestine turns upward and carries the anal with it into the tegmen. The thrust exerted by the outward growth of the intestine which dis- places the anal and also the radianal (see p. 538) must not be con- Neh T0;) D152. 2 Ref. 12, pp. 227-28. 3 Radianal. See Ref. 16, p. 333. 546 HERRICK E. WILSON sidered as a gentle pressure, nor the displacement of these plates as a gentle process, due only to their much closer association with the intestine than with the surrounding plates. The outward push of the intestine is in proportion to the strength of the calyx walls a powerful force, capable of inhibiting plate growth and of greatly distorting the relations of plate contact and position. In the recent works of Springer and Clark attention has been called repeatedly to the remarkable parallel between the ontogenetic migrations of the radianal in modern crinoids and its phylogenetic - migration in fossil forms. Since the migration of the radianal in recent forms is caused directly from its intimate association with the hind-gut in its upward growth, there can be no doubt that such an association existed in the ancient crinoids, and that the tendency for shifting the radianal gradually increased and the association became so firmly established that the radianal is now completely withdrawn from the cup in individual development. The radianal in the early Flexibilia is incorporated in the basal cycle below the right-posterior radial, and probably appeared in that position in the ancestors of the Flexibilia.2, The outward push of the developing intestine was then directed obliquely to the right against the radianal and in the succeeding stages shoved and pulled this plate upward and to the right into the posterior inter- radius and out of the cup. Furthermore, in Sagenocrinus* it permitted such an enlargement of the radianal that the right mar- gin of the posterior basal was shoved to the center of the posterior interray. This change is of especial interest in the study of basal plate evolution, as it shows one method of obtaining a posteriorly directed basal suture, such as is exhibited in all genera of Camerata having a hexagonal, tripartite base. This apparent digression from the subject of basal plate evolu- tion in the Camerata—a group in which, as far as we now know, a radianal plate never appeared—is for the purpose of bringing clearly to mind the powerful effect of the growing intestine and the presence of zones of potential weakness in the calyx. These zones of weakness lie along the posterior interradius in the radial cycle Reta rhs ao: 2 Ref. 31, p. 430, Pl. V, Fig. 9. 3 [bid., Pl. VII, Fig. 18. 3 F . ‘ ' BASAL PLATES IN CRINOIDEA CAMERATA 547 and along the right-posterior radius in the basal cycle, and it is to the latter zone that especial attention is called. Wachsmuth and Springer have noted this zone of potentiality in but one instance (description of Fig. I, No. 7, of this paper), although it seems the only zone of potentiality in the basal cycle which can logically be accounted for. Pressure from the end of the developing hind-gut must from the necessity of its position be directed obliquely to the right against the posterior interradius, and any lateral pres- sure of the gut must be directed against the right side of the calyx, the two combining to shove the right side outward and away from the left side. Thus the stress produced by this shove would naturally fall along the right-posterior basal suture, as this is the nearest suture or plane of expansion adjacent to the posterior interradius within the zone of pressure exerted by the hind-gut. ORIGIN OF THE ANAL PLATE Comparison of the cup and tegminal structures in the Bato- crinidae and Platycrinidae shows that a very long period of time must have elapsed, or very rapid evolution have taken place, before such a highly specialized form as Tanaocrinus could have originated from any of the early Platycrinidae. Since Tanaocrinus is an early Silurian (Richmond) genus, relationship can be established with the Platycrinidae only through Ordovician or pre-Ordovician ances- tors; therefore, if Tanaocrinus is related to the Platycrinidae, it must have been derived from a form having a simple, pentagonal, five-basal cup. The primary step in the evolution of this form into Tanaocrinus would be the introduction of the anal plate into the radial cycle, thus giving to the cup its hexagonal outline, and inducing in the basal-plate cycle a remarkable series of modifi- cations. The questions then arise: How do we know that the anal was a secondary and not a primary plate? At what period in the ontogenetic development of the Camerata was it interpolated in the radial cycle? Where did it originate? And what changes followed its interpolation? ~ It is generally agreed that the anal plate in the hexagonal Camerata is of secondary origin. If this is true, the statement just made concerning the ancestry of Tanao- crinus is undeniable, for by eliminating the secondary plates of 548 HERRICK E. WILSON that genus and restoring the orals an ideal larval or ancestral form will appear. If, however, the anal plate is a primary plate in the radial cycle, a different line of descent is indicated. In the ontogenetic development of the skeleton in the living Antedon the anal plate is interpolated after the basals and orals have formed closed cycles, and before the radials are laterally in contact. As the anal expands by lateral and proximal growth, it comes into contact with the posterior radials and the posterior basal; but this happens before the distal margins of the basals are completed and while the radials are still separated from each other. Since growth in the anal plate and the posterior radials does not cease upon their coming into contact, shoving of the posterior radials in the direction of the unoccupied lateral areas might be expected. This, however, does not occur. On the con- trary, the crowding results in a partial inhibition of growth in the apposed margins, and a marked asymmetry results in the outline of the posterior radials, especially in the right-posterior radial. If comparison is now made between Antedon in this stage of develop- ment and the early Camerata having a hexagonal base, a striking similarity is seen in the development of the basals, radials, and the posterior side of the calyx. The posterior radials in both forms are asymmetrical and narrower than the anterior radials, and the asymmetry is due to the diminution of the posterior side of the plates, the distance between the center of the radial facet and the plate margin being less in the posterior half than in the anterior half. If comparison is then made between the relative position of axial lobes and radial plates in pentagonal and hexagonal Camerata, a further distortion is noted in the hexagonal forms. The lobes of the canal in the pentagonal forms occupy an interradial position, while either two or three of the lobes in the hexagonal forms occupy a radial position. These facts show that there is a distorting factor present in the posterior side of the cup. When it is considered that pentamerous symmetry is the rule in Echinoderms, and that sutural symmetry based upon the hexamerous plan only appears in the basal cycle with the appearance of the posteriorly directed basal suture, there seems to be no other alternative than that the anal plate is the distorting factor, and that it has BASAL PLATES IN CRINOIDEA CAMERATA 549 developed secondarily in the radial cycle, or has migrated into that position. The origin of the anal plate is as yet an unsettled question, but there are several possibilities which may well be considered. It may have originated as a secondary plate in either the basal or oral cycle. It may have developed as one of a brachial series, or an interbrachial cycle, or it may have originated as a separate plate in the radial cycle. Origin in the basal or oral cycles is clearly out of the question, for we know of no plates in either cycle from which it could have originated. Assumption of origin as a brachial, which is Bather’s explanation for the origin of the anal plate in the Fis- , tulata,t is without foundation in the Camerata. Origin as one of an interbrachial series of the ordinary type is also improbable, for although these interbrachials are present in Tanaocrinus, Xenocrinus, and Compsocrinus, they are clearly formed at a later stage of development. Origin as a first interbrachial, that is, one of an interbrachial series interpolated between the radial plates, has been seriously considered by Carpenter’ in a comparison of Xenocrinus, and some of the dicyclic Camerata, with Antedon and Thaumatocrinus renovatus; and while such a cycle may have existed and the lateral plates have atrophied, there has been found no record of such a cycle in the monocyclic Came- rata. This, however, does not preclude the idea that the anal series may have been so interpolated, and that the lateral plates which appear in some of the dicyclic Camerata have been the result of reduplication.s Let us examine the ornamentation in the anal series of Compsocrinus and Xenocrinus, and see if this may not throw some light upon the question. Bather, in calling atten- tion to the anal ridge in the Reteocrinidae, Glyptocrinus, etc., says, “The anal ridge is connected with the ridges that unite the posterior basal to the right- and left-posterior radials, and this indicates that an axial cord passed up it to govern the motions of the anal tube.’’4 That this ridge does so indicate the presence of an anal nerve seems beyond question, for in comparing the ornamenta- tion in Xenocrinus and Compsocrinus with the nervous system of t Ref. 4, pp. 319-31. 3 Ref. 16, p. 338. 2 Ref. 10, pp. 38-406. AIRE, O, [Dp 110), 55° HERRICK E. WILSON Thaumatocrinus a striking similarity is discovered. In Thau- matocrinus the interradial arms are innervated by secondary branches of the axial cord, which originate slightly above the point of bifurca- tion of the basal cord, quickly join, and pass up the interradial arms, while the main branches pass up to the radials (Fig. 7, No. 2). In Xenocrinus and Compsocrinus (Fig. 8, No. 1) the anal nerve ridge arises at the point of bifurcation of the axial trunk ridges in the posterior basal, and there can be little doubt that the branching of the underlying nerves took place in the same manner as they do in Thaumatocrinus. The parallel here is so close that the writer was at first inclined to the belief that the interpolation of the anal series in the Camerata is of the same type as the interpolation of the interradial arms in Thaumato- crinus. In Thaumatocrinus' the ‘“‘interradial”’ radials appear very early in the ontogenetic develop- ment as narrow plates separating Fic. 8.—Diagram showing the the radials and gradually increase course of the radial and anal ridges 49 the size of the true radials. An in Comprcrins andthe DROS Chection to this form of develop Compsocrinus harrisi (based upon ment has been stated by Bather, on Wachsmuth and Springer); 2, the ground that no primitive genera Thaumatocrinus renovatus (after Car- : ° aes have been found in which the anal penter); course of nerves based upon dissection by the writer. plate appears aS a Narrow linear plate. This objection, however, is not formidable, for the change may have been a discontinuous mutation, or may have taken place during periods of retreat of . the sea. Ulrich believes that most mutations have so taken ae for he says in the “Revision of the Paleozoic Systems”: “ almost invariably we deal with the nearly finished product a a process of mutation that was begun and established before the new phase invaded areas now accessible to the student of fossil faunas.’ «See Ref. 16, p. 337. 2 Ref. 3, fifth notice, p. 37. 3 Ref. pp. 498-501. BASAL PLATES IN CRINOIDEA CAMERATA 551 The logical sequence of events based upon the Thaumatocrinus theory would be the interpolation of the anal plate in the radial cycle and serial development of the succeeding anals. But when this sequence of events is applied to the pentamerous, monocyclic Camerata difficulty is immediately encountered. Either these forms have lost their true anal plate and the anal now appearing in them is the homologue of the second anal in the hexagonal Camerata, or they have developed along a different line of evolution from the hexagonal forms. But before taking up these questions let us examine this theory of interpolation more closely. Stratigraphically the pentamerous base precedes the hexamerous base, and it seems scarcely possible that a hexagonal form which could give rise to both Glyptocrinus and Tanaocrinus could have been living in the ocean basin during pre-Ordovician and Ordovician times, and that only those forms having lost the anal plate should have migrated into the epicontinental seas during the Ordovician, while those having the anal plate were withheld until Silurian (Richmond) times. This is apparently carrying the theory of selective action too far to be believed. Again, the embryological evidence shown in Thaumatocrinus may not be reliable. We have noted that the radianal in Promacocrinus originates to the left of the right-posterior radial; furthermore, we know that the radianal appears in the more primitive crinoids in a subradial position. There has been not only a progressive upward shifting of the radi- anal in these groups, but there has been apparently a progressively upward shifting of its point of origin. Embryology does not repeat all the ancestral characteristics step by step and then eliminate them in a different fashion in producing the various genera and species; certain characters which are gradually being eliminated phylogenetically are probably, in Crinoidea, the result of a progres- sively increasing inhibition of plate development in the larva, which ends in the complete obliteration of the plate. Since embry- ology does not repeat all the ancestral stages, and does in this case permit of changes in the position of origin of a plate, there is a possibility that the “‘interradial”’ radials in Thaumatocrinus did t For a more complete discussion of this phase of development based upon a wide series of observations, see Ref. 23, Chap. III, ‘‘Recapitulation.” 552 HERRICK E. WILSON not originate in the position in which they now originate, and too much dependence should not be placed upon this character. The evidence presented by fossils is fixed, although often wrongly inter- preted, and until stronger evidence is submitted it is well to hold closely to that presented in the stratigraphic succession. Stratigraphically the pentamerous base preceded the hexamerous base, and when we consider with this the evolution of the two- basaled, hexagonal Camerata, a more logical theory of development for the anal plate is presented. In this short-lived group we have a very rapid evolution from some platycrinid stock. The anal plate in Platycrinus originates in the posterior interray between the distal margins of the radial plates, while the anal plate in Dichocrinus projects above the level of the radials and costals and bends sharply inward toward the anal tube; its distal portion is reduced, and, if it were separated from the enlarged proximal portion and slightly modified, could not be distinguished from the anal in Platycrinus. Enlargement and downward growth of the anal seems then to have occurred, and a different theory is offered for the interpolation of the anal plate. This theory for the appearance of the anal plate is, then, that the anal plate is of secondary derivation; that it was interpolated phylogenetically after closure of the radial cycle, but ontogenetically after the basal had formed a closed cycle and before the radials had come into contact. The cup in this stage of development is in a flexible condition, and readjustments can readily be made. Development of the anal plate in this position requires no true migration to come into contact with the posterior basal; portional migration, or proximal growth with distal inhibition in the younger stages, will produce the Tanaocrinus type of anal, while distal growth alone is necessary to produce the Glyptocrinus type of anal. The stimulus which kept the radials apart and permitted this downward growth of the anal plate was the demand for room on the part of an enlarging hind-gut, and it is this stimulus which has caused some of the most remarkable changes in crinoid evolution. Additional plates in the anal series were probably added as needed, for protection of the anal tube, and no slipping downward of a completed series of anals is required. In considering this change BASAL PLATES IN CRINOIDEA CAMERATA 553 we must remember that we are not dealing with a completed model in which the plates are of fixed and unchangeable size, and in which every change of plate position must be accompanied by an entire readjustment of the adjacent plates; we are dealing with a growing organism in which there is a certain amount of flexibility in adjust- ment by plate growth. If this theory for the interpolation of the anal plate is correct, the first anal plates of Glyptocrinus, Platycrinus, and Dichocrinus are homologous; and in further developing the theory for the evolution of basal plates this view will be followed. [To be concluded] DIFFERENTIATION IN INTERCRUSTAL MAGMA BASINS ALFRED HARKER Cambridge University Dr. N. L. Bowen’s comprehensive article on “The Later Stages of the Evolution of the Igneous Rocks,” issued as a supplement to the last volume’ of this Journal, will be hailed with satisfaction by all petrologists, and indeed with gratitude by those who have the misfortune not to be chemists. It contains the first serious attempt to deal with the problem of magmatic differentiation directly from the standpoint of experimental knowledge. In demonstrating how the course of crystallization may be changed by the sinking of crystals, or the straining away of liquid from crystals, or the formation of zoned crystals in isomorphous groups of minerals, the author scarcely goes beyond actual laboratory experience, and his conclusions accordingly carry a great weight of authority. When he proceeds to construct on this basis a general theory of differentiation, the element of hypothesis is necessarily introduced, and, as the author recognizes, his argument can no longer command unquestioning acceptance. It is a very interest- ing contribution to a discussion which is not likely soon to be closed. I wish to make a few remarks upon one of the subsidiary issues, which, however, touches the main theory at numerous points, viz., Bowen’s predilection for differentiation 7m situ as opposed to differ- entiation prior to intrusion. That an appreciable settling down of crystals may take place after intrusion is not to be denied, but I think that the experience of any field geologist goes to show that it is a rare and exceptional incident. Daly has given a list of about thirty stratified sills and laccolites in which such “ gravitative”’ dif- ferentiation is believed or conjectured to have occurred, but proba- bly a critical examination would dispose of many of the examples cited. In some, such as the Loch Bordan mass in Sutherland, there is not a gradual transition but a sharp boundary between the several 554 DIFFERENTIATION IN MAGMA BASINS 555 rock types. Bowen remarks that the upper acid magma, remaining fluid after the lower basic portion has wholly crystallized, may come to have an intrusive relation to the latter. This would be sufficient to explain veining of the one rock by the other; but, where an overlying sheet is separated from an underlying one by a surface of discontinuity, I can see no explanation but that of distinct intru- sions. Nor is this explanation necessarily excluded even when no sharp division is seen, for, under appropriate conditions, a transitional zone may result from partial admixture. The Sudbury laccolite is probably a case in point, though I must confess to only a limited personal examination of the mass. I found no indication of a regular “composition gradient”’ in either norite or granophyre, considered separately, while the transitional zone between them has all the characters of a hybrid rock. The sulphide ore I leave out of count, as doubtless representing a magma immiscible with that of the norite. The clear instances of gravitative differentiation in sills and laccolites of which I have direct knowledge are all in rocks which must represent very unusually fluid magmas, such as the analcime-bearing intrusions of Permian age in Scotland. They are the kind of exceptions which help to prove the rule: viz., that in an intrusive body of moderate size a prohibitive viscosity soon puts a stop to the settling down of crystals. Doubtless a laccolitic mass of very large dimensions retains its fluidity longer, but it is obviously in a great intercrustal reservoir that the most favorable conditions for this action will be realized. Bowen would apply the conception of differentiation in place to the plutonic rocks of Skye; but the facts, as I see them, abso- lutely negative such a hypothesis. The peridotite is not found at the base of the gabbro, but enveloped in the midst of it. The granite breaks through the gabbro, and, where any approach to a stratiform arrangement is apparent, does not overlie, but underlies, the basic rock. In one part the granite has been so chilled against the gabbro that its margin and the offshoots from it assume the characters of a spherulitic rhyolite. I infer that the gabbro was, in this place, not only solid but cold when the granite was intruded. The large gabbro laccolite itself is made up of numerous irregular sheets, showing differences in composition and structure, and often 556 ALFRED HARKER visibly cutting one another. In the peridotite this composite struc- ture is more strikingly exhibited, and it can be detected in places in the granite, which isa much more uniform rock. The several compo- nent sheets are not disposed in an orderly fashion in accordance with their various densities. Add to this evidence the fact that peri- dotite, gabbro, and granite all make smaller separate intrusions, some much too far away from the main complex to have any direct connection with it, and it will appear beyond dispute that the differentiation which yielded these various rocks was effected prior to their intrusion, and therefore in some large reservoir at a deeper level. Bowen does not refuse the conception of a deep-seated magma basin stratified according to density; but he seems to think it an absurdity that, on that hypothesis, the earlier intruded magmas should be drawn from the lower levels (p. 73). I will try to remove his objections. I have already urged' that in order that such a basin may have a considerable degree of permanence, as it obviously has, we must suppose some approach to thermal equilibrium between it and the surrounding crust. This implies a temperature gradient within the basin approximately like the normal gradient in the earth’s crust of the region. It implies, further, what I may call a fusibility gradient corresponding with this normal temperature gradient. Now, the separation and sinking of crystals, as pictured by Bowen, goes with a cooling-down of the magma, which terminates in complete solidification. Any intrusions drawn from the basin must therefore be consequent upon remelting. The occasion of this I presume to be a gradual rise of the isothermal surfaces, which must accordingly become more closely spaced. In other words, reheating implies a temperature gradient steeper than that to which the fusibility gradient is adjusted, and it follows that the lowest layers will first become fluid. I have not attempted to develop this view of the matter, and should welcome criticism; but Bowen’s zeal for differentiation in place has caused him to pay little regard to the possibilities of this alternative. I am wholly in accord with Bowen in the conviction that alka- line and calcic rocks are derived from the same primitive magmas t See especially Compte Rendu XII Congr. Géol. Intern., Toronto, 1914, pp. 205-8. DIFFERENTIATION IN MAGMA BASINS 557 (p. 59). My belief has been, and is, that the differentiation of these two great classes of magmas from the common stock and the separation of them—in general in a horizontal sense—constitute the first and most important steps in the evolution of igneous rocks. Why the chemical differentiation should so consistently follow these lines has been a difficult problem, and it is the more gratifying to be offered at least a partial answer to the question. Stated broadly, Bowen’s ideal scheme of differentiation leads first to a series of calcic rock types and subsequently, if continued, to an alkaline series. There are qualifications of this rough statement which I do not go into here; but in general it appears that, if a separation can be brought about at a certain well-defined stage of the progres- sive differentiation, it will be a separation between calcic and alka- line. This separation, I hold, has actually been effected on a grand scale, and I have sought the immediate cause of it in the action of crustal stresses squeezing out the residual fluid. A discussion of this suggestion from the chemical point of view would be instructive, but here Bowen disappoints expectation. He dwells on particular cases in which separation has not taken place at the stage specified, but at a somewhat earlier stage; and he throws doubt upon the existence of any general regional dis- tribution of alkaline and calcic rocks, such as Iddings demon- strated long ago. The fact that, among the younger rocks of North America, alkaline types characterize the Atlantic slope and calcic the Pacific, he would explain by supposing that erosion has exposed deeper levels on the western side of the Rocky Mountains than on the eastern. He forgets that the contrast of petrographical facies holds good for the lavas as well as for the intrusive rocks. More- over, the fact that lava flows still cover vast areas on the western side, while on the eastern they have mostly been removed, makes it difficult to accept his statement about the relative amounts of erosion. As regards the association of calcic rocks with regions subjected to powerful lateral thrust, nothing would be gained by traversing old ground again, but to Bowen or any other unbeliever I will offer just one consideration. If we examine those crystalline schists which are admittedly of igneous origin, together with foliated 558 ALFRED HARKER igneous gneisses, we find that they belong almost exclusively to the calcic branch. A few exceptions there are, and must be. A nephe- line syenite may be intruded in a line of faulting during the time of movement, as in the Langesundsfjord; or it may be crushed and metamorphosed long afterward by stresses with which it has no genetic connection, as at Loch Borolan; but these are isolated and incidental occurrences. The matter is easily brought to the test. In Grubenmann’s classification, based solely on chemical composi- tion, the crystalline schists and gneisses of igneous origin are con- tained in six of the twelve groups. The calcic rocks are in Groups I, III, IV, and V, which correspond with granites, diorites, gabbros, and peridotites. They include a rich variety of types, and col- lectively make up enormous tracts of the earth’s crust. To com- plete his classificatory scheme the author has been able to produce various types of alkaline rocks, which scantily furnish forth Groups VI and VII, but most of them are little more than petrographical curiosities. In respect of the total bulk of all known occurrences, these alkaline crystalline schists as a whole are quite insignificant as compared with any single type in the calcic division. The striking disparity here noted is only one consideration among others which points to a peculiar distribution of alkaline _and calcic igneous rocks in relation to crustal stresses. If anyone seriously believes that such things are matters of chance coincidence, there is no more to be said. It is to be hoped rather that chemists, as well as geologists, will recognize here a real significance, and will lend their help in the attempt to explain the facts, not to explain them away. STRATIGRAPHY OF THE SKYKOMISH BASIN, WASHINGTON WARREN S. SMITH Berlin, Washington WITH REPORT UPON PALEONTOLOGY AND PALEOPHYTOLOGY CAROLINE A. DUROR STRATIGRAPHY I, GENERAL AREAL DISTRIBUTION In general, the rocks of the Skykomish Basin trend in a north- south direction. This is almost exactly true of the igneous and metamorphic terranes. Only the sedimentary Swauk and the tufaceous Keechelus series vary from this general statement, for they show a trend in general west of north but bending northward on their northern prolongation. Dawson has remarked this tend- ency of all the rocks of the Cordilleran system.t The rocks lie in roughly parallel bands in the north, but are replaced southward until the granodiorite entirely takes the place of the earlier series and extends over nearly the entire width of the quadrangle. Igne- ous rocks greatly predominate in the area, metamorphics and sedimentaries being approximately equal in amount, and both being comparatively small. Volcanics make up a smaller division. These relations will be illustrated by the following classification table: Shoqualmuiereranodionitens.... .- 0-64. sees 130.90 iKeechelusyandesitic series: 2-5 44:08 o eee 32.20 versodaygramiterand: Beckler stocks). sequins HBAS LOMUSCHIS tater Aton ae 2 oer fon Meee gers ea 22.95 Swaulkeisedimentanye Series: <4 ene 56.00 Maloney metamorphic series................. 14.80 Iindexmeranodioniteme- 1 snes ase Bh, 2 SWINSS OUUSKOD QE « oo og oa docs oso saa nos AS Sob ml 1G. M. Dawson, “Geological Record of the NOESY Mountain Region in Canada,” Bu Geol. Soc. Am., XII (1901), 50. 559 560 WARREN S. SMITH In the following discussions no effort will be made to subdivide the metamorphic rocks of a sedimentary origin from those of an igneous origin, where these are so intimately associated as to make the subdivision impracticable. However, the schists of the north- eastern area are easily kept distinct from the metamorphic series of the northwestern area. On stratigraphical grounds the rocks readily fall into two groups: (1) the pre-Tertiary, and (2) the Tertiary. The division line between these is the most marked unconformity in the Cascades. We have present a schist, belonging to the pre-Tertiary, which is cut by quartz and igneous rock dikes, making the oldest or basal terrane. This is called the Easton schist, and it forms the metamorphic terrane in the northeast. No definite idea of its age can be sug- gested except that it is pre-Ordovician. Fragments of it are included in the Mesozoic batholiths, and it is more complexly folded than the Maloney (Gunn Peak) metamorphic series. The next younger series belongs on paleontologic and correlation evidence to the Ordovician. It is a series of quartzites, schists, and crystal- line limestones with associated greenstones, approximately 4,000 feet thick, outcropping in the northwest. Weaver correlates this series with the Cache Creek series as defined by Dawson. It has at least one stage less of dynamic history than the Easton and on lithologic grounds it is believed to be equivalent to Smith’s Peshastin series of the Snoqualmie area; but its fossil content identifies it as Ordovician instead of Carboniferous, as the Peshastin is called by Smith and the Gunn Peak by Weaver. Nothing can be said of the remainder of the Paleozoic history of the area. The absence of later Paleozoic and Triassic seems to be general in the region of the Cascade Mountains. In Jurassic time, however, there was a notable period of deep-seated volcanic activity, result- ing in the intrusion of the great Sierra Nevada-Cascade granodi- orite batholith, possibly the greatest of igneous intrusions.2 This batholith is represented by two terranes in the Skykomish Basin. The first is the Tye soda granite in the northeast, and the second *G. M. Dawson, Ann. Rept. Can. Geol. Surv., N.S., VII (1894), 37B-40B. ?R. A. Daly, Igneous Rocks and Their Origin (1914), p. 53. STRATIGRAPHY OF THE SKYKOMISH BASIN 561 is the Index granodiorite in the northwest. The Cretaceous is not represented but is known to occur as a marine series farther north- ward. Following the Mesozoic batholithic intrusion came a posi- tive orogenic movement, corresponding to the Laramie revolution, which left the Cascade area, at the close of the Mesozoic, high above sea-level. The Tertiary period opened with a period of continental deposi- tion in which the Swauk sandstone was deposited unconformably on the eroded edges of earlier metamorphic rocks. These arkoses consist of mingled fragments of granitic rock and of schist derived from the Paleozoic metamorphic series and from the Mesozoic batholith. Both Smith and Willis consider these arkoses to be lake deposits, but Weaver more correctly designates them as purely continental. They show conglomeratic facies and cross-bedding and vary so markedly in thickness that only the maximum figure of 4,000 feet is of any significance. They are equivalent to the lower Puget of western Washington and, on paleophytological grounds, to the Fort Union of the Montana-Wyoming areas. No definite information is recorded of the late Eocene and Oligocene, the gap being succeeded by Miocene andesitic tuff beds. These are volcanic tuff ejectamenta cut by dike- and sheet-like intrusions of andesitic composition. We have insufficient evidence to suggest where these tuffs had their immediate source, but their chemical resemblance to the Miocene granodiorite suggests forcibly that they were derived from the same magma. In later Miocene a recurrence of deep-seated vulcanism took place, resulting in the injection of the Snoqualmie granodiorite, which is the most important terrane: present in the area. This batholith, if its texture is to be accounted for, must have been covered by at least 2,000 feet of rock. In this cover late Eocene and Oligocene may be included. In latest Miocene and in early Pliocene the region was planed to a low relief; a disturbance of isostatic equilibrium followed, which resulted in the arching up of the Cascade Mountains to a maximum height of approximately 8,000 feet. Since this Pliocene uplift, canyons some 5,000 feet deep have been cut in the granodiorite, a fact bearing witness to the severity of erosion experienced by the area. 562 WARREN S. SMITH In the Pleistocene the Skykomish Basin was eroded by valley glaciers, and the only discernible post-Tertiary volcanic history is recorded in a thin layer of ash, probably drifted from Mt. Rainier by prevailing southwest winds. Post-Pleistocene time has not lasted long enough seriously to change the aspect of the topography left by glaciation. II. . PRE-TERTIARY HISTORY PRE-MESOZOIC The pre-Mesozoic is represented by two divisions, one of which is considerably older than the other, and on lithologic grounds is referred to an equivalency with the Easton schist. The later division has experienced at least one stage less of metamorphic history and is of known Ordovician age, on the evidence of fossils obtained from a cherty phase of the limestone lens outcropping in Lowe Gulch (D 2), two miles west of Grotto. No evidence of the exact age of the Easton schist has been put forth and none can be suggested for the Skykomish Basin. It comprises extremely metamorphosed and crumpled rocks generally derived from sedi- ments. This structural condition of the rocks precludes any esti- mate as to their original thickness, but does enable one to infer that they are of considerably greater age than any other Paleozoic division which is much less dynamically disturbed. MALONEY The Maloney comprises a series of metamorphosed rocks including quartzites, limestones, and schists of sedimentary origin cut by basic igneous rocks, usually best described as greenstones, which are later in age than the Maloney, but which on the evidence of their extreme metamorphism are considered to be also of Paleo- zoic age. The name Maloney is applied to a series formerly believed . to be equivalent to the Peshastin series, which is called Carbonif- erous in age. The evidence on which this correlation is made is presented in an earlier paper. Subsequent identification of fossils found in the limestone lenses shows the formation to be Ordovician in age, and therefore the name Maloney is suggested. * Warren S. Smith, ‘“‘ Petrology and Economic Geology of the Skykomish Basin, Washington,” School of Mines Quarterly, XXXVI (1915), 157. . STRATIGRAPHY OF THE SKYKOMISH BASIN 563 Miss Caroline A. Duror, in another part of this paper, has identified the following fossils as of Ordovician age: Rafinesquina deltoidea and Illaenus americanus. ‘This is the first recorded evi- dence of the presence of Ordovician strata in the Cascade Moun- tains of Washington, and both fossils are the first of their kind of Paleozoic age to be found there. With this Ordovician series begins the definitely known geologic history of the Skykomish Basin. It was a period during which the area stood approximately at sea-level, as evidenced by the fact that the limestone lenses carry marine fossils. The presence of quartzite shows that prob- ably the area was not deeply submerged, but more probably was near the continental margin. There is no record of the post- Ordovician Paleozoic. MESOZOIC There are no Mesozoic deposits at present, though there is evidence that they must have existed. The sole known event of the Mesozoic in the Skykomish Basin is the intrusion of batho- lithic igneous rocks whose structural relations identify them as Mesozoic, though there is no other positive evidence of their age. But it is altogether probable that the batholiths were intruded as a part of the great Sierra Nevada intrusion identified in California and Oregon as of Jurassic age. Daly, Smith and Calkins, Russell, Weaver, and others assert this probability. The granodiorite of which this batholith is composed is granitoid in texture, and such a texture can be established only under a thick cover of super- jacent rock. This cover was removed in post-Jurassic time, leaving the batholith uncovered at the end of the Mesozoic. Russell* and Smith? have described sedimentary rocks of Cretaceous age, from Whatcom County to the north, and it is therefore inferred that removal of the 2,000 feet or more of the cover of the Jurassic batholith was toward the north. The series, as described by Smith, under the name Pasayten, begins with a conglomerate in which tT. C. Russell, ‘“‘Cascade Mountains in Washington,” 20th Ann. Rept. U.S.G.S., Part II (1898), p. 114; G. M. Dawson, “‘ Geological Record of the Rocky Mountains in Canada,” Bull. Geol. Soc. Am., XII (1901), 84. 2Smith and Calkins, ‘‘Cascade Mountains in Washington,” 20th Ann. Rept. U.S.G.S., Part II (1898), p. 114. 504 WARREN S. SMITH granitic bowlders predominate, is 6,000 feet or more thick, and is dynamically disturbed. The Mesozoic closed with a period of se- vere orogenic disturbance, as evidenced by the structural relations Fic. 1.—Swauk arkose series. Conglomeratic facies of Tertiary rocks. This unconformity is very marked, because the earliest Tertiary sedimentaries lie at distinct angular unconformity on the pre-Tertiary rocks, a break which is well described by Smith and Calkins in the Snoqualmie area.t. The pre-Tertiary *G. QO. Smith and F. C. Calkins, Folio U.S.G.S., 1906, 2. STRATIGRAPHY OF THE SKYKOMISH BASIN 565 disturbance left the area adjacent to the Skykomish Basin on the east in a condition sufficiently elevated to make the earliest Tertiary a period of erosion. III. TERTIARY EOCENE In the Skykomish area the Eocene was a period of sedimenta- tion during which 4,000 feet of arkoses, shales, and conglomerates were deposited. About the middle of the series there are two shale formations which yield a considerable flora. The series is directly continuous with the Swauk series of the Snoqualmie quadrangle, and is also related to it on paleontologic grounds. It outcrops in a belt several miles in width, striking approximately N. 45° W. in the Eagle Creek-Beckler and the Foss River valleys. The dips vary; in the measured section, half a mile south of the Great Northern Railway, the base of the series lies nearly level, with increasing dip to the east as one goes east, and to the west as one goes west, until the top of the series stands vertical. It is an anticline whose basal beds rest unconformably on Easton schist, and whose roof is a part of the Keechelus andesite series. Miss Duror’s appended report is made on the flora collected from two horizons about 600 feet apart vertically, the lower being 1,100 feet from the base of the Swauk. Fossils numbered F 831, F 831-++50, and F 844 are from the upper beds; those numbered F 865 from the lower. With the latter is associated a coal bed some 14 inches thick on which slopes have been driven in the hope of finding minable coal. The sandstones and Pee incrdtes are cross-bedded and the fragments usually angular and never assorted (Fig. 1). It has been considered a fresh-water lake deposit. In the light of the recent developments of stratigraphy it may better be classed as in part purely continental and in part deposited by streams, prob- ably in deltas. The shales, and particularly those carrying complete specimens of Sabal, could only have been laid down in situ and must have been formed in shallow water—probably in a swamp. None of the series bears evidence of deposition in deep water, and both the angular condition and the considerable size of the frag- ments forbid their having been transported for any great distance. 566 WARREN S. SMITH Correlation.—Knowlton has done the paleontologic work on the continental Eocene of Washington. He has made the determina- tions both for the Eocene (Puget) of western Washington and for the Eocene (Swauk, Teanaway, and Roslyn) of eastern Washington. Continual reference is therefore made to him as an authority. The original report on the Swauk' showed 25 species belonging to the following genera: Lygodium, Sabal, Myrica, Comptonia, Populus, Quercus, Ficus, Cinnamomum, Prunus, Diospyros, Zisy- phus, Celastrinites, Phyllites. Of these genera, Sabal, Populus, and Ficus are reported by Miss Duror from the Skykomish Basin, and both Sabai and Ficus are index fossils of the lower Eocene (Puget), as stated by Knowlton: “The following genera have been found in the lower beds but not at all in the upper: Cladophlebis, Lastrea, Siphonites, Ficus, Eucalyptus, and Aralia.’? Turning to the western area we find that the Puget formation consists of some 10,000 feet of arkoses and intercalated carboniferous shales repre- senting the Eocene. Besides the above named, Knowlton describes Quercus, Juglans, Rhamus, Populus, and. Laurus from that series. Of these Miss Duror reports Juglans, Populus, and Laurus, and the presence of Ficus and Sabal correlates the Swauk with the lower Puget (Carbonado). Miss Duror’s report further proves the equivalency in age of the Swauk with the Fort Union of Mon- tana, North Dakota, and Wyoming.2 The Swauk sedimentary series may then be correlated with the lower Puget of western Washington, with the Swauk (lowermost Eocene) of eastern Wash- inton, and with the Fort Union areas farther east. MIOCENE Keechelus—This series of rocks of volcanic origin comprises tuffs, sheets, and dikes usually of andesitic but less frequently of dacitic composition. The tuffs predominate strongly and form beds of unknown but considerable thickness widely distributed in the Skykomish Basin. They overlie the Swauk sandstone, and the t Folio 106, U.S.G.S., 1904, p. 5. 2 Folio 54, U.S.G.S., 1899, p. 3. 3 A. G. Leonard, ‘‘ Cretaceous and Tertiary Formations,” Jour. Geol., XTX (1911), 541-43. STRATIGRAPHY OF THE SKYKOMISH BASIN 567 dikes have baked the sandstones. The Keechelus is therefore post-Swauk in age. Contrariwise, the Keechelus series has been indurated and otherwise metamorphosed by the Snoqualmie batholith to the southward; in the Snoqualmie area the Keechelus series is underlaid by a sedimentary series of sandstones and water- laid pyroclastics (Ellensburg) which contains flora called Upper Miocene by Knowlton. The Keechelus series in the Skykomish Basin, being really continuous and lithologically identical with the series of the southern area, is therefore considered Miocene in age. As has been pointed out by Smith and Calkins, a striking chemical similarity exists between the Keechelus andesite and the subjacent body of granodiorite. Both fall into the same chemical classification—tonalose. It is immediately inferred that the two rocks are consanguineous, or, in other words, that the andesitic pyroclastics were blown out from a magma that later solidified as the Snoqualmie granodiorite. We must assume 2,000 feet or more of cover, and it is suggested that this Keechelus series may very well have provided at least a part of the cover which has later been removed by processes of erosion. No estimate can be made of the thickness of the Keechelus series. It is undoubtedly widely variable and probably had a thickness of several thousand feet. \ Snoqualmie granodiorite—Into the Keechelus series was in- truded one of the younger of the known great batholithic intrusions. It has a known length of major axis of about thirty miles and is approximately two-thirds as broad. Throughout, this terrane is a massive, fresh, granitoid igneous rock which has been discovered by erosion to a vertical depth of 5,000 feet or more. It has, as was seen above, metamorphosed rocks of late Miocene age, and has sent apophyses into them, and therefore must itself be Miocene or later in age. To account for its holocrystalline nature and for the fact that it has been peneplaned, uplifted, and maturely dissected, it seems necessary to put the date of its intrusion as near that of the Keechelus as possible. The age is therefore given as late 1G. O. Smith and W. C. Mendenhall, ‘‘Tertiary Granite in Northern Cascades,” Bull. Geol. Soc. Am., XI (1900), 224; G. O. Smith and F. C. Calkins, Folio 139, U.S. G.S., 1906, p. 8. 568 WARREN S. SMITH Miocene, in accordance with Smith’s interpretation.t The only metamorphic effects experienced by this rock are the slight clouding of feldspars and the formation of a system of joints. There is a southernmost corner of another batholith, or, more likely, of a subjacently connected continuation of the same batholith, north of Grotto, and it is probable that these Tertiary batholiths form the core of the Cascade Range throughout the northern half of the state. Daly has correlated several of his batholiths with the Snoqualmie batholith. PLIOCENE There is no stratigraphic evidence of Pliocene history. It is physiographic rather. We infer that the Snoqualmie batholith had a cover in excess of 2,000 feet in thickness. This Cover was removed and the entire area reduced to one of low relief in late Miocene and post-Miocene time. In the Pliocene the area was uplifted with a broad arch of north-south trend and with certain minor warpings of transverse trend. Subsequent to this uplift, but still in the Pliocene, the area was maturely dissected by steam action. This process of peneplanation, uplift, and mature dissection is evidence of the very considerable duration of the Pliocene. PLEISTOCENE This is the age of glacial occupancy, when glaciers of the alpine type filled the valleys to a depth of several thousand feet and flowed down to their confluence with the Piedmont glacier of Puget Sound. Evidence has been put forward by many writers of two periods of glacial advance in the Puget Sound Basin, but of course the last alpine glacier to occupy the valley would have destroyed all evidence of any previous glaciation, and it can only be said that the Skykomish Basin was maturely dissected by glaciers of the alpine type in Pleistocene time. Comparatively little time has elapsed since the glaciers withdrew from the valley. The only stratigraphic evidence of this period is the accumulation to a depth of several inches of a volcanic ash which has only a 1G. O. Smith and W. C. Mendenhall, ‘‘Tertiary Granite in the Northern Cas- cades,”’ Bull. Geol. Soc. Am., II (1900), 201-28. STRATIGRAPHY OF THE SKYKOMISH BASIN 569 slight soil covering at present—an evidence of recent volcanic activity in the near-by volcanic cones. RESUME Eocene time witnessed the accumulation of 4,000 feet of arkose sandstone. This was orogenically disturbed so that it now dips at considerable angles. In the Miocene a series of volcanic tuffs and andesitic intrusives were derived from a magma which ap- proached the surface and cooled as a batholith in late Miocene. Pliocene saw the formation, uplift, and mature dissection of a peneplane. The Pleistocene was a period of glaciation lasting nearly to the present, in which the Skykomish area was maturely dissected by ice erosion. IV. STRUCTURE A glance at the map shows the tendency of all formations to trend north-south. Only the Swauk arkose series deviates from this tendency, and even this formation tends to assume a normal relation in its northward outcrop. The great igneous terrane (Miocene batholith) has its major axial trend in a direction parallel to the north-south axial trend of the Cascade Range. JOINTING There are two known systems of joints of considerable impor- tance and one or more of less prominence. The first two strike N. 45° E. and N. 70° E., respectively, and the lesser system strikes N. 80° W. It is suggested that these joints are the result of pres- sure exerted by orogenic forces in raising the Cascade peneplane to its present position. If this pressure were exerted continuously in an east-west line from the Pacific side, we should anticipate a set of joints striking N. 45° E. and a lesser set striking N. 45° W. At least one important set of joints does strike N. 45° W. in the Cleopatra Mine. But it is necessary to postulate a change in the direction of application of the force to account for the system of joints striking N. 70° E. and N. 80° W.? It seems possible that such a change may have taken place in the direction of application t A. Daubrée, Géologie expérimentale (1879), pp. 316 f. 2G. F. Becker, ‘‘Finite Strain in Rocks,” Bull. Geol. Soc. Am., IV (1893), 23. 57° WARREN S. SMITH of orogenic pressure, and this change may account for the trans- verse warping of the surface of the peneplane which is noted else- where. The importance of the development of these structural relations has an important influence on ore deposition (Fig. 2). Fic. 2.—Kimball Creek. Note jointing. Rock: granodiorite REPORT ON THE FLORA OF THE SWAUK SERIES The flora of the beds is of Fort Union age. The material examined yields fifteen genera and nineteen species, of which two are new; the other seventeen species have been recorded from STRATIGRAPHY OF THE SKYKOMISH BASIN 571 Fort Union beds. Three ferns are represented by abundant speci- mens of Asplenium and Pteris in the shales of F 831 and F 831-+-50. Genus Asplenium Asplenium magnum var. intermedium var. novum (Duror) (Fig. 3) Knowlton: ‘Fossil Flora of Yellowstone Park,” extract, Mono. XXXII, U.S.G.S., 1899, Part 2, p. 667; Pl. LX XIX, Figs. 8, 8a. Heer: Flora Foss. Arct., Vol. IV, ‘“‘Ostsibriens,” Taf. XX (A. whitbiense). This form is so named because it is intermediate in character between A. magnum (Knowlton) and A. whitbiense (Heer). The frond is not pinnate, but the lobes are cleft one-half to one-third of the distance to the rachis. The margin of the lobes is entire. These lobes are one and one-half to twice as long (along midvein) as broad and come to a rounded point. Secondary veins come off at an angle of about 45°, members of each pair being almost opposite. Each secondary bears eight to ten pairs of tertiaries, which are generally once forked. A few rare cases of simple secondaries and still fewer twice- forked secondaries were observed. The Fic. 3.—Asplenium variety differs from A. magnum of Knowlton ™agnum var. inter - in possessing almost deltoid instead of ovate Tea Nea ae i i ; (4 natural size.) lobes and in being rather larger. This form is separated from A. whitbiense of Heer, because here most second- aries fork once only, and because Heer’s form has true pinnules— the form being bipinnate. There is no sug- gestion that this form is fully even once pinnate. GENUS Pteris eee ae Pieris pennaeformis (Heer) (Fig. 4) formis. (4 natural size.) O. Heer: Fl. Pert. Helvetiae (1855), p. 38; Taf. XI, Fig. ra—d; Miocene age. Lesquereux: Tert. Flora (1878), Pp. 523 Pl. IV, Figs. 3, 4, Pseudopennae- formis, Lower Lignitic age. The Cascade specimens are quite similar to that figured by Heer, except that here only that part of the pinnae with entire 572 WARREN S. SMITH margins is seen. Secondary veins are seen to fork twice, as in Heer’s Fig. re, though from his description he found such nervation EALe: GENUS Sabal Sabal powelli (Newb.) Newberry: Proc. U.S. Nat. Mus., V (March 21, 1883), 504. Later Extinct Floras of North America, p. 30; Pl. LXIII, Fig. 6; Pl. LXIV, Figs. 1-19; Tertiary (Green River group) age of Wyoming. Palms are represented in these beds, chiefly in F 831-++50 and in F 865, by numerous perfectly preserved specimens of this type. In some cases both upper and lower surfaces of the petiole of one leaf were preserved. The forms agree in all respects with Newberry’s type. GYMNOSPERMAE Gymnospermae are represented by countless fragments of Glyptostrobus, Sequoia, and Taxodium-—this last in greatest abun- dance. The greatest number of these forms is found in the shales of F 831++50. GENUS Gly ptostrobus Gly ptostrobus ungeri (Heer) Heer: Flora Tert. Helvetiae, 1, 51; Taf. XIX, XX, Fig. 1; Taf. XLIX, Fig. 50. Newberry: (G. Europaens Brogn.): 1. Annals N.Y. Nat. Hist., IX (1868), 43. 2. Illus. Cret. and Tert. Plants (1808), Pl. XI, Figs. 6-8a. 3. Later Extinct Floras of N.Am., p. 24; Pl. XXVI, Figs. 6-8a; Pl. LXV, Figs. 3-4. Fic. 5.—A, Sequozva nordenskioldii; cone in cross- section (natural size). B, Asplenium cascadia. C,Tax- = ° : Re ceag Shaka a: No cones of this species were found, odium distichum miocenum. (3 natural size.) but the general form of these fragments is like the Figs. 1e and 1a of Taf. XX, and of Fig. 50 of Taf. XLIX of Heer. They also resemble quite closely those from Birch Bay, Washington, figured by Newberry (see 3 above) on Plate LV. Those were of Fort Union age. 573 STRATIGRAPHY OF THE SKYKOMISH BASIN GENUS Taxodium Taxodium distichum Miocenum (Heer), Fig. 5,C; Fig. 7, D. Heer: Miocene Baltische Flora (1869), p. 18; Taf. II, II, Figs. 6, 7. Fig. 3, in part; Pl. LII, Figs. 2, 3, 4; Pl. LV, Fig. s. Lesquereux: Tertiary Flora (1878), VII, 223; VIL, 73. Newberry: Later Extinct Floras of N.Am., p. 22; Pl. XLVII, Fig. 6; Pl. LI, 2 natural size) ( Fic. 6.—Laurus cascadia var. leve. D, Taxodium C, Sequoia nordenskioldit. Fic. 7.