titel Aiadiaad te” Poth h kad fe CaP FF Ai i tie an Ia eee Sp -tie {h/ Sekin eee a sett 2 Sdn itd PP POE ead te < Ss o 4 a ee it te wi | Ge Bran! Ai ay oO if ~ iS ‘it nl (nll Soe oe =, 5, rh “so ER 2 See Sis es th 7 rs aN = \ * 3 Nei! seks | g& a Page SS a y Se “No, a | 7%, Xt oe %, 3 iE i " ie an ii “< oF a a 2} BS a Ade “Lay. 9 hy fay ener = eed a ee BIOS oe AY pt a0 Si het | be ne ee Sie fee SOR 2 & k eT ee ie ne “ » . Ve rae es os BN ee af b) 7, oe) & x ty a ies eR, MF Yetiey FIG, es F 4 Le a Be, 4% noe Pi das it ene Sey inte oe eS is +7 Fe (pra, Soe N a “4 *% a on Oo (es Tl it | y y eS YY oO aS ae fox | | | } > nw | i | oY CY Ke Vo, i | | | Si aN a @ | gS ran ery wiles i an|| VL A er ci ay ) Legis Ou my hate’ vil ay rT eda ah x i FOURNAL OF GEOLOGY. Pe HOUR NAL OF GEOLOGY A Semi-Quarterly Magazine of Geology and Related Sciences EDITORS T. C. CHAMBERLIN, zz General Charge R. D. SALISBURY R. A. F. PENROSE, Jr. Geographic Geology Economic Geology J. P. IDDINGS C. R. VAN HISE Petrology Pre-Cambrian Geology STUART WELLER W. H. HOLMES Paleontologic Geology Anthropic Geology S. W. WILLISTON, Vertebrate Paleontology ASSOCIATE EDITORS SIR ARCHIBALD GEIKIE G. K. GILBERT Great Britain Washington, D. C. H, ROSENBUSCH H. S. WILLIAMS Germany Vale University CHARLES BARROIS Cc. D. WALCOTT france U.S. Geological Survey ALBRECHT PENCK J. C. BRANNER Austria Stanford University HANS REUSCH IC RUSSELE Norway University of Michigan GERARD DE GEER W. B. CLARK Sweden Johns Hopkins University O. A. DERBY Brazil /Zeesenian Insiyg VOLUME XII (3 50422) y; F Nr National es an 2 al CHICAGO Che Aniversity of Chicago Press 1904 PRINTED AT The University of Chicago Press CHICAGO CONTENES OF VOLUME JOT NUMBER I PAGE GEOLOGY OF Monapnock Mountain, NEw HampsHire. Joseph H. Perry - - - - - - - - - - - - I ON A CALIFORNIA ROOFING SLATE OF IGNEOUS OricIN. Edwin C. Eckel 15 On THE CHEMICAL COMPOSITION OF AMERICAN SHALES AND ROOFING States. Edwin C. Eckel - - - - - - - - 25 ARAPAHOE GLACIER IN 1903. Junius Henderson - - - - 30 THE APPALACHIAN RIVER versus A TERTIARY TRANS-APPALACHIAN RIVER IN EASTERN TENNESSEE. Charles H. White - - - - - 34 THE CONTACT OF THE ARCHAN AND Post-ARCHZAN IN THE REGION OF THE GREAT Lakes. A. B. Willmott - - - - - - 4o THE RELATIONSHIPS AND HABITS OF THE Mosasaurs. S. W. Williston 43 Reviews: Summaries of Pre-Cambrian Literature for 1902-1903, I (C. K. Leith), 52; Irrigation, F. H. Newell (L. H. Wood), 62; Gems and Gem Minerals, Oliver Cummings Farrington (H. L.), 63. EpiroriaAt. H.F.B. - - - - - - - - - - 65 RECENT PUBLICATIONS = ~ 2 - 5 = - - - 67 NUMBER II ARTESIAN WELL SEcTIoNS AT ITHaca, N. Y. R.S. Tarr - - - 69 Tue ROLE or PosstBLeE Eurecrics In Rock Macmas. Alfred C. Lane 83 A FRACTURE VALLEY SYSTEM. Joseph P. Iddings - - : - 94 CuSPATE FORELANDS ALONG THE BAY OF QuiInTE. Alfred W. G. Wilson 106 A CONTRIBUTION TO THE STUDY OF THE INTERGLACIAL GORGE PROBLEM. George C. Matson - - = 2 = = - - - ae ate} AN INTERGLACIAL VALLEY IN ILttNois. George D. Hubbard - - 152 Reviews: Summaries of Pre-Cambrian Literature for 1902-1903, II (C. K. Leith), 161; West Virginia Geological Survey, I. C. White (R. D. S.), 176; Geographic Influences in American History, Albert Perry Brigham; American History and its Geographic Conditions, Ellen Churchill Semple (H. H. B.), 176. RECENT PUBLICATIONS - - - - - - - - - 180 vl CONTENTS OF VOLUME XII NUMBER III Icr-RETREAT IN GLACIAL LAKE NEPONSET AND IN SOUTHEASTERN Mas- sacHuSETTS. M.L. Fuller - - - - - - - - RELATIONS OF GRAVEL Deposits IN THE NORTHERN PART OF GLACIAL Lake CHARLES, MassaAcHuseETts. Frederick G. Clapp Sgetes Tue LEOPARDITE (QUARTZ PoRPHYRY) OF NoRTH CAROLINA. ‘Thomas L. Watson - - - - - - - - - - - QuARTZ-FELDSPAR-PORPHYRY (GRANIPHYRO LIPAROSE-ALASKOSE) FROM Liano, TEx. Joseph P. Iddings — - - - - - - - CRYSTOSPHENES, OR BURIED SHEETS OF ICE, IN THE TUNDRA OF NORTH- ERN AMERICA. aJ¢ Bovbkymrelle = sget= =e =i eee eee eee A Coat-MEASURE ForeESsT NEAR Socorro, NEw Mexico. C. L. Herrick THE VARIATIONS OF GLACIERS. IX. Harry Fielding Reid - - - FAREWELL LECTURE BY PROFESSOR EDWARD SUESS ON RESIGNING HIS ProressorsHip. Translated by Charles Schuchert - - - HpIrortans, PCs. - - - - - - = - = Reviews: Grundziige der Geologie des unteren Amazonasgebietes (des Staates Para Brasilien), Dr. Friedrich Katzer (J. C. Branner), 278; The Correlation of Geological Fauna: A Contribution to Devonian Paleontology. Henry Shaler Williams (Stuart Weller), 279; The Evolution of Earth Structure, T. Mellard Reade (W. H. E.), 280. NUMBER IV A SERIES OF GENTLE FOLDS ON THE BORDER OF THE APPALACHIAN System. Edward M. Kindle - - - - - - - THe LARAMIE AND Fort UNION BEDs IN NortH Dakota. Frank A. Wilder - - . - - - - - - - - - ORBICULAR GABBRO-DIORITE FROM DAVIE COUNTY. NORTH CAROLINA. Thomas L. Watson - - - - - - - - = THE OSTEOLOGY OF THE SKULL OF THE PELYCOSAURIAN GENUS, DIMET- RODON. E. C. Case - - - - - - - - - ON THE STRUCTURE OF THE FoRE Foot oF DimEetTropon. E. C. Case ON THE PYROXENITES OF THE GRENVILLE SERIES IN OTTAWA COUNTY, Canapa. C. H. Gordon - - - - - - = - A TypicaAL CASE OF STREAM-CAPTURE IN MICHIGAN. Isaiah Bowman THE DEPOSITION OF THE CARBONIFEROUS FORMATIONS OF THE NORTH SLOPE OF THE OZARK Uptiirr. Sydney H. Ball - - - - ECLOGITES IN CALIFORNIA. Ruliff S. Holway — - - - - - EDITORIAL - : - - - = - = : = = 2 Reviews: Catalogue of the Ward-Coonley Collection of Meteorites, Henry A. Ward, 360. PAGE 181 198 215 225 232 237 252 264 276 281 290 204 304 312 316 326 335 344 359 CONTENTS OF VOLUME XII NUMBER V DESCRIPTION AND CORRELATION OF THE ROMNEY FoRMATION OF Mary- LAND. Charles S. Prosser - Se. - - - - - GRANITES OF NorTH CAROLINA. Thomas L. Watson - - - - GEOLOGICAL NOTES ON THE VICINITY OF BANFF, ALBERTA. I. H. Ogilvie - GLACIAL AND Post-GraciAL History oF THE HupsSON AND CHAMPLAIN VALLEYS. Charles Emerson Peet - - - - - - - Reviews: The Non-Metallic Minerals, G. P. Merrill (G. F. K.). 470; Geology of Miller County, Sydney H. Ball and A. F. Smith (H. L.), 470 NUMBER VI PHYSIOGRAPHIC STUDIES IN SOUTHERN PENNSYLVANIA. George W. Stose UBER DIE GEGENSEITIGEN BEZIEHUNGEN ZWISCHEN DER PETROGRAPHIE UND ANGRENZENDEN WISSENSCHAFTEN. Ferdinand Zirkel - - AN OCCURRENCE OF GREENSTONE SCHISTS IN THE SAN JUAN MounraIns, CoLorapo. Ernest Howe - - - - - - - - AN OCCURRENCE OF TRACHYTE ON THE ISLAND OF Hawa. Whitman Cross - - - - - - - - ei eae - - PHYSIOGRAPHIC PROBLEMS OF TopAy. Israel C. Russell — - - - WIDESPREAD OCCURRENCE OF FAYALITE IN CERTAIN IGNEOUS ROCKS OF CENTRAL WISCONSIN. Samuel Weidman - - - - : Reviews: The Clays and Clay Industry of New Jersey, Heinrich Ries and Henry B. Kiimmel, assisted by George N. Knapp (J. H. L.), 562; Oil and Gas. Levels, I. C. White (R. D. S.), 563; Baraboo Iron- Bearing District, Samuel Weidman (R. D. S.), 564. RECENT PUBLICATIONS - - - - - - - - - NUMBER VII THE PROFILE OF MATURITY IN ALPINE GLACIAL Erosion. Willard D. Johnson - - - - - - - - - - - SYSTEMATIC ASYMMETRY OF CREST LINES IN THE HIGH SIERRA OF CALI- FORNIA. G. K. Gilbert - - - - - - - - - THE PRoBLEMS OF GEOLOGY. Charles R. Van Hise - - - - GLACIAL AND Post-GLAcIAL HistoRY OF THE HUDSON AND CHAMPLAIN Vatieys, II. Charles Emerson Peet - - - - - Reviews: The Economic Resources of the Northern Black Hills, J. D. Irving (W. D.S.), 661; Zinc and Lead Deposits of Northern Arkansas, George I. Adams, Teel by A. H. Purdue and E. F. Burchard (W. ID Sy) LOK RECENT PUBLICATIONS - - - - - - - - - 473 485 501 510 524 551 566 569 579 589 617 665 Vili CONTENTS OF VOLUME XII NUMBER VIII THE RELATIONS OF THE EARTH SCIENCES IN VIEW OF THEIR PAOLA IN THE NINETEENTH CENTURY. W.M. Davis - - - - Notice or SomME NEW REPTILES FROM THE UPPER TRIAS OF WYOMING. S. W. Williston - - - - - = = 2 =e wn PLEISTOCENE GEOLOGY OF THE SAWATCH RANGE, NEAR LEADVILLE, COLO. S. R. Capps and E. D. K. Leffingwell _— - - - - - - THREE NEW PuystoGRAPHIC TERMS. Rollin D. Salisbury : - Own CERTAIN ASPECTS OF THE LOESS OF SOUTHWESTERN Iowa. O. W. Willcox - - - - - - - - - - - - Tuer EFrFrect OF SUPERGLACIAL DEBRIS ON THE ADVANCE AND RETREAT OF SOME CANADIAN GraciErRs. I. H. Ogilvie - - - - Reviews: A Treatise on Metamorphism, Charles Richard Van Hise (K.), 744; The Stone Reefs of Brazil, Their Geological and Geographical Relations, with a Chapter on the Coral Reefs, John Caspar Branner (T. C. C.), 748; The Copper Deposits of the Encampment District of Wyoming, Arthur C. Spencer (W. D.S.), 752; Recent Seismological Investigations in Japan, Dairoku Kikuchi (T. C. C.), 754; Earth- quakes in the Light of the New Seismology, Clarence Edward Dutton (GCG) e756: PAGE 669 688 698 7°7 716 722 Sitch ve \ 5 DG a ~ 5. es ‘. iy ; 4 x i 7h : iy 4 i d HOURNAL OF fae NV UARY- -FRRRU a) \ GEOLOGY Likely DESCRIPTION oF “Bu Rocks or Monapn sea. Andalusite schist Y.og, Fibrolite schist. (Se Quartzosd ca. Rusty, §. Granite. @) Oc regce I CijAGE oF THesk Roo SS. ATTITUDE OF THR SCHINT BE. JOINTING OF kHE AcuisyT! B. iwoe. Perlops o¢/ Mefamuos RMA SOCEOR TOE SURVA AL =) SN Fe SR a pe ey te oo) MONADNOSE “Situated | it, special interstlto geologists Ne Davis 45a tyoNof the isolated icy nising above ia base-leveled la Ais: ae surrounding pénepiain, the prateee ol tion on the Zopograpiie }iap, on wy i ak int ‘From the. Jopegraphice Ma af toy ik Bt: soe SVol X21, No. 1, ‘ 35 166? feet above mean be 2-ley elie} aye At the first Glance the mounte:a Bes apy 5 ry ‘wa Monadnock is. considerett in. Ge: OF aii te Oe a te §03, 639... This area, with a-section thgng yr cain’ . sheet of the geological map ingie athh> af od Tek. a eee CARBONIFER OWS. [F2) Quarigose, mica schist: Andalusite, fibrolite, mica schist MM Rusty graphitic mica schist P0S5T-CABBONIFEROUS, ES Gea NEESER aes Meee, Wie i mits CHIN AS EY See ae Ni ” on (a= D ye IN NE rey A ‘ ——" BPs) y . ES i) itt =ar7 a ae men we ar > = aN Se t Nis: re A RES SS: L\\ years peek a5 oe ait il mK Ag ier a ae —— =3 P és ame al} Nt i we -jAZ TS si =| VEY a=) — \VHt as WS i Soa | Saw rim eel ee THE JOURNAL OF GEOLOGY VAN OAT 1 DikepAT 10 OF: GEOLOGY OF MONADNOCK MOUNTAIN, NEW REA WIESE TRy.+ CONTENTS. A. DESCRIPTION OF MONADNOCK. B. Rocks oF MONADNOCK AREA. 1. Andalusite schist. 2. Fibrolite schist. 3. Quartzose mica schist. 4. Rusty, graphitic mica schist, 5. Granite. a) Occurrence in lobes. 6) Inclusions in. c) Alternation of, with schist along the border. @) Pegmatite in. é) Relation of, to schist. C. AGE OF THESE ROCKS. D. ATTITUDE OF THE SCHIST IN THE MOUNTAIN. E. JOINTING OF THE SCHIST. F, Two PERIODS OF METAMORPHISM INDICATED IN THE SCHIST, G. REASON FOR THE SURVIVAL OF THIS MOUNTAIN. MONADNOCK, situated in southern New Hampshire, is of special interest to geologists because it was selected by Professor Davis as a type of the isolated, residuary peaks that may be found rising above a base-leveled land surface. This mountain rises 3,166? feet above mean sea-level and about 2,000 feet above the surrounding peneplain, the plateau of southern New England. At the first glance the mountain may appear, from its representa- tion on the Zopographic Map, as a regular, single peak; but, on *Monadnock is considered in the Geology of New Hampshire, Vol. II, pp. 24, 503, 639. This area, with a section through the mountain, is represented on the fourth * sheet of the geological map in the atlas accompanying the same. ?From the Zopographic Map of the U.S. Geological Survey. Vol. XII, No. 1. I LC aa ara 2 POSTE LOA closer study of the map and of the mountain from different sides, it is seen to consist of two well-defined parts. There is a northeast-southwest ridge, about six miles in length, extending from the center of Dublin to Gap Mountain. This constitutes the eastern part of Monadnock. This ridge rises by a succession of steps from an elevation of 1,500 feet at its northern extremity to 2,800 or 2,900 at its culmination east of the summit of the mountain, and then descends by a like succession of steps to its southern extremity. The eastern slope of this ridge above the foothills is quite steep, even precipitous. The western part of the mountain, which includes the summit, rising 300—400 feet above the eastern ridge, consists of a single peak set, as it were, in the central part of the western slope of the ridge. The northwest slope of this peak is gentle; the west and southwest slopes are much steeper; while the northeast slope meets the northern half of the western slope of the ridge forming the valley of Mountain Brook. These divisions and their slopes are closely related to the underlying rock structure, and indicate that erosion is con- trolled or guided by this structure. The rock of this mountain is a banded mica schist, the band- ing being generally parallel to the present structure. The schist presents three marked variations. In the top of the mountain and in the upper part of the eastern ridge it is a gray, massive, garnetiferous, biotite, sericite schist, in which the biotite is specially noticeable because of its arrangement in bright, isolated scales, one-sixteenth of an inch in diameter, set ina fine, light gray groundmass.’ In addition, andalusite crystals, or what were once andalusite crystals, occur in this schist, some- times very abundantly, lying parallel to the present structure of the rock. The accompanying picture (Fig. 2) shows how abundantly these may occur in the schist. They are frequently five or six inches long by half an inch, or more, through. Owing to the unequal weathering, these prisms frequently appear in relief on the surface of the ledges. In the southern half of the tIn the Geology of New Hampshire this schist is called the Kearsarge andalusite schist. GEOLOGY OF MONADNOCK MOUNTAIN 3 ridge, and also in the western peak of the mountain, these prisms are now largely, if not entirely, made up of fine, glassy, colorless or white fibers of fibrolite. In the northern part of the eastern ridge, and in other parts of the mountain in limited areas, the andalusite crystals have changed to masses of white, pearly sericite scales. In the weathering these sericite masses ee Ne Ss Fic. 2.—Surface of Andalusite mica schist. are removed more rapidly than is the inclosing rock, producing long narrow cavities on weathered surfaces. Where the end of the sericite mass is exposed, the weathering is more rapid in the center, producing a cavity bordered by a sericite shell. In the southern part of the mountain the schist becomes, by alternating areas, a fibrolite schist, the fibrolite being original ; but even within this fibrolitic area appear small areas where the rock originally contained andalusites. It is impossible to draw a line correctly separating this schist into two parts, so 4 JOSEPH Hl. PERRY intimately associated are the original fibrolite and pseudomorphs after andalusite. As the schist becomes fibrolitic, it is distinctly and quite thinly laminated. It also contains garnets and tourma- line and, rarely, graphite in fine scales. In the middle of the northeast slope this schist is destitute of fibrolite and pseudomorphs of andalusite; it is characterized by a dark green mica of soapy feel, and a very fine, white sericite. The latter is in fibers as fine as the fibers of fibrolite, and occurs in the schist just as fibrolite occurs in other parts of this schist. The sericite is evidently fibrolite changed into sericite, as is proved by the finding of a mass of fibrolite partly so changed. In the general sericitization to which this schist has been subjected the fibrolite, in places, has been changed as well as the other miner- als. This sericitization indicates the permeation of this schist by potash solutions. These variants are considered as one schist, and are so colored on the geological map. The second phase of the mica schist, found in the mountain and the surrounding area, may be seen between the 1,600 and 1,700-foot levels on either side of the road on the southern slope. It is a gray, thinly laminated, finely granular, quartzose mica schist, containing, in addition to the granular, glassy quartz, a little fine, brown mica and fine, light green hornblende. This schist is cut by lamination planes and joints into thin, rectangular slabs. In position it is conformable with the fibrolite-andalusite schist above. The boundary between the two is a zone of alter- nation. This indicates either an alternation of sediments or an interfolding along the border. I judge that the former is the case here, because in other parts of this area the quartzose schist blends into, and alternates with, the third phase of the mica schist. In only this small area on the southern slope is there enough of this quartzose schist by itself to be represented as a distinct area on the map. The third phase of the mica schist occurs below the second, on the southern slope, and is the first rock met in going up the mountain road. This is a very rusty, thinly laminated, fre- quently fissile; muscovite, biotite schist which often becomes so quartzose as to be a micaceous quartzite. The extreme rustiness GEOLOGY OF MONADNOCK MOUNTAIN 5 is due to iron pyrites. Little scales of graphite are characteristic of this phase of schist, and are sometimes abundant enough to give a gray color to the unrusted rock. Fibrolite was not observed with the graphite, though in the more micaceous, or first phase of the schist, fibrolite and graphite may sometimes be seen together. This rusty phase occurs over quite an area in the southeastern part of the region represented in our