—A, B, B', Laurus cascadia. (4 natural size.) distichum miocenum. The form is Green- The Cascade specimens are quite typical. land Miocene or basal Eocene in age. 574 WARREN S. SMITH GENUS Sequoia Sequoia nordenskioldi (Heer) (Figs. 5, A, 7, C, 8, B) Heer: . Flora Foss. Arct., IL (‘‘Miocene Spitzbergens,” 1870), 36; Taf. II, Fig. 130; Taf. IV, Figs. ra, b, 4-38. Newberry: Later Extinct Floras of N.Am., p. 20; Pl. X XVI, Fig. 4. Sequoia nordenskioldi is represented by a few leafy branches occurring with the Taxodium,:but in addition by a handsome cone. The cross-section is given in Fig. 5. The form is referred to S. norden- skioldi rather than to S. langsdor fii because the leaves are very little if at all narrowed before they join and run down the stem. The cone is almost identical with that of Heer (op. cit., Taf. IV, Fig. 4a), except that these dimensions are FIc. 8.—A, Pterospermites whiter. 20X22 mm., not 16X13 mm., as B, Sequoia ease EES Gy Sapin- Tear gives and that here noileaeee dus obtusifolius. (% natural size.) ) remain on the branch. This cone is not as elongate as in S. langsdorfit. DICOTYLEDONS There are no identifiable monocotyledons found in these beds. Dicotyledons are represented by ten fairly well-preserved forms, and by fragments of many more. The genera are: Ficus, Juglans, Hicoria, Laurus, Magnolia, Populus, Protoficus, Pterospermites, and Sapindus. GENUS Ficus Ficus ungeri (Lesq.) Lesquereux: Supplement Ann. Rept., 1871, p. 7; Fig. 1. Hayden Survey, VII (1878), 195; Pl. XXX, Fig. 3. Numerous fragments in F 831+ 50 beds are referred doubtfully to this form. The open-bowed secondaries in almost opposite pairs and the “very entire” margin are identical. Noted from Green River group, Middle to Upper Eocene. STRATIGRAPHY OF THE SKYKOMISH BASIN 575 Ficus sp. ? (Knowlton) Knowlton: Rept. on Fossil Plants Associated with Lavas of Cascade Range (Western Oregon) (1898), p. 46; Pl. III, Fig. 1. In shales (F 831-++50) a specimen very similar to Knowlton’s figure, in the roundly notched margin, and the 45° angle of emer- gence of the secondaries, was found. Here also no base or tip was to be seen. GENUS Hicoria Hicoria (Carya) antiquorum (Newb., Knowlton) Newberry: Ann. N.Y. Lyc. Nat. Hist., 1X, (April, 1868), 72. Illust. Cret. and Tert. Plants (1878), Pl. XXIII, Figs. 1-4. Lesquereux: Later Extinct Floras of N. Am. (1868), p. 35, Pl. XX XI, Figs. 1-4. Tertiary Floras (1878), VII, 289; Pl. 1, Figs. 1-5; Vol. VIII, Pl. 1, Fig. 2. Knowlton: Tertiary Plants of N.Am. (1898), p. 117. Fossil Flora of Yellowstone Park, etc. This form is found in the sandy F 844 beds. It is referred to these species rather than to Juglans nigella, because the teeth are here rounded, as in Hicoria, and the secondaries are less prominent than in J. nigella. The leaf narrows gently toward the base and joins the stem (midvein) by a quarter-inch long winged “‘petiole.” The form is noted by Newberry and by Knowlton from Eocene beds (Planatus) at the mouth of the Yellowstone River. GENUuS Juglans Juglans acuminata (Heer) Heer: Flora Tert. Helvetiae, p. 88; Taf. CX XIX, Figs. 2-8. Flora Foss. Arct., V1I, ‘“‘Groénlands,” 761; Taf. LX XV. Knowlton: ‘Fossil Plants from Kukak Bay,” Alaskan Exp., 1V (1904), 1523 Pl. XXXIII, Fig. 3. The form is identified from fragments only, but these are rather numerous. The entire margin and the angle of the midrib and secondaries are very similar in these specimens and in Heer’s figures, but show considerable dissimilarity to Knowlton’s type, where the angle between the midrib and secondaries is larger and where the latter are alternately long and short and bowed. The form, according to Heer, “is spread through the whole Tertiary and possesses many synonyms.” 576 WARREN S. SMITH GENus Laurus Laurus cascadia N.Sp. (Figs. 6 and 7, A, B, B*) This specific name is given to several excellently preserved specimens in the shales of F 831. There is variation among the forms, but hardly more than to permit of naming of two varieties. The leaf is ovate lanceolate, coming to a sharp slender tip, and slightly unsymmetrical at the obtusely pointed base. There are six to eight pairs of strong secondaries, the members not directly opposite, besides a rather faint pair at the base, where the mem- bers are opposite. Secondaries come off at an angle of about 70° or less, as in Fig. 7, and are very gently bowed. The tertiaries show a horizontal parallel arrangement all the way across the leaf, so that they do not join the secondaries at right angles, except in the case of the two lower pairs. Fig. 6 shows preserved a fine network of veins between the tertiaries. The margin of the leaf is entire, and the ultimate veins border it in a series of loops. About one-quarter inch of petiole was found (Fig. 6). In Fig. 6 the breadth is 38mm., the length probably 1oomm. The two pieces figured are not parts of one leaf. In Fig. 7 the dimensions are 3280 mm., and here A and B are two sides of the impression of one leaf. A small form, very like Fig 6, but not drawn, was about 18X35 mm. L. cascada resembles L. similis of Knowlton (Rept. on Fossil Plants Associated with Lavas of Cascade Range of Oregon, Pl. V, Figs. 1 and 4) merely in the horizontal tertiaries and in the angle of the secondaries. Tip and base are, however, quite different in the two forms. The base of L. perdita comes nearest to being as blunt as that of L. cascadia. Laurus cascadia leve var. nova (Duror) and L. cascadia (type) differ largely, in that the former is proportionately broader and with somewhat heavier veins. GENuS Magnolia Magnolia nordenskioldi (Heer) Heer: Flora Foss. Arct., VII, ‘‘Groénlands,” 123; Taf. CVIII, Figs. 2, 3; IV. ‘‘Spitzbergens,”’ 82; Taf. LII, Fig. 1. This reference is made doubtfully on certain fragments from the sandstone of F 844. The size of the leaf, strength, and irregular STRATIGRAPHY OF THE SKYKOMISH BASIN Slo anastomosing of the secondaries of the veining are correct. The tertiarles, arising at right angles to the secondaries and dovetailing with the tertiaries from adjacent secondaries, are also similar to Heer’s figures. The form is noted from the Canadian Miocene— really Eocene. GENuS Populus Populus is represented by numerous more or less fragmentary remains, in which, however, many of the characters are fortunately plainly discernible. Three of Ward’s species are believed to be present, besides one of Heer’s. The three forms are noted from beds called Laramie, now considered of Fort Union age. Populus amblyrhynca (Ward) Ward: U.S.G.S. Bull. No. 37 (1887), 20; Pl. VI, Figs. 1-8. Reference of the forms to this species is made with great cer- tainty. In the Cascade form the base is somewhat flatter than in Ward’s figure, the identification resting chiefly upon the character of the thick tertiaries sent out from the inner side of the second pair of secondaries. The resemblance is closest to Ward’s Figs. 2 andor PI Vir. Populus cuneata (Newb.) Newberry: Later Ext. Floras, pp. 31, 64. Illus. Cret. and Tert. Plants, Pl. XIV, Figs, 1-4. Lesquereux: Cret. and Tert. Floras, p. 225; Pl. XLVI, Fig. 5. Dawson: ‘‘Cret. and Tert. Floras of Brit. Col. and N.W. Terr.,” Trans. Roy. Soc. Can., Sec. IV (1882), p. 32. Ward: U.S.G.S. Bull. No. 37 (1887), p. 19; Pl. IV, Figs. 5-8; Pl. V, Figs. 1-3. One specimen was so called because the first pair of secondaries here, as in the figures of P. cuneata, leave the midrib about 5 mm. up from the attachment of the petiole. Populus saddachi (Heer) Heer: Flora Foss. Arct., 1, 98; Taf. VI, Figs. 1-4; II, 468; Taf. XLIII, Fig. 15a; Taf. XLIV, Fig. 6. Flora Foss. Alaska, p. 26; Taf. Il, Fig. 5a. Lesquereux: Mem. Mus. Comp. Zodl., VI (1878), No. 12; Pl. VIII, Figs. 1-8. 578 WARREN S. SMITH This form is referred to this species because it is remarkable for its larger size. Judging from a fragment, the complete leaf was about 4X6 inches, as is that of Lesquereux in Fig. 8. There are three pairs of secondaries, the lowest very faint, the innermost very strong and straight. The margin is not preserved. Les- quereux notes this form from Upper Miocene, but one of Heer’s examples is Miocene of Spitzbergen (equal to Eocene). Populus artica (Heer) Heer: Flora Foss. Arct., IV, Taf. XXXII. The Cascade form is very questionably referred to this species. There is a single smooth, faintly veined leaf from beds F 831++50 whose base is almost cordate but otherwise agrees with P. artica, and even similar bases are to be found in Heer’s figures. Genus Protoficus (Saporta) Protoficus is represented by many beautifully preserved speci- mens from shales of F 831++50. They are apparently all of one new species. Protificus fossi N.Sp. (Figs. 9 and 10) The leaf is lanceolate, broadest just at the middle; the apex is a short, sharp point; the margin is irregularly, sharply dentate to one-third of the way to the base, then wavy to crenulate. The base is blunt to slightly tapering and not absolutely symmetrical, as seen in Fig. to. The midrib is straight and moderately strong, with seven pairs of secondaries. The members of each pair are not strictly opposite, except above, where they are strongly bowed outward. The nervation has a palmate aspect, since the two lowest pairs of secondaries come off at the top of the petiole, and the rest only above the lower half of the leaf, or even higher. Six pairs of tertiaries arise from the second pair of secondaries and form loops near the margin. Small nerves extend from these loops into the teeth. All the other tertiaries are at right angles to the midrib in parallel “ horizontal” rows. One small specimen measured Ir cm. in length by 4.5 in breadth (Fig. 10); the type (Fig. 9) is g cm. in width by 15 cm. in length. STRATIGRAPHY OF THE SKYKOMISH BASIN 579 The tertiary veination is similar to that of Ficus tiliafolia and there is also something of the same palmate look; but in Ficus the gap to the next pair of secondaries is only one-third as exag- gerated as here. The margin of FP. tiliafolia, moreover, is entire. Protoficus sellert of Lesquereux (Bull. Mus. Comp. Zool., XVI, No. 12 [1888], 50), while described as notably palmate, is more cordate at the base and the border is merely crenulate. This leaf is only 7X5.5 cm. = Ye KD YY LOG: <= tS i 4 =e) y = Kn y i <} Wy Fic. 9.—Protoficus fossi. (% natural Fic. 10.—Protoficus fossi. (% natural size.) size.) Protoficus inequalis (Newb.) (Proc. U.S. Nat. Mus. V [1882], 512) is not described as having a gap between the lower end and upper secondaries and is notably unsymmetrical at the base. The margin is merely undulate. GENUS Pterospermites Pterospermites whitet (Ward) (Fig. 8) Ward: U.S.G.S. Bull. No. 37 (1887), p. 94; Pl. XLI, Figs. 5 and 6. The species from the shales of F 831-++50 is in marked contrast to the foregoing Protoficus. The identification is rather certain, although the base and tip are wanting. The midrib is not quite so sinuous above as it is in Ward’s figures. The form is noted 580 WARREN S. SMITH from the Laramie of Montana, now conceded to be Fort Union in age. : GENUS Sapindus Sapindus obtusifolius (Lesq.) (Fig. 8C) Lesquereux: Hayden Surv., VII (1873), 266; Pl. XLIX, Figs. 8-11; VIII, 235; Pl. XLVI, Figs. 5-7. Knowlton: U.S.G.S. Bull. No. 204, p. 79. The form is represented by numerous fragments such as are shown in Fig.8,C. Better specimens (not figured) show the typical, unequal base and the alternately stronger and weaker secondaries. Similar forms are described by Knowlton from the Fort Union of Montana and North Dakota. The writer wishes to express her indebtedness to Dr. Arthur Hollick of the New York Botanical Gardens for valuable help and suggestions. REPORT ON THE FAUNA OF THE MALONEY SERIES Rafinesquina (Hall) deltoidea (Conrad) (Fig. 11, A, B, C, D), Leptaena del- toidea Conrad, Am. Geol. Rept., 1838, p. 115. Strophomena deltoidea Davidson, Foss. Brachiopoda, III (1864-71); Pl. XLII, Figs. 1-5; Pl. XX XIX, Fig. 22. Streptorhyncus (Strophonella) deltoidea (Hall, 1883), Sec. An. Rept. State Geol. of New York, Pl. XLII, Figs. 1-7. Rafinesquina deltoidea Hall, Pal. New York, VII; Part. I, p. 281; Pl. TXa, Figs. 1-5. (Same figures as in references above.) The Cascade specimens -are referred with slight hesitation to Rafinesquina deltoidea, notwithstanding the fragmentary nature of the material. There is a strong resemblance to Plectambonites (Leptaena) sericius (Sow.) var. rhombica, as figured in Davidson’s Fossil Brachiopoda, V, 169; Pl. XII, Figs. 4-7. However, the remains of the muscle scars, and more especially the concentric wrinkles, stronger near the hinge line, are characteristic of Rajfines- quina. The alternate striation, such as the Cascade specimens show, is described in both forms, “every fifth or seventh striation markedly stronger.’’ Here generally every fifth striation, though rarely every seventh, and occasionally every second, is emphasized. Such a variation, and another in the profile of the shell, is noted STRATIGRAPHY OF THE SKYKOMISH BASIN 581 by Professor M’Corg (quoted in Hall, op. cit.). In these specimens the most convex shell (Fig. 11, A, B) is not sharply flexed at any one point, while younger individuals, as in Fig. 11, C, show less arching, but a more sudden geniculation. The specimens are partly exfoliated and have a finely punctate surface. Fig. 11, A, B, and D are of pedicle valves, and Fig. 11, C of a brachial valve. This last specimen is partly an impression of the Fic. 11.—Rafinesquina (Hall) deltoidea (Conrad). A, B, and D, pedicle valves; C, brachial valve. outer surface of the shell, since the striations show as grooves; but part of the true shell remains where the muscle scars are shown. Illaenus americanus (Billings) Billings (1859, quoted; 1865, quoted and figure copied): Paleozoic Fossils, i329; Fig. 216a—d. Winchell and Ulrich: Minn. State. Surv., III (1897), Part II, 714; Figs. (from above) 20, 21, 22, 23. Fic. 12.—Illaenus americanus (Billings). A, front view; B, top view; C, side view. A single specimen, the glabella and fixed cheeks, was found associated with the Rafinesquina deltoidea. All the features on these parts agree perfectly with Billings’ figures. This specimen probably belonged to an animal one inch long, while Billings’ 582 WARREN S. SMITH forms ranged from two to three inches in length. Naturally, smaller forms have been noted, as in Grabau and Shimer, Index Fossils, p. 295, where one and one-fourth inches is given as the entire length. Fig. 12, Ais the front, Fig. 12, B the top, and Fig. 12, C the side view of the one specimen. The dotted line in Fig. 12, B gives the outline of the eye, which was destroyed in uncovering the fossil. Both of these forms are index fossils of Trenton age. Naturally, they are not restricted to one bed, but in no case have they been recorded as younger than Ordovician. The writer is indebted to Dr. Shimer and to Dr. Grabau for valuable help and advice. “PUFF” CONES ON MOUNT USU Y. OINOUYE Imperial Tohoku University, Sapporo, Japan Two days before the eruption of Mount Usu, in southwestern Hokkaido, Japan, the writer arrived at the foot of the volcano. He remained there for twelve days, watching every phenomenon, going without sleep the first five days. The first explosion occurred HOKKAIDO * SAPPORO bit ust Fic. 1.—Sketch map of southwestern part of Hokkaido on July 25, 1910, and others followed in rapid succession. Violent eruptions ceased in about ten days and the writer returned to Sapporo on August 7. He again visited the volcano in September and in December of the same year, in May and October, 1911, in 583 584 Y. OINOUYE May, 1912, and in May, 1913. Many interesting facts were ob- served which will be published later in another paper. Here attention is to be called only to certain peculiar cones formed on one of the mud flows. The main eruption of the volcano caused the formation of forty-five small explosion craters on its northern slope. These craters extend from east to west in two zones along Lake Toya," north of the volcano. During the first few months after their formation innumerable bombs and considerable quantities of sand and ashes were blown from every craterlet. From five of them mud flowed at different times, the flow from a small crater at the southern foot of the parasite cone Nishi-Maruyam being especially interesting. This crater is located on a gentle slope of about five degrees, and is 100 meters in diameter. For twenty days it intermittently threw out columns of hot water, occasionally mingled with mud, to a height of about 60 meters. Approximately two hundred eruptions occurred per day at intervals of from three to thirty minutes. A mass of mud, estimated by the writer at 230,000 cubic meters, spread out in a sheet averaging 1.5 meters in thickness, over an area 200 by 700 meters. It covered a farm, where it destroyed a thousand apple trees and other crops, and pushed three houses in the direction of the lake and finally destroyed them. The mud consists mainly of plagioclase, hypersthene, augite, magnetite, and hematite, and resembles the material of the sea sand at the west foot of Mount Usu. It differs, however, in also containing fragments, from the size of peas to that of nuts, of com- pact gray to coarse black andesite. These fragments are not usually exposed at the surface of the mud, having sunk on account of their greater size. : The materials thrown out by the crater were highly heated and sticky at the time of their eruption, and contained a great amount of water and gas. For several months the flow continued steaming, but as time passed and the moisture and gases became exhausted, it ceased, and the mass became harder and harder. A year after * Lake Toya is a depression lake, according to T. Kato (Report Earthquake Investi- gation Committee, Vol. LXII). “PUFF” CONES ON MOUNT USU 585 the eruption the surface of the flow was so hard that it was difficult to discern footprints upon it, and specimens could be obtained only with the aid of a pointed stick ora hammer. At this time the surface was flat except for low, wavy undulations and very irregular sun cracks. A year later the writer found the flow covered with thousands of small cones, each of which had an opening which was compara- BAY of VoLCANO Crater lets 5 Mud flow Fic. 2.—Map of Mt. Usu and vicinity, showing the position of craterlets and mud flow here described. tively large but of no particular shape. The cones were of different sizes, the smallest being 0.5 meter in diameter and o.1 meter in height while the largest was 3.0 meters in diameter and 1.5 meters in height. They were irregularly arranged on the flow at intervals of to to 30 meters, and were either dome-shaped or resembled a common bell with a slope of forty degrees. The cause which produced these elevations is the same as that which forms small pitted cones when any viscous substance is boiled, namely, the escape of gases or vapors through the mass 586 VY. OINOUVE and the breaking of the bubble at the surface. After the cessation of the mud flow the surface dried and sun cracks were formed. The gases near the surface rapidly escaped through these openings, but those imprisoned near the bottom of the mass were unable to do so, the upper part only having dried out. Later, by the coales- cing of the small bubbles, the remaining gases united beneath the surface in reservoirs of greater size. The accumulated pressure finally became great enough to force a passage through the mud to the surface, the sudden escape of the gas forcing the mud upward Fic. 3.—The largest “puff”? cone. Photo taken by the writer, May 16, 1912 to form cones. ‘The other mud flows in this district, being thinner, dried out more rapidly, and no cones were formed. The writer has been unable to find descriptions of any such phenomenon in the case of other mud flows, although similar elevations occasionally occur on lava flows. He therefore suggests the name ‘“‘puff cones.”’ No new cones were formed after the summer of 1912, the greater part of the gas having been expelled. Since that time weathering has begun to reduce the slopes, so that, in all probability, no trace of this fantastic phenomenon will remain after a few years. ORIGIN OF FOLIATION IN THE PRE-CAMBRIAN ROCKS OF NORTHERN NEW YORK? WILLIAM J. MILLER Northampton, Massachusetts INTRODUCTION Data bearing upon the problem of the origin of foliation in the pre-Cambrian rocks of northern New York have been gathered dur- ing the last ten years by the writer while he was engaged in the geological surveys of various quadrangles in the Adirondack Mountain region. In the attempt to explain the origin of the foliated structures of the rocks, examples and analogies from other parts of the world will be introduced, and it is hoped that the con- clusions reached may have a wider application than to the Adiron- dack region alone. This paper is not much concerned with criteria for the deter- mination of original igneous or sedimentary character of the rocks. The conclusions reached are almost wholly based upon observations made upon rocks which have been generally recognized as quite certainly either igneous or sedimentary. Rocks of rather doubtful origin are frequently met in minor quantity, but these may be disregarded in the present discussion. The strata all belong to the very ancient Grenville series, includ- ing various gneisses and schists, together with crystalline limestone and quartzite. The chief criteria for the determination of their sedimentary origin are: distinct banded structures, often showing alternating layers of widely different composition sharply separated from each other; presence of extensive bodies of limestone and quartzite interbedded with the gneisses; dissemination of graphite flakes through many of the rocks; and the very common occurrence of garnet in many of the rocks, and the less common occurrence of sillimanite. t Published by permission of the Director of the New York State Museum. 587 588 WILLIAM J. MILLER The metamorphic rocks of igneous origin, given in regular order of geologic age, comprisé the anorthosite series, the syenite-granite series, and the gabbros, all of which show quite varied degrees of metamorphism. All are intrusive rocks and younger than the Grenville. Among the criteria for recognizing their igneous origin are: preservation of original rock textures, such as the porphyritic and the diabasic; relative homogeneity in large bodies; common occurrence of distinct inclusions of older rocks; intrusive contacts, often with dikes from the large bodies penetrating the older rocks; very common occurrence of zircon and zoisite in fresh, well- crystallized grains. THE GRENVILLE AND ITS FOLIATION FOLDING OF THE GRENVILLE Character of the Grenville series —The Grenville series comprises the oldest rocks of the Adirondack region, and they: are, in fact, among the most ancient known rocks of the earth’s crust. They consist of a great mass of thoroughly crystallized sediments, such as limestones, sandstones, and shales which have been changed to crystalline limestones, quartzites, and various gneisses and schists. A more or less well-developed foliation is always parallel to the stratification surfaces which are usually distinctly preserved in spite of the crystallization. Granulation is not common. Gren- ville strata are well represented throughout the Adirondack region, their distribution being very irregular or ‘‘patchy”’ in small to large areas. They are considerably less extensive than the later (intru- sive) syenite-granite series, which latter, together with the Gren- ville, makes up the great bulk of Adirondack rocks. Neither top nor bottom of the Grenville series is known, though thicknesses of from 10,000 to 20,000 feet are actually shown in single sections, and the total thickness is doubtless much greater. Adams and Barlow report a Grenville section about eighteen miles thick in Ontario. The strata are often tilted at high angles or very moder- ately folded, and sometimes locally contorted. There is a general tendency toward a northeast-southwest strike of Grenville masses in the Adirondacks, but there are many important exceptions. FOLIATION IN THE PRE-CAMBRIAN OF NEW YORK 589 Grenville series generally regarded as highly folded and com- pressed.—It has been quite generally assumed by all (including the writer) who have carried on geological work in the Adirondack region that the Grenville strata have been severely compressed and folded as well as thoroughly metamorphosed and foliated by the compression. A few citations from the more recent publications will illustrate the ideas usually held. “There is abundant proof that the rocks have undergone great compression and have been folded and faulted on an extensive scale.”* The Grenville rocks “have been greatly compressed and intricately folded and pli- cated.’ ‘The old sedimentary rocks have undergone complete recrystallization, entirely obliterating their old textures, and, as a result of severe compression, have had a development of cleavable minerals along certain parallel planes, the mineral particles having a common orientation.’’3 ‘In pre-Potsdam time the pre-Cambric sediments had been tremendously folded and faulted and intruded at great depths.’* “After the intrusions the whole region was sub- jected to intense compression and metamorphism when the gneissic or foliated structure of all the rocks was developed.’’s An alternative hypothesis—That the Adirondack Grenville strata are more or less folded is admitted at the outset, but, in the light of recent studies, the writer doubts the interpretation of the folded, tilted, and foliated structures as due to intense lateral compression. Certain evident features of the Grenville strata and related intrusives are directly opposed to this interpretation, while all of the structural features may be much more satisfactorily explained in another way. Thus it is conceived that the originally horizontal, or at most only very moderately folded, Grenville strata were much broken up and tilted in masses great and small, and in other cases actually domed, by the irregular upwelling of the great bodies of magma (especially syenite-granite) under only very moderate lateral pressure. This alternative explanation will be 1D. H. Newland, New Vork State Mus. Bull., No. 111, 1908, p. 20. 2H. P. Cushing, zbid., No. 145, 1910, p. 9. 3 Ibid., No. 95, 1905, p. 400. 41. H. Ogilvie, 2b¢d., No. 96, 1905, p. 478. 5 W. J. Miller, zbzd., No. 170, 1914, p. 77. 590 WILLIAM J. MILLER developed at some length in its application to the Adirondack region. Evidence against intense folding of the Grenville——In spite of the assumption of severe lateral compression, no large-scale example of intense folding of the Grenville has ever been positively demon- strated in the Adirondacks, and this in the face of the fact that many hundreds of square miles have been mapped in detail. Describing the Grenville structures of the Elizabethtown—Port Henry quad- rangles, Kemp says: ‘The dips are prevailingly moderate and the ancient sediments appear to have been folded or tilted to only a moderate degree.”* Regarding the Long Lake quadrangle, Cushing says: ‘Nearly east and west strikes prevail, and the prevalent dip is southward. This either indicates comparatively little fold- ing, or else isoclinal folding, or else that the foliation does not coincide with the bedding and so does not bring out the folding. It is not possible to demonstrate which of these alternatives is the true one, though the second is very unlikely, and all the direct evidence obtainable is against the third.’ He also states that in the largest Grenville belt ‘the dips are so flat that they can seldom be made out with certainty.” The writer’ has described a structure section in the Broadalbin quadrangle four miles long across the strike of Grenville strata with dips of 20-30° to the southeast. The. exposed thickness of Grenville is about 10,000 feet with no repetition of beds due to possible isoclinal folding and no field evidence for profound faulting. Another Grenville section recently described by the writer? in the North Creek quadrangle is five miles long with a pretty uniform dip of from 40° to 50°, thus showing a thickness of some 18,000- 20,000 feet of strata. There is no evidence of repetition of strata by either folding or faulting. The Grenville is extensively devel- oped throughout this quadrangle, and all the available evidence points to only moderate deformation of the strata either by tilting or slight folding. tJ. F. Kemp, New York State Mus. Bull., No. 138, 1910, p. 85. 2H. P. Cushing, zbid., No. 115, 1907, p. 485. 3 W. J. Miller, zbzd., No. 153, roti, p. 13. 4 Ibid., No. 170, 1914, p. 15. FOLIATION IN THE PRE-CAMBRIAN OF NEW YORK 5091 According to Cushing, “the foliation strike over much of the Saratoga quadrangle is nearly east-west, and the dips are to the south and rather flat, seldom reaching 45°. As elsewhere, a great monocline of the rocks is suggested, and, as elsewhere, this makes a Grenville succession of enormous thickness, so thick as to suggest caution in the interpretation of the structure, and as to emphasize the probability of the alternative suggestion that the rocks are closely pinched and folded in a series of closed, overturned folds.’ It is, however, by no means necessary to assume that such common occurrences of monoclinal dips may be due to isoclinal folding. The breaking up and tilting of many blocks or belts of Grenville strata into general parallelism with the upwelling bodies of magma could quite conceivably have taken place under only very moderate lateral compression at most, and, in such cases, monoclinal dips are Just what would be expected. This matter will be more fully discussed below. In the Little Falls, Remsen, Port Leyden, and Lake Pleasant quadrangles, which are also mapped in detail, the Grenville is only sparingly represented, but none of the field evidence points to profound folding of the strata due to lateral compression. The recent (1913-14) survey of the Blue Mountain quadrangle by the writer has thrown important light on the structure of the Grenville series which is there extensively represented. The great Panther-Snowy mountain mass (altitude 3,900 feet) of syenite occupying the southern portion of the Blue Mountain and the northern portion of the Indian Lake quadrangles is completely bounded on the west, north, and northeast by an unbroken belt of Grenville (mostly limestone) whose strikes and dips show it to lap up on the flanks of the mountain mass of igneous rock for many miles. The curving strike of the igneous rock is also essentially parallel to that of the Grenville. It is evident that we have here a large-scale example of the raising or doming of Grenville over the surface of the great body of uprising magma, the general cover having been removed by erosion, leaving only the circumferential belt of Grenville strata. The higher portions of the syenite now rise fully 2,000 feet above the Grenville. This large-scale tilting tH. P. Cushing, zbid., No. 169, 1914, p. 30. 592 WILLIAM J. MILLER of Grenville strata is certainly not due to severe lateral compression, nor is there, in any part of the quadrangle, evidence of highly folded or compressed Grenville strata. In the northwestern part of the Thirteenth Lake quadrangle the writer has examined Chimney Mountain, which is a mass of granitic syenite rising fully goo feet above a valley on the west. Perfectly bedded Grenville rocks with dip of 50° lap over the whole western face of the mountain of igneous rock, and it seems certain that the tilt of the strata was produced by the rise of the magma. We are thus led to conclude that none of the published Adiron- dack geologic maps or available data afford any reason to believe that the Grenville strata were ever profoundly folded or compressed. There is, however, much tilting on large and small scales and some very moderate folding. Such structures may be readily accounted for simply by the irregular intrusion or upwelling of great bodies of more or less plastic magma which broke up, tilted, and lifted or domed the masses of Grenville. Grenville structure in the Thousand Islands and Ontario regions.— The Thousand Islands district forms the connecting link between the Adirondack and Canadian pre-Cambrian areas, and lies to one side of the region discussed in this paper. Having recently studied _the Thousand Islands district, Cushing says: ‘The Grenville beds are now found for most part in highly inclined condition, dips of less than 45° being relatively rare, while those approaching verticality are common. . . . . It has also been shown that the dip is not everywhere in the same direction, but that, with the general direction of strike to the northeast-southwest, the dip, while prevalently to the northwest, becomes at times southeast. . The highly tilted condition of the rock series, and the changing dips seem certainly indicative of folding.’ He then describes a prominent belt of Grenville strata which he believes has a synclinal structure. But, accepting the existence of this syncline, does such a structure prove the region to have been sub- jected to an intense force of compression? Large bodies of granite bound this Grenville belt on either side, and it is quite conceivable that the uplifting effect of the intruding masses, possibly accom- *H. P. Cushing, New York State Mus. Bull., No. 145, 1910, Pp. 109. FOLIATION IN THE PRE-CAMBRIAN OF NEW YORK 593 panied by some crowding or squeezing of the Grenville between the igneous masses, may have produced this very structure. Cushing also argues that “the general parallelism of the foliation of all the pre-Cambric rocks” affords “evidence of thoroughgoing compres- sion of much later date” than the granitic intrusions. But, as will be shown below, such parallelism of foliation is not necessarily due to severe lateral compression. It should be said, however, that in the Thousand Islands region the granitic and Grenville rocks do seem to be more strikingly arranged in parallel northeast- southwest belts than is usual throughout the Adirondacks. It is possible that considerable orogenic forces did operate across the area from the Thousand Islands region northward into Canada, where also the parallelism is notable. Recent study of the Canton quadrangle seems to indicate considerable folding there. Adams and Barlow, in their description of the Haliburton and Bancroft areas, state that the batholiths ‘“‘are elongated or arranged in lines having a prevailing direction of about N. 30° E., to which direction the strike of the rocks (Grenville) lying between the batholiths in general conforms. This direction constitutes, so to speak, the general strike of the country, and shows that its present structure has been determined, not only by the rise of granite magma, but by the presence of a second factor in the form of a tangential pressure, acting simultaneously.”* But it is not at all certain that this tangential pressure was really orogenic in character. Even a very moderate compressive force, not at all sufficient thoroughly to fold and plicate the rocks, acting upon the rising magmas would readily account for all the structural phenomena now visible. Variation of foliation strikes.—Even if we grant a very consider- able lateral compression in the Thousand Islands—Canadian region, the Adirondack area, fully a hundred miles across and to the south- east, does not necessarily come under the same category. In fact, while parallelism of syenite-granite and Grenville rock belts and foliation are common in the Adirondacks, there are so many important variations from a northeast-southwest strike that any generalization regarding such a strike of the rock belts is of little t Adams and Barlow, Geol. Surv. Can., Mem. 6, 1910, p. 16. 594 WILLIAM J. MILLER significance. A glance at the accompanying sketch map (Fig. 1) will emphasize the fact that various large areas show strikes dis- tinctly out of harmony with a northeast-southwest structure. The unpublished Lake Placid geologic map shows exceedingly variable strikes. Many variations also occur within most of the other quadrangles, more especially the North Creek, Long Lake, and Blue Mountain (unpublished). Within the Lake Pleasant quadrangle, the foliation strikes relatively uniformly northwest- Fic. 1.—Sketch map of the Adirondack region showing generalized strikes of foliation within those quadrangles which have been mapped in detail. southeast or just at right angles to the assumed force of compres- sion of the region. Papers by Professor Kemp and assistants in the thirteenth, fifteenth, seventeenth, eighteenth, and nineteenth annual reports of the New York state geologist contain many very variable strike observations in the eastern Adirondacks aside from the quadrangles of the accompanying sketch map. It thus seems clear that the Adirondack rocks show strikes which could not pos- sibly have been produced by a severe lateral pressure exerted across FOLIATION IN THE PRE-CAMBRIAN OF NEW YORK 595 the whole region, for any such pressure, great enough to produce close folding, would have produced a high degree of parallelism of strikes throughout the region. Local contortions.—Local contortions or sharp folds in the Gren- ville strata are by no means uncommon, being especially prominent in the limestones and closely associated hornblende and pyroxene gneisses. Such plications have usually been regarded as strong evidence for large-scale folding, being thought of as minor folds superimposed upon large-scale folds. Now, in the first place, it is the writer’s experience that such local contortions or plications are very largely confined to the limestone beds, which are easily the most plastic of all Adirondack rocks. In the second place, the crowding of a batholithic magma against the invaded Grenville strata, or the catching of a mass of Grenville between two batho- lithic magmas, would readily account for more or less local contor- tions or even puckering of strata without any assumption of orogenic or severe lateral pressure exerted throughout the region. The shouldering action of the upwelling magmas must have produced rather severe local pressures. Regarding the Glamorgan batholith of Ontario, Adams and Barlow say that the Grenville rocks form- ing the periphery on several sides, “being squeezed between this and the adjacent batholiths, are too highly contorted .... to display the prevailing dip distinctly.”* Evidently such structures do not necessarily call for severe regional compression. Summary.—To summarize, there is no known evidence within the Adirondack region that the Grenville strata have ever been highly folded or severely compressed, while many broad Grenville belts are known to be only very moderately folded, and many masses, large and small, are merely tilted or domed at various angles. Very locally the strata are sometimes contorted or plicated. The structural relations are therefore best explained as having been the result of slow irregular upwelling of the more or less plastic magmas, probably under very moderate compression, whereby the Grenville strata, previously deformed very little or none at all, were broken up, tilted, and lifted or domed. The stratification suriaces of the Grenville were thus swung into general parallelism t Adams and Barlow, Geol. Surv. Can., Mem. 6, 1910, p. 15. 596 WILLIAM J. MILLER with the slow-moving magmatic currents. According to this view, | individual large blocks or belts of Grenville strata, or several such blocks or belts separated by intrusive masses, with strike of intru- sive masses parallel to Grenville stratification, would be expected to show monoclinal dips; some Grenville masses were shifted around in irregularly rising magmas so as to show various strikes according to directions of movement of the magmas, and hence would not be expected to exhibit monoclinal dips; some Grenville masses were merely domed over bodies of rising magma and would exhibit more or less quaquaversal strikes and dips; while still other Grenville masses were probably bent or even considerably folded into syn- clines by being caught between bodies of magma upwelling at about the same rate. Isoclinal or close folding on a large scale would scarcely be expected. In all of this discussion it is important to bear in mind that the Adirondack intrusives occupy a much greater extent than the invaded Grenville rocks, and that, in spite of their intrusive char- acter, they everywhere seem to occupy the position of a fundamental or underlying gneiss. It appears to have been literally true that. the Grenville strata were irregularly floated on a vast body of magma, the magma in many places having either arched up or broken through the strata. ORIGIN OF GRENVILLE FOLIATION We have just shown that the Grenville strata have never been highly folded or compressed. It is therefore necessary to explain the metamorphism of the strata on some other basis than that of subjection to severe lateral pressure. The old sediments are thor- oughly crystallized, and it is certain that they have been reorganized into new minerals under deep-seated conditions; that is to say, they have undergone anamorphic metamorphism. But evidently we are here dealing with a case of essentially static, rather than dynamic, metamorphism. Origin of parallelism of Grenville foliation and stratification.— The universal parallelism of Grenville foliation and stratification is a fact of prime importance. If the Grenville and accompanying great intrusives had been subjected to compression severe enough FOLIATION IN THE PRE-CAMBRIAN OF NEW YORK 507 to develop the distinct foliation, is it not remarkable that the strati- fication surfaces have never been obliterated and cleavage developed instead, and also that the stratification and foliation are always parallel? Now, the stratification of the highly crystalline Gren- ville is remarkably well preserved. Also, unless we assume intense isoclinal folding, so that mineral elongation could everywhere have taken place at right angles to the direction of lateral pressure, the parallelism of stratification and foliation cannot be accounted for by crystallization under severe lateral pressure. We have already shown, not only that there is no positive evidence for such isoclinal folding, but also that there is much positive evidence against any more than the tilting, or, at most, very moderate folding, of the Grenville on large scales. Again, if the foliation of the Grenville were essentially a dynamic -process—that is to say, the result of regional compression after the great igneous intrusions—why should the Grenville be notably less foliated and granulated than the intrusives? (See below.) We are thus forced to the only alternative conclusion, namely, that the Grenville foliation was developed during the crystalliza- tion of essentially horizontal strata under heavy load of overlying material. Those minerals which cause the foliation were elongated during crystallization under heavy downward pressure where con- ditions of warmth and moisture were also favorable. According to this conception the parallelism of foliation and stratification is precisely what would be expected. It is quite generally assumed that static pressure, that is to say, simple downward pressure, “to the amount exerted in the upper part of the earth’s outer crust, appears to have little metamorphic effect.”* In dealing with the very ancient Grenville, however, it must be remembered that the material now at the surface was once very deeply buried. The thickness of the Grenville series in the Adirondacks is at least some miles and more than likely many miles. Adams and Barlow? have recently estimated a thickness of nearly eighteen miles for the Grenville strata in Ontario. It is also definitely known that during pre-Cambrian time the Grenville strata were subjected to tre- tL. V. Pirsson, Rocks and Rock Minerals (1909), p. 335: 2 Adams and Barlow, Geol. Surv. Can., Mem. 6, 1910, p. 33. 598 WILLIAM J. MILLER mendous erosion when at least some miles, and quite possibly many miles, in thickness of materials were removed. Thus it seems clear that much of the Grenville rock now visible was once far more deeply buried than any known body of sediments since the beginning of the Paleozoic. Conditions of downward pressure and temperature were, therefore, more than usually favorable for static metamorphism. On the basis of static metamorphism it is not necessary to account for a high degree of metamorphism, because the Grenville series, though thoroughly crystalline, is mostly only moderately foliated with relatively little granulation, and with stratification generally well preserved. It may also be suggested as a possibility that actual crystallization did not begin until an early stage in the intrusion of the slowly upwelling magmas when additional heat for regional metamorphism was supplied. Evidence from other sources.—Experimental evidence is also suggestive in this connection. Thus, Becker and Day" have proved that crystals in general have a strong tendency to grow (or elongate) most rapidly at right angles to the direction of pressure. According to Wright,? cubes of glass formed by melting together wollastonite, diopside, and anorthite heated to the state of incipient crystalliza- tion under vertical pressure, showed, under the microscope, that the three minerals crystallized with long axes at right angles to the direction of pressure. Experimental evidence, therefore, strongly supports the possible development of elongated crystals in the Grenville sediments under conditions of static metamorphism. Van Hise has suggested, regarding the parallelism of foliation and bedding in the Grenville series, that ‘‘ vertical shortening and consequently horizontal elongation below the level of no lateral stress may have begun the process.”’ The writer views this as essentially the whole process, instead of assuming, as Van Hise did, that foliation parallel to bedding continued to develop under certain peculiar conditions when the rocks were subsequently folded. The explanation of foliation parallel to bedding is greatly simplified when it is not necessary to consider severe compression of the region. t Becker and Day, Proc. Wash. Acad. Sci., VII (1905), 283-88. 7F. E. Wright, Am. Jour. Sci., 4th series, XXII (1906), 226. 3C. R. Van Hise, U.S. Geol. Surv., 16th Ann. Rep., Part I, p. 773. FOLIATION IN THE PRE-CAMBRIAN OF NEW YORK 509 Describing the metamorphism of the Shuswap pre-Cambrian series in the Canadian Rockies, Daly says: ‘It is clear that the Shuswap series has not been seriously affected by dynamic meta- morphism. ‘The strata and most of the injected granites were completely or almost completely recrystallized while the strata lay nearly flat. In some localities the effects of dynamic meta- morphism have been superposed on those due to previous static metamorphism.’ Orientation of Grenville inclusions.—Another fact of importance in connection with the origin of Grenville foliation is the occasional occurrence of well-foliated inclusions of Grenville gneisses variously oriented in the great intrusive bodies. Twenty years ago, in St. Lawrence County, Smyth, noting irregular inclusions of black gneiss in granite, said: “The two foliations, that of the black masses and of the (granite) gneiss, range from parallel with, to perpendicu- lar to, each other.’”? He also noted a similar arrangement of Grenville laminated gneiss inclusions in syenite in Jefferson County. The writer has observed similar phenomena on small and large scales at various localities. Kemp has recently noted inclusions in massive anorthosite and says: “The foliation of the fragments runs in all directions, even in an area of a few square yards. The infer- ence is drawn that the Grenville gneisses were already strongly metamorphosed when the anorthosites entered.’ It is thus clear, in spite of the usual assumption to the contrary, that the foliation of the Grenville could not have been the result of lateral pressure brought to bear after, or even during, the intrusion of even the oldest oi the great igneous masses. General absence of granulation.—Another fact favoring a process of essentially static metamorphism as opposed to that of dynamic metamorphism is that the Grenville gneisses are, as a rule, com- paratively little granulated. Some rather local granulation is to be expected because of magmatic movements, especially where Grenville masses have been crowded against the upwelling magmas. The great intrusive syenite-granite series is very notably more tR. A. Daly, Geol. Surv. Can., Transcont. Excur. C 1, Guidebook 8, 1913, p. 132. 2C. H. Smyth, 15th Ann. Rep. New York State Geologist, 1895, p. 491. 3 J. F. Kemp, Geol. Soc. Am. Bull., XXV (1914), 47. 600 WILLIAM J. MILLER granulated than the Grenville series. How is this fact to be explained if both series have been subjected to strong regional compression after the intrusions ? THE SYENITE-GRANITE SERIES AND ITS FOLIATION Character of the syenite-granite series—The main bulk of syenites and granites in the Adirondacks are regarded by the writer as facies of a single great body intrusive into the Grenville, the intrusives being much more extensively exposed than the Grenville. Perfect gradations from basic (dioritic) facies of syenite to true granite are commonly shown, a quartz syenite being the prevailing rock. As re- gards granularity, structure, and mineral composition, themembers of the syenite-granite series are very variable. The granularity ranges from fine to coarse grain, with medium grain decidedly prevalent. A porphyritic texture is sometimes well developed. Granulation is common, especially in the more acidic rocks, the feldspars generally being notably more crushed than the other minerals. In structure the rocks range from very faintly gneissoid to very clearly gneissoid or sometimes almost schistose, the foliation being accentuated by the roughly parallel arrangement of the dark-colored minerals. The minerals, especially quartz and feldspar, often show more or less flattening or elongation parallel to the foliation. In general the more highly foliated rocks appear to be most granulated. In mineral composition the range is from dioritic types rich in plagi- oclase, orthoclase, pyroxene, and hornblende; to syenite rich in microperthite, orthoclase, and hornblende or augite together with some quartz and plagioclase; to granite rich in microperthite, quartz, orthoclase, and microcline together with some plagioclase, hornblende, and biotite. Various accessory minerals in smaller amounts also occur. The color of the typical fresh syenite is greenish gray which weathers to light brown, while the fresh granite colors vary from greenish gray to light gray and pinkish gray to almost red. In common with the Grenville, the syenite-granite foliation shows a tendency toward a northeast- southwest strike, with parallelism of syenite-granite and adja- cent Grenville quite common, though there are many notable exceptions. FOLIATION IN THE PRE-CAMBRIAN OF NEW YORK 601 Northeast-southwest structure of rocks.—We have shown that the Grenville series has never been closely folded or severely com- pressed, and that its foliation was not caused essentially by lateral pressure. ‘The syenite-granite masses, being younger than an intrusive into the Grenville, cannot, therefore, have had their foliation developed chiefly by lateral pressure, though the probable existence of a very moderate lateral pressure is admitted. In spite of many important exceptions, there is some tendency toward a general parallel northeast-southwest to east-west strike of Adirondack rock masses (Grenville and syenite-granite) and foliation. One view, clearly stated by Cushing, is that ‘‘since the rock [granite] solidified it has been subjected to compression, to- gether with the Grenville rocks, giving to each a foliation parallel to the other, and elongating the batholiths in a northeast-southwest direction.’”* At another place he refers to this compression as “‘thoroughgoing”’ and of much later date than the granite intrusion. Cushing suggests the possibility of the development of ‘“‘a similar and parallel foliation’ during the solidification of the batholiths due to their shouldering pressure exerted upon the adjacent rocks during the intrusion, but he says that if any such foliation de- veloped it was obliterated by subsequent compression. | The writer’s view is that the general northeast-southwest structural parallelism was brought about by just enough tangential compression to control the general directions of the upward- moving batholithic magmas. Accordingly, the intrusive bodies were more or less elongated during the process of intrusion, and there must have been a strong tendency for large and small bodies of previously horizontal, or only slightly deformed, Grenville strata to have been caught up and arranged with their long axes and foliation parallel to the magmatic currents, while the foliation of the intrusives would also have developed, as a sort of flow struc- ture under moderate pressure, parallel to the magmatic currents. This pressure was doubtless in part due to the shouldering effect of the intrusives upon the adjacent rocks. In other words, the syenite-granite gneisses are ‘‘primary gneisses.” Thus we should 1H. P. Cushing, New York State Mus. Bull., No. 145, 1910, p. 10. 2 [bid., No. 145, 1910, p. 102. 