map, and also in the north- western part, and on the southeastern slope of Gap Mountain. In all of these areas there is no well-defined border between the first or fibrolitic phase and the third or rusty, graphitic phase; there is a blending of one into the other, and they are equiva- lents. The second phase also is only a variant of the third. Included in the area of the accompanying geological map, though not a part of the mountain, are granite masses which are closely connected with the rock structure of the mountain, and with other phenomena revealed in this study. Where this gran- ite adjoins rusty schist, it has a dark gray color, is more or less rusty on weathered surfaces, and is of medium fine, granular tex- LUKE: The quartz and the feldspar form an intimate granular mix- ture, in which the biotite is quite uniformly distributed in fine scales. Muscovite occurs in varying quantity, but is not charac- teristic of the granite, as biotite is. Fine magnetite and little, brown, wedge-shaped crystals of titanite occur in this granite along with some small particles of secondary epidote. Along the immediate border the granite sometimes contains black tourma- line. Tourmaline is, however, more frequently seen in the schist. Away from the schist the granite is lighter in color, more mus- covitic and less biotitic, contains less of the other minerals— is, in fact, more nearly a simple, medium fine, crystalline mixture of feldspar, quartz and mica. In mapping the granite this vari- ation is a good index of the nearness or remoteness of the schist border. In places the granite, in the southern part of this area, is por- phyritic, the feldspar phenocrysts sometimes measuring an inch by one quarter, and showing the Carlsbad twins. The feldspar of the groundmass is partly triclinic. The granite of the north- 6 YOST PITAL Tee RR eastern part of the area of our map is prevailingly porphyritic, and closely resembles the granite already described, except that the phenocrysts are frequently larger and sometimes show a granulated border. The granite of these areas is frequently foliated near the schist border, and parallelly to the lamination of the schist. On looking at the map, it is seen that the granite in the south- ern part of this area occurs in lobes, two of which are connected, while that in the northeastern part is in the form of a long tongue extending far into the schist, though not visibly reaching the southern granite area. These are probably parts of an extensive batholite which, possibly, extends even under the mountain. This granite incloses fragments, both large and small, of the neighboring schists. Among these may be recognized some of the light gray, quartzose mica schist thoroughly brecciated, and some of the fibrolite and andalusite schists. In the last the andalusite crystals have been generally, if not always, changed as has been described before. Inclosed in the granite may be seen prismatic masses of sericite entirely separate from, though in the vicinity of, the schist, which probably represent andalu- site crystals which were dissolved in the magma, and afterwards crystallized out and sericitized. It is difficult to decide, in some parts of this area, where to draw the boundary between granite and schist, because there is frequently a zone of alternating bands extending in the direction of the strike. Such an area is represented in the extreme western part of the geological map, and also in cross-section. The meaning of such an area is that the surface of contact between schist and granite batholite was a ragged surface— the granite having penetrated the schist at intervals, and pushed apart the vertical lamine. The erosion has brought the land surface down so as to make a section through this alternation. If the land surface had been lowered somewhat less, the rock at the surface would have been schist; whereas if the land surface had been lowered somewhat more, the rock would have been then all granite. As it is, the extension of the batholite is but a short distance below the surface. GEOLOGY OF MONADNOCK MOUNTAIN 7 In this granite are many pegmatite veins, varying from an inch to several feet in thickness, and frequently the pegmatite appears in the schists. Though there is a variation in the direc- tion of these veins, the prevailing one is northerly in the southern granite lobes. This pegmatite material shows all grades of variation from the well-defined, coarse pegmatite to pure vein quartz; and all the variations evidently had a common origin." From what has been written, it is evident that schist and granite were modified by mutual contact. The extensive seri- citization of one and the darkening of the other by the increase of biotite are the most noticeable effects. From these contact phenomena and from the schist inclusions in the granite it is evident that the granite is intrusive and younger than the schist. In rocks so thoroughly recrystallized as are the schists of our study no fossils can be expected, but the graphite found in both the rusty graphitic schist and also, though rarely, in the fibrolitic schist may point back to organic remains.?, To one acquainted with the rocks in the plateau of central Massachusetts, from Worcester to the Connecticut Valley, it is evident that these schists in and around Monadnock are but a continuation of the Massachusetts rocks, though there may be a few square miles of area between the two not yet mapped; and the conclusions that have been reached from the study of the latter are applicable to the former. After many years of study, Professor Emerson and the writer have concluded that the schists of this plateau in Massachusetts are more highly metamorphosed phases of the Carboniferous phyllite and quartzite found at Worcester. If this conclusion is correct, then these schists of Monadnock are Carboniferous, and the intruded granite is post-Carboniferous. Another fact demanding careful study is the attitude of the schist in this mountain. That the structure is not as simple as it might be is indicated by the statement on p. 639, Vol. II, of TJ. E. Spurr, “Genesis of Auriferous Quartz Veins,” Eighteenth Annual Report of the U. S. Geological Survey, Part III, pp. 311, 313. 2In the Geology of New Hampshire, Vol. Il, p. 503, a graphite mine in this schist is mentioned. 3 Geology of Worcester, Massachusetts, pp. 28, 50, 137, 139, 148, 152. 8 YOM BB al Je SAPNA the Geology of New Hampshire: ‘In structure it (Monadnock), seems to be a double synclinal. Again, on p. 24, it is stated: ‘Mt. Monadnock seems to be an isolated, contorted synclinal of andalusite mica schist.” The strike of the schist in the area east of the mountain, in the eastern slope, and in the northeast half of the ridge of the mountain, is between north and northeast, with a dip to the west and northwest. In the southern half of the ridge, as far north as an east-west line passing through the more southerly of the two houses on the mountain road, the strike is from north to almost east, with the dip to the northwest. In the top and in the northwestern part of the mountain the strike is 44) x ~ = ~ = ts A Granite Schist Granite B Fic. 3.—Section through Monadnock. Horizontal scale ggtp7; vertical scale e220; made from Topographic Map of U.S. Geol. Surv. to the northwest, with a dip to the northeast. In the extreme western part of the area here included, the strike is east and west north of the granite lobe with the dip varying from ver- tical to 40° south; and the strike is northeast on the north- west side of the same lobe, with the dip vertical. In the central part of the mountain mass, between the northeasterly and north- westerly strikes, the strike swings from one to the other through an east-west direction, with a dip of about 45° to the north. The meaning of this variation in the dip and strike is that the schist in the northern two-thirds of this mountain mass has been folded into’a synclinal having a pitch of 45° to the north; and the western limb of this synclinal is continued in an overturned anticline, with pitch to the south, around the western granite lobe. The axis of the syncline is marked, approximately by the course of Mountain Brook down the northern slope. The highest point of the mountain is not at the apex of the syncline, but is in the GEOLOGY OF MONADNOCK MOUNTAIN 9 western limb. South of the summit this syncline may be traced along the crest of the ridge, becoming narrower, as far as the east-west line of the more southerly house on the mountain ; south of this line, along the continuation of the axis of the syncline, there is a marked deviation of the strike to the east, Fic. 4.—Showing the jointing in schist. West side of Monadnock, 2800-foot level, looking southeast. The joint plane slants toward the right. The dip is toward the left. which becomes less and less as the syncline fades out to the south. ; On the transparent sheet accompanying the geological map the folds in the schist are represented. The lines of strike were first plotted, and then these curves were drawn through them. In this way the.anticline in the schist, as it folds around the granite lobe in the west, and the syncline in the mountain itself IO MOS RILEE Sal TI IIRIRNG are clearly brought out. A glance at this sheet is sufficient to convince anyone that the folding of the schist accompanied, and was due to, the intrusion of the granite; and the force of intru- sion was exerted in a northerly direction. Closely connected with the folding is the jointing in this schist, which is very noticeable in every part of the mountain, but especially above the line of vegetation. Here the rock has been broken into large blocks, perhaps 20-30 feet long by 10-15 feet thick and wide, and these are so placed in the mountain as to make a series of steps, as is shown in the accompanying illustration (Fig. 4). This picture was taken at about the 2,900- foot level, and on the western side of the mountain. Observa- tions of the direction and slant or dip of these joint surfaces were made up on the mountain where there was little or no vegeta- tion, and on different sides; and they are arranged in the table below. While there are not so many observations as there well might be, they make clear certain facts or relations. ro) 0) a @ Ha, 4 Xs 2 Fee cee Ele | er o G | os | be Velma se las ee i oe ase Side of Mountain and 3 ae BA ies | 2 | 3 [29 |8a4| See Elevation ow ng pe HOGg a8 SHE Nr yd 1 GS See | Ces) 2G. soe ee |e ce leer EAR | ARH | 35 | osn | 65 | Son | ES) 20% South ridge, southeast of rep, 2,500 level ..... 5 || ee MRA CRIN Glo Id: Negoeno: |lbocodooulldac eoda us ie Southwest of top, 2 2,850 level... *95° W |35° N E| 45° W |65°S W| 5° E | 25° W 50 80 West of top, 2,800 level .......]*N&S| 45° E | 35° W |60° S W/ E & W |Vertical 55 ae Northwest of top, 2,800 level...|N&S 45) By 50m Hy li 7Oe NeW) Cons Wa |ne75 ce) 50 North 30° west of top, 2, 800 level ... ».|-*20° BE) 40° | 32° We |60°2 S Wi6o2 E337 SE 92 si North of top, : 2 800. level . ..| 35° E | 60° E | 40° E J25° N W| 65° W | Vertical 75 95 West side of mountain, 1,900 levielltiii- . | *65° W j25° N E} 55° W |70° S W| 60° E | 4o°SE 65 85 Northwest ‘side in Marlboro trail, 2,200 level. Ba bnod dato. NAA ek ENE IBY) EAR Tee ISS NY) “soca oeulll Gooous go Northeast side, 2 200 ‘level ..... RR ID, PAGO NY | MGS 1s GSI bbe 00. || ado oe 70 Northeast side, 2,400 level ..... Fae GUE 33 Walesa Bal Gomer Hillie7ic ah, go° 40 87 Northeast side, 2,500 level ..... EOD | eo WZ | eS 15; 90° jo. Wigs° SW 75 52 Northeast side, 2,700 level ..... ge ay || SYS APS zig WA 90° Sseeh al esontS 50 In nine marked by a *, out of the twelve, the direction of joint I is approximately the same as the strike of the band- ing; or where the banding departs by local folding from paral- lelism with the side of the syncline, joint 1 is parallel to the side of the syncline. 2. The angle between the strike, or direction of joint 1, and GEOLOGY OF MONADNOCK MOUNTAIN II the strike, or direction of joint 2, varies from 40° to go”, but in five cases out of the nine is between 50° and 65°. 3. The angle between the true dip and the dip of the surface of joint I in five, out of seven, cases approximates to a right angle. The meaning of observations I and 3 is that this schist has been broken perpendicularly to the lamine into strips from one end of the syncline around to the other; and observation 2 shows that these strips have been broken crosswise, less regularly, into shorter pieces or blocks, the latter indicating a bending or twisting in these strips. There is, I think, no doubt but that these joints were produced by the bending of the schist into the great synclinal fold of the mountain, and this, as has been pointed out, accompanied the intrusion of the granite. The formation of the joints in this schist clearly indicates that this rock was in the zone of fracture when this intrusion and folding took place. The same is indicated by the brecciation appearing in some of the schist fragments included in the granite. But there is another and earlier folding evident in this schist. The accompanying illustration shows a compressed, overturned anticline occurring near the top, on the west side, which is con- spicuous for a distance of several hundred feet down the moun- tain to one ascending on that side. The part of the fold appearing in the picture is 37 feet long by 6 wide, and the apex of the fold as seen in this section points between northwest and north. This means that the fold is overturned and lies flat in the western side of the large syncline, with the apex of this small anticline pointing away from the apex of the large syncline. The formation of this small anticline, though the bending was sO severe, was not accompanied by fracture, as may be seen in the illustration (Fig. 5). Other folds, though less conspicuous, may be seen of which the same is true, and the jointing of the schist cuts through these, indicating utter independence of the one of the other. Therefore, when these smaller folds were made the rock was in the zone of rock flowage. There are indicated, therefore, two periods of metamorphism —one at a greater depth when the clastic was recrystallized and in places severely folded without fracture; and the other at less depth cz . LOS H Hal Jal, VITIINA when there was extensive sericitization, development of tourma- lines, and folding with fracture accompanying the intrusion of granite. Another point of interest in the study of Monadnock is to determine, if possible, why this rockmass has survived the pro- found erosion to which this region has been subjected. Dr. Fic. 5.—Fold in schist of Monadnock, near the top, on west side. Also shows banding in the schist. Gulliver,’ speaking of this mountain, attributes its survival to the greater resistance of the rock. In making such a comparison, it is well to bear in mind that there is an element of uncertainty in that the rocks that have been removed from above an area are not always the same, at least in this region, as those that now appear at the surface, on account of the extensive intrusions of eruptives. Dr. Haye’s? points out, in his study of the Chattanooga district, * Bulletin of the Geological Society of America, Vol. X, p. 19. ? Nineteenth Annual Report of the U.S. Geological Survey, Part I, p. 39. _ GEOLOGY OF MONADNOCK MOUNTAIN 13 that these residuals are the result of two factors, erodibility and location. In the case of Monadnock the second factor was probably quite as important as the first. It must be noticed that this residuary peak does not stand alone. The one farthest south in Massachusetts, and closely related to this one, is Asnebumskit, in the town of Paxton, rising about 400 feet above the peneplain or plateau of southern New England. It is situated about a mile east of the divide between the Connecticut and the Atlantic watersheds, and at the begin- ning of that of the Blackstone. In preglacial times it was, very likely, on the divide between the two main watersheds. This monadnock is made up of a rusty, fribrolite schist,-not to be dis- tinguished from that in hundreds of square miles of the surround- ing plateau. Next to the north is Wachusett, with two minor points associated with it, rising nearly 1,000 feet above the plateau, and situated in the town of Princeton. This is on the divide between the Ware and Nashua Rivers. It is composed of granite and the same rusty, fibrolitic mica schist, the former making up by far the larger part. There is nothing about these rocks to indicate that they are any more resisting than are similar granites and schists inthe surrounding plateau. Watatic, situated in the town of Ashburnham and rising about 700 feet above the plateau, is another monadnock situated on this divide. With the rocks of this mountain I am unacquainted. Then, crossing the state line into New Hampshire, we find Monadnock also on the same divide. There are more, though smaller, residuary peaks east and west of this; and to the north still others, even rivaling Monadnock. This is just what might be expected, even in a region made up of rocks of uniform resistance —incomplete pene- planation up toward the sources of the main streams, while it is almost complete farther to the south towards the mouths of these streams. My conclusion from the study of the rocks is that these monadnocks, situated along on the divide from central Massachusetts into southern New Hampshire, owe their survival to their position rather than to the rocks of which they are composed; for in a region made up of rocks of uniform resistance and subjected to peneplanation there would be some points up 14 S OSE? MLA PERRY near the sources of the main streams which would be the last to be brought low.’ SUMMARY. 1. This mountain is made out of a syncline in andalusite- fibrolite schist, probably of Carboniferous age. The syncline was produced by the intrusion of granite, which modified the schist already metamorphosed. 2. The schist contains andalusites changed to fibrolite and sericite, also fibrolite changed to sericite. 3. The jointing of the schist was produced by the folding; therefore the intrusion of the granite, which produced the folding, took place when the schist was in the zone of fracture. 4. There is an older folding evident, which must have been produced when the schist was in the zone of flowage. 5. This mountain and the other monadnocks to the south on the Atlantic-Connecticut River divide, probably owe their survival to their position, rather than to the greater resistance of the rocks composing them. JoserH H. PERRY. WORCESTER, MASS, *Mount Grace, in Warwick, Mass., rising about 500 feet above this plateau, and situated about six miles from the Connecticut, far to the west of the divide, owes its survival, probably, to the greater resistance of its rocks. It is made up largely, or entirely, of amphibolites, while the surrounding plateau is made up of mica schists and granite. ON, A CALIFORNIA ROOFING SLATE OF IGNEOUS ORIGIN.’ DurinG the field season of 1903 the writer. was enabled to spend several days in the study of the important roofing-slate deposits occurring north of Placerville, El Dorado county, Cali- fornia. A summary of the principal economic results of this investigation will soon appear in a bulletin? of the United States Geological Survey; while a more detailed description, with maps, will probably be issued later as one chapter in a Survey bulletin on the slate deposits of the United States. One result of the study, however, would seem to be of suf- ficient novelty and general geologic interest to be worthy of discussion in this JOURNAL, in considerable detail. This is the determination that a part of the roofing slates of the El Dorado county district have been derived, by dynamic metamorphism, from basic igneous rocks—gabbros or related types. THE CALIFORNIA SLATE DEPOSITS IN GENERAL. Location and general relations. Though roofing slate has at different times been quarried, on a small scale, in other parts of the state, the only important slate-producing area in California is located in El Dorado county. The quarries which have been opened in this district are located along a line running about N. 15° W. from Placerville, at distances of from one to six miles from that town. The location and general relations, both geo- graphical and geological, of the slate deposits and quarries, can best be understood by reference to the maps included in the “Placerville Folio” of the United States Geological Survey. The workable roofing-slate deposits of this district occur in a belt of the Mariposa slates, of late Jurassic or early Cretaceous age. The quarries which have been opened are all situated near the western boundary of this belt of Mariposa slates, where it is ™Published by permission of the director of the U. S. Geological Survey. 2 Contributions to Economic Geology, 1903. 15 16 IPIONVMUN, (On JR CIKAIBIL bordered by a large area of diabase. This diabase has been described* by Lindgren and Turner as being ‘of the age of the Mariposa slates, or older.” A number of linear areas of amphib- olite occur in the Mariposa slates. These amphibolites are described as being derived from diabase or gabbro. They are in part altered to serpentine. Previous work on the slate deposits—The ‘Placerville Folio,” No. 3, U. S. Geological Survey, published in 1894, contains the results of detailed geologic work by Lindgren and Turner in the area in which the roofing-slate deposits occur. At that date the roofing-slate industry had not assumed its present impor- tance, though all the quarries now in existence had then been opened. The existence of roofing-slate deposits is noted in the text of the folio, and the locations of the quarries are indicated on the map showing the economic geology of the area. No reference is made to the ‘green slates,’ or to the dikes cut- ting the Eureka quarry. Excellent, though brief, descriptions of the different quarries and of the condition of the slate industry at various dates are to be found in the Reports of the State Mineralogist of California, particularly in the eighth and twelfth reports. At present the most important quarry is that of the Eureka Slate Co., and this is now being worked on a large scale. This quarry is located at Slatington, about one-half mile southwest from the point where Kelsey is shown on the Placerville atlas sheet. Structural relations in Eureka quarry.—The cleavage planes of the slates in: the Bureka quarry stuke Ni25> Wa) thes dipmer the cleavage is practically vertical, with slight local variations to 80° E. or 80° W. The upper weathered beds in the quarry are overturned, by local pressure, so as to give 40° to 60° dips to the east or west, according to local conditions. This overturn- ing is evidently due merely to the weight of the overlying soil and decomposed slate, and the effects are shown only for a depth of from” 3\-to15-feets, It is vor interest, showeversasma warning against accepting dip readings taken from surface beds of the slate. *“ Placerville Folio,” U. S. Geological Survey; legend of ‘‘ Areal Geology” sheet. ROOFING SLATE OF IGNEOUS ORIGIN 17 The slate body shows rather frequent, but narrow, ‘“‘ribbons.”’ These ribbons are bands (from ;), to 4 inch thick usually, but occasionally as thick as two inches) of material differing in com- position from the mass of the slate. They are in general more siliceous than the normal slate, and do not furnish merchantable material. Their geologic interest arises from the fact that they represent differences of original sedimentation. The plane of the ribbons in a slate quarry is therefore the plane of original bedding. Inthe Eureka quarry, and, indeed, throughout the roofing-slate belt, the plane of original bedding seems to be usually within ten degrees of the plane of slaty cleavage. The slate mass is cut by a series of joints, parallel to the ‘‘grain”’ of the slates, striking N. 55° E., and dipping from 70° to $0° to the northwest. Joints across the ‘‘grain” of the slate, which would be practically horizontal, do not occur in this quarry; but many of the thin quartz seams occupy this position. Quartz and calcite occur in thin layers, filling joint spaces and occasionally cleavage spaces. Pyrite also occurs in very much flattened nodules, which were apparently parallel to the original bedding. Character of the normal slate—The mass of the Eureka quarry product is a dense, deep black slate, splitting very finely and regularly, with a smooth glistening surface much like that of the Bangor and Lehigh slates of Pennsylvania. The frequency of the ribbons, and of the pyrite nodules, prevents the slate from being serviceable as mill stock; but as a roofing material it is very satisfactory. A specimen of the black slate, free from ribbon, was selected for analysis in the laboratory of the U.S. Geological Survey. The results of this analysis, by Mr. W. T. Schaller, follow: ANALYSIS OF BLACK SLATE, EUREKA QUARRY, SLATINGTON, CALIF. Silica (SiO,) ~—- - - = = =O3.52 Alumina (Al,0;) and titanic oxide ee ) 16.34 Iron oxides (FeO, Fe,O a) Se) 57/0) Lime (CaO) - - - : ; 0.98 Magnesia (MgO) - - = - 250 Carbon dioxide (CO,) l - .86 Water \ : 18 EDWIN €. ECKEL THE ‘‘'GREEN SLATES.” Appearance and structural relations.—Perhaps the most striking feature of the quarry of the Eureka Slate Co., as seen from the old ground surface eighty feet above the present floor of the quarry, is a light green band, four feet or so in width, that extends vertically from top to bottom of the quarry, and is par- ticularly noticeable on the higher east wall. This band furnishes the ‘‘green slate”’ of the quarry men. The contrast in color between this green band and the intense black of the fresh sur- face of the rest of the slate is very striking. Viewed from the old ground level, one cannot determine whether the green band is parallel to the slaty cleavage or to original bedding; which planes, as noted earlier in this paper, commonly differ only by ten degrees or so. At first sight, there- fore, the green band might reasonably be considered to be a mere color variation, due either to original differences in compo- sition of the beds from which the green and black slates were derived, or to a later change in the color of certain beds; and this view has apparently been accepted by former observers. Closer study, however, removes this easy explanation from the list of possibilities. Even a casual examination of a green slate quarried from this band, and comparison with a slate of the normal black type, are sufficient to prove that the two slates are different in more than color; while a closer examination of the character and structural relations of the green band, when seen from the quarry floor below, suffices not only to emphasize the distinction between the green slate and the black, but to suggest a somewhat novel origin for the former. Relations of contact plane to cleavage and bedding.—On going down into the quarry and closely examining the relations of the two slates, the contact between the green and black slates is seen to be, not parallel to the ‘“‘ribbon”’ of the black slate, which indicates the plane of original bedding, but cutting the ribbon at a small angle—not over ten degrees. It is not, however, certain that the contact plane is exactly parallel to the plane of slaty cleavage, which also cuts the bedding plane at a small angle. This detail—the relation of the contact line to the cleav- > ROOFING SLATE OF IGNEOUS ORIGIN 19 age planes—has, of course, no bearing on the question of the derivation of the green slates. It may prove, however, to be of some importance in determining the probable cause of the origin of slaty cleavage in this particular portion of the Mariposa slate belt: . Disregarding this omission of data as to the relation of con- tact to cleavage plane, the fact remains that the band of green slate is not everywhere conformable to the original bedding of the Mariposa slate series. It is, therefore, highly improbable that it was an originally interbedded member of that sedimen- tary series; on structural grounds it is probable that it represents a mass of igneous rock, injected as a dike into the Mariposa series and, subsequently to its intrusion, so highly sheared as to have a very perfect slaty cleavage. This probability is increased when the chemical composition of the rock is considered. Further confirmation of this hypothesis is aftorded by an examination of the cross-section of the dike, which proves that it is not homogeneous in texture throughout, but that it varies in bands closely parallel to the contact plane. Along its con- tact with the normal black slate, the green slate is very fine- grained for an inch or so. Bordering this is a zone several inches in width, of coarser texture, and drab-green color, which is fol- lowed in turn by the typical ‘‘green slate.’ These differences in color and texture are sufficiently noticeable to be readily distinguishable by the quarry men and slate splitters. The tex- tural differences are such, in fact, that the layer immediately next to the contact is discarded as a “ribbon,” since it works unsatisfactorily. It must be recollected, however, that these layers are parallel to the contact, not to the “ribbon” or original bedding of the black slate. | Igneous rocks of the vicinity About five hundred feet west of the present quarry the western edge of the Mariposa slate is reached, a body of igneous rock limiting it in that direction. This rock is described in the ‘Placerville Folio” as diabase. )) Goes coosonce goo aod c0cr 13.84 16.45 15.47 shitamiczoxider(Gi@ snc ane cent ee spares 0.46 0.76 0.65 errousoxidey(e@)) Gaeryeccmecerseteicrcrne: 1.74 2.90 2.46 Herricjoxidey (Hes Os) arr crsstc erases « 4.00 4.04 4.03 Mimer(Ca@) Meera iceiesentoe : ceeadeniacies 5-96 1.41 Ble WMleveravesia, (MUO) co goedcaducupeadcganobe 2.67 2.32 2.45 Botashe (KAO) bes ao feccie tera s spares seecoun ees 2.67 3.60 3.25 Sodas (Nats ©) errr tre eee cue Aaicrn Serco 1.80 tO Wage Carbonvdioxidemernresciecne miei nehinee 4.62 1.46 2.64 Waterrotecombinationsaste. suet eee 3.45 3.82 3.68 Moisture belowsrlOml Gracin mrceeen 2eeel 0.89 ogy It may be noted in passing that some of the differences in composition between the Paleozoic and the later shales were, either in degree or in kind, contrary to what might have been expected, from a purely theoretical standpoint. The writer is 28 EDWIN €. ECKELE at present inclined to believe that these unexpected results are susceptible of a rational, though decidedly novel, explanation. In case this belief is supported by the results of investigations now in progress, this explanation will be presented in a later paper. Comparison of ‘average slate’? and ‘average Paleozoic shale” analyses— As thirty-five of the thirty-six slate analyses are of Paleozoic material, the ‘“‘average slate” will obviously be com- parable most directly with the composite analysis of the fifty-one Paleozoic shales. The necessity for restricting the comparison in this manner is accentuated by the fact, above intimated, that the Paleozoic and the later shales are not themselves directly comparable. The average slate contains 60.64 per cent. of silica, as against 60.15 in the Paleozoic shale. Alumina and titanic oxide together amount to 18.05 per cent. in the slate, and to 17.21 per cent. in the shale, the titanic oxide being practically the same in both. While the proportions of ferrous and ferric iron oxide in the slate and shale are reversed, the fofa/ amount of iron oxides is very close, being 6.87 in the slate and 6.94 in the shale. The lime and magnesia show the first interesting difference, and this is but slight. While their ratios are closely alike in the two MgO CaO total lime and magnesia in the slate is 4.14 per cent., and in the shale 3.73 per cent. A somewhat similar case occurs with the rocks ( ==1.688 in the slate, and 1.646 in the shale), the K alkalies. Here a =-3. 101 sin the slate, and 32564; imamethe shale; while the total alkalies amount to 4.74 and 4.61 per cent. respectively. The variations in carbon dioxide, water of combination, and moisture are so slight that they would not appreciably affect the percentages of the other constituents, with the exception of silica and alumina. Summary of results —The results of the comparison are mainly negative, but they are of value even in that way. The average slate is practically identical in composition with AMERICAN SHALES AND ROOFING SLATES . 29 the average shale. It contains sizghtly more of certain readily soluble constituents than does the average shale. This is to be accounted for by the fact that the slate is, on the whole, made up from finer materials than the shale; if it were otherwise, its cleavage would not be so perfect. During the change from shale to slate—or from mud to slate—its composition, omitting water from consideration, was practically unaltered. The only effect of metamorphism was the assumption of slaty cleavage, and this was effected without the introduction of any new constituent. - Incidentally, the statement in the preceding paper, that the green slate from Slatington, Calif., differs in composition from any normal clay slate, is entirely justified. Epwin C. ECKEL. U. S. GEOLOGICAL SURVEY, Washington, D. C. ARAPAHOE GLACIER IN 1903:! A visir to Arapahoe Glacier, in the cirque on the east side of Arapahoe Peak, west of Boulder, Col., in 1903, for the purpose of comparing its present condition with that existing at the time of the preceding examination, brought to light some interesting and important facts. In 1902 we were on the ice on three different days during the last week of August. In 1903 the visit was made on September 2, only one day being spent on the ice by the writer, in company with Professor A. H. Felger, of Denver. The summer of 1902 was particularly favorable for our first visit because the high temperature and preceding shortage of precipitation had caused the snow to melt from the surface of the entire glacier up to the Bergschrund, which was exposed as a great, gaping break clear across the face of the glacier; but in 1903, even at a great distance, it could be seen that snow still remained on the ice down to the main system of crevasses, while the Bergschrund was exposed for only a very short distance, which was perhaps due as much to the unusually cool spring and summer as to the greater fall of snow last winter. The greater part of the Bergschrund and many of the visible crevasses were partly or wholly filled with last year’s snow, though in some cases the snow merely formed a bridge instead of filling the crevasse, which made it dangerous traveling above the line of uncovered ice. On the whole, there was a slight increase in the height of the snow and ice along the north side and in the height of the ice along the terminal moraine, but at two points along the terminal moraine there has been a decided shrinkage during the year, which is of some importance. One such point is at the lake retained by the moraine, or rather where the central surface drainage system pours its waters into the lake, which is shown on the map (Fig. 2) and the photograph (Fig. 8) accompanying ‘For a general account of this glacier see this JOURNAL, Vol. VIII, p. 647. For a more detailed account see Vol. X, p. 839. 