602 WILLIAM J. MILLER expect a general northeast-southwest strike of both rock masses and foliation of Grenville and intrusives to be of common occurrence. A statement made by Smyth twenty years ago regarding black gneiss inclusions in the syenite-granite series of St. Lawrence County is significant in this connection: ‘The parallel arrangement of the neighboring bands [inclusions] doubtless results from currents in the molten magma, which would tend to produce such a result. It is probable that the breaking into blocks resulted, in part, from strains applied after the magma was in a pasty and partially crystal- lized state. The blocks were more or less widely separated, and the intervening space was filled by the magma which flowed around the blocks without destroying their angular contour, and, at the same time, often produced an obscure flow structure in the gneiss parallel to the sides of the inclusions.”' The bandlike inclusions here described by Smyth are seldom more than a few rods long, but the writer believes the principles set forth are applicable on a much larger scale throughout the Adirondack region. Such parallelism of structural features does not, therefore, demonstrate that the rocks have been thoroughly compressed subsequent to the syenite-granite intrusions. The northeast- southwest structural features here referred to are more pronounced in the Thousand Islands region than is usual throughout the Adiron- dacks, and this may be readily explained by granting somewhat greater lateral pressure during the intrusion in the first-named region. In any case it is necessary to assume only very moderate compression—far less than would have been necessary to elongate the batholiths and develop distinct foliation in them after their complete solidification. Exceptions to northeast-southwest structure-—There are many exceptions to the general northeast-southwest structural arrange- ment, and these prove that no severe tangential compression could ever have been exerted throughout the region after or during the intrusions. Among such exceptions are sharp variations in strike of groups oi inclusions of well-foliated Grenville gneiss in the intru- sives. Examples have already been cited. If the whole region has been subjected to compression thoroughgoing enough to flatten «C,H. Smyth, r5th Ann. Rep. New York State Geologist, 1895, Pp. 491. FOLIATION IN THE PRE-CAMBRIAN OF NEW YORK 603 out batholiths and develop foliation after the consolidation of the magmas, how are these sharp variations in strike of Grenville inclusions to be accounted for? According to the writer’s view, such inclusions present no difficulties, because their foliation was produced prior to the intrusions, and some fragments, especially those caught up late in the stiff, nearly consolidated magmas with poorly defined currents, would not have been swung into parallelism in the uprising magmas. Also there are important exceptions to parallelism of foliation of adjacent syenite or granite and Grenville gneisses in relatively large areas. A few examples will suffice: eastern side of Port Leyden quadrangle where Grenville with north-south strike is surrounded with syenite with strike N. 30° E.; northwest corner of North Creek quadrangle (see geologic map); near northeast corner of Lake Pleasant quadrangle (see geologic map); northwest of Indian Lake Village; one mile west of Long Lake Village; and in the Broadalbin quadrangle where the large areas of Grenville and adjacent syenite-granite show very different strikes. If the foliation has been produced by compression after the intrusions, how are such sharp differences in strike to be accounted for? Granting the writer’s conception that the Grenville was foliated prior to the intrusions, and that the syenite-granite foliation was the result of magmatic flowage, it is to be expected that the mag- matic currents would occasionally have broken across the Grenville and its foliation. Very strong evidence against the development of foliation by compression of the great intrusives is the frequent occurrence of sharp variations in the strike of the foliation, often within short distances. Examination of the Long Lake, North Creek, and Lake Pleasant geologic maps, upon which foliation strikes are plotted, shows many strikes in granite or syenite ranging from parallel to right angles to each other, often within distances of a mile or two. Similar foliation variations occur upon the writer’s Blue Mountain and Lake Placid geologic maps, not yet published. How can such foliation variations possibly be explained as due to lateral pressure ? Ii due to compression of the whole region, should not the foliation always strike essentially at right angles to the compressive force ? 604 WILLIAM J. MILLER If, however, we regard the foliation as essentially a sort of flow structure, such phenomena are readily accounted for as due to local variations in the magmatic currents. Curving strike of foliation—Still another piece of evidence, though less commonly shown, is the existence of certain broad, sweeping curves in the foliation of syenite or granite. The North Creek, Lake Pleasant, and Long Lake geologic maps show such features. When larger areas of the Adirondacks are mapped in detail, it is probable that more and better examples will be brought to light. _ An excellent case of curving oi foliation on a large scale is in the Panther-Snowy mountain mass above described as extending nearly across the southern portion of the Blue Mountain sheet and the northern portion of the Indian Lake sheet. The great mass of syenite shows an almost perfect radiation of foliation dips from its center toward the we&t, north, and east. The only reasonable explanation of such an arrangement of dips is that the foliation was produced as a flow structure in the uprising magma, the most rapid currents having been toward the center of the mass. In the writer’s opinion, such a large-scale curved arrangement of foliation strikes and dips not only cannot possibly be explained as due to lateral pressure, since the foliation would then everywhere be practically at right angles to the pressure, but also conclusively proves that no severe compression ever affected the syenite. Nearly thirty years ago, in his study of the Rainy Lake region, Lawson described a somewhat similar curved foliated structure in granite gneiss and said regarding its origin: “‘ The simplest explana- tion that suggests itself to account for the structure is that of an uprising force acting on a plastic mass (pasty magma), such force acting with greatest intensity in the vertical line which would correspond to the axis of the cone or dome.’”” A similar type of structure appears to be common in the Haliburton-Bancroft area of Ontario as described by Adams and Barlow, who say: ‘‘Within the batholiths themselves the strike of the foliation follows sweeping curves, which are usually closed and centered about a certain spot. .... From these central aieas 1A. C. Lawson, Ann. Rep. Geol. Surv. Can. (N.S.), III (1887-88), 116. FOLIATION IN THE PRE-CAMBRIAN OF NEW YORK 605 of flat-lying gneiss the dip . . . . is generally outward in all direc- tions. The batholiths, therefore, are undoubtedly formed by an uprising of the granite magma, and these foci indicate the axis of greatest upward movement, and those along which the granite magma has been supplied most rapidly.’ There are not only important variations from the general northeast-southwest arrangement of the region within the quad- rangles themselves, but also much broader variations shown by a comparison of the average foliation strikes of all the quad- rangles of the Adirondacks which have been mapped in detail. This is graphically presented by the accompanying sketch map. Such marked differences in foliation directions on large scales throughout the Adirondacks is certainly incompatible with any idea of thoroughgoing compression of the region. Thus in the Lake Pleasant, North Creek, Blue Mountain, and Saratoga quadrangles the foliation, either wholly or largely, strikes at high angles across the general northeast-southwest strike of the region, while in the Lake Placid quadrangle the strikes are exceedingly variable. If due to compression, the foliation strikes would be much more nearly northeast-southwest than they actually are. Flow structure character of foliation.—Another significant feature of the foliation should be mentioned, namely, that, while all the minerals are arranged with long axes roughly parallel to the direc- tion of foliation, the dark-colored minerals which accentuate the structure most often appear as narrow, irregular, wavy streaks which are seldom continuous for more than a few inches or a foot. In the writer’s experience this type of foliation is by far the most common in the syenite-granite series, and it is believed to be the result of magmatic flowage. Lawson has noted an exactly similar phenomenon in certain granite gneisses of the Rainy Lake region of Ontario and says: ‘‘The lines of streaking are very often not straight but are wavy or contorted, sometimes intricately so, and are evidently due to flow movements in the magma prior to its final consolidation.”? As already suggested, flow structures are locally t Adams and Barlow, Geol. Surv. Can., Mem. 6, 1910, p. 14. 2A. C. Lawson, Geol. Surv. Can., Mem. 40, 1913, P- 93. 606 WILLIAM J. MILLER very distinctly developed, especially around some of the inclusions in syenite or granite. Differences in degree of foliation.—Another important considera- tion is the frequent pronounced variation in degree of foliation in the rocks of the syenite-granite series. They are mostly distinctly gneissoid, rarely so much so as to be almost schistose, while in other cases they are so faintly gneissoid as to be practically massive. A striking feature is the frequent rapid change within a few rods or yards, from rocks which are very clearly gneissoid to others in which the foliation is scarcely discernible. Sometimes, within a foot or two, a very gneissoid zone lies between others which are only moderately foliated. In many cases there is no evidence whatever of shearing to account for these variations. It seems impossible to conceive that such abrupt foliation changes could ever have been produced by severe compression of the rocks after solidification. Such compression would certainly have brought about a much more uniform degree of foliation. According to the writer’s view, these variations are best ex- plained as due to forced differential flowage in the pasty magmas, probably after partial consolidation. Regarding the origin of igneous rock foliation, Pirsson says: “Sometimes this texture has been imposed upon the igneous rocks after they had solidified, by intense pressure and shearing, and sometimes while they were still soft, pasty, and crystallizing, by forced differential flowage, due to various causes.’”* Those portions of the magma which were forced in probably a more fluid condition between other, probably more pasty or solidified, portions would have had a more perfectly developed foliated structure. According to Leith: ‘‘Many more schists than gneisses have been proved to be the result of mashing of igneous rocks. . . . . In fact, so commonly do the igneous rocks appear when mashed to take on schistose as contrasted with gneissic structure as to raise the question whether gneisses are not exceptional results, most gneisses to be explained as igneous rocks with original flow struc- tures.’”? The evidence from the Adirondacks is in harmony with tL. V. Pirsson, Rocks and Rock Minerals (1908), p. 356. 7C. K. Leith, Structural Geology (1913), p. 103. FOLIATION IN THE PRE-CAMBRIAN OF NEW YORK 607 this statement by Leith, since anything like true schists are very rare if not wholly absent’ from the syenite-granite series. Significance of granulation.—Granulation of the rocks of the syenite-granite series is of common occurrence. Most of the mineral constituents are more or less granulated, though it is quite the rule that the quartz shows the effects of crushing less than the others. In the greatest bulk of the rock the cataclastic texture shows itself by flattened or irregular lens-shaped quartz individuals, and more or less lens-shaped broken feldspars, imbedded in a mass of small broken feldspar grains together with some crushed quartz and leaves of mica. In many cases more or less thoroughly elongated and crushed hornblende or augite also occur. This granulation has usually been regarded as proof that the rocks have been subjected to severe lateral compression and crushing after their consolidation. Thus Smyth, keeping in mind the frequent lack of crushing of the quartz, has said: ‘‘ As the quartz could hardly flow while the feldspar fractured, the conclusion is obvious, and seems to be well grounded, that, in the case of the quartz, there has been crystallization after the production of cataclastic structure in the rock.”* But does this prove the quartz to be largely recrystallized or of secondary origin? Could not movements in the magma during a late stage of consolidation, and before much quartz (the last to form) had: crystallized out, have caused granulation of the earlier-formed crystals, while the quartz would have been more or less unaffected? In explaining the origin of foliation in the granite-gneiss of the Thousand Islands region, Cushing says: “The rock has been much crushed and somewhat recrystallized under compressive stress, since it originally con- gealed.”’? Now, while some granulation and recrystallization may have taken place after the magma consolidation, it is by no means a necessary inference that the granite has been much crushed and principally foliated after it had congealed. Strong evidence against severe compression of the Adirondack region has been presented in this paper, while the best evidence points to the origin of the folia- tion as essentially a flow structure developed under moderate tC. H. Smyth, 15th Ann. Rep. New York State Geologist, 1895, pp. 488-80. 2H. P. Cushing, New York State Mus. Bull., No. 145, 1910, p. 102. 608 WILLIAM J. MILLER pressure. This being the case, it is only necessary to consider that there were movements in the slowly cooling and stiffening magma whereby the minerals already crystallized out were more or less broken and drawn out into a sort of fluidal arrangement parallel to the foliation, while the minerals last to form were much less granulated. A significant point in this connection is that, in rocks which are definitely known to have been subjected to severe com- pression, quartz is quite generally more granulated than feldspar. Both Leith? and Loughlin? have emphasized this point. Now, in the Adirondack intrusives, as we have shown, the feldspar is very commonly distinctly more granulated than the quartz, and the evidence is, therefore, opposed to deformation of the Adirondack rocks by severe regional compression. The facts that degree of foliation and granulation often vary markedly within a few feet or yards, and that the most perfectly foliated portions are often also the most highly granulated, are to be expected, because flowage in certain portions of the magma during the late stage of consolidation would produce in those por- tions not only good primary foliation but also notable crushing of the already formed minerals by the movements in the stiff, pasty magma. It seems impossible to explain satisfactorily such marked differences in degree of both foliation and granulation in the syenite-granite series except as the result of movements in the con- gealing magma. In few cases, if any, is there evidence for shearing, so that if compression of the region be assumed as the cause of the foliation and granulation, it is impossible to explain why adjacent zones often present such differences in degree of foliation and granulation. Again, the general lack of notable granulation in the oldest rocks of the region—the Grenville—is not compatible with the idea of production of cataclastic structure in the intrusives by lateral pressure, else why were the still older rocks also not proportionately affected ? Other workers have presented strong evidence for the produc- tion of a granulated or protoclastic texture in igneous rocks by some ™C. K. Leith, U.S. Geol. Surv., Bull. 239, 1905, pp. 33-34- 2G. F. Loughlin, 7bid., Bull. 492, 1912, p. 128. FOLIATION IN THE PRE-CAMBRIAN OF NEW YORK 609 such process as that outlined above. Barlow, describing the granite of central Ontario, says: “‘The movements ... . continued as the rock cooled and while it was filled with abundant products of crystallization, the movements being brought to a close only by _ the complete solidification of the rock. Evidence of protoclastic structure can, therefore, be seen throughout all the areas colored as granite or granite-gneiss on the map.’ Teall says, regarding the granite of the county of Kircudbright: “The quartz and alkali feldspar, which . . . . were the last con- stituents to solidify, are those which have yielded most to the deforming stresses. They show signs of crushing..... It is probable that the pressure acted before the rock mass had actually cooled.’ : McMahon, discussing the gneissic granite of the Himalayas, says: “It is no argument against the idea of the development of foliation before final consolidation of the granite to point to evidence of strain and mechanical action in the rock; for the existence of strain and mechanical action during the critical period in the history of the granite is an essential part of the theory itself.” He admits that the granite has been subjected to lateral pressure but says — that this does not prove the foliation to have been produced by such pressure. Weinschenk,’ explaining certain schistose Alpine granites, sug- gests that, in a somewhat advanced stage of magma consolidation, a crystalline skeleton is formed whose interstices are filled with liquid magma. Movements cause crushing of the skeleton, breaking the feldspars and bending the mica plates. Quartz, the last mineral to crystallize, is flattened out but not much broken. According to Trueman: ‘It seems not illogical to assume that the movements which were, apparently, present late in the period of consolidation should have sometimes been continued after por- tions or the whole of the rock had completely solidified. If such tA. E. Barlow, Geol. Surv. Can., Mem. 57, 1915, p. 48. 2J. J. H. Teall, Mem. Geol. Surv. Great Britain, Expl. Sheet 5, 1896, p. 43. 3C. A. McMahon, Geol. Mag., N.S., Decade 4, IV (1897), 347. 4E. Weinschenk, Congrés géol. inter., Compte rendu, Session VIII, 1 (1900), 341. 610 WILLIAM J. MILLER were the case there would result considerable recrystallization and granulation so that typical crystalloblastic or cataclastic textures might be superimposed upon that resulting from primary con- solidation.’”* The possibility of some granulation and recrystallization in the Adirondack intrusives after complete consolidation is admitted by the writer, but, in view of the evidence above presented, such processes must have had relatively little to do with the development of the textural and structural features of the rocks. Cause of mineral elongation.—Still another matter to consider briefly is the cause of the flattening or elongation of minerals in the primary gneiss. Flattening or elongation of minerals, espe- cially quartz and feldspar, are common in the Adirondack intrusives, varying from rocks in which the phenomenon is scarcely noticeable to others in which it is extremely developed. It is the writer’s experience that many such variations exist within short distances. Quartz exhibits such flattening better and more frequently than the feldspars. The writer believes that the mineral flattening or elonga- tion was caused essentially by crystallization in the magma under pressure. Trueman’ has recently presented considerable evidence to show that elongation (and presumably flattening) of mineral con- stituents by crystallization under differential pressure must often have been a very important factor in the production of foliation of primary gneisses. In the Adirondack syenite-granite series, quartz shows the effects of flattening most because it was the last mineral to crystal- lize out and hence was not subject to so many of the movements in the magmas. Loughlin presents a similar argument regarding the Sterling granite-gneiss of Connecticut as follows: ‘‘ After crys- tallization had become so far advanced that the rock became a mass of feldspar crystals (plus a small amount of quartz) with interstices filled with still fluid quartz, the feldspars would suffer strain, rotation, and slicing, and become a more or less granular lens-shaped aggregate elongated in the direction of least pressure. . . . As the interstitial quartz began to crystallize, it would be tJ. D. Trueman, Jour. Geol., XX (1912), 2 2 Ibid., XX (1912), 235-42. FOLIATION IN THE PRE-CAMBRIAN OF NEW VORK 611 obliged to take on the form of the elongated or flattened inter- stices.’”? It is not at all necessary to assume a very active lateral com- pression of the region to account for this pressure. As suggested by Cushing,’ considerable compression of the magmas must have resulted from the batholithic intrusions, which, in order to make room for themselves, exerted a shouldering pressure upon the ad- jacent rocks. It is believed that such a shouldering pressure within the magmas was sufficient, not only to cause more or less flattening and elongation of minerals during consolidation and crystallization, but also to determine to a considerable extent the directions of the magmatic currents and hence the resulting strike of the foliation. Under the very conditions of intrusion, differential pressures must have been common, thus best explaining the frequent variations in degree of flattening of mineral constituents. This view does not of course preclude the possibility of moderate lateral pressure exerted throughout the whole region during, or even aiter, the magma consolidation. Foliation of batholithic borders.—Beiore leaving this. discussion another feature of the foliation of the intrusive masses should be mentioned, namely, that they often exhibit a greater degree of foliation and granulation around their borders than in their interiors. This phenomenon seems to be best shown in the anorthosite and the gabbro, and will be discussed below. Suffice it to say here that . production of foliation and granulation in the congealing magmas affords a more plausible explanation for the peripheral distribution of such features than their production by compression of the whole region. Summary.—During the process of intrusion, which was long continued, the great syenite-granite magmatic masses were under only enough lateral pressure to control the general strike of the uprising magmas with consequent tendency toward parallel arrange- ment of syenite-granite and invaded Grenville masses; the foliation is essentially a flow structure produced under moderate pressure during the intrusion; the sharp variations of strike on large and «G. F. Loughlin, U.S. Geol. Surv., Bull. 492, 1912, p. 129. 2H. P. Cushing, New York State Mus. Bull., No. 145, 1910, p. rot. 612 WILLIAM J. MILLER small scales, and rapid variations in degree of foliation, are essen- tially the result of varying magmatic currents under differential pressure, principally during a late stage of magma consolidation; the almost universal, but varied, granulation of these rocks was produced mostly by movements in the partially solidified magma, and possibly in part by moderate pressure applied after complete consolidation; and the mineral flattening or elongation was caused by crystallization under differential pressure in the cooling magma. FOLIATION OF THE ANORTHOSITE It is not the present purpose to discuss thoroughly the origin of the structural and textural features of the Adirondack anortho- site. Only a few of the more important phenomena will be briefly considered. In general, the explanations above given regarding the foliation and granulation of the syenite-granite series apply also to the anorthosite. Character of the anorthosite-—With the exception of a few small outlying masses, the anorthosite occupies a practically unbroken area of 1,200 square miles in the central-eastern Adirondack region. It quite certainly represents a single great intrusive body which is older than the syenite-granite series. In its typical develop- ment the rock consists almost wholly of basic bluish-gray plagioclase and is very coarse-grained, the feldspars often measuring from one to several inches in length. There are several important differ- entiation variations of the great mass. One of these is coarse- grained, but carries a considerable percentage of dark minerals; another is finer-grained and more gabbroic looking, owing to dark minerals, chiefly augite and ilmenite; while still another facies consists almost wholly of white basic plagioclase, or such white feldspar and more or less dark minerals. The great bulk of the rock is highly feldspathic and practically devoid of foliated structure, probably partly because of lack of minerals favorable for its production or accentuation, while the more gabbroic (especially finer-grained) types are almost invariably well foliated, frequently excessively so. All of the varieties show more or less granulation, sometimes to a high degree, this being particularly true of the less coarse- FOLIATION IN THE PRE-CAMBRIAN OF NEW YORK , 613 grained gabbroic types. As regards amount of granulation of feldspar, it is probably not very different in anorthosite and syenite- granite. The gabbroic, well-foliated, and granulated facies are developed on a grand scale around the borders of the great anortho- site area, but similar types are often encountered irregularly dis- tributed throughout the area. Large feldspar individuals, usually unaltered rounded or lens-like cores of crystals, quite typically stand out prominently in a finer-grained, generally well-granulated, groundmass. In spite of much granulation, it seems certain that the typical original rock (before thorough consolidation) was char- acterized by a coarse porphyritic texture. Cause of the foliation and granulation.—The foliation and granu- lation of the anorthosite has been explained as due to the same severe compression of the region which is supposed to have caused similar phenomena in the syenite-granite series. According to this view, the more general lack of anorthosite foliation is considered to be due to lesser effect of the regional pressure toward the interior of the great intrusive body than around its borders. Also it is thought that coarseness of original grain and general lack of min- erals, especially dark minerals, other than feldspar have militated against such complete granulation and foliation of the rock as characterizes the syenite-granite series. Regarding severe compression after the magma consolidation as the prime cause of the foliation and granulation is, however, open to many of the same objections already discussed in connection with the syenite-granite series. It is the writer’s belief that an insur- mountable objection to the severe-compression idea lies in the fact that there are so many sudden variations in degree of foliation and granulation, and in strike of foliation, throughout the great anortho- site area. Thus, well within the area, the writer has repeatedly seen very gneissoid gabbroic facies—both coarse and medium- grained—in close proximity to gabbroic facies of similar grain with little or no foliation. All types of anorthosite also often exhibit varying degrees of granulation in close proximity. If they were caused by regional compression, why are so many portions highly foliated or granulated when others close by are unaffected? Also, if regional compression were the cause of the foliation, how are 614 WILLIAM J. MILLER the frequent very notable variations in strike, often within relatively small areas, to be accounted for ? According to the general principles outlined in connection with the foregoing discussion of the syenite-granite series, it is the writer’s conception that the foliation and granulation of the anortho- site were developed essentially by flowage or other movements in the magma under moderate pressure, mostly just prior to its com- plete consolidation. As Cushing has said regarding the Long Lake quadrangle anorthosite: ‘In some portions of the rock the feldspar crystals are more numerous, are smaller and are all arranged with their long axes parallel. This is a ‘flow structure’ due to move- ments in the mass during solidification.’ The better foliated or better granulated belts throughout the great mass represent merely places where the magmatic currents or other movements were more pronounced. The coarser-grained portions would of course have undergone less complete granulation, but coarseness of grain and absence of dark minerals would not necessarily have tended to prevent the development of foliation. Thus, in the Broadalbin, North Creek, and Lake Pleasant quadrangles the writer has observed coarse granite-porphyry, almost free from dark minerals, with highly gneissoid structure due to thorough flattening out of both quartz and feldspar, while in other cases the porphyry shows little or no foliation. It would seem, therefore, that the general absence of foliation throughout so much of the anorthosite is best explained as the result of the much more uniform intrusion of this single great body which is less involved with Grenville masses, or, in other words, to much less forced differential flowage. Because of its great size, the pressure due to shouldering effect on adjacent rocks would be relatively slighter toward the interior of the mass. Not only is the foliation well developed around the margin of the great intrusive, but it also appears to be especially well exhibited in many parts of the area in the gabbroic facies where they are close to masses or inclusions of Grenville. Just as flow structure is often best shown close to the wall-rock of, or an inclusion in, a small intrusive body, probably because of friction against the wall-rock tH. P. Cushing, New York State Mus. Bull., No. 115, 1907, p. 472. FOLIATION IN THE PRE-CAMBRIAN OF NEW VORK 615 or inclusion and consequent development of differential pressure and flowage, so, on a large scale, in the anorthosite body it is reasonable to think that foliation due to magmatic flowage would have been best developed around the margin of, or close to, masses of country rock within the great anorthosite body. In many other places, however, primary gneissoid structures may have been produced by differential flowage far from any country rock. The cataclastic texture of the anorthosite is believed to have resulted from the crushing of minerals already crystallized out of the stiff, solidifying magma by movements in the magma. The shouldering pressure exerted by the great intrusive mass in order to make room for itself must have been sufficient to have affected the whole mass until final consolidation. Kemp says, regarding the anorthosite of the Elizabethtown quadrangle: ‘The entire area has been subjected to such severe pressure and granulation that the outer borders oi the crystals are always crushed to a finely granular and whitish mass. Within this rim the bluish nuclei of the plagioclases remain. When shear- ing and dragging have been added the nuclei yield augen-gneisses.’”* It is, however, not at all necessary to assume severe regional pres- sure to account for these phenomena. Forced differential flowage in the stiff, nearly congealed magma (under pressure due chiefly to its own shouldering action) could have produced most, if not all, of the granulation and dragging effects, the “‘augen”’ being cores of what were large, probably porphyritic, feldspars in the nearly solidified magma. Moderate pressure during or even after con- solidation may possibly have operated to accentuate the phenomena. Adams says, concerning the Morin anorthosite north of Mon- treal: ‘‘The circumstance that the streaks or irregular bands (foliation), when present in the otherwise massive rock, assume no definite direction, but twist about as if owing to movements oi the rock while in a pasty condition, indicate that they have been produced by movements before the rock solidified... . . The granu- lation of the coarsely crystalline massive anorthosite, usually with concomitant development of more or less foliated or schistose structure in the way described, is undoubtedly due to movements tJ. F. Kemp, New Vork State Mus. Bull., No. 138, 1910, p. 28. 616 WILLIAM J. MILLER in the rock, resulting from pressure which acted subsequent to or possibly during the last stages of its consolidation.’ FOLIATION OF THE GABBRO The gabbro here considered is the latest Adirondack intrusive which exhibits foliation and granulation. Diabase is the only intrusive still younger. A few years ago the writer? discussed the origin of certain primary variations of Adirondack gabbro. At that time, in accordance with the usual idea, the foliation was thought to be largely a secondary structure and so was omitted from the discussion. Character of the gabbro.—Most of the gabbro is in the form of small stocks or bosses, the outcropping areas typically ranging from elliptical to almost circular, and the dimensions from a few rods to one or two miles. They are especially well-shown on the North Creek, Long Lake, Elizabethtown, and forthcoming Blue Mountain geologic maps. Most of them are of pluglike or pipelike form, with practically vertical, sharp contacts against the country rock. The - stocks exhibit many variations in composition and texture from the normal, homogeneous, dark, basic gabbro with diabasic texture, to lighter-colored rocks of. dioritic and even syenitic make-up. They also range from fine-grained to very coarse-grained with feld- spars up to an inch or more in length. The typical gabbro con- tains principally basic plagioclase, pyroxene, hornblende, biotite, garnet, and ilmenite, while orthoclase and quartz often occur in the more acidic facies. A very important feature, from the standpoint of our present discussion, is the almost universal development of highly foliated amphibolitic borders which often completely surround the stocks, while the interior portions of the typical stocks are usually non- foliated. In many cases, however, stocks seem to be wholly changed to amphibolite, or only very small cores remain. In still other cases coarse-grained gabbro shows gneissoid structure thoroughly developed throughout. As a rule the gabbro exhibits as good, ‘{ not better, foliation than the older intrusives. Often tF, D. Adams, Geol. Surv. Can., Guide Book No. 3, 1913, Pp. 17- 2W. J. Miller, Jour. Geol., XXI (1913), 160-80. FOLIATION IN THE PRE-CAMBRIAN OF NEW YORK 617 the degree of foliation varies much even well within single stocks. : Another very persistent feature is granulation which appears to be of two types, that of so-called “corrosion rims” around certain minerals, and a more generally distributed granulation. Granu- lated “corrosion rims” occur even in non-foliated gabbro with diabasic texture. Cause of foliation and granulation.—The foliation and granula- tion of the gabbro, like that of the older intrusives, are quite gen- erally regarded as secondary features brought about by the influence of regional pressure, the non-foliated, uncrushed cores of stocks being considered as portions protected from pressure influence. Granting the existence of regional compression severe enough to give rise to these phenomena, it is evident that the same pressure must have affected the older intrusives in a similar manner, but this we have proved to be not the case. It is very difficult to imagine a process of development of foliation, which boxes the compass around the borders of the gabbro stocks, by regional com- pression. Such foliation of course often strikes directly across the foliation of the older adjacent rocks, an excellent case in point being at the south end of the large stock just north of Loon Lake of the North Creek sheet. How can such phenomena be explained as due to regional pressure when it is well known that cleavage or foliation produced by this means must everywhere strike at least approximately at right angles to the direction of application of pressure? Also how are such frequent notable variations in foliaton and granulation, not oniy in near-by stocks but also within stocks, to be explained ? According to the thesis of th’s paper, the foliation and granula- ton are largely, if not wholly, primary features. There are, admittedly, some puzzling things about the foliation and granula- tion of the gabbro, but certainly they are to be much more reason- ably interpreted as caused by movements in the magma before complete consolidation. Weinschenk, in explaining schistose peripheral zones around certain Alpine granitic cores, has suggested: ‘The consolidation of the rock commenced with the separation of the dark minerals— 618 WILLIAM J. MILLER biotite and hornblende. Mica formed first in the liquid mass. At this time the orogenic pressure acted upon the peripheral zone of the magma by orienting this minera! normal to the pressure. In the heart of the viscous mass this faculty of orientation was replaced by an interior tension not directed in any particular way.’ Orogenic pressure did not exist in the Adirondack region, but if for it we substitute the pressures within the stock magmas themselves, this idea of Weinschenk affords a plausible explanation of the foliated borders. Considerable pressures must have obtained within the stock magmas which were intruded under very deep-seated con- ditions. Such pressure against the country rock, combined with the usual development of differential flowage in the magmatic borders, as already explained in this paper, would readily account for the peripheral foliated zones which were produced, no doubt, during a late stage of magma consolidation. But the conditions for magmatic pressure and flowage must often have varied a great deal, so that it is to be expected that, in some cases, even coarse- grained gabbro would exhibit primary foliation, while, in other cases, amphibolite would make up the whole mass, or, in still other cases, finer-grained, very gneissoid, and granulated belts or bands would occur in the midst of coarser, less foliated types. It should be noted in this connection that unmistakable flow structures do often occur around inclusions in the gabbro. Applying these ideas, the puzzling features of various gabbro stocks find a ready interpretation. A good example is the stock near Blackbridge in the Lake Pleasant quadrangle. For most part this is a very basic, gabbroic-looking rock, sometimes pretty massive and very coarse-grained, and at other times not so coarse, but streaked or almost banded, owing to layers of amphibolite. All phases of the rock are much granulated and distinctly gneissoid, the coarser-grained portions being least so. A diabasic texture frequently occurs. Differential flowage and other movements under pressure in the congealing magma best explain these phenom- ena. The more foliated, finer-grained belts in the midst of the «E. Weinschenk, Congrés géol. inter., Compte rendu, Session VIII, 1 (1900), 340. Freely translated from the French. 2W. J. Miller, New York State Mus. Bull., No. 182, 1916, p. 29-30. FOLIATION IN THE PRE-CAMBRIAN OF NEW YORK 619 coarser, less foliated rock may be readily interpreted as the result of a forcing of slightly more liquid portions of the congealing magma through more solidified portions. Highly developed ‘‘corrosion rims”’ are beautifully and exten- sively developed in the Adirondack gabbros.t. Their occurrence even in non-joliated gabbro with diabasic texture argues strongly for their production before final consolidation of the magma, this possibility having been recognized by other investigators. How could regional compression have caused so much of this sort of granulation without otherwise affecting the rock ? Also, if granula- tion of this sort has resulted from movements prior to solidification of the magma, why could not the more general cataclastic textures of the syenite-granite, anorthosite, and gabbro have been similarly produced? As in the older intrusives, so in the gabbro, the finer-grained more foliated portions are quite generally the most granulated, this doubtless resulting from more pronounced flowage movements in certain portions of the magma late in the process of consolidation. tW. J. Miller, Jour. Geol., XXI (1913), 168-70. THE COMPOSITION OF THE AVERAGE IGNEOUS ROCK* ADOLPH KNOPF U.S. Geological Survey, Washington, D.C. The composition of the “‘average igneous rock”’ has been com- puted by Clarke, Harker, and Washington. Clarke’s most recent estimate was published in 1915.2, The earlier computations were made by averaging the results of large numbers of analyses, and the later by averaging each constituent according to the number of determinations made, and reducing the sum to too per cent. The objection to these methods, as is well known, is that they take no direct account of the quantitative distribution of the rocks; each analysis or determination receives the same weight, regardless of the size of the geologic body that it is held to repre- sent. The force of this objection has been recognized by Clarke,s who concludes that ‘‘the whole land surface of the earth must be taken into account before the true average can be finally ascertained.” A first approximation to this true average can be reached by calculations based on data recently assembled by Daly in [gneous Rocks and Their Origin. In Table IV is given the total areas — covered by each of the rock species named and mapped in the Cordilleran and Appalachian folios of the United States Geo- logical Survey. The area occupied by any rock species divided by the total area of igneous rocks (16,728 square miles) gives a weight-factor, and this factor multiplied by the average composi- « Published with the permission of the Director of the U. S. Geological Survey. 2 Analyses of Rocks and Minerals from the Laboratory of the United States Geo- logical Survey, U.S. Geological Survey Bulletin, No. 591, pp. 18-22, 1915. 3 The Data of Geochemistry (3d ed.), U.S. Geological Survey Bulletin, No. 616, p. 26, 1916. 620 COMPOSITION OF THE AVERAGE IGNEOUS ROCK 621 tion of the species, which has been computed by Daly in Table I, gives the percentage contribution of that species to the composition of the average igneous rock. In this calculation species covering less than 2 square miles were omitted, as their inclusion would not affect the second decimal place of the result. The composition thus calculated is that of the average exposed igneous rock; whether it represents the composition of the average igneous rock of the to-mile crust depends on the verity of certain petrogenic considera- tions. The following proportional factors were used in the computa- tions: Granite, including allied porphyries....... 0.23212 Granodionite pyr ter. Cele ele tals HL2TOS Rihvioliteneane nace Bette es bk dah ee ts . 12834 INTIMeSIVE Men re tn tee eels rsa ei eee Soe . 23864 TRESS YAS Bets BO meee Coie ee ee . 20773 Quartz monzonite and allied porphyry..... .oo108 Dignity yey Peer ey el aacs .01802 Gab Drones erent ee eee ak .02225 PATOL OSC pete eee nee eet siece) 3c ehsjie kare .003II Syenite and syenite porphyry............. .00389 Non ZO mire merrier eet eee Sm se a esce gees .OO161 iNephelinersyenitemmri arts cr cerca teeerrr. .00024 Shrombsimibe meron rsa esc arte icr els a ace oie .00054 TRY aye seee TES og WEN ale ata reas Arm a . 00036 IP SCIGIOIE NCES ce oktroca's o-8 Oia ane remem eS .00436 Ey KOXEMILE Hemera. ne tsisai 2). so maka .OOOTI WDiabasewew errr yn teria Surshes sia als .01602 Wa citeury re Meee onin ie eit ie ect sucha .00536 TS PACMAG) Necture) aie crcl ocdia eke le cee ep ech ee ie .00036 Waterers tna asc Oye eter ik Gite . 00030 IPhonoliter een mele ie eee hee he .00048 iINephelinemmelilite:basalt (43 )s 2 .00018 [Gimp ungate warmest kee eis ce. ss uk ake .O00T2 otal rere sac uncucuaes eis A Cas I .00000 The result of the new calculation of the composition of the average igneous rock is given in column I of the accompanying table; for comparison, the most recent estimate by Clarke is given in column II. 622 ADOLPH KNOPF Notwithstanding the widely different methods employed in the calculations the new estimate agrees to a remarkable extent with COMPOSITION OF THE “AVERAGE IGNEOUS ROCK” I II SiOx. 2 Ue eee 61.64 60.47 WMO2.: tee eee Os73 0.80 ALLO}. inte eee m5 70 E5407 FeO; 55 Oe 2.91 2.68 1 Oe. errmreen cattndts clnises tttee B25 su5O MnO. e 8 2k aseeeaee 0.16 0.10 MeO) ociccls Deere 2°07 3.85 GCaOs* eee ee 5.06 4.88 INaO)s 735 eee 3.40 3.41 KO... Eee ee 2.65 3.03 ELO= >. Oa it 6 f 0.48 HO. : 2 ee if aah \ 1.44 BLO gs «cea 0.26 0.29 100.00 100.00 Clarke’s average. The most notable departures are the increase in silica and the relatively strong decreases in magnesia and potassa. REVIEWS The General Magnetic Survey of the Earth. By L. A. BAUER. Bull. Am. Geog. Soc., XLVI, July, 1914, pp. 481-09. Figs. 6. About the earth sphere are lines of magnetic force very similar to those of any magnetic field, their poles not quite coincident with those of the earth axis. That the magnetic needle varies from true north was discovered at least as early as the fifteenth century, when Columbus sailed west from Europe. Subsequent observations have shown in addition that there is a constant change in the earth’s magnetism, mak- ing repeated magnetic surveys necessary. In magnetic observations, the horizontal declination, vertical magnetic dip, and intensity of the attraction are measured. Since 1904 the Carnegie Institution of Wash- ington has conducted extensive magnetic surveys of the earth in which a total mileage of approximately one million miles has been traveled. Ocean surveys have been conducted in a specially constructed non- magnetic vessel. IR (Co IML The Mud Lumps at the Mouths of the Mississippi. By EUGENE W. SHAW. U.S. Geol. Surv., Prof. Paper 85, Part B, 1913. Pp. 17, pls. 3, figs. 6. The territory within a mile or two of each of the mouths of the Mississippi is characterized by large swellings or upheavals of tough bluish-gray clay, to which the name “mud lumps”’ has been applied. Many of the mud lumps rise just off-shore and form islands having a surface extent of an acre or more and a height of 5 to 10 feet, but some do not reach the water surface. Almost all occur near the bars at the mouths of the rivers. In contrast with the general structure of the delta, which is composed of thin, alternating sandy and clayey beds, the mud lumps are of thick, compact clay. On and around the clay core lies a series of faulted and folded strata of sand and silt which have been carried up from the sea bottom and deformed in the upheaval. It seems most likely that these lumps owe their origin to a squeezing of the soft layers, and an accumulation of clay from such layers in places where the pressure is less strong. This postulates a gentle seaward flow of layers 623 624 REVIEWS of semifluid clay, the flow meeting resistance particularly at the ends of the Passes, where there is an accumulation of more resistant material and a greater lack of equilibrium between the heavy land on one side and the water on the other. The report is somewhat preliminary in nature. R;, CyMe The Upper Cretaceous and Eocene Floras of South Carolina and Georgia. By EW. Berry.: U.S. Geol. Sury., Prot gen 84, 1914. Pp. 200, pls. 20, figs. 12. The Upper Cretaceous of South Carolina is represented by the Black Creek formation, which is divisible into two members, the Middendorf arkose, with certain related clays, and. a sandy, marine member. A number of localities in the Middendorf have yielded plant remains, among which are found magnolias, figs, laurels, oaks, walnuts, cinnamon, the eucalyptus, etc. The collection numbers 75 species. The climate, indicated by the types present, is subtropic, or at least mild temperate, for with little variation the flora extends to the western coast of Green- land. The Upper Cretaceous of Georgia, the flora of which is described in the second part of the paper, is confined to a triangular area lying west of the Ocmulgee River and comprises the eastward extension of the Eutaw and Ripley formations. The former contains an abundant fossil flora in its lower division, but the latter, except in the upper part, con- tains little plant life. The physical conditions suggested are in accord with the evidence from South Carolina and point to a mild, humid climate, without frosts. A small but very interesting Middle Eocene flora from Georgia is described in the third division of the paper. The Middle Eocene of Georgia is for the greater part deeply buried beneath younger sediments, but in the area lying between the Ocmulgee and Savannah rivers there are outcrops which have yielded a fossil flora of 17 species. Most of these have not been described previously and the author compares them with European Eocene and modern related types. The conclusion is reached that the Middle Eocene of this region enjoyed a much more tropical climate than is represented by any other known ‘Eocene flora. The Georgia flora was probably immigrant from the south and reached northward at least as far as latitude 33° N. Ri Coe Excursion on the Rock River of Illinois between Rockford and Dixon Price to cents and Excursion through the Rivers and Harbor of Chicago Price 10 cents These are two of the seven ~ Bulletins of the Geographic — Society of Chicago. Teachers of biology and geology will find them of especial importance for field trips, lectures, — Write for descriptive circulars. The University of Chicago Press Chicago - eet - Illinois ordering by telegraph. _ library use will find this catalogue helpful. . 354 Fourth Avenue A New Edition of our Caries of : SCHOOL AND COLLEGE TEXTBOOKS Now Ready for Distribution THE BAKER & TAYLOR CO. Wholesale dealers in the books of all publishers NEW YORK Do You Want to Know What Your Handwriting ‘Reveals? _ : \ X 7 E have published a fascinating booklet of 32 pages on the sub- ject of graphology, by William Leslie French. Every reader of this publi- cation will be interested in comparing the examples of handwriting in judg- ing character and disposition of the writer. A copy of this booklet and 12 differ- ent patterns of The Spencerian Steel Pens will be sent postpaid on receipt of ten cents. If you can do good work with a poor pen you can do better work with a good pen. Silver plated Falcon pens, smooth points, . wear long, will not corrode. SPENCERIAN PEN CO. 349 Broadway - - - New York City Asin previous issues of this catalogue, we have listed all the important textbooks on all subjects, arranged alphabetically by author. book is indicated, and a telegraph code word is given for the convenience of those The wholesale list price of each A subject index to the more important books listed is preted: The best students’ texts for the study of the English and American classics recom- _. mended for secondary schools and required for college-entrance examinations are a conveniently listed for quick reference and comparison. The text editions of the French and German classics used most widely in high schools and colleges are given space; and handy literal and interlinear translations of Latin, Greek, French, and German texts are included. Everyone upon whom devolves the responsibility of ordering books for class or It will be sent free on request. At Twenty-Sixth St. All Essentials of Optical Superiority Besides wide variety of models and_ practical improvements in equipment, genuine optical superiority is manifest in Rausch lomb Microscopes They are the product of an experience of over sixty years in every phase of optical manufacture and have withstood the most exacting tests. Model FS2 (illustrated) has our lever type of side fine adjustment, gives unusual space for object manipulation, and is finished in our new crystal black—reagent-proof and more durable than the ordinary lacquer. Price, $34.00. Simple microscopes from $2.50 up—compound models - from $18.00 up. Special terms to Educational Institutions. Write for descriptive, illustrated price lists of Microscopes ; A pes, Accessories, and Microscope Lamps Bausch £9 Jomb Optical ©. 412 ST. PAUL STREET ROCHESTER, N.Y. New York Chicago Washington San Francisco Leading American Makers of Photographic Lenses, Binoculars, parrot Lanterns (Balopticons) and other high-grade optical products. i the very latest addition to the great Remington family of typewriters. Anew — and smaller Remington—built for the needsof the professional man and thehome Mail this eaunan today \ It is the Remington | | : within ten days. $5 each. Remington Typewriter Company (Incorporated) 327 Broadway, New York Send me a Remington Junior Typewriter, | price $50, on free examination. Itis understood that I may return the machine, if I choose, If I decide to purchase it, I agree to pay for it in 10 monthly payments of FINE INKS 4%» ADHESIVES For those who KNOW Higgins’ 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 will bea revelation to you, they are so sweet, clean, well put up, and withal so efficient. At Dealers Generally CHAS. M. HIGGINS & CO., Mfrs. Branches: Chicago, London 271 Ninth Street Drawing Inks Eternal Writing Ink — Engrossing Ink Taurine Mucilage — Photo Mounter Paste Drawing Board Paste Liquid Paste Office Paste — Vegetable Glue, Etc. Brooklyn yNni¥. 2 2 4 r | A 4 A New Proposition | A TYPEWRITER BY PARCEL POST ON TEN DAYS’ FREE EXAMINATION ~ ee HEN Fi _ VOLUME XXIV ie . NUMBER 7 THE JOURNAL or GEOLOGY A a QUARTERLY EDITED By _- +, ~~ ‘THOMAS C. CHAMBERLIN AND ROLLIN D. SALISBURY With the Active Collaboration of SAMUEL Ww. WILLISTON, Vertebrate Paleontology ALBERT JOHANNSEN , Petrology STUART WELLER, Invertebrate Paleontology ROLLIN T. CHAMBERLIN, Dynamic Geology ALBERT D. BROKAW, Economic Geology ASSOCIATE EDITORS SIR ARCHIBALD GEIKIE, Great Britain JOSEPH P.IDDINGS, Washington, D.C. CHARLES BARROIS, France JOHN C, BRANNER, Leland Stanford Junior University ALBRECHT PENCK, Germany RICHARD A. F. PENROSE, Jr., Philadelphia, Pa. HANS REUSCH, Norway f WILLIAM B. CLARK, Johns Hopkins University GERARD DEGEER, Sweden : WILLIAM H. HOBBS, University of Michigan T. W. EDGEWORTH DAVID, Australia FRANK D. ADAMS, McGill University BAILEY WILLIS, Leland Stanford Junior University CHARLES K. LEITH, University of Wisconsin GROVE K. GILBERT, Washington, D.C. WALLACE W. ATWOOD, Harvard University CHARLES D. WALCOTT, Smithsonian Institution — WILLIAM H. EMMONS, University of Minnesota HENRY S. WILLIAMS, Cornell University ARTHUR L. DAY, Carnegie Institution . OCTOBER-NOVEMBER 1016 THE GENESIS OF LAKE AGASSIZ: A CONFIRMATION - - - W. A. JOHNSTON 625 THE LOWER EMBAR OF WYOMING AND ITS FAUNA - - - E. B. BRANSON 639 EVOLUTION OF THE BASAL PLATES IN MONOCYCLIC CRINOIDEA CAMERATA. III HERRICK E. WILSON 665 DISCOVERY OF THE GREAT LAKE TROUT, CRISTIVOMER NAMAYCUSH, IN THE PLEISTOCENE OF WISCONSIN - - - - = - - - L. Hussakor 685 ASSUMPTIONS INVOLVED IN THE DOCTRINE OF ISOSTATIC COMPENSATION, wi A NOTE ON HECKER’S DETERMINATION OF GRAVITY AT SEA WILLIAM HERBERT HoBsBs 690 A STAGE ATTACHMENT FOR THE METALLOGRAPHIC MICROSCOPE ALBERT D. BRoKAW 718 REVIEWS - Ci HW AR RN ER Ra PR NPSL fas ln UeM BALA ln 22 see's THE- LOEWE RS Ey yOrR CHICAGO PRESS CHICAGO, ILLINOIS, U.S.A. AGENTS THE CAMBRIDGE UNIVERSITY PRESS, London anp EDINBURGH KARL W. HIERSEMANN, LErpzic THE MARUZEN-KABUSHIKI-KAISHA, Toxyo, Osaka, Kyoto, FukuOKA, SENDAI THE MISSION BOOK COMPANY, SHANGHAI “ ” = * . THE JOURNAL OF GEOLOGY EDITED By THOMAS C. CHAMBERLIN AND ROLLIN D. SALISBURY With the Active Collaboration of A Spy be SAMUEL W. WILLISTON ALBERT JOHANNSEN Vertebrate Paleontology Petrology STUART WELLER ROLLIN T. CHAMBERLIN = Invertebrate Paleontology Dynamic Geology ALBERT D, BROKAW Economic Geology ' The Journal of Geology is published by the University of Chicago at the University Press semi- quarterly, on or about the following dates: February 1, 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. 