30 THE ARAPAHOE GLACIER 31 Dr. Fenneman’s paper. The other point of shrinkage is where the western surface drainage breaks through the moraine, Both of these ice valleys were greatly deepened, and their sides were much steeper than last year, which suggests stream erosion as the prime cause of the change in the face of the glacier at these points. However, as the surface streams flow southward, and their valleys are consequent- Morzine ly exposed to the full glare of the noonday sun and protected from cooling breezes, the confinement of heated air in the valleys is an element to be reckoned with in the problem of waste at these points. These surface streams Fic. 1 —Diagram on exaggerated scale, the on the front of the _ broken line showing change in front of glacier now glacier BAG de possi apparently in progress, from erosion and melting ble by the fact that for some distance back from the terminus the ice is free from open crevasses into which the water arising from surface melting could otherwise plunge. Wasting at these points, while apparently gaining slightly, or at least fully holding its own, at other points along the front of the glacier shows clearly how a glacier may change the shape of its terminus, and consequently of its terminal moraine, as shown by the accompanying diagram. It is probable that the confusion of moraines several miles below the present terminus of Arapahoe Glacier, where they intersect each other at all angles, is partly due to the rapid and repeated changes in the end of the former ice tongue, as well as partly to the melting out of ice blocks in the moraine, as suggested by Dr. Fenneman. Where the surface stream enters the terminal lake the ice is now worn to the ground moraine, so that the water issues from along water courses. 32 JUNIUS HENDERSON beneath the ice at the upper edge of the lake loaded with rock flour, and rises milky white abruptly through the clearer water gathered from the surface, forming a striking feature not found there in 1902, when, if it found its way into the lake at all, it was so diluted before reaching the surface that it did not attract attention. As would be expected from the deepening of the drainage valley leading into the lake, by vertical erosion, and the increase in its width by lateral erosion and melting, the area of the lake is much greater than in 1902, just as its area in 1902 was appar- ently greater than when examined by Professor Lee in the summer of 1900. Another noteworthy fact is that almost the entire front of the glacier was steeper in 1903 than in 1902, upon which subject more data and accurate measurements seem necessary before a satisfactory reason may be assigned for the change. Although the rate of flow for this glacier is unknown, it must surely be so slow as to require several years of normal or exces- sive precipitation or low temperature for its down-stream extension to recover from the effects of rapid waste during a series of warm, dry seasons. An analysis of the records of the Denver, Boulder, Sugar Loaf, Moraine, and Long’s Peak stations, as shown by the monthly statements of Mr. F. H. Brandenburg, Denver forecast official and section director of the Weather Bureau, indicates that from September 1, 1901, to September 1, 1902, the precipi- tation was far below normal and the average temperature was much above normal. On the other hand, from September 1, 1902 to September I, 1903, precipitation was considerably above normal and average temperature was much below normal. The highest of these stations is only 8,500 feet above sea-level, and the nearest is about ten miles distant, so the figures given for the stations would not exactly apply to the glacier, the altitude of which is about 13,000 feet; but the fact that all the stations record greater precipitation and lower temperature for the latter period than for the former period makes it appear exceedingly probable that the same condition would have been shown if a rec- THE ARAPAHOE GLACIER 33 ord had been kept at the glacier itself. Furthermore, the cirque containing the glacier is in full sight from the brow of University Hill in Boulder, and during the autumn of 1902, winter of 1g02— 3, and spring of 1903, the writer watched the peak as closely as possible, and was much impressed with the frequency with which clouds gathered in the glacial amphitheater when all was clear elsewhere. The snows began early in September and continued until late spring. It seems possible that the significance of the silt on the morainal bowlders in 1902 may have been misunderstood, as a similar condition was observed in 1903; and yet it is hardly pos- sible that the glacier flowed out over the moraine and then melted back to the base of the moraine between September, 1902, and September, 1903. A possible explanation suggests itself. The accumulation of snow on the glacier each year is known to be enormous, a great many feet in depth. The westerly gales are known to carry from the rim of the cirque to the glacier large quantities of dust and finely comminuted vegetable matter dur- ing every dry, windy period, which periods are common in that region and at that altitude. The accumulated snows containing such dust and vegetation cover the moraine to a great depth in winter. As the snow melts in the spring, the débris gradually accumulates at lower and lower levels, until finally it is left as silt on the tops of the bowlders and all over the moraine. Unfor- tunately, we brought none of it with us, and failed to make a minute examination of it under a good lens. The cause may be sufficient to account for the phenomena, but the subject is worthy of more extended investigation before announcing definite con- clusions. Junius HENDERSON. MusEUM, UNIVERSITY of COLORADO, Boulder, Col. THE APPALACHIAN. RIVER VERSUS A TERTIARY TRANS-APPALACHIAN RIVER IN EASTERN TENNESSEE? In a very able and interesting paper, entitled ‘‘Geomorphology of the Southern Appalachians,” by Hayes and Campbell,? con- siderable evidence is adduced in support of the theory that the drainage of the southern Appalachian valley up to late Tertiary time was through one rather large subsequent stream, occupying, in general, the position of the upper Tennessee and Coosa Rivers. This river, which ran down, as they believe, east of Lookout Moun- tain, over what is now the Tennessee-Coosa divide, and directly into the Gulf, has been called by them the Appalachian River. The evidence given for the existence of such a river is threefold: ‘(1) the perfectly base-leveled divide between the Tennessee and Coosa River basins; (2) a comparison of the volume of material eroded from the Appa- lachian valley with that of the Tertiary sediments in central Alabama; and (3) the immaturity of the Tennessee gorge through the plateau below Chattanooga.3 The three lines of evidence will be taken up in the order given above and examined briefly in the light of other facts furnished by this region, to see if the Appalachian River theory is the most tenable. The Coosa-Tennessee divide.—Iit is admitted by Hayes and Campbell that similar low divides are found between other river basins of the Appalachian valley from Pennsylvania southward; as between the Potomac and the James, and the James and the Roanoke ; but they do not believe that any great river ever flowed *The writer came to the view advocated in this paper in 1896, while engaged in the field underdiscussion. ‘The paper was written in May, 1897 as a report in a course in physical geography in Harvard University given by Professor W. M. Davis. Pub- lication has been delayed owing to a desire to make a more extended study of the problem in the field, but interest in other fields renders the making of such an exam- ination improbable at an early date; it therefore seems best to present the paper in its original form. ?C. WILLARD HayrEs AND Marius R. CAMPBELL, “Geomorphology of the Southern Appalachians,” ational Geographic Magazine, Vol. V1, pp. 63-126. 3/bid., p. 109. 34 APPALACHIAN RIVER IN EASTERN TENNESSEE 35 over those divides. The Coosa-Tennessee divide is, however, wider and if peneplained by backward-cutting small streams, those streams, they believe, should have a ‘dendritic inosculat- ing”’ arrangement. By referring to the structural sheet of the ‘Ringgold Folio”’ of the United States Geologic Atlas, it will be seen that the width of the Coosa-Tennessee divide is easily explained by the struct- ure of the region. The floor of this valley is almost entirely of Knox dolomite and a Cambrian shale both tilted at various angles, and always valley-making. It seems quite clear that, throughout this region, topography depends upon structure rather than upon the size of the streams. The width of the Coosa- Tennessee divide is limited only by the conglomerate- and sandstone-capped strata approaching horizontality. On the east, the resistant capping is Silurian; on the west, Carboniferous. A great river is evidently not necessary for the base-leveling of a region composed of the upturned dolomite and shales of this region. Cade’s, Ware’s and Tuckaleechee coves,’ east of Chil- howee Mountain, are examples of this. The floors of these coves are of deformed dolomite and Wilhite slate, a very calcareous slate, and are lowered as rapidly as the streams can cut down their channels through the Chithowee sandstone. These coves are surrounded by massive conglomerates, sandstones and slates, and the streams which drain them have their sources practically within the coves. Dendritic branching of streams could hardly be expected ina region where the beds are varied in character (alternating harder and softer) and at the same time deformed into structures approaching parallelism, as these have been in this part of the Appalachian valley. Considering the structure and elevation of the region, it is difficult to conceive how the stream habit could be other than it is. The volume of material eroded and deposited —This evidence favoring the Appalachian River theory is briefly as follows: A stream occupying, in general, the present position of the Coosa- Alabama River is held responsible for the late Cretaceous and ™“Knoxville Folio,” U.S. Geologic Atlas. 36 CHARLES H. WHITE Tertiary sediments (Ripley, Lignite, Buhrstone, Claiborne, and White limestone) deposited in an area, limited on the east by a line midway between the Alabama and the Chattahoochee, and on the west by a line similarly drawn between the Alabama and the Tombigbee Rivers ; sediments that occupy an area of 6,500 square miles and a volume of 2,340 cubic miles. The volume of material carried away from the basin of the Alabama and its tributaries between the Cretaceous and Tertiary peneplains is about 622 cubic miles; but if we add to this the volume carried from between the two peneplains in the upper Tennessee valley, the total amounts to 2,500 cubic miles. If it be admitted that the only source of the Cretaceous and Tertiary sediments in the area named was from between the Cre- taceous and Tertiary peneplains in the two river basins, and that this detritus was carried out in the direction named, the evidence is conclusive that the Appalachian River was a reality. But from the character of the coastal plain beds in this particular area, it is doubtful if the Alabama River, or its ancestor, should be held wholly responsible for the strata lying along its lower course. At least one-third of the thickness of these beds shown in outcrops is of limestone, bearing quantities of corals and other fossils.* The Alabama River cannot be held accountable for those deposits transported in solution or in a very finely divided state to the exclusion of the other rivers emptying into the Mississippi embayment. After the formation of the Nita cre- vasse in 1890, fine mud from the Mississippi River was deposited in Mississippi Sound even up to the mouth of Mobile Bay, driving out the fish and killing the oysters.2 The thickest and most important of the beds attributed to the Appalachian River, the Lignitic (g00 feet), is composed of cross-bedded sands and clays, and would seem, at first sight, more than all others, to have been deposited by this river; but a study of the whole area shows that this formation increases in thickness and in coarseness toward the west, in western Alabama and in Mississippi; while ‘EUGENE A. SMITH, L. C. JOHNSON, AND DANIEL W. LANGDON, JR., Report on the Geology of the Coastal Plain of Alabama (1894), pp. 107 ff. 2 [bid., p. 30. ~~ -« APPALACHIAN RIVER IN EASTERN TENNESSEE SH east of the Alabama River it is very calcareous and ‘‘inconspicu- ous,’* showing that these sandy sediments were brought down from the west instead of from the north. Considering the nature and disposition of these sediments, and the decided notch in the Continental shelf nearly opposite Mobile Bay, and the tendency of the Gulf current to carry sediments eastward in this region, there seems no reason for believing that the Coosa-Alabama River was ever larger than it is at present. Character of the gorge below Chattanooga.—It is claimed that this gorge is too immature to have been the channel of the Ten- nessee since Cretaceous time, or during the erosion of the upper Tennessee valley. It is admitted that erosion has progressed much more rapidly on the upturned strata of the valley, but it is thought impossible that the river could make so wide a valley along one part of its course and be held within such narrow limits lower down. The rate of erosion in tilted soluble rocks, compared to that in horizontal beds of the same nature, capped by heavy beds of sandstone and conglomerate has never been definitely determined ; but the contrast is undoubtedly strong. The Nashville basin has been eroded in strata only slightly domed, while the streams leading therefrom pass through gorges in horizontal strata hav- ing only a slight siliceous covering. It is a notable fact, in this connection, that the valley of the upper Tennessee is eroded back toward the west only just to the beginning of the horizontal strata, and where the horizontal strata have been reached the slope is as steep and the distance cut into the horizontal beds is no greater than in the Walden gorge; and moreover, the streams that run from the Walden plateau east into the great valley have proportionally as narrow gorges with as steep slopes as has the Tennessee in its gorge. If the Appalachian River existed and the valley is older than Walden gorge, these side gorges are also older, and, according to this reasoning, should be wider; for the streams, although small, apparently carry all the waste brought to them. The wide coves on the eastern side of the valley made by small streams on upturned dolomite, and pre- "Tbid., p. 148. 38 CHARLES H. WHITE sumably of the same age as the gorges on the west, have already been noted. It appears, then, that evidence favoring the existence of the Appalachian River is small, although this has been the view of some observers for more than twenty years. Long," in his report of the survey of the Tennessee and Holston Rivers in 1875, states this as his view, but gives little evidence for its support. A trans- Appalachian river.—What seems to the writer a more tenable view of the history of the drainage of this part of the Appalachian province follows. Up to the close of Cretaceous time the rivers flowed off toward the northwest from the axis of the Great Smoky Mountains, or in general, at right angles to this axis; and the more nearly base-leveled the Cretaceous peneplain became, the more the streams meandered uponit. The headwaters of these old rivers, which cut directly across the strike of the present valley strata, are now represented by the Doe-Watauga, Nolichucky, French Broad, Big and Little Pigeons, Little River, Little Tennessee, upper Tellico, Hiwassee, Ocoee, and upper Con- nasauga. After the uplift of the Cretaceous peneplain, by differential erosion the great valley began to be formed, and lateral subsequent branches pushed their way back along the strike of the valley-making strata. The uplift being greater in the northern part, the south-flowing subsequent streams, being the more accelerated, have cut back more rapidly than those flowing toward the north, and have made practically all the cap- tures in this region. This process of capturing has gone on until now all the original transverse streams have been turned from their courses across the Cumberland plateau southward into the Tennessee River, which alone maintains its course across the Walden plateau. The upper course of the original Walden gorge river has already been taken south by the Connasauga, while the Tennessee awaits a friendly pirate to conduct it south directly into the Gulf through a shorter and easier course. That the position of the master stream has been west of Lookout Mountain rather ? LIEUTENANT-COLONEL S. H. LONG, “Report Relative to the Improvement of the Navigation of the Holston and Tennessee Rivers,’ House Executive Document, Forty-Third Congress, Second Session, Vol. XV, No. 167, p. 16. APPALACHIAN RIVER IN EASTERN TENNESSEE 39 than east of it for a long period is indicated by the westward drainage on Walden plateau south of the gorge. The character of the Tertiary sediments in Alabama, as pointed out on a previ- ous page, indicates that the larger streams entered the Missis- sippi embayment west of the Alabama River. The mature swing of Walden gorge also suggests that the course of the river was established here when the region was nearly at base-level, previous to the uplift of the Cretaceous peneplain. A consideration of the failure to find any satisfactory process by which a stream, when once established in this easily eroded valley and flowing directly into the sea, could be diverted across Walden’s Ridge, together with the foregoing facts, renders the existence of a Tertiary river across the present valley and along the course of Walden gorge more probable than that of a Ten- nessee-Coosa, or Appalachian River. CuarLes H. WHITE, DHE CONTACT. (OF THE eARCH AcAN DT AND ROS ARCH A ANS IN. PEE REGION FOR si GR aay LAKES. THE accompanying map has been compiled from one published by the Geological Survey of Canada, to which was added the Adirondack region from one published in Van Hise’s Pre-Cam- bian Geology. ‘‘Archzan”’ is used to include the Huronian and Laurentian, as originally proposed by Dana. It does not here be) include the Animikie or Keweenawan, which are considered post- Archean. With this exception, Archean is here used as the equivalent of the pre-Cambrian of many writers. The purpose of this paper is to bring out the peculiar step- like arrangement of the contact between the Archean and post- Archean in the region of the Great Lakes. This is shown clearly on the map, where a heavy line has been drawn to emphasize this feature. The contact is further marked by the almost universal elevation of the Archzan several hundred feet above the succeeding formations. Beginning at the St. Lawrence on the east, we find the line of contact runs about 10° N. of W. to the southeastern corner of the Georgian Bay. The Archean fronts the Paleozoic sediments nearly always as a hill rising one hundred or more feet above them. It is true the Archean underlies the sediments, and per- haps as a fairly regular plain, but it would seem that there has been some disturbance by which a part of the Archean has been relatively depressed. The contact of the Paleozoic follows closely this line of movement covering the lower part and abut- ting against the higher." The east shore line of the Georgian Bay runs 30° W. of N., and this may be taken as the line of contact. Only in the south- east corner do post-Archean rocks appear. The Archean rises in hills a few hundred feet above the water line. The contact north of Lake Huron again runs 10° N. of W. ™Compare WILSON, JOURNAL OF GEOLOGY, Vol. XI, p. 651 ff. 40 v . ‘\\vvw v ewe vw sf uvvvee © Che, DRIVE’ WS vv vvy ViAVIn! vu, DON DYE 60 OO: SO) CONC TAN vv yoy oVvVvvuwvvuvewvy viv pvwww vv ve vevvvy vv yvvKUMY YM VY Vv vw vy vv Vvve vy vvevvud v x vy vvvvy < v \v v YP vey v v v Vv KES REGION Y ys P ARCHAAN............. POST ARCH-AAN. (I RAG GE OLOGICAL BM . ee CONTACT OF ARCHLAN AND POST-ARCHZAN 41 As before, the Archzan hills rise several hundred feet above the waters of the lake, and the post-Archezan sediments abut against them at the water’s edge. The general direction of the shore line from Sault Ste. Marie to Michipicotenis 15° W.of N. It is, however, not so regular as other parts of the coast. At Gross Cap, Mamainse and Gargan- tua the Archean projects to the west in three sharp teeth with deep bays between. At each point there occurs a small area of Keweenawan rocks lying almost flat and rising but little above the waters of the lake. The Archean hills rise, on the contrary, almost vertically for several hundred feet, and at a few miles from the shore reach as much as 1,200 feet. The line of disloca- tion in this region is not straight. but serrate. The general direction of all the north-bearing lines is, however, preserved. From Michipicoten the shore line trends due west, rising in steep, precipitous hills. No post-Archzan sediments are found in contact with the Archean, but Michipicoten Island of Kewee- nawan age is only a few miles south. From the mouth of the Pucaswa River the shore line bears about 18° W. of N., having the same high, precipitous Archean cliffs, and with no post-Archzen sediments above water level. From Heron Bay the shore line again bears about 10° N. of W..as faras Nipigon Bay. Archean hills rise perpendicularly several hundred feet above the lake. In Nipigon Bay Keewee- nawan sediments appear, with interbedded lava sheets, the whole rising several hundred feet above the level of Superior. The contact here turns again to the north, but on the map does not appear to be so definite. The Archean hills do not rise in vertical cliffs along the eastern shore of Lake Nipigoan, and in consequence the lava flows of the Keweenawan, having no definite walls to impinge against, spread out in irregular bays. The east shore of Lake Nipigon is prolonged intoa deep, narrow bay at the south, and from this a valley leads through to Lake Superior, which is believed by Parks to be an old outlet of the lake. This valley and bay are in direct line with the east shore of Lake Nipigon and are taken to represent the counterpart of the coast lines described in earlier paragraphs. The Archean to 42 A Bs. VEE MO TAS the west of the line is much lower than that to the east of it,and is separated from it by a deep valley. The Keweenawan to the east of the line on Lake Nipigon is a shallow overflow. The physical features of this line are thus similar to the others, but with the characters less marked, due probably to the fading out of the force-producing dislocation. Eastward from the St. Lawrence the contact seems to run 25° S. of E. and then east. With this region I am not personally familiar, but believe that here also the Archean rises in steep hills above the adjoining, low-lying post-Archean. The two facts brought out above, viz., the step-like regularity of the lines of contact, and the difference in altitude of the rocks of the two periods, can be best explained, I believe, by assuming a dislocation in the Archean before the deposition of the post- Archean sediments. The southern part was depressed and- probably changed in inclination, as suggested by Wilson in the paper quoted above. Sediments were deposited over this depressed part and abutted against the cliffs of the northern part. They also extended into the deeper valleys of the northern part. Perhaps even a considerable portion of the northern part was thinly covered, but, if so these rocks have since been removed. The regularity of the dislocation extends from longitude 75° to longitude 88°, and from latitude 43° to latitude 50°, or for 700 miles by 500. For the greater part of the distance this line of fracture forms the northern and eastern shores of Superior and Huron. A. B. WILLMOTT. SAULT STE. MARIE, CANADA. - January 21, 1904. Die MONS HIPS AND TABITS OF Tris MOSASAURS. THERE is, at present, no group of extinct reptiles which is better known than that of the mosasaurs or pythonomorphs ; there is no group concerning whose affinities and relationships there have been more discussion and differences of opinion; and, also, the writer may venture to add, there is no group of extinct vertebrates concerning whose phylogeny and taxonomic position there is now less ground for dispute. The mosasaurs are specialized aquatic lizards, descended from the immediate ancestors of the modern monitors, through the extinct aigialo- saurs, and which became wholly extinct near the close of Cre- taceous time. They belong among the Lacertilio and are more nearly related structurally and phylogenetically to the living monitors than are the monitors to the living amphisbaenas or chameleons. Whether the Squamata are an order, as is usually taught in text-books, a superorder, as Fiirbringer and Osborn suggest, or a subclass, as Gadow believes, scarcely affects the relations of their component parts, nor the value of the group as a whole, if all the other so-called orders of reptiles are raised to equal rank. But to make the Squamata.a superorder, as does Osborn in a recent publication,’ while leaving far more specialized groups, such as the pterodactyls or turtles, in their former ordinal posi- tions, is manifestly inconsistent. I know not what may be the object of taxonomy unless to indicate the relative degrees of divergence or of specialization of organisms; and to separate the scaled reptiles from the Rhynchocephalia by superordinal characters, while the far greater differences between the pterodac- tyls and other reptiles, for instance, are accorded an ordinal value only, is, in my opinion, unjustifiable. Nor can it be said that the differences between a lizard and a snake have nearly as t Memoirs of the American Museum of Natural History, Vol. 1 (N. 1903), p. 456. 43 44 S. W. WILLISTON much value as those between any other two groups of reptiles properly called orders. That the classification of the vertebrates will some time reach a fairly satisfactory stability is certain, but that this much-wished- for consummation is not at the present time a reality is painfully evident. Especially is the lack of a taxonomic equilibrium between the various groups of reptiles and those of the birds very apparent. In every group of organisms, where individuals and species are abundant, there is a constant, an almost unchecked, tendency to raise the lower divisions to a higher rank, without due regard to their relative values, so far as those of other groups are concerned; and this tendency will never be dis- couraged save by the general morphologist; the specialist usually lacks that breadth of view so necessary for a proper perspective. The ornithologists, especially, have been in the past a law unto themselves in this respect, without due fear of the results of their taxonomic misdeeds before their eyes. Twenty-five or thirty groups of birds have. been given ordinal rank, while it seems evident that the whole class scarcely presents as many or as wide internal differences, unless it be in a few forms, as are found in not a few single groups of reptiles that are usually called orders. But, I respectfully submit, the raising of the orders of reptiles to subclass rank, as Gadow has done, is not the proper solution of this difficulty. Unless we invent superclasses and super- kingdoms, or some entirely new classificatory denominations, an impassable wall will soon be reached, when all the orders are leveled up to subclasses, and we are no nearer a final symmetri- cal classification than we were, though there may have been many excellent opportunities for coining new names. The relationship of the mosasaurs to certain modern lizards, the Varanidae, was pointed out by Cuvier in his original descrip- tion of the famous Maestrichtian example one hundred years ago. Goldfuss, Owen, and Marsh all recognized this rela- tionship as a real one, and refused to give to these animals the rank which Cope insisted upon giving. It was Baur, how- ever, who first strongly emphasized the importance of their varanid affinities. He gave to the mosasaurs only a family rank RELATIONS AND HABITS OF MOSASA URS 45 in the superfamily Varanoidea; perhaps an extreme view, but one which more recent discoveries have, I believe, in large measure substantiated. Gadow, very recently, has gone to the opposite extreme in separating the mosasaurs into a distinct subclass away from the subclass Squamata. Cope, notwithstanding his wide acquaintance with living and fossil reptiles, urged that their ophidian affinities were stronger than their lacertilian ones. Baur believed them to be only specialized and modified varanids; while Osborn has failed to see any marked varanid affinities in them, even suggesting that they did not arise from the same common stem. In 1892 Kramberger-Gorjanovicé’ described from the Lower Cretaceous (probably Gault) of the island Lesina, near the Dal- matian shores of the Adriatic, the remains of a remarkable lizard, which he called Azgzalosaurus. While his figures have recently been shown to be incorrect in some details, and while some of his minor interpretations were manifestly wrong, he correctly assigned to his new genus and its allies an intermediate position between the true Varanidae and the Mosasauridae, suggesting that they had descended from the Varanidae and were ancestral to the Mosasauria, while from them had been derived the dolicho- saurs as a side branch. Aus unseren Betrachtungen aber folgt ganz unzweideutig, dass man die erwahnten fossilen Gattungen weder der Familie Varanidae, noch direkt den Pythonomorpha zutheilen kann. Sie sind einergemassen Collectivtypen, welche an sich Merkmale zweier Unterordungen u. zw. der Lacertilia und Pythono- morpha tragen, ... . eine Uebergangsgruppe zwischen die Unterordnungen Lacertilia und Pythonomorpha zu stellen ist. His conclusions regarding the relationships of the Mosa- sauria are the more creditable from the fact that the only information available to him at that time concerning them was incomplete, and in part erroneous. Nevertheless, so apparent were they that both Boulenger and Dollo accepted them, recog- nizing in Azgialosaurus an ancestral type. More recently, Kornhuber? has described and figured, in an ™“Rad” der sitidslavischen Akademie der Wissenschaften und Kunst (Agram), Vol. CIX, pp. 96-123; Plate I, ff. 1,2 ; Plate Il, ff. 1, 2, 4. 2 Abhandlungen der k.-k. geologischen Reichs-Anstalt zu Wien, 1901. 46 a We AE LILIS LOIN excellent way, the remains of another lizard from Lesina, closely allied to Azgzalosaurus. So evident were the relationships of his Opetiosaurus to Varanus that he was inclined at first to unite them in the same genus, Unfortunately Kornhuber was appa- rently ignorant of the recent publications by Merriam, Osborn, and the writer upon the mosasaurs. Had he been more familiar with the structure of the mosasaurs, I doubt not that he would have recognized more clearly their relationships, and would have detected certain wrong determinations, clearly apparent in his figures, and recognized by Nopcsa—the supraorbital, nasals, columella, etc. There is no supraorbital in the mosasaur skull. To Nopcsa,? very recently, is due the credit for correctly estimating the value of the various annectant characters in these and the other known Cretaceous lizards. His views, though only amplifications and confirmations of those held by other writers, are supported by such undeniably forceful arguments that they cannot be gainsaid. Aigialosaurus and Opetiosaurus especially, with other Lower Cretaceous lacertilians, were clearly in a direct ancestral line between the mosasaurs and the undoubted direct ancestors of the modern monitors, thus confirming Baur’s views and disproving those of Osborn. These semi-aquatic lizards lived, evidently in abundance and in many forms, during the latter part of the Lower Cretaceous time, at least in southern Europe. The earliest mosasaurs of which we have any knowledge are -probably those from the Cenomanian of New Zealand, and it is not at all improbable that their birthplace was somewhere near the Medi- terranean Sea, at or near the close of the Lower Cretaceous. ‘Dollo long ago expressed the opinion that the center of disper- sion of the mosasaurs was somewhere in the vicinity of New Zealand, and his views were probably not far wrong. They reached northern Europe shortly after the beginning of the Upper Cretaceous, and North America before the beginning of the Senonian. Aigialosaurus and Opetiosaurus especially —for they are closely * Beitrage zur Paleontologie und Geologie Oesterreich-Ungarns und des Orients. Vol. XV.