4] 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 follows: 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 single copies, 11 cents (total 76 cents). Patrons are requested to make all remittances payable to the University of Chicago Press in postal or express money orders of bank drafts. bsg wae The following agents have been appointed and are authorized to quote the prices indicated: . For the British Empire: The Cambridge University Press, Fetter Lane, London, E.C. boned subscriptions, including postage, 18s. 8d, each; single copies, including postage, 3s. 2d. each, For the continent of Europe: Karl W. Hiersemann, Kénigstrasse 29, Leipzig, Germany. Yearly subscriptions, including postage, M. 22.65 each; single copies, including postage, M. 3.80 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, including postage, Yen 1.50 each. For China: The Mission Book Company, 18 Peking Road, Shanghai. Yearly subscriptions, $4.00; single copies, 65 cents, or their equivalents in Chinese money. Postage extra, if mailed direct outside of Shanghai, on yearly subscriptions 53 cents, on single copies 11 cents. Mb nigel Claims for missing numbers should be made within the month following the regular month of publication. The publishers expect to supply missing numbers free only when they have been lost in transit. : ; Business correspondence should be addressed to The University of Chicago Press, Chicago, Ill. Communications for the editors and manuscripts should be addressed to the Editors of THE JOURNAL OF GEOLOGY, the University of Chicago, Chicago, Ill. Contributors are requested to write scientific and proper names with particular care and 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 leading articles will be printed as ordered, of which 50 (without covers) will be furnished grafis, the actual cost of the remainder (and covers, if desired) to be paid for by the author. The actual cost may vary, and will depend upon the amount of work in re-making the pages into forms, presswork, paper, binding, etc. Separates containing half-tones may be expected to cost somewhat more, the increase depending i ae the number of cuts and the amount of work required upon them. MR Entered as second-class matter, March 20, 1893. at the Post-office at Chicago, Ill., under the Act of March 3, 1879, VOLUME XXIV NUMBER 7 THE JOURNAL OF GEOLOGY OCTOBER-NOVEMBER 1916 THE GENESIS OF LAKE AGASSIZ: A CONFIRMATION W. A. JOHNSTON Canada Geological Survey, Ottawa, Canada CONTENTS DSR OTTO CATON Jao, Bs HS OS cao Coc ORR a IP Pec Giger 625 UpHam’s CONCEPTION OF THE LIFE-HISTORY OF LAKE AGASSIZ....... 626 DIFFICULTIES IN ACCEPTING UPHAM’S INTERPRETATION.............. 627 RECORDS OF LAKE Acassiz IN Ratny RIVER-LAKE OF THE Woops DIs- THEGUCA 4) 3 ici 2 ond wid wit o Bore, O°el aha LOLY Dadi Aan PE see eS NRE 629 CrocrArHICAT RELATIONS OF THE, DISTRICT)... 4.044 4 cence eae 629 SSL T SELES meee een Paellts, we aliren vista) ale 5. Wc le) Sud af cyallats) aha; Seheral coe erie apout g 630 LO ANMINVACEE DD BS ON Wal © HAWS Mieepeii sos: Akar. iiss cave wel oe evasela, sels aueteaiane es 630 DEP OSMSROE MUAKE WNGAGSIZ Mee ge 5) olai icici clad ts a wiotmmicle etc cvslans 631 UNCONFORMITY AT THE BASE OF LAKE AGASSIZ SEDIMENTS........ 632 BEARING OF THE LIFE-History OF LAKE AGASSIZ ON THE QUESTION OF THE CHARACTER AND CAUSE OF THE DIFFERENTIAL UPLIFT....... 636 SHURE sn Ba che cles 8 pio ators nc ea TEI Oe ge aac Uo aa 638 INTRODUCTION In a paper published in the Journal of Geology in 18096, J. B. Tyrrell stated that the results of his field work in the region lying to the west and southwest of Hudson Bay showed that— the Keewatin glacier seems to have retired northward well into Manitoba and possibly even beyond the northern limit of that province before it was joined by the eastern glacier. When they united the water was ponded between the t Published by permission of the Director of the Geological Survey of Canada. Vol. XXIV, No. 7 ‘ 625 626 W. A. JOHNSTON fronts of the two glaciers to the north and northeast and the high ground to the south and west. Thus Lake Agassiz had its beginning. Its waters rapidly rose until they overflowed southward into the valley of the Mississippi and then gradually declined as River Warren deepened its channel. By his more recent work in the region lying to the south of Hudson Bay, Tyrrell has shown that the last invasion of glacial ice in that region was by an ice sheet which advanced in a south- westerly direction and overlapped a portion of the area previously occupied by an ice sheet which he named the Patrician Glacier. This last advance of the ice extended in a southwesterly direction at least as far as the headwaters of the Severn River and in a westerly direction approximately as far as the Hayes River, where it was met by a readvance of the Keewatin glacier.’ Field work done by the writer during portions of the seasons of 1913 and rorq in the vicinity of the Rainy River and Lake of the Woods, Ontario, has brought forth evidence which confirms Tyrrell’s view that Lake Agassiz had at first a rising stage, due to the blocking of the northward drainage, and later subsided, and that, during the entire existence of the lake, the ice border was far to the north and northeast. This conception of the life-history of Lake Agassiz differs radically from that of Warren Upham, by whose work Lake Agassiz is best known, and whose interpretation has been most widely accepted. The object of the present paper is to present the evidence which confirms Tyrrell’s view as to the genesis of Lake Agassiz and to point out that the acceptance of this view has an important bearing upon the question of the character and cause of the epeirogenic movements which deformed the shore lines of Lake Agassiz. UPHAM’S CONCEPTION OF THE LIFE-HISTORY OF LAKE AGASSIZ Glacial Lake Agassiz is best known from the work of Warren Upham, the results of which were published in 1895 by the United States Geological Survey as Monograph 25. Upham’s field work in connection with the investigation of the basin of Lake Agassiz was done some thirty years ago and was largely confined to the tJ. B. Tyrrell, ““The Genesis of Lake Agassiz,’ Jour. Geol., IV (1806), 813. 2 J. B. Tyrrell, ‘The Patrician Glacier South of Hudson Bay,’’ Congrés Géologique International, Canada, 1913, Compte-Rendu (Ottawa, Canada, 1914), pp. 523-34. THE GENESIS OF LAKE AGASSIZ 627 _ western or prairie portion of the basin. At that time little was known of the extension of the lake in the northern portion of the state of Minnesota or in the adjoining portions of Canada, for much of this region was densely wooded, largely unsettled, and difficult of access. At that time, also, the general conception was that during Pleistocene time the Laurentide glacier occupied the greater portion of Central and Northwestern Canada. It was not until ‘some time later that the subdivision into Keewatin and Labra- dorean ice fields was recognized. Upham believed, as the result of his investigations, that the northward drainage of Red River valley and adjacent areas was ponded between the retreating front of the Laurentide glacier on the north and northeast and the divide on the south, that the lake had at first a small beginning in the southern part of the basin and gradually grew in size as the ice withdrew toward the northeast, and that a great series of moraines was deposited in the waters of the lake at stages of halt or slight readvance during the general retreat of the ice sheet. He found that the lake, during its higher stages, discharged southward to the Mississippi along the course of the present Lake Traverse and Minnesota River valleys. Dur- ing the operation of the southern outlet several strong shore lines of the lake were developed. As the ice retired and uplift took place, lower outlets were opened toward the northeast and other and lower beaches were developed in the northern part of the basin. He also showed that beaches which are single in the southern por- tion of the basin split into series in the northern portion of the basin and rise differentially toward the north-northeast, the highest being most upwarped and the lowest least, thus proving that differential elevation of the land went on during the existence of the lake.* DIFFICULTIES IN ACCEPTING UPHAM’S INTERPRETATION Some of the difficulties involved in accepting Upham’s inter- pretation of the life-history of Lake Agassiz were pointed out by T. C. Chamberlin. It was found by Upham that the uppermost or Herman beach was continuous for a long distance northward t Warren Upham, ‘“‘The Glacial Lake Agassiz,” U.S. Geol. Survey, Monograph 25, 1895. 628 W. A. JOHNSTON and that it overrode three prominent moraines which marked halts or readvances of the ice front. Upham supposed that the Herman beach represented the whole time of the formation of the several moraines and of the retreat of the ice front for at least 250 miles, in spite of the fact that he found the beach to be not very massive and not very notably stronger in the southern than in the northern portion. Recognizing this difficulty, Chamberlin suggested that “the whole history of Lake Agassiz may not have fallen within the period of stationary or rising crustal movement but that the early part of it may have taken place during the latter portion of the period within which the crust was being depressed.’* In this way it may be supposed that shore lines were formed at early stages of the lake but were later submerged. The uppermost Herman beach would have been formed at the time of maximum submergence. It would be all of one age and would represent a comparatively short time. Another difficulty arises from the character of the deposits laid down in the basin of the lake. Upham held that the greater part of these deposits were derived from the ice sheet and its inclosed drift—a necessary inference from his interpretation of the history of the lake. But he found that ‘‘bowlders are absent or exceedingly rare in the beaches, deltas, and finer lacustrine sediments.’? If it is true that a series of moraines was deposited in the lake, and if the sediments of the lake basin were largely derived from the ice sheet, it seems highly probable that berg deposits would form an important part and that bowlders would be included in the sedi- ments. A serious difficulty also arises if Upham’s interpretation of the mode of origin of the sediments which occur in Red River valley is accepted. Upham held that these sediments are recent fluvial deposits laid down in local depressions and on flood plains of streams after the disappearance of Lake Agassiz. The deposits, he states, ““have commonly greater thickness and extent than the underlying silt of glacial Lake Agassiz.”’3 In the southern portion of the basin they are in places underlain at considerable depths by ‘‘sheets of turf,’’4 etc., apparently indicating the presence of an old soil. The ™ “The Glacial Lake Agassiz,” U.S. Geol. Survey, Monograph 25, p. 245. 2 Ibid., pp. 183 and 2o1. 3 Ibid., p. 202. bd pee THE GENESIS OF LAKE AGASSIZ 629 great thickness and extent of these deposits and the occurrence of “sheets of turf’’ in their lower portions seem difficult of explanation on the assumption that they are ‘“‘recent fluvial deposits.” All these difficulties disappear, however, if it is considered, as the evidence seems to show, that Lake Agassiz had at first small beginnings in Red River valley and gradually rose until it over- flowed to the south, owing to a blocking of the northward drainage by an advance of the ice, and that the ice advanced only into the northern portion of the basin, so that the whole southern part of the lake was practically free from ice during the entire existence of the lake. RECORDS OF LAKE AGASSIZ IN RAINY RIVER-LAKE OF THE WOODS DISTRICT Geographical relations of the district—The eastern portion of Rainy River—Lake of the Woods district lies about midway between Lake Superior and the Red River of the province of Manitoba. The Rainy River connects Rainy Lake and Lake of the Woods and for a distance of 82 miles forms the international boundary between the state of Minnesota and the adjoining portion of the province of Ontario. The Rainy River, the main stream of the region, flows westward to Lake of the Woods, which drains northwestward to Lake Winnipeg and thence to Hudson Bay, so that the whole area lies within the Hudson Bay drainage system. ‘The altitude of Rainy Lake is 1,107 feet and of Lake of the Woods 1,060 feet above the sea, and the general altitude of the plain bordering the Red River on the west is about 200 feet lower. The southern portion of Lake of the Woods is shallow and is generally bordered by drift deposits. The divide southwest of the lake, separating the lake basin from that of the Red River on the west, is low and for some distance is less than 30 feet above Lake of the Woods. On the northwest, near Northwest Angle, the divide is also only a few feet above the level of the lake, so that the plains of Manitoba and northern Minnesota are practically continuous on the southwest and northeast into the southern portion of the Lake of the Woods basin. In southeastern Manitoba, and west of the southern portion of Lake of the Woods, the continuity of the plain’s surface is broken 630 W. A. JOHNSTON by a relatively high area which rises to a maximum height of about 200 feet above the general level of the plains. The area lying between Rainy Lake and Lake of the Woods is so deeply drift- covered that comparatively few solid-rock exposures occur. The surface has generally very slight relief, and slopes gently toward the west, so that the area really forms a portion of the eastward extension of the wooded portion of the prairie plains of Manitoba and northern Minnesota, from which it is separated by the shallow basin of the southern portion of Lake of the Woods. In the north- ern portion of Lake of the Woods and north ofa line drawn from the central part of the lake southeastward to Rainy Lake, the country is generally rocky and has comparatively little drift covering. Till sheets —There are at least two distinct till sheets in the district. The upper and younger is distinguished from the lower and older till sheet by the calcareous nature of its materials, and by the presence in it of bowlders of limestone and other rocks which are known to outcrop in Manitoba, but not in the district itself nor in the region lying to the northeast. Striae observed on the bedrock beneath the till sheet trend southeastward or eastward. These striae were not seen to be crossed by later striae, and no till was seen to overlie this till sheet. It seems evident, therefore, that the calcareous till was deposited by a lobe of the Keewatin glacier and that the area in which the calcareous till occurs was not over- ridden by an advance of ice from the northeast at a later time. The lower and older till sheet was deposited by an ice sheet advancing from the northeast. This is shown by the southwest- ward trend of striae on the bedrock underlying this till sheet and by the fact that the till contains no limestone similar to that which occurs in the upper sheet. Associated with the lower till sheet are considerable deposits of fluvio-glacial sands and gravels which also contain no limestone. No evidence was seen which would suggest that there was any considerable lapse of time between the depo- sition of the two till sheets, and it is presumed that they were nearly contemporaneous in age and were deposited during the Wisconsin stage of glaciation. Laminated stony clays.—A series of laminated clays, containing in places striated stones and bowlders, occurs in the district. The THE GENESIS OF LAKE AGASSIZ 631 clays overlie and in the eastern portion of the district also underlie the calcareous till, with which they are closely associated. In some sections there is a sort of transition upward from the unstratified till into the laminated clays; that is, in the lower portion of the clays distinct laminae of clay are separated by unstratified stony material resembling the underlying till. The stony layers at the base rarely exceed a few inches or at most a foot in thickness and rapidly die out, so that “the transition beds” are, as a rule, only 4 or 5 feet thick. The laminated clays in the district range in altitude from 1,060 up to at least 1,200 feet, but they are generally only a few feet in thickness. These clays were deposited in a glacial marginal lake which is here referred to as Early Lake Agassiz. This lake was associated with an advance of the Keewatin glacier which deposited the calcareous till in the region. The clays were in part deposited during the time of advance of the ice sheet, for in places till overlies the clays. This relation is well seen in the sections exposed in the gravel pit one and one-half miles west of Fort Frances, where 8 feet of calcareous till overlie laminated clays, which are again underlain by non-calcareous, fluvio-glacial sands and gravels. The laminated clays were also in part deposited during the time of retreat of the ice. Early Lake Agassiz was, however, largely if not wholly drained before the later Lake Agassiz came into existence, for the desposits of Lake Agassiz rest uncon- formably on the calcareous till and on the closely related laminated clays. Deposits of Lake Agassiz.—Numerous raised beaches of Lake Agassiz occur in the district, at altitudes ranging from 1,100 to 1,200 feet. ‘The strongest and best-developed beach extends north- ward for some distance from the vicinity of the town of Rainy River. The altitude of this beach near the town of Rainy River is 1,117 feet. Ten miles northeast of this locality its altitude is about 1,125, and twenty-four miles northeast its altitude is about 1,140. Higher beaches occur at various altitudes up to at least 1,200 feet. A comparatively small part of the drift-covered area lying between Rainy Lake and Lake of the Woods rises more than 1,200 feet, but in the northern portion of Minnesota immediately south of Rainy River district drift-covered areas rise considerably higher. In this 632 W. A. JOHNSTON area a number of beaches rising well above 1,200 feet have been found by Mr. Leverett, who states that bars of gravel and sand formed by the waters of Lake Agassiz occur on the highest points of Bel- trami [sland.* The lacustrine deposits of Lake Agassiz in the district occupy areas of considerable extent and are in places at least 30 feet thick. They are generally even-bedded but not strongly laminated. In places they are characterized by an irregular alternation of sandy and clayey layers and occasionally thin gravelly layers. The beds are in places more sandy toward their base than in their upper portion, and are frequently ripple-marked but not cross-bedded. The material is more oxidized than that of the older laminated stony clays, and there can be little doubt that the material was derived from erosion of land surfaces by wave and stream action. The sandy ripple-marked beds underlying clay, and the occurrence of gravelly layers interbedded with sandy and clayey layers are explained by the fact that the sediments were deposited in a rising body of water. The lacustrine beds are also characterized by the presence, in their lower portions at least, of fossil fresh-water shells. Fossil fresh-water shells also occur in some of the beach ridges at various altitudes up to 1,149 feet, or 88 feet above Lake of the Woods. Unconformity at the base of Lake Agassiz sediments.—The evi- dence found in Rainy River district, which confirms Tyrrell’s con- tention that Lake Agassiz had at first a rising stage, is based largely on the fact that the sediments deposited in Lake Agassiz rest unconformably upon the underlying deposits; that is, a period of erosion intervened after the deposition of the calcareous till and associated laminated stony clay, and before the later lacustrine sediments were laid down. ’ This is well shown in numerous sections exposed along the Rainy River and around the shores of the southern portion of Lake of the Woods. Fig. 1 illustrates the character of one of these sections which has been exposed by wave erosion on the present shore of Lake of the Woods at its southern side. At the base a «Frank Leverett, “Surface Formations and Agricultural Conditions of North- west Minnesota,” Minn. Geol. Soc., Bull. No. 12, 1915, p. 37. THE GENESIS OF LAKE AGASSIZ 633 small thickness of calcareous till is exposed, passing up into lami- nated, stony clay which is overlain unconformably by Lake Agassiz lacustrine clays containing fresh-water shells. The contact is a wave-cut plain. The lacustrine deposits above the wave-cut plain are clayey in character and evenly and thinly bedded, so that it is evident that the water must have risen to a considerable height to permit of such deposition. In places around the southeastern portion of Lake of the Woods these lacustrine deposits are at least Fic. 1.—Section exposed on south shore of Lake of the Woods, showing at the bottom calcareous till passing upward into laminated, stony clays unconformably overlain by Lake Agassiz lacustrine clays. The contact is a wave-cut plain. 30 feet thick and rise to an altitude higher than the divide sepa- rating the Lake of the Woods basin from that of the Red River on the west. Furthermore, the first strong raised beach above the level of Lake of the Woods, at the level of which the water must have stood, if not at some higher level, when the lacustrine deposits were laid down, is in the southern portion of the district from 45 to 55 feet above the level of Lake of the Woods; and this beach passes over the divide to the southwest of the lake.t Hence it follows that these lacustrine deposits were not laid down in a local lake t Minn. Geol. Survey, Bull. No. 12, 1915 (map). 634 W. A. JOHNSTON but in a body of water which covered not only the Rainy River and Lake of the Woods districts but also occupied Red River valley, and that this was the last great glacial-marginal lake in the region, viz., Lake Agassiz. | Numerous sections also show that weathering and erosion took place during the interval of erosion before the deposition of Lake Agassiz sediments. ‘This is well shown in sections along the Rainy River from one to three miles below the town of Rainy River. In places, small stream valleys were eroded and later partially or wholly filled with lacustrine deposits. ‘This relation is well seen in the small creek valley which enters the Rainy River three miles below Fort Frances. In one place, on Buffalo Point on the south- west side of Lake of the Woods, thin peaty bands occur in the lower portion of Lake Agassiz deposits. The sections exposed on the south shore of Lake of the Woods (Fig. 1) afford a demonstration that the water must have risen to a sufficient height to permit of the deposition of the fine lacustrine clays overlying the old wave-cut beach, and it is clear that these waters formed part of Lake Agassiz during a rising stage. There is evidence in the district that the waters rose through a vertical interval of at least 60 feet; for the lake clays are unconformable on the underlying sediments throughout this vertical interval. The highest shore line found in Rainy River district has an altitude of 1,200 feet. During the highest stages of the lake, practically the entire district was submerged and the highest shore line, if there had been land high enough to have received it, would have a present altitude of approximately 1,350 feet, as estimated from Upham’s determination of the highest beaches in other parts of the basin. It is not certain that the water rose to the level of the highest shore lines recognized in other portions of the basin; but it seems probable that it rose to the uppermost strong beach (Herman), because this beach, as already stated, is continuous for a long distance north- ward and is apparently all of one age. It is possible that the Milnor beach which Upham found to be traceable for only a short distance in the southern part of the basin marks a shore line of Early Lake Agassiz, but the extent of this lake or of any of its shore lines is not definitely known. This lake was largely drained THE GENESIS OF LAKE AGASSIZ 635 before Lake Agassiz came into existence and its sediments are buried beneath those of Lake Agassiz. It is at least certain that the waters of Lake Agassiz stood at one time at about the present level of Lake of the Woods, and that they later rose considerably higher. It seems probable also that the lake which preceded Lake Agassiz was almost completely drained, and that Red River valley was a land surface before the latest advance of the ice brought Lake Agassiz into existence; for the character of the deposits in Red River valley, which Upham‘ regarded as post-Lake Agassiz fluvial deposits, suggests rather that they are lacustrine deposits and that they are unconformable on the underlying sediments. Regarding these deposits Upham stated: Thus the occurrence of shells, rushes and sedges in these alluvial beds at McCauleyville, Minnesota, 32 and 45 feet below the surface or about 7 and 20 feet below the level of Red River, of sheets of turf, many fragments of decaying wood and a log a foot in diameter at Glyndon, Minnesota, 13 to 35 feet below the surface, and numerous other observations of vegetation elsewhere along the Red River valley in these beds, demonstrate that Lake Agassiz had been drained away, and that the valley was a land surface subject to overflow by the river at its stages of flood when these remains were deposited.? He also stated: “‘The deposits have commonly greater thickness and extent than the underlying silt of glacial Lake Agassiz.” It is evident that a land surface existed in Red River valley before these sediments were laid down; but it seems probable that the sediments are largely lacustrine in origin and not fluvial. G. M. Dawson, in describing the section across Red River valley near the international boundary stated that the valley is floored with a fine silty deposit, a portion of the upper layers of which may have been formed by the overflow of the river itself. He described the typical deposit as of great thickness and consisting of fine yellowish, marly, and arenaceous clay, holding considerable cal- careous matter, and effervescing freely with an acid. The great extent and thickness and the high calcareous content of the clays t Warren Upham, U.S. Geol. Survey, Monograph 25 (1895), p. 253- 2 [bid., p. 202. 3G. M. Dawson, Report on the Geology and Resources of the Region in the Vicinity of the Forty-ninth Parallel, 1875, pp. 248-49. 636 W. A. JOHNSTON would seem to show that they are lacustrine in origin and not fluvial. It seems probable that they are Lake Agassiz deposits and that they are unconformable upon the underlying sediments. It is concluded, therefore, with Tyrrell, that after the retreat of the Keewatin glacier well toward the north there was compara- tively free drainage to the north and that a later advance of the ice from the northeast was met by a slight readvance of the Keewatin glacier, which resulted in the ponding of the northward drainage and the inception of Lake Agassiz. It is not certain just how far the latest advance of the ice extended. It did not reach Rainy River district, for the calcareous till derived from Manitoba is not overlain by till derived from the northeast, and the southeastward- and eastward-bearing striae are not crossed by later striae. At Stony Mountain, near Winnipeg, southeastward-bearing striae cross striae trending nearly south, but are not themselves crossed by later striae. Tyrrell found that along the east side of Lake Winnipeg southwestward-bearing striae cross earlier striae bearing nearly southward. ‘Tyrrell? also held that the “‘ Winnipeg Moraine”’ represented by islands in Lake Winnipeg and developed in places along the western shore of the lake marked the termination of the Labradorean glacier. It seems evident that during the life of the last great glacial marginal lake of the region, viz., Lake Agassiz, the ice margin in Manitoba was at no time farther south than the southern portion of Lake Winnipeg, and that the whole southern portion of the lake was practically free from ice. Lake Agassiz was associated with a readvance of the ice sheet, chiefly of the Labradorean glacier at a very late time during the Wisconsin stage of glaciation, and its disappearance followed the final withdrawal of ice sheets from the region. BEARING OF THE LIFE-HISTORY OF LAKE AGASSIZ ON THE QUESTION OF THE CHARACTER AND CAUSE OF THE DIFFERENTIAL UPLIFT If it is true, as seems probable, that during the existence of Lake Agassiz the ice border was far north of the southern end of the lake, this fact has an important bearing on the character and tJ. B. Tyrrell, Amer. Geol., VIII, 21. ? Bull. Geol. Soc. Amer., XXIII (1911) 22 » #O0° THE GENESIS OF LAKE AGASSIZ 637 cause of the differential uplift which is shown to have taken place by the deformation of the shore lines of the lake. It is known that the whole of the southern portion of the Lake Agassiz basin was affected by uplift but that the region south of the southern outlet of the lake was unaffected, for the abandoned shore line of Lake Dakota in this region is apparently nearly horizontal.* That is, there is a sort of ‘“‘hinge-line” here. The location of this “‘hinge-line’’ was not due to “quick recovery of the crust by uplift”’ following removal of the ice from the immediate neighborhood, for the ice border was at least 250 miles north of the location of the “‘hinge-line.”’ The question also arises whether, as Chamberlin suggested, the land was being depressed during the time of advance of the latest ice sheet. It would be possible to determine this if the present altitude with respect to sea-level of the beaches which were made during the rising stage of the lake could be determined. It was found in Rainy River district that the strongest beach of Lake Agassiz apparently marks a long stand of the waters during the rising stage and again during the subsiding stage; for the beach deposits show evidence of having been partly eroded and spread out by the rising waters and beach ridges having a slightly different trend were later built on the older deposits. This would seem to show that the land was already depressed during the rising stage - of the lake, but the evidence is not very conclusive. In the case of the ‘‘fossil’’ shore line seen in sections along the south shore of Lake of the Woods (see Fig. 1), it was found that the beach main- tains the same altitude in a direction corresponding to the trend of the isobases of the beaches formed during the subsiding stage of Lake Agassiz. It is not known whether it rises toward the north- east, for unfortunately no record of its occurrence could be found in the northern part of Lake of the Woods. The evidence suggests, but does not prove, that if, as seems probable, the uplift of the land was due to isostatic readjustment following the removal of the burden of the ice sheets, there was no close sympathetic relation; but that uplift lagged? considerably * U.S. Geol. Survey, Monograph 25, p. 267. 2J. Le Conte, Bull. Geol. Soc. Amer., II (1891), 329-30; W. B. Wright, The Quaternary Ice Age, 1914. 638 W. A. JOHNSTON behind the removal of the great mass of the Wisconsin ice sheets and was only completed after the final retreat following the latest advance of the ice. SUMMARY Evidence bearing on the life-history of Lake Agassiz, found in the Rainy River—Lake of the Woods district, Ontario, confirms Tyrrell’s conclusion that Lake Agassiz had at first a rising stage. The evidence is based largely on the fact that an unconformity exists at the base of the Lake Agassiz lacustrine sediments. The lake was associated with the latest advance of ice sheets, chiefly of the Labradorean glacier, during the Wisconsin stage of glacia- tion. An earlier glacial marginal lake, which is herein referred to as Early Lake Agassiz, was associated with a lobe of the Kee- watin glacier; but this lake was largely if not wholly drained before Lake Agassiz came into existence. The latest advance of the ice into the Lake Agassiz basin did not extend farther south than the southern portion of Lake Winnipeg, so that the ice border of Lake Agassiz was at least 250 miles north of the southern end of the lake during the entire existence of the lake. The acceptance of this interpretation of the genesis of Lake Agassiz has an important bearing on the question of the character and cause of the differential uplift which is shown to have affected the region by the deformation of the shore lines. The evidence suggests, but does not prove, that if the uplift was due to isostatic readjustment following the removal of the burden of the ice sheets, there was no close sympathetic relation, but that, as Le Conte and Wright have supposed, uplift lagged considerably behind and was only completed after the final retreat following the latest advance of the ice sheets of the Wisconsin stage of glaciation. THE LOWER EMBAR OF WYOMING AND ITS FAUNA E. B. BRANSON The University of Missouri About ten years ago Mr. N. H. Brown sent me a few Helodus teeth from the Embar limestone near Lander, Wyoming, and in the summer of 1911 I had the good fortune to discover a rather abundant fish fauna in the same region and on Bull Lake Creek. In the former region the fossils occur about 25 feet from the bottom of the formation and their vertical range is little more than 3 feet, while in the latter they are found at from 35 to 38 feet from the bottom. Blackwelder? and Woodruff? have reported two distinct faunas from the Embar, the upper of which Girty? refers to the Permian, but the fauna reported here is older and entirely distinct from the other two. Collections from bottom to top of the Embar in several localities show only one species common to the fish horizon and the upper horizons. DESCRIPTION OF THE EMBAR LIMESTONE The Embar formation lies conformably, for the most part, below the Chugwater formation on the eastern slope of the Wind River Mountains and without apparent unconformity above the Tensleep sandstone. In the Big Popo Agie region, near Lander, it is about 400 feet thick, is largely limestone, and bears three and perhaps more phosphate beds. The lowest bed is 23 feet from the bottom, ranges from 1 foot to 5 or 6 feet in thickness, and bears fish remains and Orbiculoidea utahensis in abundance. This bed has been traced some 15 miles southeastward along the strike and about 5 miles nothwestward. I have examined the same bed on t Eliot Blackwelder, “‘New or Little Known Paleozoic Faunas from Wyoming and Idaho,” Am. Jour. Sci., XXXVI (1913), 177-79. 2E. G. Woodruff, “The Lander Oil Field,’ U.S. Geol. Surv., Bull. 452 (1913), pp. 12-14. 3 Ibid., p. 13. 639 640 E. B. BRANSON the North Fork of Little Wind River, 27 miles to the northwest, and on Bull Lake Creek, 35 miles to the northwest. On Bull Lake Creek it is 38 feet from the bottom and 3 feet 6 inches in thickness. This is probably the same bed that Blackwelder describes as occurring in Dinwoody Canyon 29 feet 8 inches from the bottom, 2 feet 2.5 inches thick, and containing Lingulodiscina utahensis. Wherever the bed was examined it contained fish remains, and Orbiculoidea utahensis in abundance. In Big Popo Agie Canyon a 4-inch phosphate bed, as it was measured, but probably much thicker, occurs 58 feet from the bottom, and 6 miles southeast along the strike a bed 2 feet thick occurs 54 feet from the bottom, and is followed 1o inches higher by 14 inches of phosphatic shale. About 150 feet higher a 5-foot bed of gray phosphatic limestone is present in Big Popo Agie Canyon, and 6 miles southeast along the strike a 5-foot bed occurs at the same horizon. This is the bed that Blackwelder describes as occurring 12 miles west of Lander—‘ Brownish-gray odlitic and nodular phosphate rock full of Productus subhorridus and other fossils (42.4 per cent tricalcium phosphate) ’’'—« feet 7 inches thick. This bed is readily distinguished from the lowest: bed by its fauna. It contains Producti, spirifers, and Spiriferina pulchra in abundance, while the lower bed contains none of these. The Embar limestone is well exposed in Little Popo Agie Canyon and the following description was worked out in 1913: 32. Greenish sandy shales and limestones: covered with talus and with a bed outcropping here and there.................... 75’ to 100° 31. Light-gray, highly siliceous limestone with the silica in druses as quartz and not as chert; this is the limestone that forms long dip slopes on the east side of the Wind River M ountains. . C70 uae go. Laght-gray, chenty lamestone). 225 -. 0 sc. \ WHERE DUPLICATE COPIESARENEEDED - Daus Improved Tip Top Duplicator From the Cambridge University Press THE GENUS IRIS By WILLIAM RICKATSON DYKES With 48 colored plates and 30 line drawings in the text This magnificent work brings together the avail- able information on all known species of Iris. The - most striking feature of the book is the life-size colored plates, reproduced from originals drawn from living _ plants—making it a volume of remarkable beauty as well as of great scientific importance. 254 pages, demi folio, half morocco; price $37.59, postage extra (weight rr Ibs.). The University of Chicago Press 5758 Ellis Avenue, Chicago, Illinois is an ideal assistant, always ready to reproduce clean, clear, perfect copies of letters, manuscripts, price lists, bills, in- voices, reports, drawings, menus, etc., etc.. Don’t take our word, try it yourself for Ten days’ trial without deposit If not satisfactory, return and no questions asked. Makes 150 handwritten and 50 typewritten copies in a few min- utes. Negatives require no washing out after using, and negative rolls can be used from 15 to 20 times. Complete outfit contains a continuous roll of Daus Non-Peeling Anti- Stick Negative Roll. Capsize,8}x13inches. Price $7.50 FELIX K. DAUS DUPLICATOR CO. Daus Building, 111 John Street, NEW YORK TEACHERS OF BIOLOGY AND GEOLOGY Will find the Bulletins of the Geo- graphic Society of Chicago of special value for field trips, lectures, etc. Write for descriptive circulars. a The University of Chicago Press Chicago - - - - Illinois (Sr eS BUY “ON EXAMINATION” THAT MEANS ‘* Test and try before you buy’’ - Remington Junior Typewriter Remington Quality; Remington Workmanship Remington Guaranty Produced purposely to supply the demand for a high- grade but low-priced writing machine for the home, the farm, the store, the office and the study. Cut out and post today REMINGTON TYPEWRITER COMPANY (Incorporated) 327 Broadway, New York Send me a Remington Junior Typewriter, price $50, on free examination. It is understood that I may re- turn the machine, if I choose, within ten days. If I decide to purchase it, I agree to pay for it in ro monthly payments of $5 each. Do You Want to Know What Your Handwriting Reveals? WE have published a fascinating booklet of 32 pages on the sub- ject of graphology, by William Leslie French. Every reader of this publi- cation will be interested in comparing the examples of handwriting in judg- ing character and disposition of the writer. A copy of this booklet and 12 differ- ent patterns of The Spencerian Steel Pens will be sent postpaid on receipt of ten cents. If you can do good work with a poor pen you can do better work with a good pen. Silver plated Falcon pens, smooth points, wear long, will not corrode. SPENCERIAN PEN CO. 349 Broadway - . - New York City Widely Recognized as Superior Instruments Their general use by leading educational institu- tions, both great and small, and by eminent microscopists and scientists, is acknowledgment of the superiority of B ancch lal Microscopes The hard, steady use of classroom and laboratory emphasizes their durable mechanical construction, their extreme optical precision, and their practica- bility. Models vary from simple form to wonder- fully precise compound instruments. Model FS2 (illustrated) is unusually well adapted for educational use, having sensitive, fine adjust- ment combined with curved handle-arm which allows ample space for object manipulation. Fin- ished in our new black crystal, reagent proof and more durable than ordinary lacquer. Price $34.00. Simple Microscopes $2.50 up. Compound Models, $18.00 up. Special terms to Educational Institu- tions. White for descriptive catalogs, giving illustrations and prices of all models. Bausch £9 lomb Optical ©. 412 ST. PAUL STREET ROCHESTER, N.Y. New York Washington Chicago San Francisco Leading American Makers of Photographic Lenses, Binoculars, Fraeciee Lanterns (Balopticons) and other high-grade optical _ products. P NUMBER 8 THE 2NAL or GEOLOGY EDITED BY With the Active Collaboration of ser0W" Vertebrate Paleontology ALBERT JOHANNSEN, Petrology , Invertebrate Paleontology ROLLIN T. CHAMBERLIN, Dynamic Geology ALBERT D,. BROKAW, Economic Geology ASSOCIATE EDITORS Great Britain JOSEPH P.IDDINGS, Washington, D. G. JOHN C. BRANNER, Leland Stanford Junior University RIGHARD A. F. PENROSE, Jr., Philadelphia, Pa. WILLIAM B. CLARK, Johns Hopkins University 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 DAVID, ‘Australia . Leland Stanford Junior University BERT, Washington, D.C. ! i: LCOTT, Smithsonian mstaturion mell Urey ~ NOVEMBER- -DECEMBER 1916 OLE “OF ~ ANORGANIC AGENCIES ‘IN THE DEPOSITION OF CALCIUM - - JoHN JOHNSTON AND E. D. WILLIAMSON = 729 IE STRUCTURAL “RELATIONS ae AUSTRALASL NEW GUINEA, AND - E. C. ANDREWS. 751 TERION OF THE ANTIQUITY OF THE ANGIOSPERMS EpMUND W.SINNOTT 777 eo. Bay , ATHOLITHS” OF THE HALIBURTON-BANCROFT AREA, ONTARIO, RECTLY NAMED? ~— - : »< op) Ll “FRancis M. VAN Tuyt ~ 792 ‘S OF CAPILLARITY ON OIL ACCUMULATION _. A.W. McCoy 798 PROCESS OF SULPHUR DEPOSITION- - _ = -¥, Omovve- 806 TUDE ENTS: CO TO THE STUDY OF RIPPLE MARKS é Doucias W. JoHNson 809 iy 1 H NT PU CATIONS OUNE XXIV spy tests DBS oa E UNIVERSITY OF GHIC AGO PRESS sd <°“ CHICAGO, ILLINOIS, U.S-A: AGENTS "THE CAMBRIDGE UNIVERSITY PRESS, Lonpon ANP access KARL W. HIERSEMANN, Leipzic THE en 2008 Toxyo, OSAKA, Kyoto, FUKUOKA, SENDA! THE MISSION BOOK COMPANY, SHANGHAI EDINBURGH THE JOURNAL OF GEOLOGY EDITED BY THOMAS C. CHAMBERLIN AND ROLLIN D. SALISBURY - With the Active Collaboration of f SAMUEL W. WILLISTON ALBERT JOHANNSEN Vertebrate Paleontology Petrology STUART WELLER ROLLIN T. CHAMBERLIN Invertebrate Paleontology Dynamic Geology ALBERT D. BROKAW Economic Geology The Journal of Geology is published by the University of Chicago at the University Press semi- quarterly, on or about the following dates: February 1, March 15, May I, 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. 4] Postage is charged extra as follows: 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 single copies, I1 cents (total 76 cents). J Patrons are requested to make all remittances payable to the University of Chicago Press in postal or express money orders or bank drafts. The following agents have been appointed and are authorized to quote the prices indicated: For the British Empire: The Cambridge University Press, Fetter Lane, London, E.C. Yearly — subscriptions, including postage, 18s. 8d, each; single copies, including postage, 35. 2d. each. For the continent of Europe: Karl W. Hiersemann, Ko6nigstrasse 29, Leipzig, Germany. Yearly subscriptions, including postage, M. 22.65 each; single copies, including postage, M. 3.80 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, including postage, Yen 1.50 each. For China: The Mission Book ‘Company, 18 Peking Road, Shanghai. Yearly subscriptions, $4.00; single copies, 65 cents, or their equivalents in Chinese money. Postage extra, if mailed direct outside of Shanghai, on yearly subscriptions 53 cents, on single copies II cents. Claims for missing numbers should be made within the month following the regular month of publication. The publishers expect to supply missing numbers free only when they have been lost in . transit. Business correspondence should be addressed to The University of Chicago Press, Chicago, III. Communications for the editors and manuscripts should be addressed to the Editors of THE JOURNAL OF GEOLOGY, the University of Chicago, Chicago, Ill. Contributors are requested to write scientific and proper names with particular care and 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 leading articles will We printed as ordered, of which 50 (without covers) will be furnished gratis, the actual cost of the vena (and covers, if desired) to be paid for by the author. The actual cost may vary, and will depend upon the amount of work in re-making the pages into forms, presswork, paper, binding, etc. Separates containing half-tones may be expected to cost somewhat more, the increase depending upon the number of cuts and the amount of work required upon them. Entered as second-class matter, March 20, 1893, at the Post-office at Chicago, II]., under the Act of March 3, 1879. VOLUME XXIV NUMBER 8 THE lOWRNAL OF GEOLOGY NOVEMBER-DECEMBER 1916 THE ROLE OF INORGANIC AGENCIES IN THE DEPOSI- TION OF CALCIUM CARBONATE JOHN JOHNSTON anv E. D. WILLIAMSON Geophysical Laboratory, Carnegie Institution of Washington, Washington, D.C. Organic agencies are doubtless the predominant occasion of the deposition of calcium carbonate, yet certain inorganic factors may not safely be left out of account. The mode of action of the former, which is in part a biological question, we shall not enter into, but we shall limit ourselves to a discussion of the effects producible by variation of certain inorganic factors which affect directly the solu- bility of calctum carbonate. The question of the concentration of calcium relative to the limiting saturation concentration of calcium carbonate under the particular conditions—in other words, the rela- tive degree of saturation with respect to calcium carbonate—has not received adequate consideration; this is largely the consequence of faulty data and of contradictory and erroneous statements regarding the solubility of calclum carbonate under various conditions. It is our purpose to direct attention to the quantitative effect, as deduced from laboratory study, producible by variation of those factors which, by affecting directly the degree of solubility of calcium car- bonate, induce its precipitation from a solution saturated with it; and to emphasize the fact that many of the points now ambiguous may be settled by means of systematic and accurate investigation Vol. XXIV, No. 8 720 730 JOHN JOHNSTON AND E. D. WILLIAMSON of a certain group of properties of sea-water, properties which, more- over, are of high importance in connection with certain biological problems. The mode of treatment is similar to that employed by Stieglitz,t who, at the instance of Chamberlin, carried out series of calculations to ascertain the proportion of CaCO, which one might expect to find in gypsum that had been deposited from solutions saturated with respect to both CaSO, and CaCO, at different partial pressures of CO, in the atmosphere in contact with the solution. The principles, therefore, are not new, though the point of view differs somewhat; and we now have the advantage of more extensive data than were available in 1907. The data bearing on the solubility of pure calcite have been col- lated and discussed at length in two recent papers,? to which the reader desirous of fuller information on the chemical side is referred. As it would lead too far to discuss here all details of the solubility- product constant and of its mode of calculation, we shall give only the established conclusions which are pertinent to the present dis- cussion, premising that a symbol inclosed within brackets represents the concentration (expressed in moles per liter) of that particular ionic or molecular species. 1. Inasolution at a fixed temperature saturated with pure calcite, the solubility-product—i.e. [Catt] [CO], the product of the respec- tive concentrations of calcium-ion and carbonate-ion—is a constant, independent of the proportion of free CO, in the solution and of the presence of other salts. This characteristic solubility-product con- stant is to be carefully distinguished from the solubility which, as ordinarily measured, is the concentration of total calcium in a solu- tion in equilibrium with solid calcite; and this calcium is associated with bicarbonate and hydroxide (and with any other anion present, e.g., chloride or sulphate) as well as with carbonate—indeed, under ordinary atmospheric conditions but a small fraction of the total tJ. Stieglitz, ‘The Relations of Equilibrium between the Carbon Dioxide of the Atmosphere and the Calcium Sulphate, Calcium Carbonate, and Calcium Bicarbonate in Water Solutions in Contact with It,” in ‘‘The Tidal and Other Problems,” by T. C. Chamberlin et al., Carnegie Inst. Publ. No. 107 (1909). 2J. Johnston, Jour. Am. Chem. Soc., XXXVII (1915), 2001, hereinafter desig- nated for convenience as op. cit.; Johnston and Williamson, ibid., XXXVIITI (1916), 975. AGENCIES IN DEPOSITION OF CALCIUM CARBONATE 731 calcium is ever associated with carbonate. The fact of the con- stancy of this solubility-product in presence of solid calcitet enables us to calculate, with all the accuracy required for the purposes of this paper, the solubility of calcite under any specified conditions, e.g., in presence of calcium-ion or carbonate-ion from whatever source derived, provided only that we can ascertain what these ionic concentrations actually are. 2. The concentration of H,CO, (‘“‘free” CO,)in solution is regu- lated by the partial pressure (P) or proportion of CO, in the layer of atmosphere in contact with the solution, and conversely; and, for a given value of P, it diminishes with rising temperature, since the absorption coefficient (solubility) of CO, diminishes. _3. Ata given temperature the total solubility as usually measured —1.e., the total concentration of calcium in the solution—varies with the concentration of H,CO, (hence with P), owing to the fact that the latter determines the proportion of carbonate-ion COj, hydrocarbonate-ion HCO7, and hydroxide-ion OH™ in accordance with definite mathematical expressions; and since the product [Cat*] [CO] remains constant [Ca**] must vary inversely as [CO§]. The presence of other salts also affects this total solubility; so long as pure calcite is the stable solid phase in equilibrium with the solution, the magnitude of this effect is readily calculable, since the several concentrations always adjust themselves until the solubility- product [Cat*] [CO%] attains its characteristic value. 4. The solubility-product constant of calcite diminishes with rising temperature; it is not affected to an appreciable extent by change of hydrostatic pressure. | The mathematical expressions are given below: [H.CO,]=cP [Cat : [CO>|= XK. (in presence of solid calcite) [HCOs}/[CO>]=/[H.CO,] =lcP [OH—}?/[COF]=m/|[H.CO,]=m/cP where c, K,,/, and m are constants at any given temperature.? We may note, moreover, that the free CO, and the total CO, (ie., ‘Similar remarks apply, mutatis mutandis, to impure calcite or to aragonite; to this point we revert later. 2 For their values and significance, see Johnston, of. cit., p. 2011. 732 JOHN JOHNSTON AND E. D. WILLIAMSON [H,CO,]+-[CO ;]+[HCO7]) determine [OH™], the degree of alkalinity (or acidity) of the solution; and that no change can be made in any one of these quantities without affecting each of the others. Accordingly the solubility of calcite is significant only if the con- centration of free CO, is controlled and measured, for changes in the latter, such as may easily occur, exert a large influence on the amount dissolved.t' This is evident from Table I, which gives the solubility TABLE I SOLUBILITY OF CALCITE AT 16° FOR VARIOUS VALUES OF P CO, IN THE ATMOSPHERE EXPRESSED FREE CO; or SOLUBILITY OF Bet INSOLUTION, ae Parts: As Partial _|As Parts per 10,000 SASS aCO; PER Pressure P (by Valimes MILLION MILLION 0.0001 TO 0.18 44 - 0002 2.0 * 360 55 -00025 2.5 -45 59 - 0003 3.0 55 6 3 00035 Bub 64 , 66 ©.0005 aie) 0.90 75 at 16° for various values of P not far removed from the proportion normally present in the atmosphere (about 3 parts per 10,000). Calculation shows that except for very small partial pressures of CO, the calcium in solution is associated almost entirely with bicarbonate—thus even when P is only 0.0005, the proportion as carbonate is only about 2 per cent, whereas when P is 1.0, the pro- portion is less than 1 part in 30,000; nevertheless, carbonate is still the solid phase which separates out, an excellent example of the fact that it is the solubility relations and not the “affinity” relations in solution that determine which of the possible stable solid phases shall appear. t Neglect of this factor or, in general, a failure to secure equilibrium conditions is responsible for erroneous statements in the literature. For instance, the solubility as given by Treadwell and Reuter (Z. anorg. Chem., XVII [1898], 170) is not a real solubility at all; acceptance of their figure (238 parts per million) has led several writers astray. Cf. op. cit., p. 2009. Thus on this basis J. C. Jones (Science, XX [1914], 829) concluded that the waters of the Lake Lahontan basin are only about one-twentieth saturated with CaCO. AGENCIES IN DEPOSITION OF CALCIUM CARBONATE 733 The change of solubility with temperature, the proportion of CO, being constant, is evident from Table II, which contains values interpolated from the curve expressing the observations by Wells, as well as the molar absorption-coefficient (c) of CO, and the calcu- lated value (K¢) at each temperature. There is a slight error involved in identifying KZ with the solubility-product constant [Cat*] [CO%] except at temperatures close to 18°, because in calcu- lating K; we have—for lack of better knowledge—assumed that the ratio (mr) of the first to the second ionization-constant of H,CO, is independent of the temperature; nevertheless, since these values of Ko were obtained from actual measurements of solubility, they enable one to calculate? for any temperature up to 30° the solubility of calcite under any conditions of CO, pressure or salt-concentration.3 TABLE II THE SOLUBILITY OF CALCITE UNDER ATMOSPHERIC CONDITIONS (P=0.00032), AND THE SOLUBILITY-PRODUCT CONSTANT AT SEVERAL TEMPERATURES Temperature Solubility ot Calcite eens Solubility, treduct : per Million ee Ki,Xro! fo) 81 0.0765 1 9} 5 75 .0637 1.14 Io 70 -0535 1.06 15 65 0455 0.99 20 60 .0392 0.93 25 56 .0338 0.87 30 52 0.0297 0.81 From the foregoing it follows that in order to decide definitely if a natural water is saturated with respect to calcite one must know: (a) the concentration of free CO, in the water, (6) the temperature, (c) the concentrations of the other constituents present. Of these the third is the only one which has in general: been satisfactorily ascertained, but it is only of subsidiary importance; experimental data on the two important factors are commonly either lacking or ™R. C. Wells, Jour. Wash. Acad. Sci., V (1915), 617. 2 For the mode of calculation see Johnston, op. cit., p. 2011. 3 This holds only so long as calcite is the stable phase. If the salt-concentration is such that some other carbonate (e.g., a double carbonate) is the stable phase, the appropriate constant must be employed in place of that characteristic of calcite. 734 JOHN JOHNSTON AND E. D. WILLIAMSON untrustworthy.t On the other hand, the concentration of free CO, in any water, at a given temperature, can be calculated by means of the known absorption coefficient of CO,, if the proportion of CO, in the atmosphere with which it has been in contact is known;? and as at the present time this proportion is usually close to 3 parts in 10,000 the corresponding solubility of calcite in natural waters should be close to the values given in Table II. Consideration of the published analyses from this standpoint leads to the conclusion that the surface layers of the warmer portions of the sea (in so far as they have been investigated), as well as many river waters,‘ are substantially saturated with calcite. Murray,’ in adverting to this question, states the opinion that ‘the ocean as a whole remains just about saturated for calcium carbonate’’; but this statement is without doubt too sweeping, except in the sense that the concentra- tion of CaCO, throughout the ocean is probably as great as it is in the warm surface layers. But there isalsomoredirectevidence. Thoulet® t The titration methods which have usually been employed for the determina- tion of free CO.—and to some extent of combined CO.—are altogether untrustworthy, since the results depend upon the amount of indicator added and upon other factors which have not been adequately controlled. This question is discussed at length in another paper (J. Johnston, Jour. Am. Chem. Soc., XX XVIII [1916], 947). Cf. also Morgulis and Fuller, Jour. Biol. Chem. XXIV (1916), 31. 2 With regard to the solubility of COz in a sea-water see C. J. J. Fox, Trans. Faraday Soc., V (1909), 68. 3See F. W. Clarke, Data of Geochemistry; but especially a paper by E. Dubois, “The Amount of the Circulation of CaCO; and the Age of the Earth.” (Proc. Acad. Wetenschappen Amsterdam [1901], pp. 43-62). Cohen and Raken (ébid. [t901], p. 63) have determined directly the solubility of CaCO; in an artifical sea- water at 15° and found about 55 parts per million; but their method of experiment is not unexceptionable and would tend to yield low results; they also conclude that this sea-water is saturated with CaCO;. Wells also (Jour. Wash. Acad. Sci., V [x915], 621) points out that the amount of carbonate carried by the Mississippi River diminishes steadily as it flows southward, i.e., in the direction of rising temperature. 4Indeed, the amount carried by many rivers is much in excess of the true solu- bility, indicating that some of it is in suspension. Where such a river reaches the sea, the salts cause the flocculation of this and any other suspended material, and in this way induce the formation of deposits there. 5 Murray and Hiort, The Depths of the Ocean (1912), p. 180. 6 J. Thoulet, “‘“Etude bathylithologique des cétes du Golfe de Lyons,” Annales de I’ Institut Océanographique, IV, fasc. VII (1912), pp. 32-35: AGENCIES IN DEPOSITION OF CALCIUM CARBONATE 735 studied silt grains taken from various parts of the Gulf of Lyons, and observed upon them films of calcium carbonate which had been precipitated during the process of sedimentation; this shows, therefore, that the water of the Gulf of Lyons is substantially saturated with CaCO;. Recent experiments of A. G. Mayer™ show that the sea-water about the coast of Florida is likewise substantially saturated, for shells exposed to it for a year lost no significant weight. Moreover, the investigations of T. W. Vaughan? on coral reefs ‘‘ show that submarine solution is not effective there [about Florida] as all the bays, sounds, and lagoons are being filled with sediment,” a con- clusion which accords “with the conclusions reached by numerous investigators in the Pacific, which are that the more or less continu- ous walls inclosing lagoons have been formed by constructional geo- logic processes and that lagoon channels and atoll lagoons are not due to submarine solution.” | The evidence just presented leads us, therefore, to the opinion that the surface layers of the ocean, except in the polar regions and within currents of cold water—in other words, the warmer portions of the ocean water—are substantially saturated with CaCO;. We wish to point out specifically, however, that this belief cannot be regarded as established (or indeed disproved) until trustworthy deter- minations of the several quantities concerned have been made; indeed, to emphasize the necessity of such investigations is the prime purpose of this paper. But in this connection it may be remarked that a permanent deposit of limestone can hardly result unless (1) the final solution locally in contact with it is saturated, or (2) the pre- cipitated carbonate is protected from the water by an organic tissue or otherwise, or (3) the process of deposition is rapid, in water circu- lating very slowly or not at all, under which conditions re-solution by diffusion is very slow. In this paper, which is dealing primarily with the chemical argu- ments, it would be out of place to take up the geologic lines of evi- dence which indicate that the ocean as a whole is not saturated with CaCO,, for this point is not at issue; but we may fitly advert to the tA. G. Mayer, Proc. Nat. Acad., II (1916), 28. 2T. W. Vaughan, Am. Jour. Sci., XLI (1916), 133. See also his earlier papers, especially in the Year Books of the Carnegie Institution of Washington. 736 JOHN JOHNSTON AND E. D. WILLIAMSON chemical arguments which have been adduced in favor of this posi- tion. Thus Tolman writes: As direct evidence that the ocean is not saturated with calcium acid car- bonate, we find (1) of the many hundred bottles of the Challenger’s samples of sea-water, from all depths and collected at all temperatures, kept several years, only one or two showed deposit of lime.?_ (2) Sea-shells from the bottom of the Pacific show corrosion and re-solution.3 The Pteropod shells are not found below fifteen hundred fathoms, and two thousand eight hundred fathoms is the limit for the globigerina ooze.4 (3) Thoulet found by actual experiment that sea-water will dissolve calcium carbonate from shells, corals, etc. (4) Usiglio, studying the evaporation of the Mediterranean water at Cette, found that no precipitate was formed until the specific gravity of the sea-water increased from 1.02, the specific gravity of the unevaporated water, to 1.0503, when the first precipitation begins, composed largely of calcium carbonate with ferric oxide.® Let us now consider these arguments severally. (1) That samples of surface water did not deposit CaCO, on standing is not good evi- dence one way or the other unless conditions were carefully con- trolled, for change of temperature or of concentration of CO, would influence the result. (2) This shows, of course, that the lower (and colder) layers are not saturated; to this point we revert later. (3) Reference to Thoulet’s paper shows that his work proves nothing as to the point in question, for neither the temperature nor the partial pressure of CO, was controlled. Indeed, he writes: “In the case of marble . . . . and of coral, the loss [of weight] in sea-water was negative. This result arises from the fact that small algae appeared . . . . the weight of which confuses the result.” (4) This observation also is no proof, in view of the well-known fact that solutions of calcium carbonate exhibit a great tendency to super- saturation when no solid CaCO; is already present; therefore it shows only that, when the density was 1.05, the degree of super- saturation had become so great that precipitation took place. From tC, F. Tolman, Jour. Geol., VII (1899), 604. 2 Challenger Reports, p. 221. 3 Jour. Geol., I, 504. 4 Challenger Reports, p. 221. 5 Comptes rend., CVIII, 753. 6 Encycl. Brit., XXI, 220. AGENCIES IN DEPOSITION OF CALCIUM CARBONATE 737 this discussion, then, it follows that three of these arguments are not conclusive as to the point at issue. Let us now consider the modes in which CaCO, may be precipi- tated. We shall for convenience arrange them under three heads, which, however, cannot be sharply differentiated: (1) direct evapo- ration of the water; (2) through organic agencies; (3) change of conditions, especially of temperature and concentration of free CO,, these being the predominant inorganic factors. 1. By direct evaporation.—When natural waters evaporate, CaCO, is commonly (though not necessarily) the first substance to be deposited, and may be very largely precipitated before any of the other salts separate;' the more soluble salts, moreover, will tend to be leached out of such deposits. But since all such deposits are of obvious origin and of minor importance, they need not detain us further. 2. Deposition through organic agencies —The agencies which come under this category are of the greatest importance and are predomi- nantly responsible for the deep-sea deposits, yet little as to their mode of action can be definitely stated until more is known about the biologic processes involved. This question is altogether beyond the scope of this paper; we shall mention merely two established effects of organic agencies, reverting to them later: viz., the abstrac- tion of free CO, from fresh water by growing plants,? and the production of ammonia in sea-water by decaying organisms or by bacteria. Both of these effects disturb the equilibrium in a solu- tion originally saturated with CaCO,, the former by diminishing the concentration of H,CO,, the latter by increasing the concentration tCf. Van’t Hofi’s Ozeanische Salzablagerungen. 2 Various references to these effects are given in Clarke’s Data of Geochemistry under ‘“‘Limestone.”’ See especially C. A. Davis, Jour. Geol., VIII (1900), 485, 494; IX (1901), 401. According to Murray, the calcareous algae common in the warmer oceans no doubt secrete their skeletons in the same way. See also S. T. Powell, ‘Effect of Algae on Bicarbonates in Shallow Reservoirs,” Jour. Am. Waterworks Assoc., II (1915), 703. 3 This question has been discussed recently by G. H. Drew (Carnegi (Institution Publ. No. 182 [1914], p. 7), and by Kellerman and Smith (Jour. Wash. Acad. Sci., IV [1914], 400), and so need not be treated here. Many decaying organisms and bac- teria (as well as the respiration of animals) produce COz, and to this extent they would act as an adverse influence on the deposition of CaCO. 738 JOHN JOHNSTON AND E. D. WILLIAMSON of CO; directly,’ the net result in either case being the precipita- tion of an amount of CaCO, which could readily be calculated by means of the equilibrium equations if the amount of CO, abstracted, or of ammonia produced, were known. But by this we do not imply thatan organism cannot secrete calcium carbonate except from a solu- tion already saturated with it. Nevertheless the possibility is open that the effects just considered may sometimes be in reality examples of the changes in conditions to be next considered—that the organism may be merely the agency which localizes the process, the mechanism which occasions the precipitation. It may even be that certain bacteria are abundant where CaCO, is being precipitated because there they can easily secure material—particularly CO.,—needed for their life-processes; on this. basis they would be concomitants, rather than causes, of the deposition of the carbonate. 3. Change of conditions.—The important physico-chemical factors are temperature, and concentration of free CO,, of the water; in com- parison with these two all other such factors are entirely subsidiary. As an illustration of the magnitude of the effect producible by change of these factors, a change in the proportion of CO, in the air in actual contact with the solution from 3.3 down to 3.0 parts per 10,000—a change? which may occur at the present time—of itself decreases the solubility from 65 to 63 parts per million, and so will cause the ulti- mate precipitation of the corresponding quantity of CaCO, from a solution already saturated with it. A similar amount will be deposited? if the temperature of the saturated solution rises about 2° C., the proportion of CO, in the air remaining constant; under these circumstances the concentration of free CO, in the water falls, and its diminution is responsible for the larger part of the diminished t The production of free ammonia causes an increase in the concentration of OH and therefore an increase in [CO 5], since the quotient [OH }/[CO ] is constant when P is constant, in accordance with the equation already given. It may be noted that the production by the decaying organisms of an ammonium salt such as NH,NO; or (NH,)2SO, would tend to increase the solubility of CaCO; and hence would not favor its deposition. 2 The range of variation in the course of geologic time has in all probability been very much greater than this with correspondingly larger possible consequences. 3 This of course implies that supersaturation does not take place; but in the sea supersaturation is highly improbable by reason of the great abundance of appropriate nuclei always present. AGENCIES IN DEPOSITION OF CALCIUM CARBONATE 739 solubility, since the diminution of the solubility-product constant of calcite with temperature is, as we have seen, proportionally less than that of the absorption coefficient of CO,. The abstraction of CO, from a saturated solution results ulti- mately, then, in the deposition of CaCO,, no matter what the agency which abstracts the CO,. This agency may be a diminished pro-. portion of CO, in the air, or a higher temperature, or both; or it may be organisms which make use of the CO, in their vital processes, or the production by bacterial action of ammonia, which indirectly achieves the same result; or, in short, it may be any way in which the concentration of CO, may possibly be diminished. Conse- quently, if the surface layers of the sea are saturated, as we believe they are, precipitation of CaCO, will be brought about wherever any of the foregoing agencies are operative. In this connection two points which are the consequence of accepting erroneous chemical data are to benoted. Thus Davis‘ in his excellent work on marls has made a slight slip. He observed that on bubbling oxygen gas through a‘solution containing CaCO, the latter was precipitated, and he attributed this effect to a specific action of the oxygen; but any other gas would have produced the same effect, which was actually due to the sweeping out of the CO, from the solution. Nor is it necessary to consider, when ammonia is being produced, whether it appears as hydroxide or as carbonate, or whether there is a subsequent metathesis with calcium sulphate or chloride or some other reaction; in either case the net result can be predicted immediately from a consideration of the effect of the added substance upon the concentration of calcium-ion and of carbonate- ion,? and of the magnitude of the product [Cat*] [CO%] in relation to its precipitation value. To some this procedure may appear complicated; in reality, while it pays no heed to those easily derived arithmetical equations so often considered as representing reactions, it takes into account the several equilibria which must be adjusted, tC. A. Davis, Jour. Geol., VIII (1900), 487. 2 Change of concentration of CO% affects, and is affected by, the concentration of HCO, and OH, these being all dependent variables; see Johnston, of. cit. 3 Provided that calcite is still the stable solid phase in equilibrium with the solu- tion; cf. footnote 3, p. 733. 740 JOHN JOHNSTON AND E. D. WILLIAMSON and is the only procedure which will yield accurate results and lead to correct conclusions. Moreover, a comprehension of this question is desirable because this apparently complex equilibrium is typical of what takes place in many other systems, aqueo-igneous and igne- ous as well as aqueous; it is in but very few cases, however, that we know even what molecular species are important factors in the equilibrium, and in still fewer is any information available as to the quantitative relations at equilibrium. There has, moreover, been considerable misapprehension as to the réle of hydrostatic pressure in increasing the solubility of CaCO,; thus in a recent paper Daly' writes: ‘On account of the higher temperatures and lower bottom pressures (pressure increasing the solubility of the carbonate) of the shallower water we should expect the rate of chemical precipitation of calcium carbonate at the bottom to be concentrated in the neritic (epicontinental) and shallower bathyal regions.” And many smiliar statements relative to the effect of hydrostatic pressure might be quoted. As a matter of fact, the hydrostatic pressure? acting on the water is of itself an abso- lutely negligible factor; thus water a mile below the surface of the sea will hold in solution an amount of CaCO, which does not differ “by an appreciable quantity from the amount the same water at the surface will hold, provided that the concentration of free CO, and the temperature be the same in both cases. The increased solubility with depth in the ocean is due entirely to the lower temperature of the water and to the increased proportion of free CO,, but not at all to the increased hydrostatic pressure there prevailing. The only pressure which does affect the solubility is the partial pressure, i.e., the proportion, of CO, in the layer of atmosphere? in ™R. A. Daly, Bull. Geol. Soc. Am., XX (1909), 156; also in “Geology of North American Cordillera,” Memoir No. 38, Geol. Survey of Canada, Part II (1912), p. 651. Italics are ours. 2 Increase of hydrostatic pressure decreases the solubility of some substances. In any case the effect is very small indeed; its magnitude and direction can be calculated if the appropriate data on volume changes are known. Cf. Johnston and Niggli, Jour. Geol., XXI (1913), 504, where references are given. 3 The proportion of CO, in the air increases, ceteris paribus, as we pass from higher to lower levels; but this is a factor of no moment to the present discussion because the diffusion downward through the water is in all probability very slow in comparison with the natural circulation of the water. AGENCIES IN DEPOSITION OF CALCIUM CARBONATE 741 contact with the solution, for there is a definite and quickly attained equilibrium between the proportion of CO, in the adjacent atmos- phere and the concentration of free CO, in the water, the factor of proportionality being the absorption coefficient (the solubility) of CO, in the solution at the particular temperature. It is true that water at depth can hold more CO, in solution, if it gets hold of it, for in that case bubbles of CO, gas cannot form until its virtual pressure just exceeds the hydrostatic pressure; but slow diffusion upward would tend to equalize the concentration at various depths. In the ocean, on the other hand, the content of CO, is only what it was able to absorb when at the surface, supplemented by that which has been produced by organic processes—the latter being in all probability but a small fraction of the whole in deep water.t How- ever this may be, it is manifest why the water at depth should con- tain more CO,, for its present low temperature, retained from its polar days, establishes the fact that when at the surface in contact with the atmosphere it was cold, and lowering of temperature increases very markedly the amount of CO, which water can absorb through contact with an atmosphere containing a constant propor- tion of CO,.?. This fact, combined with its present low temperature —for, as we have seen, lowering of temperature of itself increases the solubility of CaCO,—suffices to account for the well-known fact that all shells and tests disappear in the depths of the ocean. Now let us revert to the consequence of abstraction of CO,, and consider what will happen when, in the course of the oceanic circu- lation, this cold water, which carries more CO, and more CaCO, than the warmer surface waters,‘ reaches the surface and is slowly warmed. t Buchanan (Proc. Roy. Soc., XXII [1874], 483) writes: “‘Down to nearly 2,000 fathoms life is still abundant; below this depth, however, the amount rapidly decreases till, at about 2,800 fathoms, it is, for carbonic acid producing purposes, practically extinct.” 2 Thus in contact with any atmosphere, water (or a dilute salt solution) absorbs about twice as much CO: at o° as at 20°. 3 See the report of the Challenger expedition or the work of Sir John Murray. In the present connection it is immaterial whether these shells consist of calcite or aragonite, although assertions to the contrary may be found. 4This appears a necessary prerequisite, no matter what be the mechanism of precipitation. Dittmar, in his article in the Encyclopaedia Britannica, states that there is a slight but indubitable increase in concentration of calcium with depth. Moreover, 742 JOHN JOHNSTON AND E. D. WILLIAMSON In the first place, it will gradually lose CO, to the air, the residual concentration of free CO, being dependent at any moment upon the temperature of the water and the proportion’ of CO, in the air at that place. The consequence of this loss is that the amount of cal- cium in solution will at some point exceed the concentration which the water is able to hold in solution—r, in other words, the product [Cat*] [CO] reaches its characteristic precipitation value—where- upon precipitation? sets in, and continues thereafter so long as the temperature continues to rise. This process is without doubt taking place now in tropical and subtropical regions wherever and whenever the necessary conditions are fulfilled. It has been correlated’ with the abundant bacterial and planktonic life found under such cir- cumstances, and there would seem to be little question that the organ- isms are a factor in the process, if only in the sense of catalyzing it. But may it not be, in some cases at least, that the organisms are abundant there because of the abundance of the CO, available for their life-processes in such water? For it is to be borne in mind that the precipitation of CaCO, is accompanied by the setting free of an equivalent quantity of CO, which, if not used up in the sea, will pass into the atmosphere. Be this as it may, the physico-chemical factors are in themselves competent to account for the precipitation’ of CaCO, on a large scale, and the prerequisite conditions for deposition by this means do not differ materially from the postulates required for precipitation by bacterial action or by organisms generally. Buchanan (Proc. Roy. Soc. London, XXIV [1876]) writes: “There is usually more COz in waters taken from the bottom and intermediate depths than in surface water; but if regard be had to the temperature of the water, it will be seen that there is but little difference in the amount in waters of the same temperature from whatever depth they have been derived.’ It is to be observed that these determinations all refer to low latitudes; conditions in the Polar regions may well be different. t All experiments indicate that this proportion departs in general very little from 3 parts in 10,000, except in or near large towns. Off the west coast of Greenland, however, amounts up to 7 parts in 10,000 were observed by Krogh (Meddelelser om Gronland, XXVI [1904], 409). 2 Supersaturation is under these conditions obviously a negligible factor. 3G. H. Drew, Carnegie Inst. Publ. No. 182 (1914), p. 7; Kellerman and Smith, Jour. Wash. Acad. Sci., IV (1914), 400. See also recent papers by T. Wayland Vaughan. 4 Likewise for its re-solution. AGENCIES IN DEPOSITION OF CALCIUM CARBONATE 743 Indeed, there are several facts which point to a parallelism between the amount of lime secreted by organisms and the degree of satura- tion of the sea with respect to CaCO,; thus the animals of the warm seas secrete more lime, on the average, than the same types in cold seas;t and, according to Murray,” ‘“‘on the whole, lime at the present time appears to be accumulating toward the equator.”’ These obser- vations directly corroborate the idea that solubility is a significant factor even in the secretion of lime by organisms; that the decreasing abundance of calcareous organisms toward the polar regions is a question not only of the decrease of general vitality (rate of growth and of reproduction) with lowering of temperature, but also of the decreasing capacity’ of the organism to secrete CaCO, from colder sea-water, this being associated with the fact that, though the con- centration of lime is no smaller in the colder water, the degree of unsaturation is greater the colder the sea-water. According to Murray, ‘‘a limited amount of purely inorganic pre- cipitation does, indeed, take place in coral reefs and some shallow water deposits and in the Black Sea.”4 Now it has been argueds that chemically precipitated limestones are due to the production of ammonia by decaying organic matter; according to this view such limestones could form only when conditions were such that a long- continued process of persistent decay was possible. According to the view emphasized in the present paper—and, be it noted, this is primarily a chemical, rather than a geological, question—chemical tSee citations from the Challenger Reports, in Chamberlin, Jour. Geol., VII (1899), 576-77. 2 The Depths of the Ocean, p. 180. “In very deep water, even within the tropics, the calcareous shells do not accumulate on the bottom, being apparently removed through the solvent action of sea-water, and with increasing depth the Globigerina ooze passes gradually into another pelagic type, usually Red Clay” (p. 164). “Pteropod ooze is limited to the tropical and subtropical regions, usually in the neighborhood of oceanic islands and on the summits and sides of submarine elevations; it is found in relatively shallow water, and covers a relatively small extent of the ocean floor” (p. 167). 3Tt would be of interest to know if these calcareous organisms could secrete CaCO; from colder water kept saturated with calcium carbonate. 4 The Depths of the Ocean, p. 178. 5 Most recently by R. A. Daly, Bull. Geol. Soc. Am., XX (1909), 153; in more extended form in Memoir No. 38, Geol. Survey Canada (1912), pp. 643 f. 744 JOHN JOHNSTON AND E. D. WILLIAMSON precipitation would take place wherever, and so long as, a current of water saturated with calcite was being warmed. ‘These views are not at all mutually exclusive; but their implications differ, and it ought to be possible to decide by appropriate observation and deduction in how far either has been a dominant cause on a large scale. The magnitude of the scale of this presumed process of precipi- tation through purely inorganic agencies depends primarily upon the rate of circulation and upon the amount of calcium carried by this water rising to the surface. We shall now consider the competence of these agencies as geologic factors. In doing so let us suppose that a cold current of sea-water is not saturated with CaCO, until it has reached a temperature of 15°, and that this current after traveling 1,000 kilometers (600 miles) has attained a temperature of 20°; further, that the water in this stretch of 1,000 km. is changed 10 times a year, corresponding to a current speed of 1,000X 10/365 or 27.4 km.aday. Now this water in being warmed from 15° to 20° would precipitate 5.4 parts CaCO; per million by weight, or 2 parts per million by volume; on these assumptions,’ therefore, in the course of a year the mean thickness of the deposit (presuming that all of the precipitate finally settles to the bottom within this stretch) would be 2/1,000,000 of the depth of the current. Hence if the depth of the current is 100 m., the average deposit over the whole area would be, on the specific assumptions just mentioned, 2 mm. yearly.2 This estimate is probably a minimum, particularly because we have supposed that deposition would take place over the whole area, whereas in reality deposition would be localized (e.g., if there is, as is likely, a more rapid warming up at some places) so that the deposits actually formed would be thicker. Moreover, the deposit would be redissolved whenever the current is underlain by colder unsaturated water; therefore actual deposits belonging to this cate- gory should occur only in localities bathed by currents which are «The numerical values adopted were chosen as being reasonable; in any case these calculations will serve as an illustration, and anyone may make similar calcu- lations using whatever numerical values he deems most consonant with the facts. 2 This corresponds to about 5,000 tons CaCO; per square kilometer per year, or to a thickness of about 8 inches in a century. AGENCIES IN DEPOSITION OF CALCIUM CARBONATE 745 warm and hence comparatively shallow and rising in temperature as they proceed. Consequently, if deposits of CaCO, are being formed in this way—and there is no direct evidence at hand which contradicts this view—it should be possible to correlate the position and rate of formation of such precipitated deposits with other things by means of series of bathymetrical observations on the manner of flow of the currents, the temperature of the water, and, above all, the concentration of free CO,. As regards the latter, it may be said that the methods hitherto in vogue are very faulty indeed,‘ and that systematic, accurate determinations of the free CO, (which can be made by proper choice of method) are very much to be desired, not only on account of their bearing on the present question, but because an accurate knowledge of the concentration of free CO, is of high importance? in connection with many biological problems, both theoretical and practical. Nor would the establishment of this presumed correlation between deposits of CaCO, and the physico-chemical conditions pre- vailing in the ocean be of importance only in relation to present-day formations belonging to this category; it would also be of use in interpreting past deposits of this character which have persisted and in co-ordinating them with other factors. For the rate of formation . of such deposits (including the limiting case of zero rate) depends obviously upon the mode and rate of circulation of the ocean and the amount of calcium carried by the water rising to the surface, secondarily, therefore, upon the amount of calcium carbonate brought down by the rivers to the sea; all of which depend ultimately upon the physiographic and petrologic character of the land surface, upon the magnitude of the seasonal variations and regional differ- ences, upon the climate over the whole earth, and upon the propor- tion of CO, in the atmosphere. It would lead too far to discuss this question in all its bearings; in order to show the importance of the t See Johnston, Jour. Am. Chem. Soc., XX XVIII (1916), 947. | 2 The free and combined CO:z and the alkalinity of the solution are not independent variables, a fact often forgotten; and doubtless many effects ascribed to a change in alkalinity are due equally, or primarily, to a change in the CO? equilibrium in the solution. 3 The discussions of Chamberlin (Jour. Geol., VII [1899], 545, 667, 757), Tolman (ibid., p. 585), Krogh (Meddelelser om Gronland, XXVII [1904], 334), and others, require some revision in the light of data available since that time. 746 JOHN JOHNSTON AND E. D. WILLIAMSON last factor—apart altogether from the influence of CO, as an agency disintegrating the rocks—we have calculated the concentration of the free and combined CO, in sea-water at three temperatures for several proportions of CO, in the atmosphere. The specific assumptions made in these calculations are: (a) that the molar absorption coeffi- cient (c) of CO, is the same as ina 0.6 N (3.5 per cent) solution of sodium chloride; (b) that the water is always saturated with respect to calcite, so that we are justified in using the solubility-product constant (K() corresponding to the temperature; (c) that the degree of ionization of the carbonate is 0.6, a value which is probably high rather than low.’ On this basis the formula becomes A=total CO,=cP+V 11200cK/P/o.6, where the first term represents the free CO, and the second the total combined CO,, each expressed in moles per liter; whence by multi- plication by the factor 0.044 one obtains the result in grams CO, per cubic meter (parts per million) as given in Table III. TABLE III THE CONCENTRATION OF FREE (f), COMBINED (0), AND ToTaL (A) COz—ExPRESSED IN GRAMS COz PER CUBIC METER (PARTS PER MILLION)—IN SEA-WATER AT SEVERAL TEMPERATURES AND SEVERAL PARTIAL PRESSURES (P) OF COz IN THE ATMOSPHERE; CALCULATED ON THE BASIS OF THE SPECIFIC ASSUMPTIONS MEN- TIONED ABOVE 10° 20° 30° CO: in ADJACENT c=0.0463 €=0.0335 c=0.0260 ATMOSPHERE AS Kco=1.06X10—8 Ko=0.93X1078 Kco=0.81X107—8 Partial Parts per eon Become f 6 A if b A tf b A ©.00025 | 0.51 | 81.5 82 | 0 .0003 0.61 | 86.6 87 | 0 -00025 | 0-71 || OL: xr 92 | 0 .003 6512 |1870 |) 1935 4. 0.03 61.2 |402. 463 |44 The figures for total CO,, derived in this way, are in substantial agreement with the results of analyses of sea-water; in any case the t Murray and Hiort (The Depths of the Ocean [1912], p. 175) estimate the aggregate degree of ionization in sea-water to be 0.9; but this is undoubtedly much too high. AGENCIES IN DEPOSITION OF CALCIUM CARBONATE 1747 relative values for the different conditions are probably good even if the absolute values are inaccurate. According to the table, the concentration of total CO, in water at constant temperature varies practically as the cube root of P, for the small values of P; in other words, a change of 3 per cent in the present proportion of CO, (e.g., from 3.1 to 3.0 parts per 10,000) will produce a change of but 1 per cent in the concentration of total CO, in the sea-water. Likewise under present conditions (i.e., P=0.0003) the total CO, in the ocean decreases about 1.5 per cent of its value for each degree of rise in temperature. At the higher pressures of CO, the proportion of free CO, in the water becomes relatively much more important; but a hundred-fold increase in the proportion in the adjacent air would cause only a fivefold increase in the total CO, in the sea. In this estimate and in the subsequent discussion, be it noted, the assump- tion is implicit that the water is continuously saturated with CaCO, at 15° for all values of P, which in turn implies that conditions were such that the rivers transported to the sea sufficient lime to achieve this. On this basis, therefore, if the present amount of CO, in the atmosphere were increased a hundred fold, the total amount of CO, in atmosphere and ocean would be only six times as much as it is now; the conditions of equilibrium always being such that a chance in the proportion of CO, in the atmosphere is minimized, not through a permanent change in the proportion of free CO, in the sea (and of its alkalinity), but ultimately by means of the precipitation or solution of a definite quantity of CaCO3. Let us now make a computation of the ratio of the total amount of CO, in the whole ocean to that in the whole atmosphere, this being, of course, a measure of the capacity of the ocean to regulate the pro- portion now present in the atmosphere. We assume again that the ocean as a whole would be saturated with CaCO, if its temperature were 15°,! and that its mean depth is 3,600 meters;? on this basis the mean amount of CO, under each square meter of surface of the sea is 813,600 gm. or 290 kg. The mean amount above each square « This is just equivalent to the assumption that the average proportion of CaCO; throughout the ocean is that which corresponds to its solubility at 15°, about 60-70 parts CaCO; per million for values of P not far removed from 0.0003. Ci. Table I. 2 See Encycl. Brit., article “Ocean.” 748 JOHN.JOHNSTON AND E. D. WILLIAMSON meter of the earth’s surface (sea and land together) is 3 kg.; for if the proportion of CO, in the air at the earth’s surface is 3 parts per 10,000, the proportion in the whole atmosphere is 2 in 10,000 by volume, hence 3 in 10,000 by weight, or 3 kg. per square meter. Consequently, since the ocean covers about 71 per cent of the total surface of the globe, the ratio total CO, in ocean __ 290X0.71_¢ total CO, in air 3 ae In other words, the ocean contains about 70 times™ as much CO, as the air, on the basis of the assumptions specified above. On this basis the total CO, now present in the ocean and atmosphere com- bined would form a layer of CaCO, only about 17 cm. thick over the whole globe, or about 86 cm. (nearly 3 feet) over one-half the present land area; likewise if the amount of CO, in the atmosphere were 100 times as much as at present, the corresponding values would be slightly more than 6 times as large, namely, 110 cm. over the globe, or 550 cm. (18 feet) over one-half the present land area. The possible deductions, however, must remain uncertain until series of simultaneous accurate determinations of free and total CO., temperature, and salinity in the sea at various depths and in different localities shall have been made. The precipitation of CaCO, in forms other than calcite——Besides calcite, which is the stable crystalline form of CaCO, under all ordinary conditions, there are two unstable crystalline forms, aragon- ite and yu-CaCO,;, which may precipitate under certain circum- stances. This whole question is discussed at length in another paper, to which the reader desirous of further information is referred ;? we shall here merely recapitulate the conclusions relevant to the present discussion. The existence of the yu-form in nature has not been definitely established, possibly by reason of the fact that some of the criteria which have been used to differentiate the several forms of CaCO, have not been unexceptionable, possibly on account of its t This estimate is higher than that (27 times) of Krogh (Meddelelser om Gronland, XXVI [1904], 420) or that (55 times) given in Chamberlin and Salisbury’s Geology, II, 661, which see, with respect to the whole discussion. 2 “The Several Forms of Calcium Carbonate,” Johnston, Merwin, and Williamson, Am. Jour. Sci. (4), XLI (1916), 473. AGENCIES IN DEPOSITION OF CALCIUM CARBONATE 749 instability, for in presence of water the u-form transforms to calcite in a few days. Calcite also appears as spherulites and as “‘amor- phous” CaCO,; but there is little question that the divergent prop- erties of the latter are due entirely to its fineness of grain, i.e., to its extent of surface in proportion to its mass. Consequently the only form other than calcite which we need consider here is aragonite. Apparently aragonite is formed in nature (a) through organic agencies (e.g., in certain shells), (6) by deposition from hot springs, (c) when an isomorphous carbonate is present to serve as nucleus, and (d) by chemical precipitation in saline waters, even at ordinary temperatures, under circumstances which we are unable to specify except by saying that the presence of sulphate appears to be a favorable factor. But pure aragonite cannot persist for any length of time in presence of water and calcite, hence only in special cir- cumstances will it be found persisting in the sea. There is, however, the possibility that aragonite may take up in solid solution enough material to bring its own solubility below that of calcite, and hence in the saline solution in equilibrium with the solid solution to render the latter stable with respect to calcite; on this basis it is possible that such impure aragonite may persist in contact with sea-water under certain circumstances, although when exposed to the action of meteoric waters it would soon transform to calcite. However this may be, the circumstance that CaCO, precipitates otherwise than as calcite would not of itself affect appreciably anything stated in this paper, since the whole effect would be that ensuing upon the substi- tution for the solubility-product constant of calcite of the corre- sponding value for the other form, the latter being certainly no more than twice as great as the former; so the precipitation of the less stable forms is therefore of only subsidiary importance in the present connection. Summary.—Though organic agencies are predominantly respon- sible for the deposition of calcium carbonate, yet the purely inorganic factors should also be taken into account in discussions of the mode of deposition: In this paper emphasis has been laid on one point which has not received adequate recognition; namely, the concen- tration of calcium relative to the limiting saturation concentration of calcium carbonate under the particular conditions, or, in other 750 JOHN JOHNSTON AND E. D. WILLIAMSON words, the relative degree of saturation with respect to calcium car- bonate in the ocean. The importance of this factor is obvious if we recollect that the chance of a permanent deposit is, ceteris paribus, greater the more nearly saturated the surrounding water is; its neglect is doubtless due to the erroneous and misleading statements as to the solubility of CaCO, which have been prevalent. The solu- bility under specified conditions can now be calculated with the requisite accuracy; it is affected materially by variations of tem- perature and of concentration of free CO, such as occur in nature. For example, a change in the proportion of CO, in the adjacent air from 3.2 to 3.0 parts per 10,000, or an increase of temperature of 2° C. would result ultimately in the precipitation of about 2 gm. CaCO, from every cubic meter of a solution saturated with it. Comparison of the solubility as calculated with the available ana- lytical data indicates that the warmer surface layers of the sea are substantially saturated with respect to calcite, and consequently that precipitation is to be expected wherever the water is being warmed or is losing CO,, or both, and this independently of any other agencies. Indeed, these inorganic factors must be considered no matter what may be the agency inducing precipitation; for example, there is ground for believing that calcareous organisms are more abundant the more nearly saturated with CaCO, the water is. The view here advocated, that a somewhat greater réle be assigned to the inorganic factors than has hitherto been usual, does not conflict with other views—it merely shifts the emphasis a little; nor does it con- flict with any facts that have been definitely ascertained. Its precise importance can be determined only by accurate determination of temperature, salinity, and particularly of concentration of CO,— free and total—of the water carried out systematically over the ocean; the results of such an investigation, properly carried out, would have an important bearing on many outstanding biological, as well as geological, problems. NOTES ON THE STRUCTURAL RELATIONS OF AUSTRALASIA, NEW GUINEA, AND NEW ZEALAND E. C. ANDREWS Sydney, New South Wales, Australia Introduction Previous Work Done and Scope of Present Work Growth of Australasia from Pre-Cambrian to Recent Time Evidence of the Ore Deposition Conclusion INTRODUCTION The accompanying brief note is an attempt at the co-ordina- tion of our increasing knowledge of the structural development of Australia and the neighboring islands. The ideas given in this note are intended only as a temporary viewpoint from which to consider the work of the great pioneers of geology in Australia and as an inference or tentative hypothesis to stimulate interest in those magnificent field problems in Austral- asia, New Guinea, New Caledonia, and New Zealand, which call so urgently for solution. In this way it is hoped that the scheme here proposed will serve as a rough clue to the unraveling of certain vexed questions in the stratigraphic and structural history of Australia. Several difficult points need explanation before any simple account of the building of Australia would be possible. Thus in the discussion of the Devonian it must not be forgotten that folds supposed to be of this age occur in Northwest Australia. Highly altered rocks of unknown age and of large area occur also in North- ern and Northeastern Queensland, and the occurrence of these has not been explained in the present note. Then again, it must not be forgotten that our knowledge of some of the Permo- Carboniferous rocks, such as the Gympie of the Queensland geolo- gists, is far from satisfactory. 752 752 E. C. ANDREWS Again, it is not known how many of the observations of the older workers in connection with the strike of folds were merely local and how many were conducted on a large scale. Acknowledgments.—The writer is deeply indebted to Pro- fessor Leo A. Cotton and Dr. W. N. Benson, of Sydney University, for their kindness in reading the report in manuscript and for supply- ing additional information as to literature on the pre-Cambrian and Ordovician, and for kindly criticism of the notes on the Devo- nian, the Permo-Carboniferous and the Trias-Jura. Previous workers.—lIt is the desire of the writer at this stage of our scientific development in Australia to draw attention to the work of the pioneers of geology in the great island continent. Australian pioneer geologists, in common with Australian explorers and miners, and in common also with American pioneers, have breathed the inspiration of their own mighty surroundings. Fore- most among the pathfinders of geology in the country under consideration—men who crossed trackless wastes and endured untold discomforts in their pursuit of knowledge—were W. B. Clarke, R. Daintree, and A. R. C. Selwyn; others who followed in the track of these giants, but who nevertheless bore much of the heat and burden of the day and were either worthy successors or contemporaries of the pioneer trio, were H. Y. L. Brown, J. E. Carne, T. W. E. David, W. Howchin, R. Logan Jack, A. Gibb Maitland, R. Murray, S. Stutchbury, R. Tate, W. H. Twelvetrees, and C. S. Wilkinson. Among them also must be named the paleontologist, R. Etheridge, Jr., whose labors in the cause of Australian paleontology have done so much to simplify the task of the field workers. A whole group of younger enthusiasts have built and are today building on the work laid down by these pioneers. PREVIOUS WORK DONE AND SCOPE OF PRESENT NOTE The earliest definite statement known to the writer concerning the building of Australia as a whole was made by T. W. E. David." In this detailed account Professor David’s descriptions imply « Presidential address on ‘The Growth of Australia,” Proc. Linn. Soc. N.S. Wales, 1893, pp. 547-607. AUSTRALASIA, NEW GUINEA, AND NEW ZEALAND 753 the growth of Australia as from west to east. In 1911 the same _ writer amplified his earlier statement and said: ‘Since the close of Paleozoic time Australia has been subjected to broad warps, but not to true folding except in the direction of New Guinea, where Cre- taceous, and even early Tertiary, strata are highly folded. New Guinea is thus a new fold region; and even in Australia tectonic movements are newer as New Guinea is approached” (p. 59). H. I. Jensen? also, in a later note, discussed the gradual growth of the eastern portion of the continent. Like David’s earlier reports, this paper of Jensen’s is important and suggestive. Jensen approaches the problem of Eastern Australian history also from the viewpoint of “‘petrological unity.” He, however, considered that the folding of the Permo- Carboniferous sediments was sporadic and had died out practically in Northern New South Wales. In 1914 David’ presented an epitome of Australasian geology. In this he said: ‘‘The latest folding to which the earth’s crust in Australia has been subjected belongs to late Carboniferous time”’ (p. 256). He qualified this, however, by the statement: “‘The strata in the Permo-Carboniferous system are either perfectly horizontal or disposed in broad open troughs and arches. Only in the case of the strata of Drake and Undercliffe in New England and the Ashford areas [New South Wales] and the Gympie area in Queensland, are the strata of this system highly disturbed near granite intrusions”’ (p. 267). The reader is referred for a consideration of this statement to the discussion of the field evidence in connection with the Permo- Carboniferous. It will then be seen how incomplete is our knowl- edge of the ageof the sediments of Eastern Australia lying to thenorth of Sydney, and hence how great the need for caution to be exercised in coming to any definite conclusion as to the scheme of structure. =T, W. Edgeworth David, “Presidential Address,” Proc. Roy. Soc. N.S. Wales, IQII, pp. 15-76. 2‘The Building of Eastern Australia,” Proc. Roy. Soc. Queensland, July, 1911, Pp. 149-98. 3T. W. Edgeworth David, “The Geology of the Commonwealth” (Federal Handbook), Brit. Assoc. Adv. Sci., Australian meeting, 1914, pp. 241-325. 754 E. C. ANDREWS Origin and scope of present note-—The idea of writing a paper similar to the present one was conceived as far back as 1905-6, when the writer was surveying an area of folded sediments in Northern New South Wales. Previous workers had considered these beds as belonging to the older Paleozoic because they were strongly folded, whereas the beds of known Permo-Carboniferous type in Australia at that time were either horizontally bedded or only moderately domed. David, however, in connection with these beds, had pointed out as far back as 1893: ‘‘I have, however, lately. come to the conclusion that the whole of the Paleozoic sedimentary rocks of the Vegetable Creek district, provisionally classed by me as Upper Silurian or Devonian, are referable to the Gympie horizon’ [presumably Carboniferous.—E. C. A.]. During the progress of the survey these beds were discovered to contain many characteristic Lower Marine (Permo-Carboniferous) fossils, as probably also some Upper Marine types. The area of these Permo-Carboniferous types was proved to extend far to the north and west afterward by the field work of Carne and the writer.? In Southern Queensland, near Warwick, these two observers found Lower Marine rocks folded in the most complicated manner,3 while in 1908 during a visit to Mount Morgan, about 500 miles north of the New South Wales border, they saw rocks indicated on the Queensland geological map‘ as of the same age as the Warwick types, also highly folded. In 1901 Mr. C. Hedley and the writer traced rocks at intervals from a little north of Townsville to near Cooktown, all highly con- torted, and all shown on the Queensland geological map as of the same age as the Warwick beds. ‘This caused the writer to consider t “Presidential Address,”’ Proc. Linn. Soc. N.S. Wales, 1893, pp. 586-87. 2 J. E. Carne, “The Tin-Mining Industry of New South Wales,” Mineral Resources No. 14, Dept. Mines, Sydney, N.S.W., 1911, pp. 54, 70, 71; E. C. Andrews, “‘The Drake Copper and Gold Field, N.S. Wales,” Mineral Resources No. 12, Dept. Mines, Sydney, N.S.W., 1908, pp. 3-11. 3 Drake Report. See plate opposite p. ro. 4R. L. Jack and R. Etheridge, Geology of Queensland, Brisbane (by Authority), 1892, plate 60. AUSTRALASIA, NEW GUINEA, AND NEW ZEALAND 755 the significance of the peculiar problem of the Permo-Carboniferous, inasmuch as the whole of the work of the field officers of the geo- logical survey of New South Wales had proved that the Permo- Carboniferous strata south of the Hunter River (lat. 33° S.) lie almost horizontally. The independent testimony of the ore deposits of New Zealand and Australasia was then examined, and the problem of Austral- asian growth was formulated in the following terms: During the progress of geological time folding movements in Australasia retreated north and east, while ore deposition moved parallel with these movements. The growth of New Zealand does not appear to be known defi- nitely, but the New Guinea and the New Caledonian movements appear to have opposed the Australian direction of growth. A study both of structure and of ore deposits suggests that New Zealand, Australia, and New Guinea have had independent origins. GROWTH OF AUSTRALASIA FROM PRE-CAMBRIAN TO RECENT TIME Pre-Cambrian —The greater portion of Australia, which stretches to the west of a line drawn from the southwest of Tas- mania to the center of North Queensland,’ is composed of pre- Cambrian schists, gneisses, granites, and allied rock types. The dominant strike of the foliations is northwest and southeast, approximately, with a marked tendency to show large local, or even regional, corrugations in the eastern portion of the area. This is well shown on David’s map accompanying his report of 19117 to the Royal Society. It is possible that at the close of the pre- Cambrian period in Australia the land surface extended across the southeastern or even the eastern portion of the continent. This is suggested, not only by the schists of Cloncurry in Northern Queensland mentioned by Woolnough, but also by the pres- ence of great masses of schists and gneisses of unknown age in Eastern Victoria extending northward into the Cooma district. 2W. G. Woolnough, Bulletin of the Northern Territory, No. 4, Dept. External Affairs, Melbourne, 1912, p. 51. 2 ‘Presidential Address,” Proc. Roy. Soc. N.S. Wales, t911. See large map accompanying the paper. zs UN AL 120 ee" bye za gen Yah 4 Metall he rit Ie aa Hh Oi York Song, OM ag Lay ae i) { A ch» prot Luatinye y ‘ in {imate A, « Dey 2 /Tpetvery yh he t Lond cre rave Range Desert i tt tie ayes youth kyr’ \ olan Le 5 ~ i | | sdmipalty [2 O18! Matthios! gqutin I! lata Torre oe Bismarck Archies ago pstharaday) Keron Siauittr waminonr i re Maley, 4 siqynen SEP \ Gc Dypertion gy! of ork |, cnarlotle fey Musa Mu Setawan ver jn? i A Alle Cyoroate! pyich I Cavpenthos % pe Fycyialtery- = dptiren poled! Giz | pellag ylt eA gy f al ) = hitart } LM achat { aGdonia opi (Frenchy AA “n Wo Si NEMS NEV aime (ireken Mid) Py» Finder | m With v Port Minter {) My W. 4 on Jackson Te Port Makquarte 2 ager ar pt nya \ Vinton Pron vw Oe a 0 “shure a a car’ Macquarte Bg NSWISL —— of 175 Gowen Ld — len At Leelee va 756 E. C. ANDREWS Browne,’ however, inclines to the belief that a portion of this area, at least, is of Ordovician age. Other schist masses exist in Queensland. The possibility of sediments and other rock masses being molded onto, or being wrapped round, these resistant blocks is thus sug- gested. Cambro-Ordovician.—Since the momentous pre-Cambrian period the greater portion of the area mentioned appears to have been a positive or buoyant element till the present day. A great negative area appears to have existed at this time over Eastern Tasmania, Victoria, and New South Wales. It is possible, however, that a positive element existed in this period in Southeastern Victoria and New South Wales. The Cambrian sediments are more in evidence on the western strip of this area, while the Ordovician are common on the southeastern and eastern portions. It is possible that the Ordovician sediments of the more eastern areas are conformable to the Cambrian, but there is an unconformity between the shallow water forms of the two in the MacDonnell Ranges of Central Australia. Silurian.