(1903), Plates V, VI. RELATIONS AND HABITS OF MOSASA URS 47 allied —so far as the structure of the skull is concerned, present a most striking resemblance to the mosasaurs, save perhaps in the absence of pterygoid teeth. They have the same peculiar form of the quadrate, similar teeth, similar frontal and parietal bones, nares, and, most remarkable of all, the same peculiar hinge in the lower jaws, unknown in other-reptiles. Were the vertebre and ribs to be found in the Kansas chalk unassociated with limb bones, they would almost without hesitation be referred to the mosasaurs. The hands and feet have the same strong resemblances in their form, but there is no hyperphalangy, a character but feebly indicated in some of the mosasaurs. In the elongated limb bones, and especially in the presence of a true sacrum, the real differences between these reptiles are seen. On the other hand, it is especially in those characters wherein they resemble the mosasaurs that they differ from the monitors. Like the mosasaurs, they have but seven vertebrez in the neck, a number found in no other scaled reptiles. It is very certain that they had webbed feet, though the claws had not dis- appeared. That the varanids are a very old group, dating at least as far back as the Jurassic, seems highly probable, especially so in the light of the recent discovery of an iguanid lizard by Broom in the Karoo beds of South Africa. As was long ago shown by Boulenger, Beddard, and others, the Varanidae are an isolated group among modern lizards, distinguished not only in the structure of their skeleton, but also by many other characteristic differences in their anatomy. In my opinion, there are no more striking examples of evo- lution presented in all vertebrate paleontology than that of the aquatic mosasaurs of the Upper Cretaceous, through the semi- aquatic aigialosaurs of the Lower Cretaceous, from the terrestrial varanoids of the lowermost Cretaceous or Upper Jura. That the snakes originated from the lizards much earlier than we have any definite knowledge would seem very probable. We have no certain evidence of them prior to the latter part of the Upper Cretaceous, that is, the Laramie; but, since it now seems almost certain that at least two distinct phyla of the lizards lead- 48 S. W. WILLISTON ing to modern forms had become well differentiated by the close of the Jurassic, or shortly thereafter, it is not at all unreasonable to suppose that the snakes had branched off as early as, or earlier than, the beginning of the Lower Cretaceous. The relationships of the chief groups of the Squamata may be expressed as follows: Order Squamata (Lepidosauria). Suborder Sauria (Lacertilia). Superfamily Platynota. Family Varanidae Gray. Pleistocene—recent. Genus Varanus Merrem. India, Australia, Africa. Family Dolichosauridae (Kramb.) Nopcsa. Cretaceous. Genera: Dolichosaurus Owen, Acteosaurus Meyer, Pontosaurus Kramb., Adriosaurus Seeley. Family Aigialosauridae (Kramb.) Nopcsa. Lower Cretaceous. Genera: AZzgialosaurus Kramb., Carsosaurus Kornh., Opetiosaurus Kornh., Mesoleptus Carnalia. Superfamily Mosasauria. Family Mosasauridae Gervais. Upper Cretaceous. Genera: Mosasaurus Conyb., Clidastes Cope, Platecarpus Cope, Tylosaurus Marsh, Baptosaurus Marsh, Plioplatecarpus Dollo. Hlainosaurus Dollo, Prognathosaurus Dollo, Phosphorosaurus Dollo, Brachysaurus Williston. Superfamily Kionocrania (true lizards). Trias— recent. Superfamily Amphisbaenia. Oligocene — recent. Superfamily Rhiptoglossa (chameleons). Recent. Suborder Serpentes (Ophidia), Laramie Cretaceous— recent. It is a remarkable fact, for which there has never been any adequate explanation, that the mosasaurs are wholly unknown in the juvenile condition. Altogether, throughout the world, more than three thousand specimens of these animals have been brought to light during the past century, and every one of them, so far as my own knowledge goes—and I have seen many hun- dreds—and so far as anything that has been published would indicate, is supposed to be of an adult or fairly mature animal. Not a single specimen which can be suspected to be embryonic is known. I hardly think that the same can be said of any other group of extinct aquatic animals. In the Cretaceous of Kansas the young of the plesiosaurs are fairly common, perhaps every third or fourth specimen showing immaturity. I have found RELATIONS AND HABITS OF MOSASA URS 49 them in almost every stage of adolescence, and the embryos have been discovered in Europe. Furthermore, I have often sought in vain for the remains of young mosasaurs in the stom- ach contents of different animals in the Kansas Cretaceous, and in the bone-beds, which rarely occur in that formation. Under such apparently favorable circumstances as are presented by the chalk deposits of western Kansas, the seemingly entire absence of all remains of young mosasaurs is inexplicable. Half-grown forms do occur, but none that are very young. As I have pre- viously remarked, it is certain that all mosasaurs did not die of old age. Indeed, the many hyperostosial mutilations of ante- mortem origin indicate only too well the fierce struggles the mosasaurs had with the carnivorous enemies of their own and other kinds. If the mosasaurs were exclusively marine animals, it would seem almost certain that they were not viviparous, as were the ichthyosaurs, and probably also the plesiosaurs, and as are some modern lizards. As Fraas has remarked concerning the European ichthyosaurs, if one searches carefully, he will find in many Kan- sas specimens the remains of the skin and stomach contents, but never has there been found anything which has the faintest suggestion of mosasaur embryos. No aquatic reptiles of the pres- ent time lay their eggs in the water. The sea snakes are vivipa- rous, or at least all available information concerning them gives viviparous habits. The sea turtles and the crocodiles lay their eggs upon the beaches, the latter guarding their nests and young. Doubtless the crocodiles of the past had the same beach-laying habits, suggesting that they were never inhabitants of the open oceans. The mosasaurs must have been practically helpless upon land; stillit is notimpossible that they may have frequented the beaches for the deposition of their eggs, though it is highly improbable that they gave any attention or care to either their eggs or their young. That the eggs of the mosasaurs were more numerous than are those of the terrestrial lizards of the present time is not to be supposed. The waters in which the mosasaurs flourished swarmed with highly predaceous fishes, sharks, and plesiosaurs, to say nothing of the hordes of their own kind; and, 50 S. W. WILLISTON unless the eggs were very numerous, or unless the young were jealously guarded by the parent, the young reptiles must have stood very little chance in the fierce struggle for existence. It is, of course, possible that the shallow waters of the bays and estuaries may have afforded sufficient protection for the young mosasaurs, but this is doubtful, in the entire absence of all remains of such animals in marine deposits. It seems more prob- able that the mosasaurs were brought forth, perhaps alive, in fresh water, that the females ascended the rivers to breed, and that the young remained in such protected places until fairly able to care for themselves. That the mosasaurs, as also the aigial- osaurs, were in part denizens of fresh water may be, perhaps, one reason for the great relative abundance of their remains in the deposits of inland or protected seas. Their occurrence in Kan- sas associated with great numbers of small turtles, pterodactyls, and birds would seem to be fairly good evidence that they were more littoral than pelagic in their habits. That mosasaur remains have never been found in fresh-water deposits is not at all conclusive in controversion of this hypoth- esis. We know very little indeed, if anything at all, of the fresh- water life that existed during the times when the mosasaurs flourished. Possibly the Belly River deposits may furnish some evidence bearing on this hypothesis, though I suspect that the mosasaurs had disappeared in America before the time of those deposits. It will be of interest to record here certain additional facts concerning the structure of the mosasaurs. A most extraordi- narily complete specimen, referred to Holosaurus abruptus Marsh, discovered by Mr. E. B. Branson in western Kansas the past summer, has a slight emargination of the coracoids, finally: and conclusively demonstrating the invalidity of this character in the separation of the genera. The scales of this form, as apparently also of other species of Platecarpus, are without carina. The pubes meet ina perfect symphysis, not as they are always figured in text-books. The ilia were suspended vertically, and the depth of the body posteriorly is greater than the restora- tions indicate. The upper ends of the ilia were in relation, not RELATIONS AND HABITS OF MOSASA URS 51 with the first of the pygal vertebra, as I have contended, but with the second or third at least, as Dollo believed. There were more than ninety vertebre in the tail. The skeleton measured more than twenty feet in length as it lay in the chalk, with nearly every bone in its proper position. The entire skeleton has been brought to the laboratory in chalk slabs, and when it is finally prepared will, I believe, add a number of new facts to our already full knowledge of these remarkable animals. S. W. WILLISTON. REVIEWS SUMMARIES OF PRE-CAMBRIAN LITERATURE FOR 1902-1903. I. [ Continued from Vol. X, p. 913.| (Gg TGS iro) ae Tet ARTHUR KEITH. ‘‘ Description of the Cranberry Quadrangle of North Caro- lina and Tennessee.” Geologic Atlas of the U. S., Cranberry Folio, No. go, U.S. Geological Survey, 1903, pp. I-9. Keith describes and maps the geology of the Cranberry quadrangle of North Caro- lina and Tennessee, along the junction of the Piedmont Plateau and Blue Ridge. Archean, and doubtful Algonkian, rocks occupy all but the northwest corner of the area. The Archzan rocks are mapped and described under the heads: Carolina gneiss, Roan gneiss, soapstone, Cranberry granite, Blowing rock granite,and Beech granite. The Carolina gneiss is the oldest rock of the ridge and consists of interbedded mica-schist mica-gneiss, and fine granitoid layers. The Roan gneiss consists of hornblende- gneiss, hornblende-schist, diorite, with some interbedded mica-schist and gneiss, all cutting the Carolina gneiss. Soapstone, resulting from the alteration of peridotite and pyroxenite, occurs in bodies closely associated with the Roan gneiss and probably of the same age. The Cranberry granite is the most extensive formation in the district, occurring chiefly in the mountain districts. It consists of granite and of schist derived from granite, and cuts the Roan gneiss and Carolina gneiss. All of the before-named rocks are cut by the Blowing rock gneiss and the Beech granite, which are considered to be of the same age. Four formations are classed as doubtful Algonkian. These are: Linville meta- diabase —an altered greenish diabase and gabbro; Montezuma schist—a blue and green epidotic schist, probably altered basal, and amygdaloidal basalt; Flattop schist, — a gray and black schist, probably altered andesitic rocks ; meta-rhyolite — a grayish meta-rhyolite and rhyolite porphyry. The first of these appears to be the lower part of a surface flow, and the last three are of surface volcanic nature. T.L. Watson. “ Copper-bearing Rocks of Virgilina Copper District, Virginia and North Carolina.’ Pudletin of the Geological Society of America, Vol. XIII (1902), pp. 353-76. Watson describes the copper-bearing rocks of the Virgilina copper district of Virginia and North Carolina and shows the adjacent rocks to be pre-Cambrian meta- morphosed andesite associated with corresponding volcanic clastics. All are col- lectively referred to as greenstones, and are thought to be similar to greenstones described as occurring along the Atlantic coast region from eastern Canada to Georgia, and from Alabama to the Lake Superior region. 52 REVIEWS 53 ARTHUR KEITH. ‘Geology of the Piedmont Plateau Area of the Washing- ton Quadrangle.” Geologic Atlas of the U. S., Washington Folio, No. 70, U. S. Geological Survey, IgoI, pp. 2, 3. Keith describes and maps the geology of the Piedmont Plateau of the Washington quadrangles. Igneous rocks of Archean age are mapped under the following headings ; biotite-granite, soapstone and serpentine (altering trom peridotite and pyroxenite), gabbro, meta-gabbro, diorite and meta-diorite (including granite, gneissoid granite and schistose granite), and Carolina gneiss (including mica-gneiss, mica-schist, and small bodies of granite, schistose granite, and diorite). Inage these rocks rank in the order named, the Carolina gneiss being the oldest. Also the relative areas of the groups nearly correspond with their ages. T. NELSON DALE. ‘Structural Details in the Green Mountain Region, and in Eastern New York.” Azdlletin of the U. S. Geological Survey, No. 195, 1902. Dale sketches structural details in the Green Mountain district and in eastern New York, suchas folds, cleavage, joints, and faults, some of them in the pre-Cambrian rocks. F. J. H. MERRILL. ‘ Metamorphic Crystalline Rocks of the New York Quadrangle. Geologic Atlas of the U. S., New York City Folio, No. 83, U.S. Geological Survey, 1902, pp. 3-5. Merrill describes the metamorphic crystalline rocks of the New York quadrangle. Of these only one, the Fordham gneiss, is of pre-Cambrianage. The petfography of this gneiss is described. No opinion is expressed as to its sedimentary or igneous origin or as to its Algonkian or Archean age. C. H. SMyTu, Jr., AND D.H.NEWLAND, “ Report on Progress Made During 1898, in Mapping the Crystalline Rocks of the Western Adirondack Region.” Eighteenth Annual Report of State Geologist for 1898 pub- lished in F2fty-second Annual Refort of the New York State Museum (for 1898), Vol. II, 1900, pp. 129-35. Smyth and Newland report progress in the mapping of the crystalline rocks of the western Adirondack region. Inclusions of hornblende schists found in the more acid gneisses of the region are believed to afford important evidence as to the origin of the gneisses. Also light red granitoid gneiss has been found intrusive into a gray gneiss, indicating, as before held, that all the gneisses are not of the same age, Certain of the gneisses are found to be younger than, and intrusive into, certain schists associated with the limestones. H. P. CusHInG. ‘Recent Geologic Work in Franklin and S. Lawrence Counties, N. Y.” Nineteenth Annual Report of State Geologist for 1goo, published in /ifty-third Annual Report of the New York State Museum, Vol. I, 1902, pp. r 23-95. Cushing discusses recent geological work in Franklin and St. Lawrence counties, N. Y., and concludes : 1. That the Adirondack anorthosite is cut intrusively by an augite syenite, which is therefore younger. 54 REVIEWS 2. That, while the larger part of the augite syenite of the Adirondacks is in such situation with respect to the anorthosite as to render impossible any determinations of relative age, its general character is so uniform throughout that it is exceedingly prob- able that it is all of the same approximate age and consists of intrusions from the same source. 3. That at their borders these syenites pass over into granites, part of which at least cut the syenite eruptively, and are therefore younger. 4. That the syenite grades into granite on the one hand, and into gabbro diorite onthe other, and apparently into anorthosite as well. 5. That the three together, anorthosite, syenite, and granite, form a great eruptive complex in the heart of the Adirondack region, and that all are younger than the (in part at least) sedimentary Grenville rocks. H. P. CusuHinc. ‘“ Pre-Cambrian Outlier at Little Falls, Herkimer Co., N. Y.” Nineteenth Annual Report of State Geologist for 1900, published in Fifty-third Annual Report of New York State Museum, 1900, pp. 183-95. Cushing describes a pre-Cambrian outlier at Little Falls,in Herkimer county, N. Y., and points of difference with the syenite of the Adirondacks. A. W.G. Witson. “The Laurentian Peneplain.’’” JOURNAL OF GEOLOGY, Vol. XI (1903), pp. 615-69. Wilson describes the Laurentian peneplain of the great pre-Cambrian shield of Canada and adjacent portions of the United States. ‘lhe peneplain is an ancient one, which has undergone differential elevation, has been denuded, and subsequently slightly incised around the uplifted margin. At several places on the margin, as exposed today, the dissection may be regarded as submature. The date of the major development of the peneplain is not determined, but may be pre-Ordovician or Creta- ceous. Around the southern margin between Montreal and Winnipeg there are traces of a peneplain (or probably more than one) of still earlier date, upon which pale- ozoic sediments were laid down, and which has been uncovered by processes of degra: dation and denundation since the differential uplift of the latest peneplain. S. WEIDMAN. “The Pre-Potsdam Peneplain of the Pre-Cambrian of North- Central Wisconsin.’ JOURNAL OF GEOLOGY, Vol. XI (1903), pp. 289-313. Weidman describes a pre-Potsdam peneplain of the pre-Cambrian of north-cen- tral Wisconsin and shows the same to slope gradually to the south, where itis covered by Paleozoic sediments. Several monadnocks stand above the pre-Potsdam peneplain. Extensive clay deposits near the contact of the Paleozoic and the pre-Cambrian are believed to have developed during the pre-Potsdam base-leveling. Comment.— The peneplain described by Weidman is perhaps to be correlated with the pre-Paleozoic peneplain described by Wilson as appearing about the periphery of the great pre-Cambrian area of Canada, with a slope inclined to the great pene- plain of the pre-Cambrian interior. It is of interest also to note that evidence of pre- Cambrian base-leveling has been described by Crosby near Manitou, Col., and that Crosby has called attention to the widespread occurrence of such a plain in North -America.