—At the close of the Ordovician there was a very powerful folding movement. Wherever the Ordovician occurs in New South Wales or Victoria, it is strongly folded and altered. The new land surface was carried far to the north and east by this fold- ing movement. Ordovician sediments occur quite near the coast about 100 miles south of Sydney, and they outcrop within 120 miles (lat. 33° S.) of Sydney in a direction west-northwest. Thence to the pre-Cambrian outcrops of the more western areas they may be seen in many places, exposed by the stripping of their Devonian cappings.? In the majority of the localities observed the strike of the sediments is west of north. During the Silurian the old negative area which had been occu- pied by the Cambro-Ordovician sediments once more sank, and the sea transgressed far to the west, almost to Broken Hill (long. 141° tW.R. Browne, ‘‘The Geology of the Cooma District, N.S. Wales,” Jour. Proc. Roy. Soc. N.S. Wales, XLVIII (1914), 172-222. 2E. C. Andrews, “The Canbelego Gold and Copper Field,” Mineral Resources, No. 18, Dept. Mines, N.S. Wales, 1913. See maps and sections. AUSTRALASIA, NEW GUINEA, AND NEW ZEALAND 757 E.), not nearly so far west, nevertheless, as the Ordovician sea had transgressed. ‘This sea was shallow in places and full of islands. As in the Cambro-Ordovician period, sandy sediments and con- glomerates also were deposited in the west, while great areas of coralline limestone were deposited in the eastern portions. Much of the area colored on the geological map as Devonian in the west of New South Wales may be found hereafter to be Silurian or Cambro-Ordovician in age. No fossils have been found in these beds, and they have been referred to the Devonian because of their lithological resemblance to the eastern Devonian quartzites and sandstones. ‘The strikes of the sediments are similar to those of the Ordovician. Devonian.—A strong movement of folding closed the Silurian and ushered in the Devonian sedimentation. The Devonian prob- lem in Australia is complicated much in the same way as are the Carboniferous, the Permo-Carboniferous, the Trias-Jura, and the Tertiary. The work of the pioneer geologists suggested that there were two, if not three, divisions in the Devonian period, with an unconformity between two of the sets of sediment. Mr. W. S. Dun has made a study of the Devonian in Australia and he has supplied the following notes for this report. He states that the Buchan and Bindi sediments in Victoria appear to be of Middle Devonian age, and that they are the equivalents, in great measure, of the Murrumbidgee beds in Southern New South Wales, the two groups containing types of fossils in common. In this case, however, Mr. Dun points out that it is probable that after detailed examination, Lower Devonian sediments would be found developed in these regions passing into Middle Devonian. The Upper Devonian series of sediments are characterized by the forms Lepidodendron australe, Spirifer disjuncta, and Rhyn- chonella pleurodon. The Upper Devonian series occur both at Mount Lambie and at Tamworth (New England). In the latter locality, however, Spirifer disjuncta and Rhynchonella pleurodon do not appear to have been found. Sussmilch,? in dealing with the Devonian, says: “An alternative explanation of the relations between the Lower and Upper Devonian tC, A. Sussmilch, Geology of New South Wales, 1914, pp. 77-80. 758 E. C. ANDREWS formations, however, suggests itself, and that is that the two forma- tions were deposited more or less contemporaneously, the former in an open but comparatively shallow continental sea, at some dis- tance from a shoreline; the latter in the shallow coastal waters of the same sea’’ (p. 80). He suggests that the marked differences so well established by R. Etheridge, T. W. E. David, and W. S. Dun between the faunas of the two formations would be due in such case to the unlike environments. To Benson’s field work, however, we are indebted for one definite piece of knowledge which may be expected to help in clearing up the tangle which has gathered round the Devonian in Eastern Australia. He' showed that the Carboniferous in New England is actually con- formable with the Devonian in that region, the sediments of each age being strongly folded, the strike of the folding being north- northwest approximately, as traced for 200 miles at least. During various geological surveys in the western, southern, and northern parts of New South Wales, the writer has noted that the Devonian sediments vary in appearance and structure, and the results of those observations would suggest that in very great measure the Devonian sea transgressed the area of folded Silurian sediments as far west as the Darling, without extending, however, as far in that direction as had the Silurian sea. A move- ment of folding apparently occurred in the Devonian which affected the eastern portions of Southern New South Wales strongly, being more marked as a whole in the northern portion of that area than in the southern, and more marked in the east than in the west. This movement may have been revived still later, with a tendency to cause Australia to grow northward and eastward as it had at the close of both the Silurian and Ordovician periods, the movement of sea transgression to the west and south being less during each suc- ceeding period. This brings us to a mention of the long zone of weakness extend- ing from a point somewhat south of Sydney to Queensland in a direction slightly west of north. The great negative area which had received the Ordovician and Silurian sediments had been changed to a positive element with the close of the Devonian sedimentation in the south and west. ‘The negative area by this time had shifted = W.N. Benson, Proc. Linn. Soc. N.S. Wales, 1913, pp. 490-517. AUSTRALASIA, NEW GUINEA, AND NEW ZEALAND 759 to the position mentioned above, and such a zone of weakness appears to mark the boundary of two geological provinces in Eastern Australia. Benson has shown that in this heavy area the Devonian and Carboniferous accumulated conformably, none of the series apparently being folded until the close of the Carboniferous. Carboniferous sediments are believed by some geologists to exist in Australia south of this zone, but Mr. W. S. Dun, in a personal communication, has informed the writer that in his opinion the fossils from such sediments are to be referred to the Devonian rather than to the Carboniferous. It is advisable, at this stage, to consider the general scheme of folding for the Devonian in Eastern Australia, inasmuch as what obtains for the Devonian in a general way, as regards its structure, is true also of the Silurian and the Permo-Carboniferous with this difference, that the analogies of form in rocks of the various periods considered are to be sought in areas which succeed each other to the north-northeast approximately in succession of time. Thus if the Silurian has been folded strongly over a large area, it may be found that the strongest folding of Devonian might be expected to be found north and east of the southern Silurian folds, whereas in certain areas of the strongest Silurian folding the Devonian is to be found bedded almost horizontally. Thus in Tasmania the Devonian is missing; in Victoria it is folded, apparently in two series separated by an unconformity; in Eastern New South Wales it is strongly folded, whereas in Western New South Wales it occurs as a series of gentle rolls and folds, with small areas, however, exhibiting local nipping or sharp folding within the complex basement of Cambro-Ordovician and Silurian.” Reference to forms very similar will be made in the chapter dealing with the Permo-Carboniferous. It may be mentioned here that a peculiar occurrence of so-called Devonian sediments has been recorded from Northwestern Aus- tralia by H. V. Woodward.2 This observer mentions Devonian E. C. Andrews, “Canbelego Gold and Copper Field,” Mineral Resources, No. 18, Dept. Mines, N.S. Wales. See maps and sections. 2 Report on Gold Fields of the Kimberley District (by Authority), Perth, 1891, p. ro. Quoted from T. W. E. David’s “Presidential Address,” Proc. Linn. Soc. N.S. Wales, 1893. 760 E. C. ANDREWS sediments at Kimberley (Northwestern Australia), which are said to possess a strike almost northeast and southwest, and a dip of 70°, while so-called Carboniferous sediments lying immediately above are almost horizontal. Carboniferous.—Benson’s great contribution concerning the conformability of the Carboniferous with the Devonian in North- eastern New South Wales allows us to infer that the Devonian south of lat. 33° S. was folded prior to the tilting of the New England Devonian, and it suggests also that not only Middle Devonian but also Upper Devonian is to be expected in this New England series. Permo-Carboniferous—The Permo-Carboniferous period was a most interesting one in Australia, but only the salient points dealing with its history are here recorded so far as they deal with the main thesis of this report. A great period of submergence is indicated over wide areas throughout peripheral Australia, but the strong folding to which certain sediments of this age were subjected at the close of the sedi- mentation was confined to a relatively narrow area within Eastern Australia north of the Hunter River, lat. 33° S. Thus the sedi- ments of this age in Tasmania and Western Australia are almost horizontal; in Victoria they appear to be flexed only in the neigh- borhood of Tertiary faults or monoclinal folds; in Southeastern New South Wales they exhibit corrugations scarcely recognizable; in the coastal region 100 miles north of Sydney they are moder- ately domed,? whereas to the west and southwest they are almost horizontal. In Northeastern New South Wales, as shown by Carne, Woolnough, and the writer, the Permo-Carboniferous sedi- ments are much folded and intruded by granite, whereas at a dis- tance of 200 miles to the west the strata lie almost flat. In Southern Queensland the Permo-Carboniferous is intensely folded, as men- tioned elsewhere in this report. tE. C. Andrews, “‘ Yalwal Gold Field,’ Mineral Resources No. 9, Dept. Mines, N.S. Wales, 1901; L. F. Harper, “‘The Southern Coal Field,” Memoir No. 7, Dept. Mines, N.S. Wales, 1916. 2T. W. E. David, “‘The Hunter River Coal Measures,’ Memoir No. 4, Dept. Mines, N.S. Wales, 1907. 3 J. E. Carne, “Western Coal Fields,” Memoir No. 6, Dept. Mines, N.S. Wales, 1908. AUSTRALASIA, NEW GUINEA, AND NEW ZEALAND 761 Farther north, as, for example, at Gympie' and Mount Morgan, the sediments are strongly folded and are Permo-Carboniferous in age. Along the north coastal area of Queensland there is a very long belt of sediments which are highly contorted and which appear to be Permo-Carboniferous. Nevertheless, less than 100 miles inland the Permo-Carboniferous‘ dips at an angle of 12° only. All this indicates that the close of the Permo-Carboniferous period was accompanied by a strong folding movement in areas north and slightly east of the areas affected by the great closing Carbonif- erous movement in New England. Additional evidence of this is adduced when dealing with the regions of ore deposition in Australia. The strikes of these foldings may be considered as subparallel to the Carboniferous lines of Benson, namely, northwest to north- northwest. It may be pointed out here also that the area of Permo- Carboniferous sediments in Tasmania, South Australia, and Western Australia is very large; nevertheless the beds there are horizontally bedded. Trias-Jura.—lIn this period there appears to have been a tend- ency for the old heavy, or negative, area of Central-Eastern Australia to sag again, or for the long zone of weakness separating New Eng- land from the land to the west and south to be broadened. In the northeastern portion of New South Wales and in Southern Queens- land another area of sagging received a great thickness of Trias- Jura sediments. ‘The geographical conditions under which the two sets of sedi- ment were deposited differed in certain well-marked features. This t Jack and Etheridge, Geology of Queensland (by Authority), 1892, pp. 72-84. B. Dunstan has also produced detailed geological maps of this area in a late number of the Queensland Survey Publications. 2 Jack and Etheridge, Geology of Queensland, p. 508. 3 Lionel C. Ball, ‘‘Wolfram, Molybdenite, and Bismuth at Bamford, Northern Queensland,” Queensland Mining Journal, 1914, p. 568. Mr. Ball has made a more definite statement as to the Permo-Carboniferous age of the beds in this district in a recent communication to the writer. 4]. H. Reid, “Permo-Carboniferous at Bett’s Creek,” Queensland Government Mining Journal, July, 1914, pp. 408-12. 762 E. C. ANDREWS has been indicated most clearly by Carne,’ who calls attention to the fact that massive conglomerates, coal seams, and abundant fossil trees characterize the northern sediments as to their lower members, while tuffs and sandstones without heavy conglomerates, coal seams, or abundant tree stems characterized the southern and western sedimentation (Triassic). Cross-bedded sandstones of warm-brown color and intercalated black shales characterize the southern, middle, or Hawkesbury series, while cross-bedded sandstones are very common in the northern or Clarence series. In both series the later stages of the Trias-Jura appear to be dark shales in the main. It is possible that the great folding at the close of the Carbonifer- ous in Northeastern New South Wales was responsible, in great measure, for the heavy conglomerates of the Clarence series as well as of the Triassic of the Upper Hunter valley, and it is probable that very high land barriers separated the two sinking areas during a moderate part, at least, of the period. This might be expected to have caused variations in local floras. As an example, the Clarence series contains a characteristic fossil, namely, Taeniopteris Dain- ireei, whereas it is absent from the Hawkesbury. On the other hand, however, Taeniopteris Daintreei is found in the Victoria Trias-Jura, so that Carne and others consider the Clarence to be of different age from that of the Hawkesbury. The southern or Hawkesbury (Triassic) area does not appear to have been dominated to the west and south by high land, inas- much as the adjacent and subjacent Permo-Carboniferous in those directions does not appear to have been disturbed except by a gentle movement of sagging. ‘The earlier period of the Trias-Jura appears to have been one of moderate to fair precipitation, but the middle period appears to have been subarid. In the Sydney district massive cross-bedded sandstones predominate in these middle beds, with relatively thin layers of dark-gray shales. In places layers of grit and subangular pebbles are interspersed with large blocks of these dark-gray shales, all mixed confusedly, apparently marking J. E. Carne, “Western Coal Field,” Memoir No. 6, Geol. Survey, N.S. Wales, 1907, pp. 26-41; see also E. F. Pittman, Aun. Report N.S. Wales, 1880, p. 244. Quoted by Carne, of. cit., p. 26. AUSTRALASIA, NEW GUINEA, AND NEW ZEALAND 763 periods of short-lived floods (sheet floods) which broke up the clays and mixed them with the pebbles and grits carried downstream by the local cloud-bursts or heavy rains. This appears also to have been the opinion of Professor H. E. Gregory, of Yale, from an examination by him in 1916 of the Sydney and Blue Mountain exposures. The upper portion of the period appears to have been one of greater precipitation in which actual lakes were in existence. At Sydney, and a little south of that area, the Triassic beds dip inland at a very gentle angle, but, as Carne has shown, the whole southern area of these sediments has a dip averaging only from 1° to 2°. In the northeastern part of New South Wales, however, the Mesozoic coal measures and the conglomerates dip from 10° to 20°, while in Southern Queensland they have been still more disturbed. In the western portions of Eastern Australia, however, as also in Western Australia, they lie practically horizontal. There appears to be no consensus of opinion among Australian geologists as to the origin of the Hawkesbury beds. Rev. J. E. Tenison Woods considered them to be of wind-blown origin’ with lakes and swamps between the dunes. In an unpublished paper R. S. Bonny considers them to be of estuarine origin. On the whole they may be said to be continental in origin, being formed in a sinking area mainly by water strains in a rather dry period. Cretaceous.—The Cretaceous period marked a spilling over of the ocean with the formation of great epicontinental seas, especially during the Upper Cretaceous period. The area most affected was the northern portion of the old heavy area separating Eastern and Western Australia. It is probable that, during the Upper Cre- taceous, the epicontinental sea extended from the Gulf of Carpen- taria to the Southern Ocean. The eastern area occupied by the Triassic sediments, however, consisted of dry land during the Cretaceous. At the close of the period the whole center of Australia appears to have been raised to a moderate height above sea-level. Dunstan and Richards have recorded pronounced folding (40°-55°) “The Hawkesbury Sandstone,” Proc. Jour. Roy. Soc. N.S. Wales, XVI (1882), 53-90. 764 E. C. ANDREWS of Lower Cretaceous rocks on the coast of Queensland at some dis- tance north of Brisbane.t , Reference will be made later to this local evidence of folded Trias-Jura and Cretaceous rocks along the coast of Central and Northern Queensland. Tertiary pertod.—The Eocene sea was not large and appears to have been confined to areas, relatively small, in the north and south of the continent. Indeed, the continent as a whole, except in the northeast, appears to have been growing in size subsequently to the close of the Cretaceous, although a submergence, postdating the recent Glacial period, appears to have isolated New Guinea and Tasmania from the mainland. It is as if there has been a general tendency in Australasia and New Zealand to move in a vertical direction in post-Cretaceous time, the movement being subject to two great laws: 1. That elevation, or vertical movement, of the land was empha- sized in an easterly direction, as from Western Australia to New Zealand, due allowance being made for the lagging behind differ- entially of the two great and relatively heavy portions, namely, Central Australia and the suboceanic mass separating Australia from New Zealand. 2. That the uplifts after the widespread peneplanation of the Cretaceous period did not proceed continuously, but were saltatory in their action, and, moreover, that the periods of time separating these uplifts became less as the present time approached, and that, nevertheless, the amounts of individual uplifts became greater as the periods marking the pauses between the uplifts became less in duration. This has given rise to great “‘valley-in-valley”’ struc- tures owing to the interrupted work of the streams. Thus in Australia, during what appears to be the Cretaceous period, great peneplains were formed in the land areas lying east and west of the Cretaceous sea, and only the hardest rock structures remained to show the existence of former plateaus or hills. In the ~ various Tertiary divisions of time the streams carved valleys with *B. Dunstan, Queensland Government Mining Journal, December, 1912; H. C. Richards, ‘‘The Cretaceous Rocks of Woody Island,” Queensland Aust. Assoc. Adv. Sci., Melbourne meeting, 1913, pp. 719-88. AUSTRALASIA, NEW GUINEA, AND NEW ZEALAND’ 765 widths so great as to appear as local peneplains, although they are only very broad, shallow valleys in whose bases other broad and shallow valleys have been excavated. The great uplifts of the later Kosciusko period allowed the streams to form profound can- yons which receded along these older shallow valleys. In other words, the main Tertiary land history has consisted of repeated elevations with stream revivals. During one or more of the Ter- tiary divisions of time, particularly in what may be the Miocene, the land appears to have sunk with the formation of lakelike expanses along the stream courses and the burial, later, of deep-. river deposits beneath basalt floods covering thousands of square miles in Eastern Australia. This led to great modifications in the stream drainage, but the dominating lesson of the repeated revival of stream action must not be overlooked, the modifica- tions due to lava floods being only an incident in the great geographical unity of Australia in Tertiary and post-Tertiary times. New Guinea.—lf attention be turned, however, to the north- eastern part of Australia, it will be found that as geological time progressed, the area occupied now in part by New Guinea was built to the south and west. An excellent epitome of the main features of structure known to date has been supplied by Professor T. W. E. David. Schists outcrop at very high altitudes along its northern portion, while strongly folded Cretaceous Strata are reported to occur at the highest altitudes in the north, their steep dips ending abruptly against a high and deeply dissected plateau surface. For 50 or 60 miles inland from the south, the area consists of middle and late Tertiary strata, all intensely folded, and all beveled off by a high plain, probably one of submarine erosion. The knowledge of this strong orogenic movement in late or closing Tertiary time and the excavation of a plain of erosion within folded sediments of this age was established by Carne while doing pioneering work in the oil industry.” t “Geology of Papua” (Federal Handbook), Brit. Assoc. Adv. Sci., Australian meeting, 1914, pp. 316-25. 2J. E. Carne, Bull. of the Territory of Papua, No, 1, Dept. External Affairs, Melbourne, 1913, pp. 19-29. 766 E. C. ANDREWS David’s conclusion is: In regard to the broad tectonic features of Papua it may be suggested, very tentatively, that the mainland of Australia has functioned as a “foreland massif,” Torres Straits, the Gulf of Carpentaria, the Arafura Sea, and the deep Mesozoic and Tertiary basins with their thick strata as a Senkungsfeld. Pos- sibly the crystalline schists forming a great part of the backbone of the island have played the part of an inner, or riick-land massif which has helped to roll up the Mesozoic and Tertiary sediments. In passing, it may be mentioned that this simply raises the question again as to the origin of the forces of crumpling. Do they act from the land as suggested by Suess in his discussion of the Asiatic framework, or do they act from the oceans? If the move- ments be assumed to act as from Central Australia toward the oceans, then it is difficult to understand the stability and rigidity of such central area of force. If the source of energy is suboceanic and directed toward the continents, then it is difficult to explain the growth of Australia north and east, while that of New Guinea appears to be south and west, unless, indeed, it be assumed gratui- tously that the later foldings in Northeastern Australia are simply the expressions of orogenic movements dying away in a south- westerly direction from the Pacific. Even so the intense contor- tions evidenced in the Miocene and Pliocene beds might be expected on the northeastern aspect of New Guinea rather than on the south and southwestern. It would seem, indeed, as though each negative or heavy area had played a part in the movements. New Caledonia—In New Caledonia the Mesozoic sediments have been intensely folded, especially on the western and south- western aspects, and the overfolding appears to have been directed toward Australia, according to Peletan, Depiet, Piroutet, and others as quoted by Suess.?__ New Zealand.—In turning to a consideration of New Zealand we meet with a certain amount of disappointment, inasmuch as there is no consensus of opinion among the workers on certain funda- mental points. Thus a glance at Dr. J. W. Gregory’s map in the article on New Zealand in the eleventh edition of the Encyclopaedia tT, W. E. David, “Geology of Papua,” op. cit., pp. 324-25. 2 Die Antlitze der Erde (Eng. tr.), IV, 314-15. AUSTRALASIA, NEW GUINEA, AND NEW ZEALAND 767 Britannica suggests that this island group was built principally as from southwest to east and north, or at any rate that with the prog- ress of geological time folding movements retreated to the north by east. Marshall, however, in a personal communication, under date of April, 1916, states that much of the New Zealand Jurassic has been confused with the Maitai (so-called Carboniferous) by older workers. Marshall, however, adduces sound reasons for considering New Zealand as being the true boundary of the Pacific Ocean‘ in that portion of its area. Cotton in a recent paper states that the “most profound deformation of this vast sedimentary group [Paleozoic and Mesozoic] took place in Later Jurassic or Early Cretaceous times.”? He also states that the average trend of the strike of this older mass appears to be west of north (p. 245). And again he writes: “It is apparent that during the period of their deposition [that is, the Tertiary Andes] a great part of the site of the present islands of New Zealand was continually submerged”’ (p. 247). 2 Cotton also speaks of orogenic movements in the Pliocene in the northern and more eastern portion of the group, and it is known, moreover, that great volcanic activity has taken place in the northeastern portion of the group with the formation of important gold deposits. In a personal communication dated August 22, 1916, Cotton writes: The early geological history [of New Zealand] is much obscured by the later happenings—a great deal more so, it would appear, than is that of Australia. We cannot even be sure that we have any considerable area of Paleozoic rocks. The small areas of Ordovician and Silurian in northern Nelson we can be certain of, but we know nothing whatever of the relations of these, either to each other or to rocks of later Paleozoic or Mesozoic age. It is the opinion of the present director of the Geological Survey that the greywacke rocks extending southward along the West Coast are of Aorere (Ordovician) age; but they contain no paleontological evidence of age and are part of the “Maitai” system of other writers. As for the ‘‘Maitai” rocks throughout tP, Marshall, “Presidential Address,” Geological Section Aust. Assoc. Adv. Sci. Sydney, XIII (1911), 90-99. 2 ‘“‘ The Structure and Later Geological History of New Zealand,” Geol. Mag. Lon- don, No. 624, June, 1916. 768 E. C. ANDREWS New Zealand, there seems to be no reason now for classing them as Paleozoic. As regards the Manapouri rocks of southwestern Otago, they may, of course (with the exception of some intrusives), be very ancient; but their relations to other systems are absolutely unknown. It may be that this is an upfaulted block from which a Mesozoic cover has been removed. So far as I know there is no evidence of later formations having been folded against it. The remarkable flat-lying schists of central and eastern Otago are, again, of indefinite age. Marshall regards them as metamorphosed Mesozoics. He traces a transition to the unaltered ‘‘ Maitais,”’ but in eastern Otago, along the junction of the schist and greywacke rocks there is a complex of faulted blocks (greywacke now forming the surface in some and schist in others) which had, there can be no doubt, been planed down before the deposition of what I call the ‘‘covering strata.” Later faults, which affect the cover also, have sometimes followed the lines of the older breaks, but have reversed the throw. As to the direction of folding in New Zealand I have formed no opinion. The latest or Kaikoura folding was accompanied by the formation of great reverse faults in the northeastern part of the South Island, and these hade to the northwest. Many small reverse faults in the Wellington neighbourhood, which intersect (?) Triassic rocks and were perhaps developed during the Mesozoic period, hade in the same direction. One question of great importance is that of the source of the enormously thick ‘‘Maitai’”’ sediments, which consist, from end to end of New Zealand, almost universally of the little-worn detritus from acid igneous rocks. Evi- dently these deposits accumulated not far from a great land mass, but I know of no evidence as to the position of that land mass. Apparently the New Zealand area sometimes formed a part of the continent, for at a number of places there are deposits containing Mesozoic plants. So far as I know there was no strong folding accompanying the formation of the Hauraki gold deposits, but there have been considerable “block” movements since. EVIDENCE OF THE ORE DEPOSITION It is proposed here to see what light may be thrown on the possible structural relations or differences of Australasia, New Guinea, and New Zealand, by a study of the peculiarities of ore deposition in certain areas within these regions. In this connection it is proposed to deal principally with one set of minerals only, namely, the tin group, although conclusions equally interesting would have been forthcoming from a consideration of the gold and copper, together with the silver-lead and zinc groups. Thus with regard to gold it would have been possible to elaborate with a wealth of detail the knowledge that the gold deposits of AUSTRALASIA, NEW GUINEA, AND NEW ZEALAND 769 West Australia are found in the great area of pre-Cambrian rocks there developed, and, moreover, that they occur in belts arranged more or less parallel and relatively narrow in width, although in certain localities they appear as small isolated areas or patches; that these narrow and well-defined belts have a general northwest and southeast direction, with divergences in certain instances of several degrees on either side of this direction; that the ore deposits in these belts or zones, owing to certain activities, do not crop out in long and unbroken lines, but are cut up into relatively short lenticles, arranged en echelon. Table I gives the approximate values of the several metals mined in these countries. TABLE I v APPROXIMATE TOTAL VALUES, IN MILLIONS or PouNDS STERLING, OF THE MorE Important Metats MINED IN AUSTRALASIA AND NEW ZEALAND Gold Copper Silver Zinc Tin | Wolfram | Bismuth Pate West Australia..| 73.00 | II.00 |.......|....... TROOUM EN sceage ell oeee ae opera] ones ee North Merritony.|>) 2:10) |) ©.20) |... . esc] oe... Ov sslelONOAe lige sake ni Grp cease Sowa Amelie. || se.Croy | o,ero I" Menai} [bibaig o4 llace do oclsoaucdcledecosolloucoane Tasmania....... RoGe || Wi GO | Ooo! loco ous T/2)500})|) OnOSi lO. O2N| eae ciee Wictoriay.)s)..2 4. BOOROOM ON 2I5iiI| NhOsOOO| neice = Or SO OOS |e. kate alee reee rte New South Wales........} 62.00 | 13.50 | 74.00 | 12.00 | 10.50 | 0.25 | 0.166] 0.10 , Queensland..... SOnCOM rT SON | 2)425) ||leleysvers 8.50) | 10.80) | ont25|) Ons iINewsZealand eee |MSicHooi MOOD ills Saks tas. teal peiee Vacs es [oust ciel iseetaess teenies The general direction of these auriferous belts almost everywhere coincides with the strikes of the schists, which, with one or two exceptions, invariably form the matrices of the gold-bearing reefs. ... . The quartz reefs are of two distinct types, namely, white quartz reefs and laminated quartz and jasper veins approaching very closely the hematite-bearing quartzites which invari- ably form a conspicuous feature in most of the gold fields of the State. Some of the laminated quartz veins range from almost-pure quartz, through banded jaspers, with crystals of magnetite, to bands appearing to the eye to be vir- tually pure haematite. The quartz reefs, of what may be called the massive types, occur plentifully in both the schists and the granites.t Like the gold deposits of Western Australia those of the North- ern Territory and of South Australia appear to be pre-Cambrian tA, Gibb Maitland, ‘“‘Mining Fields of Australia” (Federal Handbook), Brit. Assoc. Adv. Sci., Australian meeting, 1914, pp. 447-48. 770 E. C. ANDREWS in age, although there are certain indications of relative youth in the gold deposits of the Northern Territory and South Australia. Eastward, in the old trough lying within New South Wales and Victoria, now filled with Cambro-Ordovician and infolded Silurian sediments, occur the most important gold deposits of Australasia, especially the famous saddle reefs of Bendigo, Ballarat, Canbelego, and other localities. Immediately to the east and north lie the ore deposits beyond the Hunter zone of weakness where Benson’s line of serpentine occurs with its gold deposits of Carboniferous age. Beyond, but parallel, or subparallel, with these, are the great gold fields of the closing Paleozoic period in New England and Eastern Queensland, as, for example, at Hillgrove, Gympie, Mount Morgan, and the Palmer. It might be mentioned that, although no gold deposits appear to have been formed in Australia since that momen- tous period, nevertheless the important gold deposits of the North Island of New Zealand are of late Tertiary age. It might be men- tioned here that the gold deposits of Southwestern New Zealand appear to occur in Paleozoic rocks. Or it would have been interesting to enlarge upon the facts connected with the copper deposits of Australia: how in the west they are of pre-Cambrian age, according to Maitland and his geological staff; how the nature of the deposits there suggests deposition at a great depth below the old land surface; how the copper deposits of great but of unknown age, in the Northern Territory, South Australia, Western New South Wales, and Tas- mania, as, for example, at Wallaroo, Moonta, Burra Burra, Cobar, Nymagee, and Mount Lyell, do not appear to be dependent upon ordinary igneous rock types, but, from an examination of the reports of Ward, Jack, J. W. Gregory, and the writer, they appear to be the equivalents themselves of igneous rocks because of their peculiar mineral assemblages; how with these famous deposits might be mentioned the great Broken Hill deposit of silver-lead and zinc which is apparently a replacement of schists by garnet, rhodonite, feldspar, and sulphides, owing to the action of vapors arising along a shear zone; how the arrangement en echelon of these metalliferous areas and the individual ore lenses within such areas must be sig- 1 E. C. Andrews, “Broken Hill Lode,” Economic Geology, October, 1908, pp. 643-45. AUSTRALASIA, NEW GUINEA, AND NEW ZEALAND 771 nificant in the extreme. It would be instructive also to tell how, in New England, the copper and the gold which were introduced during the Carboniferous folding of Benson occur in the same deposits as a rule, as also do those of the closing Paleozoic both in New England and in the more coastal portions of Eastern Queensland (examples, Drake and Mount Morgan); how also in the Carbon- iferous of New England the copper and gold depend upon the ser- pentine belt for their existence, whereas in the Permo-Carboniferous they are related to lamprophyric dykes and basic granitic types. In New Guinea the copper deposits appear to be in very ancient rocks, whereas in New Zealand copper is practically absent. The tin group of minerals—Turning, however, from these inter- esting points to the tin-wolfram—molybdenite-bismuth group of minerals in Australia, it may be noted that all four may occur together in certain ore deposits in this continent, but as a rule the deposits of commercial importance may be classed under two main heads. Thus tin is frequently associated with wolfram, whereas molybdenite is associated with bismuth. Should molybdenite and bismuth be associated with other minerals of the group, the prefer- ence is for wolfram rather than for tin. Indeed, the minerals asso- ciated with tin and wolfram, such as tourmaline, topaz, beryl, and quartz with rutile, are practically unknown to the writer in con- nection with molybdenite deposits. All these minerals in Australia—tin, wolfram, molybdenite, and bismuth—are associated with siliceous granites or their equiv- alents. In New South Wales the typical tin-wolfram granites range from 75 to 79 per cent silica, while the typical molybdenite- bismuth types range from 72 to 74 per cent silica. These various granites may be distinguished easily by their peculiar vegetation, and appear to have been the hosts of the tin-molybdenite minerals in Australasia. The vapors which conveyed the minerals of the tin group to the marginal portions of the granites, preferably the roofs, or upper and lateral portions, appear to have varied in their power of penetration. Thus the tin and wolfram deposits, with their boric and fluoric associates, are found in many places at slight distances from the siliceous granites themselves, in rocks such as slate, basic igneous rock, or quartz-porphyry. Always, however, T72 E. C. ANDREWS the tin minerals may be seen to be intimately related to the siliceous granites. The molybdenite deposits are almost always within the marginal development of the siliceous granites, while tourmaline, topaz, and allied minerals are characteristically absent. Contact deposits of molybdenite in Australia, as, for instance, at Yetholme (New South Wales), are rare. Although these granites in Australia accompanied strong folding movements, and although ore deposits in that continent appear to have been dependent upon strong folding phenomena, nevertheless it must not be inferred that all periods of folding in Australasia have been associated with the formation of ore deposits on a commercial scale, but simply that all ore deposits of com- mercial importance in Australasia are intimately related in some way to periods of folding. This statement refers, naturally, only to deposits of the metallic minerals. a) Western Australia: The vast area of Western Australia con- sists, In the main, of highly altered rocks of pre-Cambrian age. These schists and allied types are intruded by siliceous granites and allied rocks, which also are considered to be pre-Cambrian in age. ‘‘The old granite rocks are traversed by many large ice-like Quartz) Feels.9.5. a: These older granite rocks ... . . formithe matrices of the tin and allied deposits of the state.” This mineral has been found to the extent of about 14,000 tons in Western Australia, while wolfram is subordinate in amount. Molybdenite has been recorded in small scattered flakes from this area. b) Northern Territory: The rocks of the Northern Territory are extremely old, probably pre-Cambrian in many places. Tin and wolfram to the values respectively of £400,000 and £40,000 approxi- mately have been won from the Northern Territory. Molybdenite has been reported, but it has not been worked as yet. c) South Australia: The ore deposits of South Australia are very old. Tin, wolfram, and molybdenite have been found in this state, but the amounts won are too negligible to be considered. tA. Gibb Maitland, “Mining Fields of Australia” (Federal Handbook), Brit. Assoc. Adv. Sci., Australia meeting, 1914, pp. 446-47. AUSTRALASIA, NEW GUINEA, AND NEW ZEALAND 773 d) Tasmania: The tin, wolfram, molybdenite, and bismuth deposits of Tasmania are considered to be of closing Silurian or early Devonian age.’ The tin production exceeds £12,000,000 and the wolfram £50,000 in value. Molybdenite has not been worked, but bismuth to the extent of about £200,000 value has been won. e) Victoria and Southeastern New South Wales: In Victoria the age of the tin, wolfram, molybdenite, and bismuth deposits is not known definitely. The value of the tin won is slightly less than £1,000,000, that of the wolfram about £5,000, while molybdenite and bismuth have been found only in very small quantities. Probably Victoria and Southeastern New South Wales form | one geological province, and in the latter area the tin and allied minerals may be considered as of post-Devonian and of pre- Permo-Carboniferous age. ‘Tin is relatively rare, but molybdenite and bismuth are abundantly represented. f) Northeastern New South Wales and Eastern Queensland: The northeastern portion of New South Wales appears to be a province geologically distinct from that of the southeastern portion of the state, and the tin, wolfram, molybdenite, and bismuth deposits found there appear to be closing Paleozoic in age. These deposits are confined to a strip less than 150 miles from the coast. The commercial molybdenite and the bismuth occur within the eastern zone, while the commercial tin occurs within the western .zone. The small deposits of the far west, near Broken Hill, for example, apparently are of very early Paleozoic age, and they really belong to the South Australian region or province. The value of the tin won from New South Wales exceeds £10,500,000, the wolfram values approximate £200,000, the bis- muth £150,000, and the molybdenite about £100,000. In this connection it should be remembered that until 1902 molybdenite was considered as an impurity in the bismuth, its =W. H. Twelvetrees, “The Scamander Mineral District,” Bull. No. 9, Geol. Survey, Tasmania, 1911, p. 23-24; other official reports of great interest dealing with the subject of mineral deposits in Tasmania are by L. K. Ward, Loftus Hills, and L. L. Waterhouse. 774 E. C. ANDREWS inseparable associate in New South Wales; thus great amounts of the molybdenite have been lost. Queensland, in its eastern portion, should be considered as belonging, probably, to the same geological province as New England, or Northeastern New South Wales. The tin, wolfram, molybdenite, and bismuth deposits are found only within the eastern strip of the state, and their age appears to be the close of the Permo-Carboniferous. The granites and mineral associations of the two areas are almost identical also. Thus this great province of Eastern Queensland and New England, which has yielded the bulk of the world’s supply of molybdenite, lies on a great flat arc having a general trend of northwest to north-northwest. These ore deposits are associated with strong movements of folding, the age of which appears to be closing Paleozoic. The approximate values of the tin, wolfram, molybdenite, and bismuth won from Queensland are respectively £0,000,000 to £10,000,000, £1,000,000, £250,000, and £150,000. It would thus appear that the deposits of the tin and molybdenite group of minerals in the great geological province of Western Australia, South Australia, and the Northern Territory are of great age, but that they are almost negligible in commercial value. It is not known, however, what proportion of this absence is due to removal by erosion of the upper portions of the granites. The deposits of this group in Tasmania may be of closing Silurian or early Devonian age, the tin values ‘being very large, but wolfram, molybdenite, and bismuth are unimportant; the deposits of the geological province of Southeastern Victoria and Southeastern New South Wales are important and are post-Devonian and pre-Permo-Carboniferous in age; while the deposits of the prov- ince of New England and Eastern Queensland, forming a coastal fringe to Northeastern Australia, are highly important from a commercial point of view and appear to be closing Paleozoic in age. t Official reports have been written on the tin and molybdenite areas of New South Wales by T. W. E. David, J. E. Carne, and the writer, while Professor Leo A. Cotton has published reports on the tin of New England in the Proc. Linn. Soc. N.S. Wales XXXIV (1909), 738-81; Cotton intends to continue the study of tin genesis in Australia in the near future. AUSTRALASIA, NEW GUINEA, AND NEW ZEALAND 775 All of these Australasian deposits are intimately related to strong movements of folding, accompanied by intrusions of very siliceous granite. No molybdenite, bismuth, wolfram, nor tin of any commercial importance whatever appears to have been found in New Guinea, New Caledonia, or New Zealand, although molybdenite and allied minerals have been recorded as curiosities in older Paleozoic granites in New Zealand. Neither are there in New Zealand any important copper deposits similar to those which are so intimately associated with the gold, tin, and molybdenite in Australasia. CONCLUSION It is therefore permissible, perhaps, to infer that each of the three great groups, namely Australia, New Guinea, and New Zealand, is a distinct geological province, but whereas in New Guinea the movements appear to have opposed the Australian growth with a tendency to fill the intervening negative area; on the other hand the growth of Australasia and New Zealand appears to have been intimately related in some manner, as though each had grown sympathetically in response to some simultaneous dominat- ing agency. The folding action ceased in the Australasian area long before it did so in the New Zealand area. The foldings in New Guinea also were maintained right into recent geological time. Here again the ore deposits proclaim the independence of the three centers. The oil fields of New Guinea suggest the Burmese or Malaysian origin of the New Guinea lines of structure,’ and in a similar way the tin-wolfram—molybdenite-bismuth group of miner- als appears to mark the real limits of Australasia. A little of the molybdenite group occurs in the New Zealand area, in the very old rocks, but the group as a whole, with its grand suite of siliceous granite horsts, may be said to end at the east side of Australasia. Moreover, as the folding movements retreated east and west, with progress of time they appear to have passed away finally to the northeast from Southwestern Australia toward New Caledonia. tT, W. E. David, ‘Geology of Papua” (Federal Handbook), Brit. Assoc. Adv. Sci. Australia, 1914, p. 320. 776 E. C. ANDREWS It is therefore permissible, perhaps, to infer that the Tasman Sea is of great age, especially in its more southern portions, inasmuch as it appears to have been a barrier to common or related ore deposition between Australasia and New Zealand through the ages. This of course does not imply that Australasia and New Zealand have not been closer together in the past, nor that Australasia has not extended considerably farther to the east in former times, especially in its northeastern portions; it simply suggests that some great agency which controlled the growth of Australasia and New Zealand appears to have admitted a negative or relatively sunken area from early times in the region of the Tasman Sea, and that this agency had faded away to epeirogenic movements in the Austral- asian area while yet it was vigorously folding the New Zealand rocks. All this appears to be in harmony with the general contention of Marshall* who maintains that New Zealand, and not Australia, lies on the real border of the Pacific. Marshall, however, approaches the subject from a point of view entirely different from that taken in the present note. tP, Marshall, ‘Presidential Address,” Geological Section, Australian Assoc. Adv. Sci. Sydney, XIII, (1911), 90-99. A BOTANICAL CRITERION OF THE ANTIQUITY OF THE ANGIOSPERMS EDMUND W. SINNOTT Connecticut Agricultural College As to the origin of the angiosperms, that group of seed plants which is now such a dominant element in the earth’s vegetation, we know almost nothing. They first appear as fossils in the deposits of the lowest Cretaceous in eastern North America, Alaska, Green- land, and Portugal, but just where they actually originated, and how long ago, are still matters of great uncertainty. The aim of the present paper is to throw a little light on the antiquity of this great plant group by studying the rate of evolution displayed by its members. | Evolution has not been a uniformly rapid process. The fact that plants recognized as “primitive”? and others recognized as “recent’’ exist together at the present time makes it evident that certain vegetable types have changed but little throughout long geological periods, whereas others have for one cause or another become altered much faster. The degree of inherent ‘“variability”’ and the frequency of hybridization have doubtless been influential in determining this rate of change, but a more important factor perhaps than either seems to be the length of the generation or period from seed to seed. A plant in which this cycle is completed in a year or two is able to multiply its generations more rapidly, and thus to accumulate heritable changes much faster, than one which requires a longer time for the attainment of reproductive maturity. This length of generation is definitely correlated with the growth habit of the plant, being greatest in trees—which usually reach an age of from fifteen to twenty years (in many cases much more) before bearing fruit—less in shrubs, and shortest of all in herbs, where one or two seasons from seed suffice to produce a fruiting plant again. Ina given length of time, therefore, 777 778 EDMUND W. SINNOTT a herbaceous species will pass through a much larger number of generations than a woody one, and will consequently tend, other factors being equal, to become changed in type much more rapidly. We should thus expect the herbaceous element in the vegetation to have been evolved at a much faster rate than the woody element. The establishment of this as a fact, taken with what we know as to the history and present numerical status of herbs and woody plants, will provide us with a valuable clue as to the antiquity of the angiosperms. That herbs are indeed subject to more rapid changes than any other plant type is indicated by the fact that the first local species and genera to develop in a region subsequent to its isolation have apparently almost always been herbs. This is well illustrated by a comparison of the floras of temperate North America and of Europe. On these continents today there are many local or ‘‘endemic’”’ genera which are limited in their distribution to one or to the other. Certain of these are evidently “‘relict’’ endemics, isolated survivors of types once much more widely disseminated. They may be recognized from the fact that they stand without near relatives in the floras; and many of them, such as sassafras and hickory, occur as fossils on both sides of the Atlantic. These relicts doubt- less constitute a very ancient floral element, and it is significant that among them are practically all the genera of trees and shrubs which are local to either North America or Europe. The majority of the endemic genera, however, seem to belong to quite a different category, for they occur in groups of from three to twenty genera, the members in each of which are closely related to one another, each group apparently to be looked upon as a separate center of evolution and the nucleus of a new family. The genera centering around Lesquerella in the Cruciferae, around Eriogonum in the Polygonaceae, around Godetia in the Onagraceae, around Pent- stemon in the Scrophulariaceae, and around Solidago in the Compositae, are a few of the sixty or more such groups in the dicoty- ledonous flora of North America, and there are as many in Europe. These “indigenous”? endemic genera most probably had their origin on their respective continents, since a free interchange of plants between America and Europe was interrupted, presumably THE ANTIQUITY OF THE ANGIOSPERMS 779 in the Early or Middle Tertiary; for had they existed before that date in anything like their present numbers and importance, it is highly unlikely that they would now be represented in the floras of both hemispheres. During the time since the isolation of the two continents, and while the rest of the flora have remained unchanged or have been developing endemic species merely, these plants have evidently undergone much wider changes, until they have finally given rise to new generic types. We are thus forced to conclude that the indigenous endemic genera constitute the most rapidly evolving members of their flora; and it is significant that they include practically nothing but herbaceous species— surely excellent evidence that the herb changes in type more rapidly than the tree or the shrub. Further evidence pointing to the same conclusion is presented by a study of the distribution of herbs and of woody plants in the modern scheme of botanical classification, for herbs are found to occur in larger groups than woody plants, their genera containing more species and their families more genera. Monotypes and very small genera and families are very much less common among herbs than among woody plants. These facts are what one might expect on the supposition that herbs are changing faster than the rest of the angiospermous vegetation, for the more rapid production of new forms leads to the building up of larger aggregations, and enables genera or families which have become reduced in size through extinction to repair these ravages quickly. A study of the structure, distribution, and ancestry of herba- ceous angiosperms’ indicates that they have been evolved in com- . paratively recent times from a woody ancestry, and have undergone practically their whole course of development since the beginning of the Tertiary. As opposed to this rapid change among herbs, we know from fossil evidence that very many woody genera have existed with very little alteration for a much more extended period than the length of the Tertiary—a convincing demonstration of the slowness with which trees and shrubs undergo evolutionary change. Almost all our woody genera bear evidence, in present IE, W. Sinnott and I. W. Bailey, ““‘The Origin and Dispersal of Herbaceous Angiosperms,”’ Annals of Botany, XXVIII (1914), 547-600. 780 EDMUND W. SINNOTT distribution or fossil remains, of a considerable degree of antiquity. To corroborate this testimony as to the relative rapidity of evolution in herbs and in woody plants, data as to their actual rate of change today would be highly desirable; but this is very difficult to obtain. As far as differences in ‘‘variability,” using the term in its broadest sense, are concerned, the two growth forms seem nearly equal. In both there are many highly variable types and many of great constancy. In the floras of three repre- sentative regions—Eastern North America, Australia, and Ceylon— the proportion of varieties and named forms among the woody species is found to be practically the same as among herbs. Nor is there a radical difference between the two in the extent of cross- pollination by insects, although in temperate regions this is some- what more common among herbs than among trees and shrubs. The difference in length of generation to which we have called attention is probably the most important factor in determining the rate at which they have evolved. To whatever cause we may attribute it, however, there seems to be little doubt that during the evolution of the angiosperms the primitive, woody element has been developed very much more slowly than the more recent, herbaceous one; and it is this differ- ence which gives us a hint as to the antiquity of the whole group. We find in the angiosperm flora today (dicotyledons alone con- sidered) over 4,200 genera of trees or shrubs, as opposed to only 2,600 genera of herbs. We may be reasonably sure that practically all of these 2,600 genera of herbs have been developed since the beginning of the Tertiary; and if we assume that herbs are pro- ducing new types only twice as fast as trees and shrubs—surely a conservative estimate—we must believe that only about 1,300 woody genera have been evolved during the same time. The evolution of the 4,200 genera of woody plants at present existing, to say nothing of the great numbers which have been lost through extinction (by which trees and shrubs have suffered much more than herbs), would therefore require a period at least thrice the length of the Tertiary. If the common assumption that the Tertiary was approximately as long as the Cretaceous is correct, THE ANTIQUITY OF THE ANGIOSPERMS 781 the origin of the angiosperms would thus be thrust back to a date much earlier than the beginning of the Cretaceous. Of course such an estimate is hypothetical in the extreme; but by indicating that the history of the woody members of the group extends back over a period many times as long as that during which herbs have existed, it serves to give us a clue as to angiosperm antiquity, and it emphasizes the fact that our present huge array of trees and shrubs, types very slow in changing, must have required an enormous length of time for their evolution. There is evidence, moreover, that evolution took place even more slowly in former times than it does at present, since flower-loving insects, to the agency of which many attribute the rapid development of the angiosperms, did not appear on the scene, at least in numbers, till the dawn of the Tertiary.