* rW. O. Crossy, “ Archzean Cambrian Contact Near Manitou, Col.” zdlletin Geo- logical Society of America, Vol. X (1899), pp. 141-64. REVIEWS 55 REGINALD A. Day. “Variolitic Pillow-Lava from Newfoundland.” Amert- can Geologist, Vol. XXXII (1903), pp. 65-78. Daly described variolitic pillow-lava from Newfoundland, and calls attention to the widespread occurrence of this or similar rocks, frequently called ellipsoidal green- stones, in Minnesota, New Brunswick, California, and Michigan. ’ R. W. Etxts. “The Progress of Geological Investigation in Nova Scotia,’ Proceedings and Transactions of the Nova Scotian Institute of Science, Vol. X, Part 4 (Ig0I—1902), pp. 433-46. Ells summarizes the progress of geological investigation in Nova Scotia. L. C. GRATON. ‘On the Petrographical Relations of the Laurentian Lime- stones and the Granite in the Township of Glamorgan, Haliburton County, Ontario.”” Canadian Record of Science, Vol. 1X (1903), pp. 1-38. Graton describes in detail the petrographical relations of the Grenville limestones and granite in the township of Glamorgan, Haliburton county, Ontario. His conclu- sions are of importance as bearing on the relations of limestones and gneisses over other extensive areas in eastern Canada, the Adirondacks, and New Jersey. He summari- zes his conclusions as follows: The district exhibits a development of Grenville limestone pierced by intrusions of gneissic granite which contain masses of dioritic rock. Considerable deformation took place during the intrusion. Between the limestone and the granite is a highly brecciated zone, holding large amounts of lime-rich silicates which are eminently characteristic of contact metamor- phism. ; Diagenesis took place. _ Toa great extent, however, the elements, other than the lime necessary for the formation of these minerals, came from the intrusion and its accompanying exhala- tions. The metamorphism, then, was largely also metasomatic, In the gray gneisses and in the granite are dark basic masses which represent fragments broken off from the limestone series and floated away into the igneous mass. ‘They have been still more highly metamorphosed than the rocks from which they came, and have been more or less dissolved and changed in character by the granite. In other words, they have been partially “ granitized.” The gray gneisses, which have the composition of quartz diorites, may represent an intermediate phase of this “ granitization’’—between the inclusions and the granite. This theory’ may account for the large amount of plagioclase feldspar found in the gran- ite itself. R.W.ELuLs. “ Report on the Geology of Argenteuil, Ottawa and Part of Pon- tiac Counties, Province of Quebec, and Portions of Carleton, Russell and Prescott Counties, Province of Ontario.”” Axnual Report of Geological Survey of Canada, Vol. XII (1899), New Series, pp. I j—138}. Ells maps and describes the geology of Argenteuil, Ottawa, and part of Pontiac counties, province of Quebec, and portions of Carleton, Russell, and Prescott counties, province of Ontario, covering most of what has long been known as the Original Lau- rentian district, and summarizes previous work in this district. Archean rocks, 56 REVIEWS occupying most of the area north of the Ottawa River, are mapped as crystalline lime- stone, gneiss, quartzite, anorthosite, granite-gneiss and porphyry. In the text the limestones with the quartzites and gneisses associated with them are described as sed- imentary and are classed as Grenville, and the underlying gneisses and granite-gneisses are described as of igneous origin and are called Fundamental complex —in this clas- sification following previous writers. It is evident that the rocks of the Grenville series are decidedly newer than those of the Fundamental division. As forthe numer- ous and often large area of red granite-gneiss, many of these are undoubtedly of more recent date than either of the others since they clearly cut both the gneiss and limestone. While in some points the newer granite-gneiss presents features similar to the Fundamental division, as in the foliation of certain portions, there is, over large areas, a marked difference in their aspect in the field. A. OSANN. “ Notes on Certain Archzean Rocks of the Ottawa Valley.” Annual Report of the Geological Survey of Canada, Vol. X11 (1899), New Series, pp. 10-840. Osann makes a detailed petrographic description of the crystalline rocks in the Original Laurentian area of the Ottawa Valley. WILLET G. MILLER. “Lake Temiscaming to the Height of Land.” Report of the Bureau of Mines, Ontario, 1902, pp. 214-30. Miller publishes geological notes taken on a canoe trip from Lake Temiscaming northward to the height of land. Special attention was paid to occurrence of miner- als of commercial value, and no mapping was attempted. He finds various kinds of igneous rocks, both plutonic and volcanic, such as granite, syenite, diorite, olivine diabase, quartz-porphyry, and others of less importance. In addition to these, most of the metamorphic fragmental rocks characteristic of the Huronian occur, among which may be mentioned quartzite, slate graywacke, and different varieties of the pyroclastic series, ash rocks, and agglomerates. The popular belief that the height of land in this district represents the highest point of the surface from which sediment was derived for the formation of deposits of later age which lie to the southward is scarcely based on fact. He found what appear to be thick deposits of Huronian conglomerate and other water-formed material resting on the surface close to the height of land. It is evident from this that the surface level must have changed con- siderably since Huronian times, and that what is now the height of land may have once been a comparatively low-lying area. L. L. Botton. ‘Round Lake to Abitibi River.” Report of the* Bureau of Mines, Ontario, 1903, pp. 173-90. Bolton reports on the geological reconnaissance from Round Lake north to the Abitibi River in the district of Nipissing. Laurentian granite was seen near both the southeastern and southwestern corners of Eby. Elsewhere Huronian rocks are exposed. Of these there is a considerable variety, many of which are of fragmental origin. The following types were seen: diorite, diabase, brecciated conglomerate, slate, graywacke, hornblende schist, etc. As the rock outcrops of the district explored are, as a rule, separated by areas of sand, swampy, or clayey soil, the relations of the different types could seldom be worked out. REVIEWS 57 A. P. COLEMAN. “The Sudbury Nickel Deposits.” Report of the Bureau of Mines, Ontario, 1903, pp. 235-303. Coleman describes and maps the nickel deposits near Sudbury, Ontario, and incidentally discusses the geology of the region. The probable succession and age of ‘the rocks of the district is as follows, in ascending order: Dikes of diabase. Keweenawan(?) , Younger Granite. Nickel-bearing eruptive; norite; micropegmatite; granite. Animikie (?) or Upper Huronian (?) — Oval area of tuffs, sandstones, and slates overlying the preceding. Laurentian.—Granitoid gneiss. Green schists and greenstones. Upper Huronian Arkoses, quartzites, and graywackes. It can hardly be said that the precise age of any of these groups of rocks is known, though they probably range from the base of the Upper Huronian to the Keweenawan, including the Laurentian as later than the Upper Huronian. No rocks undoubtedly of Lower Huronian age are known from the nickel district proper ; though the ranges of banded silica and magnetite extending through Hutton and Wisner townships to the north of the nickel area evidently belong to the Lower Huronian.t!' The latter rocks occur entirely inclosed, so far as known, in granites and gneisses, generally considered Laurentian, and have not been found in direct connec- tion with the rocks here described. The fact has been brought out that all of the nickel deposits are either on the basic edge of a great eruptive band, which at the opposite edge becomes a quartz syenite or granite, or in dike-like offshoots, often, however, interrupted by other rocks projecting from the southeastern basic edge of the great gabbro band. This band has been found to outcrop in a great oval, the north and south sides of which have been known respectively as the North and South nickel ranges. The structure is synclinal, and the center is occupied by Animikie or Upper Huronian rocks. There are two different types of deposits represented in the mines of the district : those along the southeastern margin of the main range, often crowded into bay-like indentations of the adjoining rock; and those strung out along the narrow off-shoots from the main range, as Peters suggests, ‘‘like sausages on a string, but with a long piece ot string between the sausages.”* Among the former class are the Creighton, Gertrude, Elsie, Murray, and Blezard mines; among the latter, the Copper Cliff, Evans, Frood and Stobie, and the Victoria and Worthington mines. Perhaps a third variety should be distinguished for the Vermilion mine, which contains rich nickel and copper ores, but has no visible association with a band of gabbro, having, however, been formed probably by hot circulating fluids proceeding from such a band. The final impression left is that the marginal type of deposit is in the main of plutonic origin, aqueous work being relatively unimportant; that in the offset type plutonic is generally more important than aqueous action, though one example, that of the Worthington, suggests more complete rearrangement of the materials by circulating water; thus forming a transition to ordinary vein deposits wholly due to water action, as at the Vermilion mine. t Report of the Bureau of Mines, 1901, p. 186. 2? Mineral Resources of Ontario, p. 104. 58 REVIEWS C. K. Leiru. ‘Moose Mountain Iron Range.” SRefort of the Bureau of Mines, 1903, pp. 318-21. Leith describes the Moose Mountain iron range in the township of Hutton and district of Nipissing. Iron formation consisting of magnetite, of banded magnetite and quartz, and of magnetite, associated with amphibole and epidote, occurs in bands and lenses in a complex of basic igneous rocks characterized by uniform abundance of amphiboles. Some of the greenstones are basal and some intrusive into the iron- bearing bands. Intrusive into the greenstone and probably into the iron formation are granite masses. Closely associated with the iron formation, but with relations unknown, is a pyritiferous graywacke. The ores and associated rock as a whole are in general similar lithologically to the Vermilion iron-bearing district of Minnesota, although showing many points of difference. Comment.—Further field work in 1903 in adjacent areas indicates that a great graywacke and conglomerate series rests unconformably against the rocks of the iron range, thus adding another point of similarity of this range to the Vermilion iron range of Minnesota. L. C. Graton. “Up and Down the Mississaga.” efort of the Bureau of Mines, Ontario, 1903, pp. 157-72. Graton reports on a geological reconnaissance along the Mississaga River and east and west along Niven’s baseline in the district of Algoma. Laurentian granites occupy all of the area north of township 188, where was found a greenish slate con- glomerate belonging to the Huronian. W.G. MILLER. “Iron Ranges of Northern Ontario.” Refort of the Bureau of Mines, Ontario, 1903, pp. 304-17. Miller gives a résumé of the occurrence of iron ore in northern Ontario, and incidentally discusses their geological relations. A. P. COLEMAN. “Iron Ranges of Northwestern Ontario.” Report of the Bureau of Mines, Ontario, 1902, pp. 128-51. Coleman gives results of an examination of the iron ranges of northwestern Ontario, principally the Mattawan, Atikokan, Steep Rock Lake, and other districts along the Canadian Northern Railway, the Slate Islands in Lake Superior, and near Dryden on the Canadian Pacific. The description of the details of the districts contains but few references to general stratigraphy and correlation, but at the end a general classification of the iron ores of Canada is given. ‘To the upper part of the Lower Huronian (Archean of the U. S. Geological Survey) are referred the siliceous and sideritic iron ranges occurring in practically every iron-bearing area in Ontario, but being mined at only one place, at the Helen mine in the Michipicoten district. To the lower part of the Lower Huronian are referred the magnetite lenses in green schists of the Atikokan district and the titaniferous magnetite, occurring as segrega- tions in basic eruptives, especially gabbro. Yo the Grenville series “probably Huronian” are referred the magnetite and hematite ores associated with bands of crystalline limestone and gneiss of eastern Ontario. Tothe Animikie or Lower Huronian (Upper Huronian of the U.S. Geological Survey) are referred impure siderite and hematite occurring in the neighborhood of Thunder Bay and also near Algoma. To the Pleistocene are referred the bog and lake ores and postglacial magnetic sands occurring widely in Ontario and especially in the eastern part. REVIEWS 59 A, P. COLEMAN. ‘Nepheline and Other Syenites Near Port Coldwell,” Ontario. American Journal of Science, Vol. CLXIV (1902), pp. 147-55. See also Report of Ontario Bureau of Mines for (1902), pp. 208-13. Coleman describes nepheline and other syenites near Port Coldwell, Ontario, and calls attention to their widespread distribution in Ontario and the United States. W.G. MILLER. ‘‘Nepheline Syenite in Western Ontario,’ American Geologist, Vol. XXXII (1903), pp. 182-85. Miller describes bowlders of nepheline syenite near Sturgeon Lake, northwest of Lake Superior, indicating the occurrence of rocks of this character in the pre- Cambrian rocks farther north. R. G. MCCONNELL. ‘Note on the So-called Basal Granite of the Yukon Valley.” American Geologist, Vol. XXX (1902), pp. 55-62. McConnell describes the granite gneiss of the upper part of the Yukon Valley, extending from the Nordenskiold River in a northwesterly direction across the White River valley to the Tanana, and down this stream to near the mouth of the Delta River —a total distance of about 380 miles—-and concludes that a part of the gneisses at least must be regarded as intrusive through, and therefore younger than, the clastic schists associated with them. It is still possible, however, as the work done so far has been largely of an exploratory character, that older gneisses may be present in the district, but no evidence of this was obtained in the course of the investigation. O. H. HERSHEY. ‘Structure of the Southern Portion of the Klamath Mountains, California.” American Geologist, Vol. XXXII (1903), pp. 231-45. Hershey discusses the structure of the southern portion of the Klamath Mountains, of California. The oldest rocks in the mountains west of the Sacramento River are the Abrams mica-schists, 1,000 feet thick, and overlying it the Salmon hornblende-schist, known to be at least 2,500 feet thick, both of them supposed to be of pre-Cambrian age, probably Algonkian, and possibly Archean. The Abrams mica-schist is a sedimentary rock, and the Salmon hornblende-schist is a metamorphosed volcanic ash. The Klamath schists form the central ridge of the Klamath region. They are bordered on the west by a great, unsymmetrical geosyncline, and on the east by the western limb of another great geosyncline. The first geosyncline is limited on the west by another belt of schist, chiefly the Abrams mica-schist, which forms the South Fork Mountain and is prolonged northwestward to and probably across the Klamath River near Wichiper. The sandstones of the Coast Range region adjoin this schist belt on the west. According to Mr. Diller, toward the north, approaching the Klamath River, long narrow belts of schist alternate with narrow belts of sandstone, the latter dipping eastward as though going under the schists. This apparent anomaly is evidently due to a series of faults. It is further evident that the Coast Range formations have buried the western portion of the schist belt which may extend, immediately under the sandstone, far toward the coast. The eastern schist belt emerges from beneath the Cretaceous sandstones and shales in the Sacramento valley west of Ono, with a width of eight miles and 60 REVIEWS gradually increases as it advances northward to a maximum of about twelve miles west of Scott Valley. Southward from the Trinity River, the pre-Paleozoic area is occupied chiefly by the Abrams mica-schist, the hornblende-schist being confined to narrow strips, but northward from the Trinity River the hornblende-schist spreads out and finally nearly excludes the mica-schist as in the valley of the South Fork of the Salmon River. Still farther north, in the mountains west of Scott Valley, the mica-schist has again asserted its supremacy. O H. HERSHEY. ‘Some Crystalline Rocks of Southern California.” 4 mer?