‘ All this makes it highly probable that these now dominant seed plants did not begin their existence in the early Cretaceous, where they first appear as fossils, but that they had already undergone a long course of evolution before that time. Indeed, the external features, and more particularly the internal anatomy, of these earliest fossil angiosperms are not at all those of primitive types, but exhibit a considerable degree of specializa- tion.? To regard such plants as having sprung suddenly into being from gymnospermous ancestors is to overtax the imagination of even an ultra-mutationist. As to why the earliest members of the group apparently failed to be preserved we cannot be sure, but evidence is at hand that they were upland forms which would tend less frequently to become fossilized. This predilection of primitive angiosperms for an equa- ble, reasonably cool climate, if it can be proved, will lead us to look back to the era of low temperatures in the Jurassic, or perhaps even to as remote a period as the cataclysmic refrigeration of the Permian, for the date when the first angiospermous stock began to be differentiated from its gymnospermous ancestry. The botanical evidence is therefore overwhelmingly in favor of the conclusion that angiosperms existed for a considerable period t Handlirsch, Die fossile Insecten. 2M. C. Stopes, ‘‘Petrifactions of the Earliest European Angiosperms,” Phil. Trans. Royal Society, B, 203, pp. 75-100. 782 EDMUND W. SINNOTT previous to the Cretaceous, although this cannot be said to be absolutely proved till they are brought to light as fossils from the earlier periods of the Mesozoic, a discovery which diligent search may reasonably be expected to yield. The establishment for the angiosperms of an antiquity greater than that usually accorded them at the present time will be of some importance geologically, since the occurrence of fossil members of the group in a given formation will no longer be regarded as a demonstration of the post-Jurassic age of the latter. ARE THE “BATHOLITHS” OF THE HALIBURTON- BANCROFT AREA, ONTARIO, CORRECTLY NAMED? W. G. FOYE Harvard University The large areas, composed essentially of banded red gneiss, which are found throughout the Haliburton-Bancroft area have been called by Adams and Barlow “‘batholiths.”* These appear on maps of this region as circular or oval masses more or less com- pletely surrounded by sediments or schists of sedimentary origin. The stratification of these sediments follows in strike the boundaries of the adjacent gneiss. Moreover, within the gneissic areas are layers of amphibolite or gray gneiss which conform in dip and strike to this same boundary. ‘The gneissic areas, therefore, may be described as domes of red granite gneiss containing gray gneiss and amphibolite in layers striking concentrically to points more or less fixed within the mass and dipping quaquaversally at angles which vary from 37° to 45°. In his earlier writings, F. D. Adams stated three views as to the origin and method of emplacement of the ‘Fundamental Gneiss.”’ 1. The Fundamental Gneiss may be the remains of a primitive crust which was penetrated by great masses of igneous rocks and subjected to suc- cessive dynamic movements. The Grenville series may be an upward con- tinuation of the Fundamental Gneiss under altered conditions, marking a transition from a primitive crust to normal sediments. 2. The Grenville series may be considered as distinct from the Funda- mental Gneiss and reposing on it unconformably, being a highly altered series of clastic origin; the Fundamental Gneiss having some such origin as sug- gested above or being an older intrusive series of still more highly altered sediments. 3. The fundamental Gneiss may be considered as a great mass of eruptive rock which has eaten upward and penetrated the Grenville series, while the 1 Geol. Surv. Can., Mono., VI (1910), 12. 783 784 W. G. FOVE Grenville series represents a series of altered sediments of Laurentian, Huro- nian, or subsequent age." Adams in the same article stated that the last hypothesis was untenable. The world-wide distribution of the Fundamental Gneiss (forming, as it does, wherever the base of the geological column is exposed to view, the founda- tion upon which all subsequent rocks are seen to rest) is opposed to this view as is also its persistent gneissic or banded character.? Later, in 1897, Adams altered his earlier view. He writes: The batholiths are undoubtedly formed by an uprising of the granitic magma from below, and these foci indicate the axes of greatest upward movement. ‘These centers are not all areas of most rapid uplift, however. On the contrary, the gneissic foliation in some cases dips inward in all direc- tions toward the center, thus marking them as places where the uprise of the magma was impeded, that is to say, places where the overlying strata have sagged down into the granite magma.3 Fic. 1.—Map of the corundum syneite district of Craigmont, Ontario. A striking fact concerning Black, limestone; white, amphibolite; these so-called batholiths is that dashed, gneissic granite; dotted, gneissic granite with amphibolitic inclusion. they do not cut across the struc- ture of the invaded rocks, a fundamental characteristic by which post-Cambrian batholiths are recognized. While it is true that there are bodies within the district which cut across the structure of the country rock, they are unusual, and concordant relationships are much more common. A glance at the map (Fig. 1) which shows the corundum syenite district of Craigmont, Ontario, makes clear the concentric arrangement of the sedimentary rocks within the gneiss areas. t Journal of Geology, I (1893), 330-32. 2 Op. cit., 332. 3 Am. Jour. Sci., III (1897), 173-80. “ BATHOLITHS” OF HALIBURTON-BANCROFT AREA 785 The granite was intruded between the layers of limestone. As in the present case, long narrow layers of limestone are often found isolated in the gneiss. These layers are in parallel bands and the strike of their stratification conforms to the strike of the gneissic structure of the surrounding granite. If the boundary between gneiss and pure limestone is sought, it will invariably be found that there is a transitional contact zone. The distinction between areas which may be designated as “‘ Gneiss with amphibolitic inclusions” or ‘‘Amphibolite” or ‘Limestone invaded by much gneiss” depends upon the degree to which the granite has invaded the limestone and altered it to amphibolite. In general, on crossing the strike from limestone to amphibolite, there is a gradual transformation of one rock into the other. The amphibolite in turn is transitional to red gneiss through the inter- mediate stage of gray gneiss. Xenoliths of amphibolite within the gneiss are in no degree so abundant as stringers of amphibolite varying from a few centimeters to a meter in diameter and the schistose structure of which conforms to the gneissic structure of the granite and the stratification of the limestone. Adams’ attributes the parallel arrangement of these bodies to movements of the granite after mtrusion. He conceives that the limestone blocks, stoped from the roof of the batholith, were softened by heat and pulled out into lenses by flowage. The parallel banding of pre-Cambrian rocks is not a local fea- ture, illustrated only in the rocks of the Haliburton-Bancroft area. _ It is, rather, characteristic of most pre-Cambrian terranes. The interbanding of gneiss of igneous origin with sediments is shown by Lawson in his study of the Lake of the Woods. Hégbom? has described similar relationships which are shown by the rocks about Upsala, Sweden. The gneiss of the pre-Cambrian of the Adirondacks is so mingled with limestone and other sediments that for years it has been a mooted question whether to consider it of igneous or sedimentary origin. It forms lenses and sheets in the sediments, or traverses them so irregularly that an exact interpretation is difficult. 1 Can. Geol. Surv., Memoir No. 6 (1910), 73-78. 2 Bull. Geol. Instu., Univ. Upsala, X (1910-11), 39. 786 W. G. FOYE However, C. H. Smyth, Jr.7 and H. P. Cushing? now consider them igneous. The gneisses of the Highlands of New Jersey may be described in similar language. These are considered by W. S. Bayley? and C. N. Fenner‘ to be sediments invaded by granite. G. M. Dawson writes as follows concerning the Shuswap Ter- rane of British Columbia: The Shuswap rocks proper evidently represent highly metamorphosed sediments with perhaps the addition of contemporaneous bedded volcanic materials: . .... These bedded materials are, however, associated with a much greater volume of mica-schists and gneisses of more massive appearance, most of which are evidently foliated plutonic rocks, and are often found to pass into unfoliate granites. The association of these different classes of rocks is so close that it may never be possible to separate them on the map over any considerable area... .. A distinct tendency to parallelism of the strata or foliation with adjacent borders of the Cambrian system has been noted in a number of cases. This might imply that the foliation was largely produced at a time later than the Cambrian, but materials of some of the Cambrian rocks show that the Shuswap series must have fully assumed their crystalline character before the Cambrian period. Jt seems, therefore, probable that the foliation of the Shuswap rocks may have been produced rather beneath the mere weight of superincumbent strata than by pressure of a tangentical character accompanied by folding R. A. Daly, in a recent report on this same series, states that it has been .injected by innumerable sills and laccoliths. He concludes: The extraordinary prevalence of sills and other concordant injections is explained by the extreme fissility of the Shuswap sediments and greenstones. This feature is due to static metamorphism.$® Two hypotheses are offered, therefore, to explain the parallel banding of pre-Cambrian rocks. In the Haliburton-Bancroft area, Adams conceives that, in the process of intrusion by magmatic tN.Y. State Mus., 41st Ann. Rept., II (18099), 469-97. 2 Bull. No. 115, N.Y. State Mus. (1907), 451-531. 3U.S.G.S., Raritan Folio, No. 191. 4 Journal of Geology, XXII (1914), 594 fi. 5 Bull. Geol. Soc. Am., XII (1901), 63-64. 6 Ann. Rept. Dept. Mines, Can. Geol. Surv. (1911), 3-12. “BATHOLITHS” OF HALIBURTON-BANCROFT AREA 787 stoping, blocks of limestone were torn from the roof of the invading granite batholiths and elongated parallel to the contact of the granite and limestone by movements of the granite as it con- solidated. _ Daly believes that static metamorphism produced planes of weakness within the Shuswap series and that sills of granite were intruded along these planes. Fenner,’ discussing the method of intrusion of the granites of the New Jersey Highlands, states that in his opinion gaseous emanations from the granite magma penetrated the sedimentary rocks along planes of weakness and prepared the way for the intrusion of granitic fluids. The intrusion of these fluids produced banded gneisses. It is, of course, entirely possible that intrusion in the Haliburton- Bancroft region took place by magmatic stoping and that this region is not analogous to the others described. The rock types vary within the several areas and there is, necessarily, a corre- sponding change in the structural relations. The limestones of the Grenville series would undoubtedly be more altered by the meta- morphic effects of the granite than the quartzose rocks of the New Jersey Highlands and the Shuswap series. It seems to the writer, however, that the facts shown by the study of the Glamorgan gneissic area favor the theory of intrusion by parallel penetration along planes of weakness rather than the theory of intrusion by magmatic stoping. If intrusion took place by magmatic stoping, the following conditions must be postulated. The objections to each of these conditions are noted. 1. The blocks from the roof of the 1. Intrusion by magmatic stoping batholith were stoped off and elongated usually produces an irregular molar parallel to the contact. contact. The igneous rock cuts the sediments. Though the blocks were elongated parallel to the contact, this would not, except by chance, be parallel to the stratification of the sediments, and yet this is the relationship of the banded structure of the rocks of the Haliburton-Bancroft area. * Journal of Geology, XXII (1914), 594 ff. 788 2. Adams postulates that the elonga- tion of the blocks occurred in the later stages of batholithic intrusion as the granite solidified. 3. The limestone blocks, stoped from the roof of the granite batholith, floated and so were elongated by the movements of the granite parallel to its contact. W. G. FOYE 2. It would seem necessary that this should be true; for a hot, fluid magma, if too hot would melt the blocks and incor- porate them into a homogeneous magma. If it were too cold it could not elongate them. The necessary conditions for the production of parallel elongation, there- fore, is a narrow temperature range within which the blocks remain viscous. This would be found, it would seem, at a more or less constant distance from the molar contact of the intruding batholith with the country rock. As the magma progresses upward, the central heat of the batholith must likewise progress upward and hence the parallel banding of the batholith produced at any stage would be destroyed in a later stage by the complete solution of the blocks into a homogeneous magma. The so-called “batholiths” are in all stages of dis- section yet the parallel structure is persistent from center to edge. The structure is not, therefore, a border phenomenon as Adams’ theory would demand. 3. Daly? has shown that limestone blocks at high temperatures are much heavier than fluid granite. Hence these blocks should sink and leave a clear con- tact which would be_ progressively attacked by the hot granite magma. This would not give rise to the parallel structure observed. If, however, they floated they would impede the attack of the granite magma at the contact and their solution or partial solution and elongation would cause an enormous loss of heat and render the further upward progress of the batholith very difficult. It is estimated by Adams and Barlow that 20 per cent of the “‘batholithic’”’ areas consist of gray gneiss and amphibolite. estimate is low for the districts visited by the writer. This Not only these rocks but also bands of pure limestone are often found near tI gneous Rocks and Their Origin (1914), 202. “BATHOLITHS” OF HALIBURTON-BANCROFT AREA 789 the center of the gneissic areas, where, it would seem, by Adams’ _ theory, that the pure gneiss of the intruding batholith should be found. This fact makes it easy to believe that the granite was intruded along planes of slight resistance, and that the limestone terrane of the Grenville series became an immense steam pack, at the time of the intrusion of the granite, with layers of gases followed by fluid granite alternating with layers of limestone. The pre- Cambrian granites were probably accompanied by an immense amount of pneumatolytic gases. The loss of these gases at higher levels due to decreasing pressures accounts for the gradual lessening of the interaction of granite and limestone away from the main granite mass, while within the gneissic areas the retention of the gases allowed the granite to effect a complete change of the lime- stone to gray gneiss and amphibolite. At certain places the granite failed to penetrate great lenses of the limestone. The gases from layers of fluid magma at the top and bottom of these lenses meta- morphosed their borders but failed to affect their centers. Lenses of pure limestone were, therefore, preserved in the midst of gray gneiss and amphibolite. It has been inferred that the intrusion of the granite occurred along planes of weakness. These, as the structure now shows, were parallel to the stratification of the limestones. Daly and Dawson have stated that in the Shuswap area these planes were _ due to static metamorphism. Fissility produced in this way would be less apparent in limestones than in the quartzose rocks of British Columbia. However, the fact that the gneissic structure of the Laurentian gneiss is parallel to the stratification of the Grenville series would favor the view that the granite solidified under condi- tions of stress similar to those which produced the parting planes along which it was intruded. A vertical dike near Baptiste Lake, west of Bancroft, Ontario, shows horizontal schistose structure similar to certain dikes described by Daly* in the Shuswap area. Adams and Barlow? ascribed the gneissic structure of the Laurentian gneisses to the pressure of intrusion of the granite magma. This, t Cf. to figure in Guide Book No. 8, Part II, Internat. Geol. Cong., Can (1913), 130. 2 Can. Geol. Surv., Memoir No. 6 (1910), 78-81. 790 W. G. FOYE however, would not explain a phenomenon such as that shown by the dike just described (Fig. 2). It is believed that the fissility which allowed the granite to intrude the sediments and the gneissic structure of these granites were both results of a persistent force, the static pressure of the overlying sediments. The Grenville series is said by Adams and Barlow to be approximately 50,000 feet thick. This series compares with the Shuswap Terrane which is 30,000 feet thick. The elongation or compression of the amphibolitic layers and the presence of amphibolitic inclusions may be explained as easily by the theory of Daly and Fenner as by that of Adams. The granitic gases and fluids must have had their origin at certain definite points. At these points they were pushed upward and sideways along planes of easy parting and a pine-tree structure was produced. In general, the increase of material due to the addition of granite Fic. 2.—A vertical dike showing hori- would produce a doming at the zontal schistose structure. center of intrusion with qua- quaversal dips away from these points. However, the subsidence of the magma on cooling might very possibly cause a collapse of the dome and irregular dips would result. The mechanism of lit-par-lit intrusion, as explained by Fenner,” is dependent on the fluxing power of the pneumatolytic gases given off by the granite. These go before and prepare the way for the later intrusion of the granite magma. The prevalence of lit-par- lit rather than batholithic intrusions in pre-Cambrian terranes may be due, therefore, to the greater abundance of magmatic gases in the earlier periods of the earth’s history. The vast amounts of pegmatitic granite associated with pre-Cambrian areas lends support to this theory. t Journal of Geology, XXII (1914), pp. 594 fi. eee “BATHOLITHS” OF HALIBURTON-BANCROFT AREA 791 The facts presented above do not mean that cross-cutting bodies are lacking in the Haliburton-Bancroft areas. They are found but are by no means as common as concordant injections. It seems fair to conclude, therefore: (1) that the so-called “batholiths”’ were formed by the concordant injection of granite into a fissile limestone terrane; (2) that this fissility was produced by the pressure of the overlying sediments; (3) that the layers of Fic. 3. ‘‘Stromatolithic” structure (C. N. Fenner, Jour. Geol., XXII [1914], 596). limestones, lying between layers of molten granite, were permeated by the pneumatolytic gases and fluids given off by the granite and transformed to amphibolites or gray gneisses; (4) that the concordant injection of the granite produced the dome-like char- acter of the “gneissic’”’ areas; (5) that the term ‘‘batholithic” does not describe the true character of these areas and the term “stro- matolithic’’! is suggested in its place (Fig. 3). t From Greek orpwua, “a layer,” and Avos, ‘a stone.”’ The noun “Stromatolith” may be defined as a rock mass consisting of many alternating layers of igneous and sedimentary rocks in sill relationship. A CONTRIBUTION TO THE OOLITE PROBLEM FRANCIS M. VAN TUYL University of Illinois INTRODUCTION At the present time there are two prevalent theories of odlite formation, namely, the inorganic, or chemical precipitation theory, and the organic theory. Prior to the year 1890 the inorganic theory was generally agreed to and it is to this day the most widely accepted of the two. In the year mentioned, however, Wethered' pointed out a close relationship between the concretionary structure of the calcareous algae Girvanella and that of true odlite, and showed that certain so-called odlites of the Carboniferous and Jurassic of England really consist, in part at least, of rounded calcareous masses secreted by this organism, since they possess in addition to the concretionary structure the vermiform tubules which characterize the genus. But in this and again in a succeeding paper, entitled ‘‘The Formation of Odlite,” which appeared in 1895,? Wethered was unable to demon- strate the presence of the Girvanella tubules in typical odlite spher- ules showing both radial and concentric structure, although he was led to believe that these were also of algal origin. Following closely upon Wethered as a champion of the organic theory came Rothpletz, who published a paper on the origin of odlite in 1892.3 This investigator upon studying the recent odlites of Great Salt Lake found that where these were still in the water they were usually covered by a bluish-green algal mass consisting of the cells of Gloeocapsa and Gloeothece, forms which are known to secrete carbonate of lime; and, when the odlite grains and rodlike t Quar. Jour. Geol. Soc. London, XLVI, 270-83. 2 Ibid., LI, 196-209. 3 Botanisches Centralblatt, No. 35, pp. 265-68 (English translation by F. W. Cragin, American Geologist, X, 279-82). 792 A CONTRIBUTION TO THE OOLITE PROBLEM 793 calcareous bodies on the shore were dissolved in acid, they all yielded dead and shriveled fission algae. Rothpletz, therefore, concluded that the odlites of Great Salt Lake are the product of lime-secreting fission algae, and that their formation is proceeding day by day. Furthermore, a study of the recent and near-recent odlites of the Red Sea showed these also to contain minute grains of organic material suggesting fission algae. But these differ from the Great Salt Lake odlites in that their nuclei always consist of sand grains and in that their concentric structure is less well developed. They also. possess small vermiform canals filled with calcite, which are interpreted as imprisoned algae of another type. Rothpletz also remarks that certain elongated corpuscles pos- sessing odlitic structures, which he interprets as organic, occurs in the Lias limestone of the Vilser Alps, and concludes as follows: “‘Accord- ing to the present stage of my researches, I am inclined to believe that at least the majority of the marine calcareous odlites with regular zonal and radial structure are of plant origin; the product of microscopically small algae of very low rank, capable of secreting lime.” In spite of these discoveries by Wethered and Rothpletz, later students of the odlite problem have tended to drift back to the inorganic theory and to regard the association of odlites with algae as accidental. Thus Linck’ has shown by experiment that odlites similar to natural ones may be produced artificially by the action of sodium carbonate and ammonium carbonate on the calcium sul- phate of sea-water. He points out that these carbonates are formed by decomposition of animal and plant tissues in the sea, and favors the view that odlites have been formed in this way. That natural odlites can be formed chemically is demonstrated by Vaughan,’ who points out that odlitic structure is now being developed in the calcareous muds precipitated through the agency of bacteria off the coasts of Florida and the Bahamas. In a recent review of the whole question of odélite formation, T. C. Brown? has endeavored to substantiate Linck’s conclusions t Neues Jahrb., Beil. Bd. 16 (1903), pp. 495-513- 2 Jour. Washington Acad. Sci., II (1913), 302-4. 3 Bull. Geol. Soc. America, XXV (1914), 745-80. 704 FRANCIS M. VAN TUYL and to discount the importance of the algal theory. To quote from him: ‘The dead algal cells in the Salt Lake odlite are regarded as cells which had selected the odlite as a point of attachment. They became imprisoned within it by the further accretion of aragonite by chemical precipitation.” He suggests that the decay of the attached algae furnishes Na,zCO; which acts as a precipitating agent and thereby aids the growth of the odlite. As regards the importance of algae in the production of the odlites of Great Salt Lake, future studies may be expected to throw additional light on the problem. Microscopic examination of these by several investigators has failed to reveal any indications of algal structure in the calcareous grains themselves. On the other hand, they exhibit highly developed radial and concentric structure. THE PRAIRIE DU CHIEN OOLITE Some time ago the writer had occasion to examine microscopi- cally a siliceous odlite which marks the base of the Ordovician in northeastern Iowa, and found to his surprise that the odlite grains of this showed undoubted algal structures. The bed in question constitutes the so-called transition member between the Prairie du Chien dolomite and the Saint Croix sandstone. With reference to this bed Leonard, in his “Geology of Clayton County,” says: The lower Magnesian is not marked off sharply from the underlying Saint Croix, but there is a transition from the one to the other through from fifteen to twenty feet of calcareous sandstone or siliceous odlite. The rock is com- posed of clear rounded grains of quartz cemented by lime carbonate. In some beds this cementing material is quite abundant, in others there is only enough to hold together the grains. The ledges vary in thickness from a few inches to two or three feet. This siliceous odlite is well exposed in an old quarry in the river bluff one and one half miles above North McGregor. The transition beds are also seen in the section at Point Ann, just below McGregor. Here there are alternating layers of sandstone and limestone and some odlite similar to that described above.! A bed of similar character and thickness has been described by Calvin? as occurring at the same horizon in Allamakee County, which lies directly north of Clayton. The writer has examined t Iowa Geol. Survey, XVI (1905), 239-40. 2 Ibid., IV (1894), 61. —oo A CONTRIBUTION TO THE OOLITE PROBLEM 795 the member at the Point Ann exposure only, and the samples here described and figured are entirely from that locality. Microscopic examination of the rock shows it to consist of imperfectly preserved siliceous oélite grains in a dolomitic matrix. The history of the rock is briefly as follows: Subsequent to the formation of the oélite, dolomitization set in, transforming the cal- careous matrix completely, and many of the calcareous odlite grains either wholly or in part, to dolomite. Alteration then ceased and silicification of the unchanged, or only partly changed, odlite grains ensued. The irregular areas of dolomite within the interiors and the frayed-out borders of many of the silicified odlite grains are in this way accounted for. The structure of grains which were com- pletely dolomitized prior to silicification is almost entirely obliter- ated, and these are often only with difficulty distinguished from the matrix. The odlite grains range from 0.1 mm. to 1.13 mm. in diameter, and when well preserved show, in addition to the concentric and radial structure, minute sinuous, enwrapping fibers very similar to the tubules which characterize the Girvanella type of calcareous algae. A comparison of the microphotographs of the odlite grains with that of Girvanella problematica Nicholson, described and figured by Rothpletz, in his memoir entitled ‘“‘ Ueber Algen und Hydrozoen im Silur von Gotland und Oesel,’™ will bring out this striking similarity (Figs. 1-6). It should be recognized that the interwoven fibers of the odlite have been partly obliterated by silicification. Doubtless these con- sisted of hollow tubules filled with calcite, like those shown by Girva- nella problematica prior to silicification. The fibers of the organism of the odlite have an average diameter of 0.015 mm. which agrees very closely with the diameter of the tubules of Girvanella problematica, which varies from o.o1 to 0.018 mm., according to Rothpletz. Typically the well-preserved odlite grains consist of an inner structureless nucleus, followed by a narrow intermediate band showing radial structure, and this again by an outer band bearing 1Kungl. Svenska Velenskapsakademiens Handlingar, Band 43, No. 5 (1908), TAL I bikes, 1 Fic. 1.—Microphotograph of Girvanella problematica Nicholson. About X42. After Rothpletz. Fic. 2.—Microphotograph of peripheral section of a silicified odlite grain from basal Ordovician at McGregor, Iowa. About X45. Fic. 3.—Cross-section of another grain from the same locality. About X45. Note the well-developed algal structure in the outer portion and the band showing radial structure within this. The interior is not preserved. Fic. 4.—Imperfectly preserved odlite grain. About X45. The interior and peripheral portions of the grain were replaced with dolomite, with obliteration of struc- ture, prior to silicification. Fic. 5.—Silicified grain showing well-developed radial structure but with algal fibers nearly obliterated. About X45. Fic. 6.—Another grain showing fine concentric structure but with no distinct algal fibers preserved. About X45. A CONTRIBUTION TO THE OOLITE PROBLEM 707 sinuous fibers. In some instances, however, the two outer bands grade gradually into each other without any distinct line of demarka- tion; or indeed the radial structure may be entirely wanting and the concentric structure may continue into the nucleus. The fibers are best shown in peripheral sections of the grains. .In these they appear to enwrap the bodies. Some of the grains, however, show little or no trace of algal fibers, but there is convincing evidence that this fact has resulted in most, if not all cases, from the obliteration of original structures as an accompaniment of silicification. All stages of such obliteration may be traced under the microscope. SOME EFFECTS OF CAPILLARITY ON OIL ACCUMULATION’ A. W. McCOY The University of Oklahoma All rocks in the upper crust of the earth contain pore space. The percentage by volume of this space varies from a fraction of 1 per cent in the case of most fresh crystalline rocks? up to 4o per cent in some sandstones. Below ground-water level these openings are more or less saturated with water, which moves about from points of higher to points of lower pressure. The movement of water thus entombed does not exactly follow hydrostatic laws, as can be observed by the small loss of head in artesian flow. For example, an instance is cited} by Van Hise4 where water traveled under ground 150 kilometers with a loss of only 50m. in head. This shows that the movement was very slow (perhaps a few feet per year), for the friction through the porous stratum was almost nothing. In the case of water moving in large openings, such as pipes, friction is an important factor. A somewhat similar example was observed by the author in Missouri, where the loss of head by flow in the Roubidoux sandstone was about 200 ft. in 75 miles. A theoretical means of comparison with the observed facts is to note the size of the openings in the rocks. All tubular openings less than 0.508 mm. are capillary. There- fore, by geometrical proof, it can be shown that sandstones with uniform rounded grains of less than 2mm. in diameter, would contain mainly capillary openings. Rocks with uniform rounded grains, regardless of the size of grain, contain about the same t A paper read before the Geologic Conference of Oklahoma, January 7, 1916, at Norman, Oklahoma. 2 Van Hise, Monograph, U.S.G.S. 47, p. 125. 3G. P. Merril, Rocks, Rock-Weathering and Soils, p. 198. 4 Monograph, U.S.G.S. 47, p. 587. 5 Alfred Daniell, Text Book of Physics, p. 315. 798 EFFECTS OF CAPILLARITY ON OIL ACCUMULATION 799 amount of pore space, and this is greater than in rocks which have varying-sized and angular grains. Most rocks are made up of particles irregular in shape and less than 2 mm. in diameter, con- sequently the movement of underground water must be greatly affected by capillary action, and evidently the forces of static capil- larity must be overbalanced before movement can take place. For that reason a discussion of Poiseuille’s law of flow in capillary tubes has been omitted, and the conditions of static capillarity are thought to be of first importance. The phenomenon of capillarity—that of a column of liquid rising or being depressed by a small opening—is due to two causes: (z) the surface tension of the liquid, and (2) the fact that the mate- rial of which the tube is composed has a greater or less adhesion for the liquid than the cohesion of the liquid itself. Surface tension is the force at the surface of a liquid, which tends to make the liquid contract, and can be expressed by the following formula: a) jp eS ; 277 COS @ where r equals the radius of the tube; 4, the height of liquid standing in the tube; gq, the density of the liquid; g, the accelera- tion of gravity; and a, the angle of contact between the liquid and the tube. Surface tension is a linear function of the absolute temperature,* and that for water can be expressed by: b) T =o. 21(370—1) ? where ¢ equals the temperature Centigrade. Pressure causes some change in surface tension, but presumably small. ‘For changes in the properties of water induced by pres- sure of, say, 1,000 atmospheres are usually similar in magnitude and direction to those observed when a relatively small quantity of a salt is dissolved in it; and the surface tension of such dilute (0.5 N or less) solutions differs by only a small percentage from that of pure water.’ * Knipp, Physical Review, XI, 151. 2 Johnston and Adams, Journal of Geology, XXII, 9. 3 [bid. 800 A. W. Mccoy Different substances have different surface tensions, which can be calculated by means of formula a) with the necessary observed factors. For instance, crude oil at 20° C. has an average surface tension of about 25 dynes per cm.;? water at 18° C. about 75 dynes; and mercury at 20° C. about 540 dynes.? Surface tension also varies with the nature of materials in surfacial contact. For instance, the surface tension of mercury when in contact with water is different from when in contact with air. Unfortunately, a number of such different values are not recorded, so that this discussion is limited to liquids in contact with air. It is necessary that the adhesion of the material in the tube be either greater or less than the cohesion of the liquid, otherwise there would be no chance for surface tension to display itself. When adhesion is less than cohesion, depression in the liquid results, as in the case of mercury and glass; when adhesion is greater than cohesion, there is a rise in the capillary tube. If adhesion greatly overbalances surface tension, the liquid surface may break and the liquid mount up the sides of the vessel, as in the case of some light oils in a low porcelain cup. Consequently, before one liquid will replace another in capillary openings the replacing liquid must not only have a greater surface tension but also a greater adhesive power for the material of which the tube is composed. Capillary force according to equation a) is a function of surface tension, contact angle, diameter of pore space, density of liquid and acceleration of gravity. In the case of water-air surface the con- tact angle is o, therefore (cos a) equals 1; the density of water is 1; so the equation resolves itself into: h=kT/r, where k equals 0.00204. Starting with a temperature of 15° C., at a depth of 100 m., the capillary pressures shown on p. 801 are computed from the above formula. Pressures are recorded in kilograms per square centimeter. The following calculations show, first, that capillary pressures decrease with depth on account of the increase in temperature; * Washburn, A.I.M.E., L, 831. 2 Tait, Properties of Matter, p. 264. EFFECTS OF CAPILLARITY ON OIL ACCUMULATION 8ot secondly, that above 750 m. capillary pressure in openings of 0.01 micron is greater than the combined rock and hydrostatic pres- sures; therefore capillarity is most important in the upper 3,000 ft. of the earth’s crust; and thirdly, that above 5,000 ft. one liquid of greater surface tension and adhesion for the tube material should readily replace a weaker liquid in small openings; or in other words, the liquid of less surface tension should be concentrated in the larger openings. CAPILLARY PRESSURES UNDER VARYING CONDITIONS* CAPILLARY PRESSURE FOR Ewe Hyprostatic Pore DIAMETER OF Rocz HOSE ATG AND CAPILLARY METERSt PRESSURE ie es RESSURE aneconis tooMc o0.o1rMc TOON ereveace Io 03 306 27 316 BOO ry cision te 50 03 204 135 344 TIOOO seta sia sa) 100 027 278 270 378 ROO mbes reieiat: 200 025 250 540 450 * Johnston and Adams, op. cit., XXII, 13. t An increase of temperature of 1° for every 30 m. was used to obtain these results. Capillary phenomena can take place in openings of o.o1 micron, as shown by Bakker,’ where he concludes that the minimum size of capillary openings is a few times the diameter of the molecule. According to Whitney,? mud contains more than 10,000,000,000 particles per gram. If these were perfectly round particles, so that the pore space could be a maximum, the diameter of the ‘individual would be about 3 microns. Therefore the maximum openings would be about 0.5 micron. Clay used in the following experiments was made up of particles which varied from 1 to 5 microns in diameter, as measured by a microscope. The openings then at a maximum would be a fraction of a micron. Now, since the openings in mud are evidently less than 1 micron by both of the above methods of approach, it has been assumed for the following hypothetical problem, that in compressed shales where the particles are not round nor of equal size the openings are diminished to ©.OI micron. 1 Zeitschrift fiir physikalische Chemie, LXXX, No. 2, 129. 2U.S. Dept. Ag., Weather Bureau, Bull, 4, p. 73- 802 A. W. McCOY Capillary pressure of 300 atmospheres means that water will enter the pore spaces above static water level until the pressure in the pore tubes, due to the weight of the column of liquid above or otherwise, is equivalent to 300 atmospheres pressure; or that, if the water is held back by a gas or liquid of less surface tension, it will accumulate a pressure in the said gas or liquid proportional to the difference in capillary pressures for that temperature and size of opening. | The following assumptions have been made for a hypothetical problem: (1) there exists a cavity or series of connected open- ings, larger than o.5 mm., under a strip of rock 10,000 ft. wide and 1,000 ft. thick. The openings in the rock above are as small as o.or micron, and filled with water; (2) the material below the cavity is an oil shale in which the openings are 0.01 micron, and that water is in the lower part of this shale under sufficient head to make it rise to the level of the bottom of the cavity. The water will drive the oil into the open cavity with a pressure equal to the difference in the capillary pressures of oil and water for that size of opening. This amount for the given temperature of 15° C. and openings of 0.01 micron is approximately 200 atmos- pheres, or about 400,000 lb. per sq. ft. The weight of the rock column above is approximately 150,000 Ib. per sq. ft.; and that of the full water column would be less than 62,000 lb., because the column cannot possibly act upon a full square foot, but only upon the area of pore space, for convenience say 50,000 lb. Now the resultant pressure upon the rock above the cavity is 400,000 minus (150,000 plus 50,000), or 200,000 lb. per sq. ft. This pressure acts as upon a beam fixed at both ends. The capillary water above prevents the rising of the oil into the rock, but in turn affords no downward pressure on the oil in the opening, other than the weight of the hydrostatic column, as has been accounted for in the above assumptions. The deflection for a beam fixed at both ends with a uniform load may be expressed by the following formula: wii d=" 34E1 : EFFECTS OF CAPILLARITY ON OIL ACCUMULATION 803 where d is the deflection; w, the uniform load; 1, length of beam; £, the modulus of elasticity; and I, the moment of inertia. Substituting the values for a beam of rock 10,000 ft. wide, 1,000 ft. deep, 1 ft. broad, with E equal to 6,000,000 lb. per sq. in. (the value of granite), and I equal to 643/12 or (1,000)3/12, the equation resolves itself into the following: 200,000 X 10,000 X 10,000 10,000 X 10,000 12 384. X 6,000,000 X 144 X 1,000 X 1,000 X 1,000 or approximately 72 ft. This means an anticline with a dip each way from the crest of about 1 degree. EXPERIMENT I Statement.—An open glass cylinder (3 in. in diameter, and 8 in. in length) was placed in a pan of wet sand, so that the sand filled the lower one-third of the cylinder. The water had free access from the sand in the pan to the sand in the cylinder. Then a layer of oil-saturated mud was placed in the cyl- inder upon the wet sand; this mud occu- pied about one-third of the cylinder and was above the level of the water in the pan. The cylinder was then filled with dry sand, and the top sealed with a tube attachment toa closed barometer. Read- ings of the mercury were taken before sealing and compared with a standard barometer in the same room. Results.—The water migrated up- ward about 1 cm. into the mud and the oil moved about the same amount into the dry sand. The mercury had risen within 24 hours, about 243 cm. over the atmospheric pressure as compared with the barometer; it then remained stationary. The oil also migrated down into the wet sand and collected in some of the larger openings. Pry send Oi shale Wet sond Wet Sond connected with C closed Saromerver : Tube conzecting jar Experiment [ and bafomete; DM 89H% EXPERIMENT 2 Statement.—A (2-in.) layer of wet sand was placed between two layers of oil in a (8 in.X4 in.X4 in.) rectangular glass box. The sand layer was ar- ranged in an arched manner so that the artificial anticline dipped about 30 degrees to either side. The sand grains in the top of the curve were small (all passing a 4o-mesh sieve), while those in the troughs were comparatively coarse (none passing a 10o-mesh sieve). The top was sealed with paraffin and water 804 A..'G. Mccoy was allowed to enter the box through openings at the lowest horizon of the sand. This water level was never as high as the top of the curve in the sand. Results —The water entered the mud in both directions from the sand layer and replaced about an inch strip of the oil in the mud. The oil moved into the coarser grains of sand and within 24 hours 2 Ql shale there was an oil pool in both Fine sand . . . & Gebescdmiing worer Experiment 2 synclines on either side of a water- & Parattin seal’ | . 5 . F Lines showing replece- filled anticline. Later, the oil ment of of/ wy shale began to move out of the openings which admitted water from the outside, and collected upon the surface of the water. EXPERIMENT 3 Statement.—A (3-in.) layer of oil mud was placed in a (round 14-in. diam- eter) pan, which had a number of small holes in the bottom. A circular lens of dry sand (3 in. in diameter and 3 in. thick) was fitted down in the center at the top of the mud. The surface was leveled as carefully as possible and covered with a }-in. layer of paraffin. This pan was then set in a pan of wet sand, so that the water level stood about 1 in. below the top of the mud in the first pan. A tens of ary sond Results—After two weeks the 3 ov shore” *”” C Wet sand He also dis- cusses the nomenclature of ripple marks at much length in a paper published in 1904,° and elsewhere quotes Lieutenant Damant, R.N., 1H. T. De la Beche, The Geological Observer (Philadelphia, 1851), p. 506. 2H. C. Sorby, “On the Structures Produced by the Currents Present during the Deposition of Stratified Rocks, ” Geologist, April, 1859, p. 141. 3A, R. Hunt, “On the Formation of Ripple-Mark,” Proc. Roy. Soc. London, XXXIV (1882), 2, 18. 4A. R. Hunt, “The Descriptive Nomenclature of Ripple-Mark,” Geol. Mag., N.S., I (1904), 411. 5A. R. Hunt, “On the Action of Waves on Sea-Beaches and Sea-Bottoms,” Proc. Roy. Dublin Soc., N.S., TV (1884), 261-62. 6A. R. Hunt, “The Descriptive Nomenclature of Ripple-Mark,”’ Geol. Mag. N.S., I (1904), 410-18. 812 STUDIES FOR STUDENTS as having observed ripple marks while diving at depths of 60 and 7O Teel. In 1883, the year following the publication of Hunt’s earliest paper cited above, there appeared three important essays on ripple marks: one by De Candolle on “Rides formées 4 la surface du sable déposé au fond de l’eau et autres phénoménes analogues’’; another by Forel on ‘‘ Les rides de fond étudiées dans le lac Léman”’ and a third by Darwin “On the Formation of Ripple-Mark in Sand.” De Candolle produced ripple marks artificially by experi- menting, not only with sand and various substances in powdered form covered by water, but also with liquids of varying viscosity, covered with water and other liquids.?, Regarding sand or powder mixed with water as a viscous substance, he concluded from his experiments that ‘‘when viscous material in contact with a fluid less viscous than itself is subjected to oscillatory or intermittent friction, resulting either from a movement of the covering fluid or from a movement of the viscous mass itself with respect to the covering fluid, (1) the surface of the viscous substance is ridged perpendicularly to the direction of friction, and (2) the interval between the ridges is directly proportional to the amplitude of the friction-producing movement.” That ripple marks depend on simple friction alone, and not on any change of level in the covering liquid, such as occurs during wave action, De Candolle proved by an experiment with a rotating disk submerged in a tank of water. After submerging the disk and mixing an insoluble powder in the water, the apparatus was left until the powder settled on the disk and floor of the tank as an even film, and the water came to rest. An oscillatory rotary movement then applied to the disk caused radiating ripples to form upon it, while no ripples formed on the stationary bottom, and the surface of the water remained quiescent. The author concludes that the formation of ripples in sand, whether under currents of air or under water currents, is identical in origin with the formation of water ripples under moving air. If the cur- tA. R. Hunt, “Facts Observed by Lieut. Damant, R.N., at One Sea-Bottom,”’ Geol. Mag., N.S. WW. (1908), 31-33. HO. Gle Candiaite: ‘Rides formées 4 la surface du sable déposé au fond de l’eau et autres phénoménes foalosncs! Archives des sciences physiques et naturelles, 3° Sér., IX (1883), 241-78. ae anil STUDIES FOR STUDENTS 813 rent moves always in one direction we have intermittent friction due to varying velocities. Otherwise we have oscillatory friction due to alternating change of direction. Current ripples result from the first type of friction, oscillation ripples from the second. Forel in his excellent essay on “‘Les rides de fond étudiées dans le lac Léman”” sets forth the mature results of studies which had been briefly mentioned by him in three communications of earlier date.» Abandoning his first theory, that the formation of ripple marks is dependent in part upon the vertical pressure of water waves upon the bottom,’ Forel reached the following important conclusions as the result of many careful observations and experi- ments: (1) Current ripples are asymmetrical and migrate with the current like ordinary sand dunes, whereas oscillation ripples are stationary and symmetrical. (2) Each oscillation ripple is really a composite of two current ripples, resulting from the action of two currents moving alternately in opposite directions, each cur- rent attempting to form the ridge into a current ripple migrating with it, but being defeated when the return current tries with equal force to shape the ridge into a current ripple directed in the opposite sense. (3) The length of the water body has no direct effect on the spacing of the ripples. (4) Other things being equal, the ripples are more closely spaced with increasing depth. (5) At a given depth, and with other conditions uniform, the ripples are more widely spaced with increase in coarseness of sand grains. (6) Ripples once formed do not experience a change in spacing as a result of diminishing amplitude of oscillation of the water, although the original spacing does depend upon the amplitude of oscillation, as pointed out by De Candolle. (7) For any given coarseness of sand grains there is a certain mean velocity of the oscillating currents which will produce ripples; lower velocities 1 F, A. Forel, ‘‘Les rides de fond étudiées dans le lac Léman,” Archives des sciences physiques et naturelles, 3° Sér, X (1883), 39-72. 2F, A. Forel, “‘La formation des rides du Léman,” Bulletin de la Société Vaudoise des sciences naturelles, X (1870), 518; ‘Les rides de fond,” ibid., XV (1878), P.V. 66- 68; “Les rides de fond dans le golfe de Morgues,” tbid., 76-77. 3F. A. Forel, “La formation des rides du Léman,” Bulletin de la Société Vaudois des sciences naturelles, X (1870), 518; “Les rides de fond étudiées dans le lac Léman,” ‘Archives des sciences physiques et naturelles, 3° Sér., X (1883), 40. 814 STUDIES FOR STUDENTS will fail to move the sand grains, and hence cannot build ripples, while higher velocities agitate the whole mass of sand so violently that no ripples can form. (8) The formation of ripples is initiated by some obstacle or inequality on the surface of the sand, behind which sand grains accumulate in the eddy caused by its presence; this leaves a furrow on either side of the initial ridge, .due to the abstraction of sand accumulated in the ridge; and these furrows in their turn cause additional ridges to develop on their outer mar- gins, and so on. (g) In a given locality, ripple marks almost always form with the same spacing, regardless of the varying intensity of winds and waves affecting the water body; this is in consequence of laws 7 and 6 stated above. (10) The depth at which ripple marks may form is limited by the depth to which wave action may extend with sufficient energy to move the bottom sands; hence it depends on the size of the waves, and therefore in part indirectly on the size of the water body; in the Rhone, the limiting depth is a few decimeters; in Lake Geneva, some ten meters; and in the ocean, from 20 to 188 meters, according to Lyell and Siau. Forel revised De Candolle’s law regarding the relation of ripple spacing to the amplitude of the friction-producing movement to read: “The breadth of the ripples, or the distance from one crest to another, is the length of the path followed during a single oscillation by a grain of sand freely transported by the water.’’ The length of this path varies directly as the horizontal amplitude of the oscillatory movement of the water, directly as the velocity of that movement, inversely as the density of the sand, and inversely as the size of the sand grains. Darwin’s paper ‘‘On the Formation of Ripple-Mark in Sand” is especially noteworthy for its careful analysis of the vortices which are so important a factor in the construction of the ripples. When symmetrical oscillation ripples were subjected to the action of a steady current, Darwin noticed that not only did sand grains migrate up the weather slope of ‘each ripple with the current, but that they also ascended the lee slopes, apparently against the cur- rent. This proved conclusively the existence of vortices. Darwin *G. H. Darwin, “On the Formation of Ripple-Mark in Sand,” Proc. Roy. Soc. London, XXXVI (1883), 18-43. STUDIES FOR STUDENTS 815 then proceeded to study the vortices by watching the movements of a drop of ink released from the end of a fine glass tube at that point in the water where the action was to be observed. In this manner the vortices associated with the alternating currents which produce oscillation ripples were analyzed with a high degree of precision, and much light was thrown upon the method of ripple growth. Darwin concluded that “the formation of irregular ripple marks or dunes [current ripples] by a current is due to the vortex which exists on the lee side of any superficial inequality of the bottom; the direct current carries the sand up the weather slope and the vortex up the lee slope. Thus any existing inequali- ties are increased, and the surface of sand becomes mottled over with irregular dunes.” The intermittent friction which De Candolle adduced is not essential in this explanation of current ripples. Oscillation ripples of regular pattern are changed by a continuous current into regularly spaced current ripples; but a uniform current cannot of itself initiate regularly spaced ripple marks. ‘‘Regular ripple mark [oscillation ripples] is formed by water which oscillates relatively to the bottom. A pair of vortices, or in some cases four vortices, are established in the water; each set of vortices corresponds to a single ripple crest.” Forel’s con- ception of an oscillation ripple as a composite of two current ripples formed alternately by oscillating currents is regarded as correct; but his law for the relation of ripple spacing to amplitude of oscillation is believed to require some modification. Further studies of ripple-forming vortices were made by Mrs. Hertha Ayrton, the results of which were not published until 1910.t With the aid of well-soaked grains of ground black pepper, or of particles of potassium permanganate dissolving and coloring the water while the latter was in oscillation, she observed the formation of vortices and endeavored to explain the mechanics of their growth. Although she expressed disagreement with the conclusions of Darwin and others on certain points, most of her results afford essential confirmation of their main contentions. Some doubt must attach to certain of her deductions, such as 1H. Ayrton, “The Origin and Growth of Ripple-mark,” Proc. Roy. Soc. London, Ser. A., LX XXIV (1910), 285-310. 816 STUDIES FOR STUDENTS one to the effect that no ripple-forming vortex occurs in the lee of an obstacle over which a steady current is passing, and that hence ‘‘a steady current is unable either to generate or to maintain ripple mark.”’ The British Association Reports for the years 1889, 1890, and 1891 contain three papers by Reynolds on the action of waves and currents in model estuaries, in which are some valuable observa- tions regarding what may well be termed giant tidal ripples. While experimenting with artificial tidal currents, Reynolds dis- covered that current ripples were formed in the model estuaries. By making due allowance for the difference in size between the model estuaries and those in nature, he concluded that real tidal currents ought to produce very large current ripples, possibly 7 or 8 feet in height and 80 to 100 feet apart.2, Some years later Vaughan Cornish discovered natural tidal ripples of the same type as those produced artificially by Reynolds, having a height of 2 feet and an average distance of more than 37 feet from crest to crest. In two later papers Cornish described giant tidal ripples more fully, and illustrated their essential features with a large series of beautiful photographs.4 Some of these ripples have a height of nearly 3 feet above the intervening troughs, and a distance between crests of from 66 to 88 feet in extreme cases. The giant ripples are often covered with ordinary ripple mark, and while Cornish recognized that the larger forms were produced by the continuous steady flow of tidal currents, he was at first inclined to invoke pulsatory currents in order to explain thesmaller ripple mark.5 t Osborne Reynolds, “Report of the Committee Appointed to Investigate the Action of Waves and Currents on the Beds and Foreshores of Estuaries by Means of Working Models,” Rept. British Assoc. (1889), pp. 327-43; ibid. (1890), pp. 512-34; ibid. (1891), pp. 386-404. 2 [bid. (1889), p. 343. 3 Vaughan Cornish, “On Tidal Sand Ripples above Low-Water Mark,” Rept. British Assoc. (1900), Ppp. 733-34- 4 Vaughan Cornish, “‘Sand Waves in Tidal Currents,” Geogr. Jour., XVIII (1901), 170-202; ‘“‘On the Formation of Wave Surfaces in Sand,” Scottish Geogr. Mag., XVII (1901), I-11. 5 Vaughan Cornish, “On Tidal Sand-Ripples above Low-water Mark,” Rept. British Assoc. (1900), p. 733; ‘“‘Sand Waves in Tidal Currents,” Geogr. Jour., XVIII (1901), 197-98; ‘‘On the Formation of Wave Surfaces in Sand,” Scottish Geogr. Mag., XVII (1901), 8. it ed 2 STUDIES FOR STUDENTS 817 - This theory seems to be a survival of De Candolle’s erroneous idea that “intermittent friction” is essential to the production of current ripples, and is practically abandoned by Cornish in his more recently published book on Waves of Sand and Snow. Gil- more described tidal ripples on the Goodwin Sands having a height of ‘‘two or. three feet.’ It should be noted that all of the giant ripples referred to above belong to the asymmetrical type; they are true current ripples. So far as I am aware no giant oscillation ripples have ever been observed along modern shores. It may be doubted whether tidal currents could form symmetrical ripples, notwithstanding Rey- nold’s suggestion to the contrary. The flow and ebb of the tide constitute an oscillating current, it is true; but the currents are often of unequal force. Where equally strong, each current per- sists long enough to remodel the ridges formed by the preceding current, giving them an asymmetrical form appropriate to the current operating last. On the other hand, Gilbert has described structures in the Medina sandstone formation of New York which he believed to be giant ripples of the symmetrical type, formed by oscillating currents due to wave action. In dimensions these ridges were similar to the average examples of tidal ripples described by Cornish, having a height of from 6 inches to 3 feet, and a dis- tance from crest to crest of from 10 to 30 feet; but their nearly symmetrical form did not suggest a similar origin. Gilbert reached the tentative conclusion that they were formed by waves 60 feet high in deep water of a broad ocean. This conclusion was criti- cized by Fairchild, who advanced convincing arguments in support of the opinion that the forms in question were beach structures, possibly successive beach ridges built on the strand.* Branner ? Vaughan Cornish, Waves of Sand and Snow (London, 1914), pp. 289-90. 2 John Gilmore, Storm Warriors, or Lifeboat Work on the Goodwin Sands (London, 1874), pp. 108-9. 3 Osborne Reynolds, “Report of the Committee Appointed to Investigate the Action of Waves and Currents on the Beds and Foreshores of Estuaries by Means of Working Models,” Rept. British Assoc. (1889), DP. 343- 4G. K. Gilbert, ‘“‘Ripple-Marks and Cross-Bedding,” Bull. Geol. Soc. Amer., X (1899), 135-40. s H. L. Fairchild, “Beach Structure in the Medina Sandstone,’ Amer. Geologist, XXVIII (1901), 9-14. 818 STUDIES FOR STUDENTS suggested that they might represent fossil beach cusps seen in cross-section.” In 1911 A. P. Brown published a paper entitled ‘“‘The Formation of Ripple-Marks, Tracks, and Trails,”’ in which he endeavored to show that asymmetrical ripples (current ripples) are formed by deposition, whereas symmetrical ripples (oscillation ripples) result from the erosion of a formerly smooth botton, consequent upon the rippling of overlying water by wind action.? His con- clusions do not appear to be supported by a sufficient body of con- vincing evidence, and are opposed by theoretical considerations and by the great body of experimental data already referred to. In presenting his theory this author makes no reference to the many previous investigations of ripple marks of all kinds, the important results of which have been summarized above. Ripple marks have repeatedly been discussed in connection with the interpretation of fossil ripples found in sedimentary rocks. We need mention but a few of these discussions in the present con- nection. As early as 1831 Scrope described fossil ripple marks found on slabs of marble, and explained them as due to the oscil- latory movements of shallow water. Darwin, starting from the very questionable assumption that a great ebb and flow of the tide is essential to the formation of numerous ripples, concluded that the presence of a large number of ripple marks in a geological formation indicates with a considerable degree of probability that the tides of early times rose higher than those of today.4 Van Hise figured and described one type of oscillation ripples, and empha- sized their value as criteria for determining the original altitude of steeply inclined strata.s tJ. C. Branner, editorial note, Jour. Geol., [IX (1901), 535-36. 2A. P. Brown, “The Formation of Ripple-Marks, Tracks, and Trails,’ Proc. Assoc. Nat. Sci. Philadelphia, LXIII (1911), 536-47. 3G. P. Scrope, ““On the Rippled Markings of Many of the Forest Marble Beds North of Bath, and the Foot-Tracks of Certain Animals Occurring in Great Abuf- dance on Their Surfaces,”’ Proc. Geol. Soc. London, I (1831), 317-18. 4G. H. Darwin, “On the Geological Importance of the Tides,’’ Nature, XXV (1882), 214. 5C, R. Van Hise, ‘‘Principles of North American Pre-Cambrian Geology,’ Sixteenth Ann. Rept. U.S. G. S., Part I (1896), 719-21. = ee eC STUDIES FOR STUDENTS 819 Spurr showed that where continuous deposition takes place from a current which constantly maintains asymmetrical ripples on the surface over which it flows, the forward movement of the ripples combines with the deposition of heavier and larger fragments in the troughs and lighter particles on the crests to give a peculiar type of false bedding in the resulting formation.t Jaggar criticized Spurr’s conclusions on the ground that his own experiments and observa- tions indicated that ripple marks could not be produced in hetero- geneous material;? but Spurr met the criticism with a fuller discussion of the matter in which his original contention is well sus- tained. A short time previously Sorby had described a somewhat similar phenomenon in a paper’ printed almost exactly half a century after the publication of his first account of ripple marks, already cited. From an examination of the “ripple-drift” type of false bedding in rocks, Sorby believed that one could “ascertain with approximate accuracy, not only the direction of the current and its velocity in feet per second, but also the rate of deposition in fractions of an inch per minute.”> Additional discussions of fossil ripple marks are cited by Kindle in his paper referred to at the beginning of this article, but need not be repeated here. 1 J. E. Spurr, ‘False Bedding in Stratified Drift Deposits,’ Amer. Geologist, XIII (1894), 43-47. 2T. A. Jaggar, Jr., ‘Some Conditions of Ripple-Mark,” Amer. Geologist, XIII (1894), 199-201. 3J. E. Spurr, ‘Oscillation and Single-Current Ripple Marks,” Amer. Geologist, XIII (1894), 201-6. 4H. C. Sorby, “On the Application of Quantitative Methods to the Study of the Structure and History of Rocks,” Quart. Jour. Geol. Soc. London, LXIV (1908), 180-85. 3 Ibid., pp. 181, 197-99. REVIEWS Geology of Saratoga Springs and Vicinity. By H. P. CusHinc and R. RUEDEMANN. New York State Museum; Bull. No. 160, ro14.» Pp. 177, pls..20, figs: 17, maps"2. Scientific interest regarding Saratoga Springs and vicinity centers about its mineral waters, and this report has been published in response to a demand for detailed information on local geological conditions. Rocks of Pre-Cambrian, Cambrian, and Ordovician age outcrop in the area. The Paleozoic rocks are divided into deposits of eastern and western troughs, characterized by different sets of formations. The western trough was being eroded in Lower Cambrian times, but in the east the Georgian is the only Cambrian present. The rocks of the western division are horizontal or nearly so, but in the east the beds are intensely folded and crumpled. Two great normal faults with a number of branches cross the Saratoga quadrangle. These are known to be genetically connected with many of the mineral springs. A unique feature is the Northumberland volcanic plug. It outcrops just north of Schuylerville as a knob of extrusive rock and is unlike any other igneous rock in the state. It has been connected with one theory for the origin of the mineral springs, but unfortunately the authors were unable to determine with certainty whether the rock is in place or not, and are in doubt in regard to calling it a volcanic neck or a fragment of a surface flow. It was planned to have Professor Kemp write a chapter for this bulletin on the origin of the mineral waters but his results were published in an earlier report. The authors are not convinced that Kemp’s con- clusions are justified by the field evidence. Kemp holds that the mineral waters, in part at least, are of magmatic origin. He cites as proofs their local occurrence, the volcanic neck, the large amount of free CO., and the almost complete absence of sulphates. The authors believe that the absence of carbonated waters to the north is due to lack of shale covering and resulting dilution with surface waters. They hold that the volcanic knob furnishes no evidence of igneous activity of sufficient recency to justify connecting it with present-day juvenile waters. The abundant CO, may come from deeply buried impure limestones and 820 REVIEWS 821 shales. The absence of the SO, radicle does not dismiss the possibility of connate waters as a source of the mineral salts. The sulphates originally in the connate waters may have been lost as the waters moved along toward the surface by some such chemical reaction as the precipita- tion of gypsum by the action of sodium sulphate on calcium carbonate in the presence of free CO,. Thus the chief problem that this quadrangle offers is held in question still, but this is not due to lack of skill or painstaking effort on the part of the authors of this report. It is a worthy contribution to the geologi- cal literature of this state. W.B.W. Genesis of Pyrite Ores of St. Lawrence County. By C. H. Smrru, JR. New York State Museum, Bull. No. 158, 1912, pp. 143-82. Figs. 29. Under the most favorable conditions, definite conclusions regarding the geneses of ore bodies cannot always be drawn, and when these are _ found in bodies of rock as highly metamorphosed as the Grenville series many complications arise. In this area, pyrite is widely disseminated, but the ore bodies are associated only with “‘rusty gneisses thought to be metamorphosed impure sandstones and shales.” The writer believes that the metallizing period was subsequent largely to the main period of metamorphism, and was brought about by magmatic emanations permeating the gneisses and replacing with pyrite certain minerals which are usually very stable. These emenations came from the abundant intrusions after active move- ment of the magmas had ceased. It is not stated that the pyrite all came from the magmas. In fact, to explain the association of the ore bodies with the gneisses alone, it is suggested that only the sulphur was of igneous origin, and that the iron was furnished by the meta- morphosed sediments. To cover minor occurrences of pyrite three addi- tional periods of formation are postulated, but are not considered to have been of importance in determining the ore bodies. Additional points of interest are found in the lack of association of the ore bodies with gabbros, as some authors have stated in other areas, and possible genetic relations of pyrite with associated graphite. The explanation of the ore bodies strikes one as quite involved, but the author assures us that it is in very small proportion to the complexity of field problems and conditions. W. B. W. 822 REVIEWS Geological History of New York State. By WitttamM J. MILLER. New York State Museum, Bull. No. 168, 1914. Pp. 130, pls. 52, figs. 4o. This bulletin is a brief summary of the geological history of the state. It was the intention of the author to presuppose no scientific knowledge of geology on the part of his readers, and that the work should be in the nature of a textbook. A few pages in the introduction are devoted to geologic processes and throughout the context an effort is made to define technical terms. The reviewer does not believe the author succeeded in making the report sufficiently non-technical to be popular with laymen. It will serve better as a reference book for geologists who wish a brief statement of some of the larger phases of the region’s history. The report is illustrated with many excellent photographs of uncon- formities and other structural and physiographic features which abound in the state. Unfortunately this report, in common with other New York reports, does not contain a table of contents and its value as a reference book is impaired thereby. W. B. W. Origin of Hard Rock Phosphates of Florida. By E. H. SELLARDS. Florida Geol. Survey, Fifth Annual Report, 1913, pp. 23-80, pls. 9, map I. The hard rock phosphates are found chiefly as bowlders and irregular fragments in a formation of Pliocene age that the author has named Dunnellson. The formation is rather heterogeneous but a phase of light-gray sands is the usual matrix in which the phosphate rocks are imbedded. Theories generally advanced to explain these deposits have involved some form of guano alteration. The author believes the real source of the phosphate was from phosphoric acid derived from the disintegration, in situ, of overlying beds. The acid was borne downward by ground- water, and replaced limestone, or was chemically precipitated. No reactions are suggested for the latter process. The deposits are asso- ciated with clay lenses and other conditions that interfere with the free circulations of ground-waters. It is suggested that the presence of precipitating agents may be the important factor here rather than the retardation of ground-water circulation. The shattered and hetero- — REVIEWS 823 geneous character of the formation is explained by the caving in of solu- tion cavities and their subsequent refilling. The theory presented seems to explain the larger features of the phosphate deposits, but the report should be considered a statement of progress of investigation, rather than the last word in explanation of the deposit. W. B. W. Water Supply of Eastern and Southern Florida. By E.H.SELLARDS. Florida Geol. Survey, Fifth Annual Report, 1913, pp. 113- 288, pls. 5, figs. 17, map 1. This report covers in detail an area of twenty-two counties in which, for the most part, the artesian waters may be tapped by flowing wells. This area includes the outer rim of counties along the eastern, southern, and southwestern borders of the state. The principal aquifer is the Vicksburg limestone of Oligocene age. Underlying the whole state, this formation is exposed in the central part and dips beneath younger formations to the east and south. These younger beds have not been well differentiated and some wells may obtain water from them, but strong flows are from the Vicksburg. The structure includes a low anticline with its axis dipping gently to the east in the central part of the state. The water-bearing horizon is 100 feet below the surface along the coast, and near the crest of the anti- cline. In the northeast corner of the state the wells are from 300-400 feet deep and at the southern extremity from goo-1,000. The gentle dip of the strata does not furnish strong pressure in any locality and a head of 25 feet is rather exceptional. Local topography affects the distribution of the flowing wells. In some areas there has been great development of the artesian water supply. There are not less than 500 flowing wells in the city of Jackson- ville. Statistics covering recent years show a progressive loss of flow from the wells in this city. Much of the artesian water of the state is not potable on account of mineral salts, chiefly sodium chloride. This is notably true in the southern part. All the underground water of the state is very generally charged with hydrogen sulphide, but its use for domestic purposes is not prevented thereby. A small area of flowing wells in the western part of the state is not treated in detail in this report. : W. B. W. 824 REVIEWS Geology of North Creek Quadrangle. By Wrtt1am J. MILLER. New York State Museum, Bull. No. 170, 1914. Pp. go, pls. 14, figs. 9, map 1. This quadrangle lies wholly in Warren County, New York, in the southeastern Adirondacks. It is of geologic interest chiefly because of certain rock types and structures. At the present time no rocks of later age than Pre-Cambrian are present, but Paleozoic outliers just off the map seem to prove that late Cambrian and probably early Ordovi- cian sediments have been removed. The Grenville series makes up the meta-sedimentary rocks, and the author believes the evidence favors their Archeozoic rather than Proterozoic age. This series has a lime- stone member of remarkable thickness, 10,000-12,000 feet, and below this is 3,000 feet of quite pure quartzite. About 60 gabbro outcrops are shown on the map, usually with elliptical ground-plans. Their form is that of small stocks or bosses, rather than dikes. The author believes these gabbro occurrences furnish strong evidence in favor of Daly’s magmatic stopping and assimilation hypothesis. The igneous masses were not intruded by pushing aside the country rock, but rather by a process of replacement. Marked primary variations in the gabbros and the presence of inclusions as xenoliths are cited in support of this theory and seem to make a strong case. Garnets are present in the area in quantities of some economic importance. Some of the occurrences are attributed to the assimilation of Grenville sediments, and subsequent crystallization from “original magmas.” This use of the term “original magma” for a magma that has assimilated considerable quantities of sediments is questionable. W. B. W. The Waterlian Formations of East Central Kentucky. By W. C. Morse and A. F. Forerste. Kentucky Geol. Survey, Bull. No. 16; ° Pp) 76; figs.-s. | Stratigraphic relations of the Mississippian beds in Kentucky are of interest for economic reasons. In Ohio and West Virginia forma- tions in the Waverly series are oil- and gas-producers and their extension into Kentucky is a fact of considerable importance. This report covers twelve counties in the east-central part of the state. The beds were traced southward from known sections in Ohio, REVIEWS 825 making the correlations fairly certain. The sections show that toward the south the sandstones of the oil horizons of Ohio rapidly grade into shales. Even the shales of the Bedford and Berea formations become very thin although they do not disappear. These changes are unfavor- able for oil-bearing sands in the southern part of the state. A number of changes in correlations in formations of the Waverlian are made from those given in earlier Kentucky reports and in the Richmond Folio. The latter part of the bulletin treats of the possi- bilities of these beds in producing building stones and clays. If a map of the area covered by the report had been given it would have made part of the discussion more intelligible. W. B. W. The Geology of the Rolla Quadrangle. By WattacE LEE. Missouri Bureau of Geology and Mines, XII. Pp. 117, pls. 12, figs. 17, maps 2. The area covered by this report is in the central Ozark region of Missouri and includes Phelps and Dent counties. The strata described include the Gasconade, Roubidoux, and Jefferson City formations. The general horizon is of interest because it includes part of the Ozarkian and the Canadian of Dr. Ubrich’s classification. The author follows the usual classification, placing these beds in the Upper Cambrian. A few erosion remnants of Carboniferous age are found in the northeastern part of the area. An interesting structural feature is found in a number of sink areas. The author believes it was developed from the caving and subsequent filling of solution cavities. The economic products of this quadrangle are negligible and the chief value of the report lies in its contribution to the general stratigraphy of the region. W. B. W. Glass Sands of Oklahoma. By Frank BurtraM. Oklahoma Geol. Survey, Bull. No. 10, 1913. Pp. 91, pls. 8, figs. 3. Approximately one-half of this report is taken up with a general de- scription of the glass industry. As the author is a chemist he has treated chemical processes in glass production rather fully. Notable glass sand deposits of the state are limited to three areas: the Arbuckle Mountains, southeastern Oklahoma, and near Tahlequah, 826 REVIEWS in the northeastern part of the state. The greatest deposits are in the Arbuckle Mountains, in the Simpson formation of Ordovician age. This formation is from 1,200 to 2,000 feet thick, and the sands outcrop at four horizons. Five sections give an average thickness for the glass sands of 248 feet. The supply of raw materials seems almost inexhaust- ible but transportation facilities are lacking in most localities. In south- eastern Oklahoma the Trinity sandstone at the base of the Cretaceous carries commercial quantities of good glass sand, and several localities are readily accessible. In the northeast the Burgen sandstone, which has been correlated with the St. Peters, carries a 50-foot bed of high- grade glass sand, but is too remote from railroads for present develop- ment. Analyses of these sands show that they compare very favorably with deposits now being worked in adjacent states. Having a marked advantage in the use of natural-gas fuel, Oklahoma sands should prove strong competitors for the glass market of the central Mississippi Valley. W. B. W. Inland Lakes of Wisconsin. By Epwarp BIrRGE and CHANCY Jupay. Wisconsin Geol. Survey, Bull. No. 27, 1914. Pp. 132, figs. 8, tables 4, maps 29. A large portion of the data in this report has been published in various bulletins scattered through a dozen years. It seemed desirable to gather this material in a single volume, together with additional data not published hitherto. No lakes occur in the southwest or driftless area, and all the lakes of the remaining three-quarters of the state are of glacial origin. In general the lake basins were formed in four different ways: by melting of blocks of ice imbedded in the glacial débris, by damming of preglacial valleys, by interlocking of terminal moraine deposits, and by inequalities in deposition of ground moraine. The total number of lakes runs into the thousands, but only the larger ones are described. There are 21 hydrographic maps, and with each is a brief report on the geology and topography of adjacent regions and the origin of the lake basins. Tables give data on the locations of the lakes, their size, and the depth and shape of their basins. Lake Winnebago with an area of 215 square miles is by far the largest in the state. Few of the lakes exceed one hundred feet in depth. The United States Geological Survey has ——E——————<— << lh ee REVIEWS 827 estimated the total lake area of the state at 810 square miles. The authors believe that twice this amount is more nearly correct. W. B. W. _ Preliminary Report on Tertiary Paleontology of Western Washington. By Cuartes E. Weaver. Washington Geol. Survey, Bull. INOW TS.59r2. Pp. 80, pls. 15. A Tertiary invertebrate marine fauna of 246 species is listed in this report. Eighty-four of these are new species and are described and figured for the first time. The fauna is very largely pelecypods and gastropods. Lower Eocene rocks are absent. The Upper Eocene fauna totals 79 species. The Oligocene fauna is limited to 10 species. A detailed report will supplement this bulletin later and treat more fully of the stratigraphic and structural relations. W. B. W. Geology of East Central Oklahoma. By L. C. SNiDER. Okla. Geol. Survey, Bull. No. 17, 1914. Pp. 25, pls. 2, fig. 1. The area treated in this report includes all of Haskell County and portions of five adjoining counties. It deals with structural features almost entirely and the stratigraphy given follows United States Geologi- cal Survey reports. About twenty anticline and syncline axes are plotted. Well- drillers may locate the axes of anticlines roughly from this map and supplement it by detailed work in each locality. For the convenience of many who have not access to the annual reports of the United States Geological Survey, the report includes a map and descriptions of the principal folds in a region adjacent on the southwest. A number of wells are producing gas in these two areas, but oil wells of importance have not been reported. W. B. W. Ponca City Oil and Gas Field. By D. W. OHERN and R. E. Gar- RETT. Okla. Geol. Survey, Bull. No. 16, 1912. Pp. 30, pls. 2, 1a 3 The Ponca oil and gas field is located in north-central Oklahoma near the Kansas line. It produced gas only until 1911 when the first oil 828 REVIEWS well was brought in. Thirty producing oil wells were operating at the time this bulletin was written. The report describes the formations of Lower Permian and Pennsyl- vanian age that outcrop in the Ponca City area, and also those under- lying that outcrop to the east and west. The structure of the Ponca City anticline is shown by a contour map on the surface of the Herington limestone. It is the opinion of the authors that many of the wells labeled “dry” are not deep enough to test their localities. Some holes do not go down 1,000 feet, and few below 1,600; but the approximate position of the lowest oil sand is much deeper, and the anticline will not be tested thoroughly until wells have reached the Tucker sands at a depth of nearly 3,500 feet. W. B. W. The Mineral Springs of Saratoga. By JAmMes F. Kemp. New York State Education Department, Bull. No. 517, 1912. Pp. 79, figs. 8, tables 7. There are few problems more difficult for geologists than those con- nected with the origin of mineral springs. The district centering at Saratoga Springs has long been famous for its mineral waters, and this report has been prepared in response-to the very general interest regard- ing them. The report takes up briefly a historical sketch of the springs, the local geology, and a general description and classification of ground- waters. The chemical composition of the water is known by analyses of three different periods, 1838, 1871, and 1905. These show a total of ten acid and twelve basic ions. The most abundant salt is sodium chloride followed by calcium, magnesium, and sodium bicarbonates. The waters carry an average of two or three volumes of CO, in solution. The sulphate ion is practically absent. The author rejects any theory that attributes the springs to connate waters, the absence of sulphates being the strongest chemical evidence against such theories. The same geological section is faulted in many other places in the Hudson and Champlain valleys, yet even uncar- bonated brine springs are lacking elsewhere. The author’s conclusion is that many of the mineral constituents, as the haloids, sodium car- bonate, and the carbonic acid gas, are from deep-seated sources. The tendency of dying volcanoes to give off abundant CO, and the occurrence REVIEWS 829 within ten miles of the only purely volcanic rock in New York, Vermont, or western Massachusetts support this theory. The carbonated waters take on calcium and magnesium carbonates from the Little Falls dolomite on their upward journey. This conclusion accords with the marked tendency of economic geologists in the last decade to lay greater stress on the importance of magmatic emissions. W. B. W. Coal Resources of District No. I (Longwall). By Girpert H. Capy. Illinois Coal Mining Investigations, Bulletin No. 10, Urbana, t915. Pp. 149, pls. 9, figs. 27, tables 24. The Longwall District, comprising Bureau, Putnam, Marshall, La Salle, and Grundy counties and the adjacent parts of Livingston, Kankakee, and’ Will counties, an area of about 1,700 square miles, contains nearly six billion tons of available coal and is one of the foremost districts of the state in economic importance. This bulletin is con- cerned with the stratigraphic and structural geology of the region, the economic geology of the coals and accompanying strata, and with the working data developed. The important beds are Nos. 2, 5, 6, and 7, of which No. 2 has been extensively mined. These coals have been studied in a large number of mines. The character of the coal beds and their general structure have been worked out in detail, and many sections through the productive coal measures have been tabulated. In addi- tion to its value in connection with the coal resources, the bulletin is of general interest in that it contains an outline of the geology of the La Salle anticline, including Starved Rock, Deer Park, and the surrounding country. Aves Be Coal Resources of District No. VII. By FRED H. Kay. Illinois Coal Mining Investigations, Bulletin No. 11, Urbana, 1915. Pp. 233, pls. 4, figs. 47. District No. VII comprises Macoupin, Madison, St. Clair, Christian, Montgomery, Bond, Clinton, Washington, Perry, Moultrie, Shelby, Fayette, Marion, and parts of Sangamon, Macon, and Randolph counties, an area of about 7,000 square miles, containing coal estimated at more than forty-five billion tons in bed No. 6 alone. The stratigraphy of the coal measures has been carefully studied, and numerous sections have been measured and tabulated. Some interesting structures in the coal 830 REVIEWS beds have been noted, and illustrated by a number of diagrams.- The rocks of the area are confined to the Pennsylvanian, with comparatively simple structure. No. 6 is the only coal bed producing important quantities of coal. An De Be Notes on Geology of the Gulf of St. Lawrence. By J. M. CLarK. New York State Museum, Bull. No. 158, pp. 111-20. The author treats briefly of the geology of Entry Island. Of chief interest is his description of a type of topography which he calls “‘dem- oiselle.” The relief is due to numerous mammiform hills rounded into softly contoured domes of striking symmetry. These domes are caused by the erosion of small laccoliths. W. B. W. a _ —— RECENT PUBLICATIONS —SxirF, F. J. V. Annual Report of the Director to the Board of Trustees, for the Year 1914. [Publication 181, Field Museum of Natural History, Report Series, Vol. IV, No. 5. Chicago, January, 1915.| .—Smito, W. D. The Mineral Resources of the Philippine Islands for the Year 1913. [Division of Mines, P.I. Bureau of Sciences. Manila, 1914.] —SPENCER, M. L. J. Données Numériques de Cristallographie et de Minér- alogie. Tables annuelles de Constantes et Données Numériques de Chemie, de Physique et de Technologie. [Paris: Gauthier-Villars et Cie; Chicago: The University of Chicago Press, 1914.] —STANSFIELD, E., AND CARTER, F. E. Products and By-Products of Coal. [Canada Department of Mines, Mines Branch No. 323. Ottawa, 1915.] —STEPHENSON, L. W. The Cretaceous Eocene Contact in the Atlantic and Gulf Coastal Plain. Shorter Contributions to General Geology, 1914-J. [U.S. Geological Survey, Professional Paper 90-J. Washington, 1915.] , AND VEATCH, J. O. Underground Waters of the Coastal Plain of Georgia. And a Discussion of the Quality of the Waters, by R. B. Dote. [U.S. Geological Survey, Water-Supply Paper 341. (Prepared in co-oper- ation with the Geological Survey of Georgia.) Washington, 1915.] —SutTerR, H. Revision of the Tertiary Mollusca of New Zealand, Based on Type Material. Part I. [New Zealand Geological Survey, Department of Mines, Palaeontological Bulletin No. 2. Wellington, 1914.] Revision of the Tertiary Mollusca of New Zealand, Based on Type Material. Part II. [New Zealand Geological Survey, Palaeontological Bulletin No. 3. Wellington, 1915.] —TyrrELL, J.B. Gold-bearing Gravels of Beauce County, Quebec. [Trans- actions of the American Institute of Mining Engineers. Toronto, 1915.] Gold on the North Saskatchewan River. [Canadian Mining Insti- tute Bulletin, February, 1915. Toronto, March, ro15.] —U.S. Bureau of Mines. Fourth Annual Report of the Director of the Bureau of Mines to the Secretary of the Interior. For the Year Ended June 30, tg14. [U.S. Bureau of Mines, Washington, 1914.| —Wape, A. The Supposed Oil-bearing areas of South Australia. [Geo- logical Survey of South Australia, Department of Mines, Bulletin No. 4. Adelaide, 1915.] —Washington Academy of Sciences, Journal of the. Vol. V. (Baltimore: Waverly Press.| —Washington University Studies, Vol. II, Part I, No. t. [St. Louis, July, 1914.] 831 832 RECENT PUBLICATIONS —WEGEMANN, C. H. The Coalville Coal Field, Utah. [U.S. Geological Survey, Bulletin 581-E. Washington, 1915.] —Whit_ErR, H. J. Soils of Massachusetts and Connecticut with Especial Reference to Apples and Peaches. [U.S. Department of Agriculture, Bulletin 140. Washington, April 5, 1915.] —WILKMAN, W.W. Kvartara Nivaforandringar i Ostra Finland. Deutsches Referat. [Bulletin No. 33 de la Commission Géologique de Finlande. Helsingfors, April, 1912.] —Wituams, M. Y. The Ordovician Rocks of Lake Timiskaming. [Canada Department of Mines, No. 1542, Museum Bulletin No. 17, Geological Survey, Geological Series No. 27. Ottawa, 1915.] —wWisconsin Academy of Sciences, Arts, and Letters, Transactions of the. Vol. XVII, Part II, Nos. 1-6. [Madison, 1914.] —Woop, B. D. Stream-gaging Stations and Publications Relating to Water Resources, 1885-1913. Part VIII. Western Gulf of Mexico Drainage Basins. [U.S. Geological Survey, Water-Supply Paper 340-H. Wash- ington, 1915.] ———. Stream-gaging Stations and Publications Relating to Water Resources, 1885-1913. Part IX. Colorado River Basin. [U.S. Geo- logical Survey, Water-Supply Paper 340-I. Washington, 1o915.] Stream-gaging Stations and Publications Relating to Water Resources, 1885-1913. Part X. The Great Basin. [U.S. Geological Survey, Water-Supply Paper 340-J. Washington, 1915.] —Wood, H. O. The Hawaiian Volcano Observatory. [Reprinted from the Bulletin of the Seismological Society of America, Vol. III. No. 1, March, 1913. Hawaiian Volcano Observatory.] The Seismic Prelude to the 1914 Eruption of Mauna Loa. [Reprinted from the Bulletin of the Seismological Society of America, Vol. V, No. 1, March, 1915. Hawaiian Volcano Observatory.] VNDEX TO VOLUME XO PAGE Acadian Triassic, The. By Sidney Powers : he als EOS; 254 Andrews, E.C. Notes on the Structural Relations pemmenn Australasia, New Guinea, and New Zealand Appalachian Geosyncline, The Upper Deemer Delta ae thes By Joseph Barrell. Review by V. O. T. : Archean Geology of Rainy Lake Re-studied, The. By Andres C. Lawson. Review by T. T. Q. : Arkose Deposits, The Geological Significance atl Genetic Clsstetion of. By Donald C. Barton : Aroha Subdivision, Hauraki, Auckland, The Grates of the. By I. Henderson, assisted by ir A. Bartrum. Review by T. T. Q. ; Australasia, New Guinea, and New Zealand, Notes on the Structural Relations between. By E. C. Andrews Average Regional Slope, A Criterion for the Gubdinscion a Old hadi Surfaces. By Leopold Reinecke . ; ‘ Barrell, Joseph. The Upper Devonian Delta of the Appalachian Geo- syncline. Review by V. O. T. Barton, Donald C. Notes on the Disinertation of Granite in Bee ———. The Geological Sos and Genetic Classification of Arkose Deposits Bastin, Edson S. Geology ai fe Pitchblende Ores of leoterader Review by V. O. T. B Bauer, L. A. The General Magnetic Sure of the Earth. Review by RoC ME. Becker, George F., saat ARETE L. Dey. Noten on the ee Borce ef Growing Crystals Berry, E. W. The Upper Gieiaceats and Bieeene Bloras! of South Carolina and Georgia. Review by R.C.M. . Beyer, S. W., and H. F. Wright. The Road and Concrete Matenals of Iowa. Review by W.B.W. . Bighorn Dolomite of Wyoming, Origin of el By, Eliot Blaceweldedt Review by R. C. M. : Birge,, Edward, and Chancey naa, Talend Takes of Wieconsia: Review by NWaTBaEW 1 Blackwelder, Eliot. Origin of the Eeenore ipolomite a ees Review by RACSVE: : Bliss, Eleanora F., and Anna I. Paine, Relation of the WwW seeanckon Mica-Gneiss to the Shenandoah Limestone and to the Octoraro Mica-Schist, of the Doe Run—Avondale District, Coatesville nat: rangle, Pennsylvania. Review by V. O. T. : ' 833 751 407 309 834 INDEX TO VOLUME XXIV Botanical Criterion of the Antiquity of the asec akaae A. ByE. W. Sinnott : Branner, John C. ('Gesioen Hlsmentar: Preparada con igience especial aos Estudiantes Brazileiros e 4 Geologia do Brazil. Review by R. T. C. hirnlaer th or eared ———. Orville A. Derby : Branson, E. B. The Lower Embar oF Wrong and Tes Eauee Brokaw, Albert D. A Stage Attachment for the eee: i Mi- croscope , and Leon 2 Sraith Zen Weather of a Harulende Gabbro Buddington, A. F. Byraphyiitizetion: Pinsticdt ion! ane Silicification of Rocks around Conception Bay, Newfoundland : Butters, R. M. Permian of ‘“Permo-Carboniferous” of the acter Foothills of the Rocky Mountains in Colorado. Review by V. O. T. Buttram, Frank. Glass Sands of Oklahoma. Review by W. B. W. Cady, Gilbert H. Coal Resources of District.No. I (Longwall). Re- view by A. D. B. : Caimanoidea Visheri, A New Grocadilan From the Olipecene of South Dakota. By Maurice G. Mehl Calcium Carbonate, The Role of Inorganic Rpenee in the Deposition of. By John Johnston and E. D. Williamson Calvert, W. R., A. L. Beekly, V. H. Barnett, and M. A. Pishel. Galeer of the Standing Rock and Cheyenne River Indian Reservation North and South Dakota. Review by V. O. T. : Cape Lisburne, Alaska, The Jurassic Flora of. By F. H. Kaowlont Review of R. C. M. : Capillarity, Some Effects of, on Oil Kectaulgtian: By Ales W. McCoy Cement Materials and Industry in the State of Washington. By Solon Shedd. Review by W.B.W. . Central Ross-shire, Geology of the. By B. NT: Baath, L. Ww. Eineeman E. M. Anderson, J. Horne, C. B. seen and R. G. Carruthers. Review by T. aT. Go Chabot, Georges. Le Revernant Ebiide sur une région dene Se Jura méridional. Review by R.C.M. oe : Clark, J. M. Notes on Geology of the Gulf of St. Barren Revie by WaBoWs. sh . et! 2 Pere eae Ph.D. Gecloay. Physical hel Eustaneat Review by See ae Geulogie anes! By Gharles Sehuchere Review by Coal Resources of Beier NG. I (enerevally:| By Gilbert H. Cady. Review by A. B. D. . one ae ei of District No. VII. By Bred 5B Kay: Review or Collins, W. H. The Eunonien Bontatinns a Thickening Region Canada. Review by T. T. Q. aE hee ee 726 INDEX TO VOLUME XXIV Corries, with Special Reference to Those of the Campsie Fells. B _ J. W. Gregory. Review by R. C. M. ; . i Crawford, R. D. Geology and Ore Deposits of the Moverch aiid Tomichi Districts, Colorado. Review by V.O.T.. . Cretaceous Deposits of the Eastern Gulf Region, and Species of Receen from the Eastern Gulf Region and the Carolinas. re L. W. Stephenson. Reviewby R.C.M. . . Cristivomer namaycush, Great Lake Trout, Discovery as OG, in fie Pleistocene of Wisconsin. By L. Hussakof Cushing, H. P., and R. Riedemann. athe of ee Spring aad Vicinity. Review by W.B.W. . Day, Arthur L., and George F. Becker. Note on the Linear Force of Growing Crystals Pe NERA One Ne Uestae Td eas nantes Derby, Orville A., By John C. Biatiner Disintegration of \Granite in Eeypt, Notes on the, By Donald C. Barton Dewoe s ‘ Wane ; ; : Eakin, Henry M. A Geologic Reconnaissance of a Part of the Rampart Quadrangle, Alaska. Review by V. O. T. . Earth, The General ee Survey of the. By L. A. eee Review by R. C. M. Eastern Gulf Region, Greceeis Drarests of ine atl Spent: i iexoeyea from the Eastern Gulf Region and the Carolinas. oY W. L. Stephenson. Reviewby R.C.M. .. Ellipsoidal Lavas in the Glacier National Park, Montane: By Tae caster D. Burling Embar, The Lower, of Wyoming and its Fat auna. By, E. B. aiason Florida, Origin of Hard Rock Phosphates of. ar E. H. Sellards. Review by W.B.W. ... Foerste, A. F., and W. C. Morse. The Woveda Fomintions of East Central Kentucky. Review by W. B. W. : Foliation in the Pre-Cambrian Rocks of owners New Yon Origin of. By William J. Miller Fossils from Limestone of Steeprock Tare Gntane Notes on. By Charles D. Walcott. Review by T. T. Om Foye, W. G. Are the ‘“‘Batholiths” of the Haliburton- Bancroft en Ontario, Correctly Named? . Fuller, Myron L. The eoneys of (oes Tela: New oi, Revi iew by WeOsels rey! : Garrett, R. E., and D. W. Ohern. Ponca City Oil and Gas Field. Review by W. B. W. 5 Geologia Elementar. Preparada con H tcreren eSepial a aos Estudi. antes Brazileiros e 4 Geologia do Brazil. Por John C. Branner. Review by Rade 'C. Bilete ‘ 836 INDEX TO VOLUME XXIV Geological History of New York State. By William a3 Miller. Review by W.B.W. . Geological Map of Tene Compiled by Olaf Pp. qeskane (A. H Purdue, State Geologist). Review by R. D. S.. Geologische Beobachtungen in Spitzbergen. Ergebnis der W. FE ne schnerschen Vorexpedition nach Spitzbergen 1910. By Professor H. Philipp. Review by T. T. Q. : Geology and Geography of a Portion of aeaiet Gains panies By Alfred Reginald Schultz. Review by V.O.T. . . Geology and Ore Deposits of the Monarch and Tomichi Diseaes Colorado. By R. D. Crawford. Review by V. O. T. : Geology of Central Ross-shire, The. By B. N. Peach, L. W. nie eae E. M. Anderson, J. Horne, CB: annie. and R. G. Carruthers. Review by T. sie Q. ; 2” aah EES cil eee eae Geology of East Central Oklnharia: By L. C. Snider. - Review by WBS Wee F 5, Wt TC cee Geology of Long Telia’ New ote The. By Myron L. Fuller. Review by V. O. T. : Une Geology of North Creek underage: By William l. Miller. Review by W.B. We 2. Geology of Saratoga Sas and aay, By H. P. Gicking a R. Ruedemann. Review by W.B.W. . Geology of Southeastern Ontario, The Pre- Cannan By Willet G. Miller and Cyril W. Knight. Review by T. T. Q. : Geology of Steeprock Lake, Ontario, The. By Abdrew C: Lawaeh Review by T. T. Q. Geology of the Aroha Sabdinedens aarakt Pmcelantl The. Ey J. Henderson, assisted by J. A. Bartrum. Review by EO Geology of the Gold Belt in the James River Basin, Virginia. By Stephen Taber. Review by T. T.Q. . Geology of the Gulf of St. Lawrence, Ndtes on. ee qf: M. Clark. Review by W.B.W. : Geology of the Pitchblende Ores at Golowda : By Baan S. acces Review by V. O. T. , Geology of the Rabbit Ears Region, Routt, Grand and Tackenn Canes Colorado, Reconnaissance of the. By F. F. Grout, P. G. Worcester, and Junius Henderson. Review by V. O. T. sia. etree Geology of the Rolla Quadrangle, The. By Wallace Lee. Review by Wo BOW. oc Geology of the Standing Rock and ‘Ghoseane River (adh Reservatianes North and South Dakota. By W. R. Calvert, A. L. ae V. E Barnett, and M. A. Pishel. Review by V.O, T. . Geology, Physical and Historical. By H. F. Cleland, Ph.D. Review by-AvG. i Geomorphologie na Quast ive lone aes Sane ehirses Die By Axel Hamberg. ReviewbyR.C.M. ... Glaciology of the South Orkneys: Scottish National Anearette oes dition. By J. H. Harvey Pirie. Review by T. T. Q. : Glass Sands of Oklahoma. By Frank Buttram. Review by W. B. W. INDEX TO VOLUME XXIV Gletscher des Sarekgebirges und ihre Untersuchung, Die. By Axel Hamberg. Review by R. C. M. Gletschereises, Uber die Barallelewrakesa: aes, By Axel Hamberg Review by R. C. M. : 5 Gold Belt in the James River asi “Witeeiey Geclony eo ite! By Stephen Taber. Review by T.T.Q.. . ; Grain of Igneous Rocks, The. By A. C. Lane. Revien by R. C. M. Grandfield District, Oklahoma, Reconnaissance of the. By Malcolm J. Munn. Review by V. O. T Bis er a are Se a ic Granger, Walter, and W. J. Ginelain Paleocene Deposits of the San Juan Basin, ‘New Mexico. Review by V. O. T. ola ge 2 ae One ears Granite in Egypt, Notes on the Disintegration of. By Donald C. Barton Great Lake Trout, CaS vom: meray cushy ieeovery oh ie, in the Pleistocene of Wisconsin. By L. Hussakof Gregory, J. W. Corries, with Special Reference to Those of the Campsi Fells. Review by R. C. M. : Grout, F. F., P. G. Worcester, and Junius Henderson: Recon anes of the Geology of the Rabbit Ears Region, Routt, Grand and Jack- son Counties, Colorado. Review by V. O. T. Growing Crystals, Note on the Linear Force of. By George F. Bestes and Arthur L. Day. . Gulf of St. Lawrence, Notes on Geology of he. By J: M. Clark. Review by W.B.W. . Gypsum Deposit in Iowa, A New. ‘2B y G. F. Kay: Bevery by 1 C. M. Harker, Alfred. Differentiation in Intercrustal Magma Basins Haliburton-Bancroft Area, Ontario, Are the “Batholiths” of the, Cor- rectly Named? By W. G. Foye Hamberg, Axel. Die Geomorphologie und Quartargeologie des carck gebirges. Review by R. CINE =. Die Gletscher des See eepirees wind ifr Untersuchung Review by R. C. M. : Kurze Ubersicht der Gletscher Schwedens. Reon by R. C. M. ! sn pine ute eee?) Uber. die Parallelstruktur fles eibeererecee! Review by RCS Mis. Fhe 4 Hawkins, A. C. Toekatone Roanaton ae the Triassic of New Jey and Pennsylvania. Review by RAC eM: Haynes, Winthrop P. The Lombard Overthrust nnd ‘Related ices logical Features f Henderson, J., assisted by I: A. Bartram: "The Goleen of the Aedha Subdivision, Hauraki, Auckland. Review by T. T. Q. Henderson, Junius, F. F. Grout, and P. G. Worcester. Reconnaissance of the Geology of the Rabbit Ears Region, Routt, Grand and rire son Counties, Colorado. Review by V. O. T. Heinrich, M. On the Structure and Classification of the Stromato: poroidea UNE a: CAN pte 5X ; : Lita Peg 838 . INDEX TO VOLUME XXIV Hobbs, William Herbert. Assumptions Involved in the Doctrine of Isostatic Compensation, with a Note on Hecker’s Determination of Gravity at Sea Hornblende Gabbro, Zonal Weaheuns of a. RB, Albert D. Broan and Leon:P. Smith >": Huronian Formations of Pesiskarning Reion: Canada: The. By W. H. Collins. Review by T. T. Q. : : Hussakof, L. Discovery of the Great Lake Trout, Gestomer namay- cush, in the Pleistocene of Wisconsin . Hydrothermal Alteration, Studies in. I. By E. A. Sienheeean Igneous Rock, the Composition of the Average. By Adolph Knopf. Inland Lakes of Wisconsin. By Edward Huge and Chancey Jueey Review by W.B.W. Intercrustal Magma Basins, Paiecennntionlt in. Ey. Alfred Haricer Isostatic Compensation, Assumptions Involved in the Doctrine of, with a Note on Hecker’s Determination of Gravity at Sea. By William Herbert Hobbs Lee 1) ae Cee Re James River Basin, Virginia, eat of the Gold Belt in the. on Stephen Taber. Reviewby T.T.Q.. . Johnson, Douglas W. Studies for Students. Cautabicions to ene Study of Ripple Marks . Johnston, John, and E. W. Witliamncene The Réle aE Tngreanes Agen- cies in the Deposition of Calcium Carbonate Johnston, W. A. The Genesis of Lake Agassiz: A Gonsimacn Jonas, Anna I., and Eleanora F. Bliss. Relation of the Wissahickon Mica-Gneiss to the Shenandoah Limestone and to the Octoraro Mica-Schist, of the Doe Run—Avondale District, Coatesville Quad- rangle, Pennsylvania. Review by V. O. T. . Juday, Chancey, and Edward pee Inland Lakes ee Wicconen Review by W.B.W. Jurassic Flora of Cape epee Aina The. By F. H. Knowles! Review by R. C. M. : pt 8 ak ae Kanawha County, West Virginia. By Charles E. Krebs. Review by We Ve a ge Karakoram Glaciers Gneneco ith Breese, Penecrally of ‘Aiiuents Features of. By William H. Workman. "Review by R. C. M. Kay, Fred H. Coal Resources of District No. VII. Review Py AD Bs ke Kay, G. F. A New ‘Gy patie eeostn in nia Review ie R. Cc. M. Kemp, James F. The Mineral Springs of Saratoga. Review by Wo BoW.-8 Knight, Cyril W., and Willet G. Miller. The eres Cambern Geology of Southeastern Ontario. Review by ES TO: Knopf, Adolph. The Composition of the Average igneous Rocks! Knowlton, F. H. The gee Flora of se uh Tt Alaska. Review by RCoM: 5 : PAGE 690 200 726 685 180 620. 826 534 690 518 809 729 625 309 826 208 300 98 829 724 828 412 620 208 INDEX TO VOLUME XXIV fa BW ae Kanawha County, West Virginia. Review by Kurze Ubersicht der icles: Salinpadlans, By Axel Members Review by R. C. M. i Hv Aiur cn eka sen Lahee, Frederic H. Origin of the Lyman Schists of New Hampshire Lake Agassiz, The Geneis of: A Confirmation. By W. A. Johnston Lane, A.C. Mine Waters. Review by R. C. M ; Temperature of the Copper Mines. Review by R. Cc. M. ———. The Grain of Igneous Rocks. Review by R. C. M. Lawson, Andrew C. The Archean Geology of Rainy Lake Re-scudied! Review ony BS 1054 0)3 : ———. The Geology of Stecprock Bakes One! Review by T. T. Q. ow ee The Geology of the Rolla Quadrangle. Review By Leonard, A.G. The ree. Wiecoucin Drift oi Noreh Dakota! : Le Revermont, étude sur une région karstique du Jura méridional. By Georges Cabot. Review by R. C. M. Lincoln County, Wyoming, Geology and Geography of a Portion oh By Alfred Reginald Schultz. Review by V. O. T. ; Linear Force of Growing Crystals, Note on the. By, Genre F. Becker and Arthur L. Day .. Lockatong Formation of the Triassic of New Taser andl Penney are By A. C. Hawkins. Review by R. C. M. Lombard Overthrust and Related Geological Features, The. By Winthrop P. Haynes St ee ee Long Island, New York, The Geslory, of. By NIG ron itp Fuller. Re- view by VeO ne: Lyman Schists of New Hampshire Origin Bn the. By Frederic H. Lahee. bi noe ane elas McCoy, Alex. W. Some Effects of Capillarity on Oil Accumulation Magnetic Survey of the Earth, The General. By L.A. Bauer. Review yen CVs ate meh eax emus iey SPE, Marshall, P. Oceania. Reweu ye R. Cc. M. : Mehl, Maurice G. Caimanoidea Visheri, A New Grocoailinn fo the Oligocene of South Dakota Metallographic Microscope, A Stage Avartnaatt for the. By Albert D. Brokaw : : ; : Mexico, The EN soeta phy, a iB Wermen N. Tinaer E Miller, Willet G., and Cyril W. Knight. The Pre-Cambrian Gecloey of Southeastern Ontario. Reviewby T.T.Q.. . Miller, William J. Geological History of New York State. Review by W. B. W. Sh Calan GE North Creek @uacrancls Revere by W. B. W. —— —, Origin of Foliation in the Pre-Cambrian Rocks of Northern New York SESS POET ee BT ac hs aaa Mine Waters. By A.C. Lane. Review by R. C. M. 840 INDEX TO VOLUME XXIV Mineral Springs of Saratoga, The. By James F. Kemp. Review by PAGE W. B. W. oes Mississippian Eeachenpadn of a Mississippi Valley Base The. By Stuart Weller. Review by J. W.B. . 95 Molengraaff, G. A. F. On Oceanic Deep: oy Deposits a Cantedl Borneo. Review by R.C.M. . 405 Monarch and Tomichi Districts, Geology ma Gre inenostes of the: By R. D. Crawford. Review by V. ©. T; : 408 Monocyclic Crinoidea Camerata, Evolution of the Basal Plates in. By Herrick E. Wilson . f ; . 488, 533, 665 Morse, W. C., and A. F. Rocrier The Waverlian F ormations of East Central Kentucky. Review by W: B. W. 824 Mud Lumps at the Mouths of the Mascccinae The. Be Raves W. Shaw. Review by R. C. M. : 623 Munn, Malcolm J. Reconnaissance of the Grandeeld Dicmiee Okla- homa. Review by V. O. T. TET es New Gypsum Deposit in Iowa,A. ByG.F.Kay. ReviewbyR.C.M. 724 Niagaran Formations of Western Ohio, The Classification of the. By Charles S. Prosser ; 2, ae eee Nochmals zur Frage der Glagaibidenpent in re Rhén. Erwiderung auf die Ausfiihrung von A. Penck und Ed Briickner. By Hans Phillip. Review by R. C. M. ae eer Meee me) fi ee TS North en seaneee ee of. By William J. Miller. Review by W. B. W. sy _* : Nan .1cAfiReny. *-y. AAA? Ve pe a Aa : ee pita, A FY VAY Y Wraneta® ation EAI le nf pninak ogy va “ & 2 A at 36 aR a= = a "a f Wien Whale ae Paral w ba Tien 2 me 2 f Seece! fh Rey ohne iW 2 HEE HTT aeeee ROLY Civic sWPanknul gn ae ator t8 - q Ps NN Anne A : = : i 9) Le Nea. olotal 1 "s Seerier aa ony ie am agp Oa ahha AA Rae . 4 ve ms rte F HVT -= cake, te Tale Rees bas oT atl ean A aa q af veNINa \, Na \ hers SAA aA BS pA ; ite AM, svonennnne ltl atitonatiantees CULUSS rag “TENT | PTL TT AD ag Th = iguana age > ver Sh eh did tpt d te a th Am - ‘Toms ile Bi ahah naan « pv ‘i apapake: abP BAAR? Nf XB 4: ALDER ER | pana Aaiy a .28, 068! j 71? ; Ways: /Il ubedg ee n TTT Yee NHS My, yuculdtie 3 9088 01367 0138 a £3 fy A. id a2 fl a i PALL | WARS s 4 am) WE n be el iy Re af | gaasas. te ~~ ey 1D Da» = ak a” The Mt Th eet. wh, a Van Mees: *, 4 a be # a Neue Swit f tigi rT mA asses... @ AMIR Chee tg! BARE Ee an oe “a9 Pas ; San" e= T11yreear s aoe Ua Tt {tery me “sa. pr’ le ig viata in anna a \> 1 ern f AaAneas ’ rire ALY ay) tee nds a & z a h 5. areas 74g ba tos © 4 4 ARON € Pa “ae == ae, Lary een Kans ageaaGhe, ADa PP eapcentt larry ye mews | Tre TTT ULF pa RA AS TCTNCSO aces a ODS SOT pee aes Me WaT TEN Fis hoi ae rePLTT CG © 4 ry ee ) i ian wi “pygiP” “ er Gok oy. OSM ' * ane eT iat Lywunyi aS = i] nt yaa pid ON Mea heir we ee _-=S ihe ao OS ar ~ —) 57 ee eos yas ~ awe ner o- e> 3 eT Bi Pi, . j yt v ana BUTT a o,”' Pen at 3 corbuun’ a," m6