- can Geologist, Vol, XXIX (1902), pp. 273-90, Hershey describes the results of a brief examination of the Fraser Mountain and Sierra Pelona regions, and portions of the Tehachapai, Sierra Madre, and San Bernardino ranges, together with quite an extended section of Mohave desert, all comprised in the counties of Los Angeles, Ventura, Kern, and San Bernardino of California. The crystalline rocks are discriminated under the following heads: (1) “The Pelona Schist Series;” (2) “The Gneiss Series;” (3) “The Rocks of Fraser Mountain and Vicinity;”’ (4) “The Mesozoic Granites;” (5) “The Ravenna Plutonic Series;” (6) “The Gneiss Near Barstow;” (7) ‘“ The Quartz- ite-Limestone Series of Oro Grande;”’ (8) ‘‘ The Schists in Cajon Pass.” The Pelona schist series and the adjacent gneisses, the rocks of Fraser Mount- ain and vicinity, and the gneiss near Barstow are tentatively correlated with the Abrams schist of the Klamath region in a general way, and are considered pre- Paleozoic, perhaps in part Archzean and in part Algonkian. W. LinDGREN “The Gold Belt of the Blue Mountains of Oregon.” Zwen- ty-second Annual Report of the U. S. Geological Survey, Part II, 1g00- 1901, pp. 551-776. Lindgren describes and maps the geology of the gold belt of the Blue Mountains of Oregon. Gneiss, referred to the Archzean, occurs northwest of Blue Mountain above La Belleview mine. BAILEY WILLIS. ‘Stratigraphy and Structure, Lewis and Livingston Ranges, Montana.” Budletin of the Geological Society of America, Vol, XIII (1902), pp. 305-52. Willis describes and maps the stratigraphy and structure ot the Lewis and Liv- ingston Ranges of the Front Range of the Northern Rocky Mountains of Montana and Alberta. Lewis and Livingston Ranges consist of stratified rocks of Algonkian age, as determined on fossils which were found by Weller in the lowest limestone of the series, and identified by Walcott as probably being Beltina danai, the species of crus- tacean discovered in the Grayson shales of the Belt Mountains. The Algonkian series consists of limestone, argillite, and quartzite, classified in five formations, and aggre- gating about 12,500 feet in thickness. The formations are the Kintla argillite, Sheppard quartzite, Siyeh limestone, Grinnell argillite, and Appekunny argillite. There is apparent conformity throughout. The series is so situated with reference to other rocks that no lower or upper stratigraphic limit could be determined. Dr. G, M. Dawson classified with strata as Cambrian, Carboniferous, and Triassic, but it is believed that he mistook certain local overthrust faults for unconformities, and was misled by lithologic resemblances. REVIEWS 61 Igneous rocks occur sparingly in the Algonkian series. An intrusive sheet of diorite is extensive in the upper limestone formation and an extrusive flow of diabase caps it. The Algonkian strata form a syncline whose axis trends west of north. South- western dips vary from 5° to 30°. Northeastern dips are generally 30° to 40°, and locally approach or pass verticality. Minor flexures within the syncline are very broad and low. The northeastern limit of the fold is an eroded margin; the southwestern is an anticlinal axis whose western limb is in part eroded, in part thrown down by a normal fault along North Fork Valley. Syncline and anticlines are closely related to valley and ridge respectively, and this relation extends to heights of peaks. Along its eastern margin the oldest Algonkian formation rests upon Cretaceous rocks. The outcrop of this abnormal contact is deeply sinuous throughout the stretch from Saint. Mary Lake to Waterton Lake. The structure is described as an over- thrust fault, on which the Algonkian series has moved northeastward relatively over the Cretaceous rocks. The displacement on the thrust surface is 7 miles or more, and the vertical throw is estimated at 3,400 feet or more. The thrust surface dips from 0° to 10° southwestward, and strikes variously from north to north 60° west. Thus it is warped, and this warping is found to determine the general outline of the eastern face of the . Rocky Mountains, particularly the prominence of Chief Mountain, and the relative position of the Lewis Range, en echelon to the Livingston. W.H. WEED. “Geology and Ore Deposits of the Elkhorn Mining District, Jefferson County, Montana.” Twenty-second Annual Report of the U.S, Geological Survey, Part II, 1g00—Ig01, pp. 399-549. Weed describes and maps the geology of the Elkhorn mining district of Montana. Doubtfully referred to the Algonkian are the Turnley hornstones. The lower division is 200 feet thick and consists of shale metamorphosed to a very dense hornstone com- posed of light brown biotite and quartz. A bed of impure iron ore 20 to 30 feet thick occurs in the middle lower part of the formation. The quartzitic hornstones overlie the basal beds just noted and are 200 feet thick. Vhe rocks, though well bedded, are very dense and hard, and are of a gray-black color, so that they closely resemble the andesites. In color, composition, and relation to the overlying quartzite the rocks correspond to the red Spokane shale of the Belt terrane seen at Whitehall, 20 miles south, at Townsend to the east, and at Helena on the north. W. S. TANGIER SMITH. “Geology of the Hartville Quadrangle of Wyom- ing.” Geologic Atlas of the U. S., Hartville Folio, No. 91, U. S. Geological Survey, 1903, pp. 1-6. Smith (W. S. Tangier) describes the geology of the Hartville quadrangle of Wyoming. The Whalen group, assigned to the Algonkian, consists of gneisses, schists, quartzites, and limestones, all very schistose, the schistosity standing nearly vertical. These rocks occur principally in the northeastern part of the quadrangle. Quartzites and micaceous schists form the greater part of the exposed rocks of the Whalen group, and in places they grade into each other, so that no definite separation can be made. Some of the quartzites are more or less calcareous. Iron ore occurs within and near the contact of the limestones and schists of the Whalen group on the west side of Whalen Canyon. Information at hand is not sufficient to decide whether there are several ore-bearing horizons or a single horizon repeated by folding. Ore is being mined at Sunrise. 62 REVIEWS F. L. RANSOME. ‘Ore Deposits of the Rico Mountains, Colorado. Twenty- second Annual Report of the U. S. Geological Survey, Part II, 1go0— 1901, pp. 229-397. Ransome describes the ore deposits of the Rico Mountains, Colorado, and inci- dentally summarizes the geology of the area, referring the reader to a previous paper by Cross and Spencer? for further details. The Algonkian rocks consist of quartzites and schists, exposed just north of Rico and in the canyon of Silver Creek. They appear as fault blocks, in the heart of the dome, thrust up from below into the later beds. WHITMAN Cross. ‘Geology of the Silverton Quadrangle, Colorado.” From Bulletin No. 182, U. S, Geological Survey, 1901t, by F. L. RANSOME. Cross summarizes the geology of the Silverton quadrangle, Colorado, for Ransome’s bulletin on the economic geology of this quadrangle. Algonkian quartzites and schists appear beneath the volcanics where the Animas River and the Uncompahgre River and its tributaries cut through the volcanics. Irngation. By F.H.Newery. New York: T. Y. Crowell & Co. Mr. NEWELL’s book is written from the economic and social point of view, and emphasizes the importance of irrigation as a national problem. For this reason the general reader, as well as the person directly concerned in home-making in the arid West, will read the work with interest. Guided by his extensive experience with problems of irrigation, the author contrasts the present scanty occupation of the western two-fifths of the United States with the possibilities for home- making when the present water’ supply shall have been properly con- served. At the present time 7,300,000 acres are under irrigation, while the natural water supply is sufficient for ten times that acreage. The success already attained in this small fraction of the area abundantly justifies the national expense already incurred, and becomes the basis for urging national aid in bringing greater areas under irrigation. Cer- tainly the addition of 60,000,000 acres, equivalent to two states the size of Pennsylvania, to the present productive area of the public domain is an expansion in the right direction. The fact that these lands capable of irrigation are distributed, oasis like, through regions which must always yield but scanty returns, and that these areas have a calculable productivity equal to the best land in humid states, are «WHITMAN Cross AND A. C. SPENCER, “Geology of the Rico Mountains, Colo- rado,” Twenty-first Annual Report of the U.S. Geological Survey, Part I1, 1899-1900, pp. I-165; summarized in JOURNAL OF GEOLOGY, Vol. X (1902), p. 910. REVIEWS 63 convincing arguments for national control, and for protection from speculative monopoly. The author lays much stress upon the fact that problems arise from irrigation that cannot be successfully handled by individuals or even by states. The setting apart of forest lands for the regulation of the water supply; the building of reservoirs for impounding the head- waters of streams; the adjusting of water rights on streams that cross state lines; the establishing of experiment stations ; and the investiga- tion of a wide range of conditions of water supply and the adaptation of crops to climate and soil—these are subjects for an authority which can act in a disinterested way for all concerned. As the book is intended for popular reading, it is in no sense a manual, though the practical man will find that fundamental principles have been so clearly stated, and happily illustrated by photographs and diagrams, that he can judge intelligently concerning his own particular conditions, and avoid expenditure on ill-advised schemes. The manner in which the author deals with the questions of artesian water, the building of dams and ditches, the use of windmills as a source of power, the methods of measuring water, and the means of conducting water to land in a great variety of situations, must appeal to the common-sense of every practical farmer. Sixty-two plates and ninety-four diagrams admirably supplement the lucid text. Among the cartograms are a number that show in a striking way the relative size of western states as compared with the Atlantic states, and are well calculated to impress the reader with the vastness of the area with which the book deals. If the book were supplied with definite references to the wide literature of the field, it would be of more use to students; but as it is, it furnishes an excellent introduction to the subject. te Ee Woon: Gems and Gem Minerals. By OLivER CUMMINGS FARRINGTON, Pu.D., Curator of Geology, Field Columbian Museum. Pp. 229-+xii. Chicago: A. W. Mumford, 1903. In this book it has manifestly been the intention of the author to make the treatment of the subject as non-technical as possible. At the same time, scientific terms have been used whenever these were neces- sary to give the matter accuracy and definiteness. The subject asa whole has been discussed from the mineralogical standpoint, each gem being considered under the mineral species to which it belongs. Fol- 64 REVIEWS lowing this idea, ruby and sapphire are treated under corundum; emerald and aquamarine, under beryl, etc. At the outset a brief discussion of the nature of gems is given, and the characteristics or qualities for which they are prized are enumerated. Following a few pages devoted to the geographical and geological occurrences of gems, the more common methods of gem-mining are described. Since the coloring of gems is one of the most essential features of their value, the significance and meaning of the color ele- ments are considered, and a list of gems is given arranged according to colors. ‘The subjects of luster and hardness come next, followed by a table showing the hardness of gem minerals. Methods for the determination of specific gravity are described, since herein is a reli- able means of distinguishing between gems of different kinds, and of separating false from real stones. Next comes in turn, a discussion of the optical properties, electrical properties, phosphorescence, and fluorescence of gems. The crystal form of gems is then considered to some extent, since this characteristic often affords a ready method for their identification. Methods of cutting and mounting gems are given in some detail, and a number of line engravings have been prepared showing the usual forms of cutting. Next come chapters on the valu- ation and price of gems, imitation gems, and how to detect them, superstitions regarding gems, and birth-stones. Following the general matter noted above, the individual gem minerals are considered, the first being the diamond. In the discus- sion of the diamond its characteristics are pointed out, and the diamond fields and the famous diamonds that have been discovered are described, along with much other matter of general interest. Ina similar fashion, the several gems afforded by the mineral species corundum are dis- cussed; and then come in turn spinel, beryl, chrysoberyl, tourmaline, topaz, garnet, opal, and all other minerals, as well as some substances of animal and vegetable origin which have been used to any degree for purposes of adornment. The book is neatly printed and bound, and contains a large num- ber of half-tones and line engravings, as well as sixteen full-page illustrations in color of rare excellence. In bringing together in com- pact form so very much interesting matter concerning our gem miner- als, Dr. Farrington has performed a service that will be greatly appreciated both by the mineralogist and the general reader. lela I, 1D ORI AE: Tuat the rapid growth of special scientific nomenclatures is a serious burden is felt by every scientific worker. Any effort, accordingly, toward harmonizing conflicting usage should be and is welcome. Not every such effort necessarily produces final results, but if systematically pursued, it can hardly fail to elimi- nate some confusion. In the work of the United States Geolog- ical Survey decisions are constantly required of questions relating to the naming and correlating of geologic formations. In many instances the available evidence is so conflicting or so meager as to preclude final judgment. Nevertheless, if geologic work is to go on, and maps are to be made, a definite usage must in each case be authorized. These decisions establish precedents, which from time to time receive formal statement by the director and become rules. Such a code, if we may borrow the legal term, was published in the Zenth Annual Report of the Survey, and in the 7wenty-fourth Report is republished, revised, and enlarged by the incorporation of the precedents established in the last thirteen years, together with certain other changes recommended by the committee charged with the revision. In the new rules there are many minor and some major changes from the old. The action of the committee has been conservative in some directions and radical in others. In part the changes are seemingly retrogressive, though it is to be remembered that a wise progression never hesitates to abandon a position which experience has proved untenable. The return to g the use of “Tertiary,” ‘Quaternary,’ ‘Triassic,’ and the adop- tion of “Ordovician” as a systematic term, with the recognition of the quadruple Lyellian divisions of the first-named are move- ments which will bring the publications of the survey into closer harmony with those of other organizations, and are fully war- ranted by the developments of the last decade. The extension of the criteria for the recognition of formations so as to include 65 66 EDITORIAL physiographic data, and to allow fossils to be used for discrimi- nation as well as for correlation, will, it is believed, be generally approved. To meet the practical difficulties of mapping, litho- logic units smaller than formations may, when sufficiently important, be separately mapped as members or lenses. An effort is to be made to conform in the general plan of mapping to the logical categories of (2) sedimentary, (4) igneous, (c) met- amorphic rocks. Surficial rocks of all ages are treated as a subclass of sedimentaries, but are to be distinguished by patterns in mapping on a genetic basis. The stratified rocks of the Archean and Quaternary are given distinctive colors. There are many other changes apparent when the old and new rules are compared. Rules of nomenclature will not, unfortunately, be consistently applied if the interpretation be left entirely to each individual worker. To meet the necessities of the present case a Committee on Geologic Names has been constituted, to consider and decide the various difficulties which will inevitably arise in the varied work of the Geological Survey. This committee is charged with the inspection of all papers written by any member of the Survey corps, and as part of its work keeps a complete card catalogue of all formation names proposed or used in writings relating to American geology. The closer co-operation of the various individuals and organ- izations concerned in the advancement of geologic science in this country is surely desirable, and much misunderstanding and unproductive effort can certainly be eliminated if common usage of geologic formation names can be brought about. While a committee from the Survey, representing as it does only a part, even though the larger part, of American geologists, can in the nature of the case, have no authority over the publications of geologists not belonging to its own corps, yet it is hoped that general appreciation among geologists of the advantages of so doing will induce individual and independent workers to avail themselves of its functions, andto conform, when possible, to the usages of the large body of their geological colleagues gov- erned by its decisions. ale Je iB, RECENT PUBLICATIONS. —Apams, Geo. I., GirTy, GEO. H., AND WHITE, DAVID. Stratigraphy and Paleontology of the Upper Carboniferous Rocks of the Kansas Sec- tion. [Department of the Interior; U. S. Geological Survey, Washing- ton, 1903. | —BrIGHAM, A. P. Geographic Influence in American History. [Ginn & Co., Boston, 1903. ] —Co Le, Burt. Storage Reservoirs on Stony Creek, Cal. [Water Supply and Irrigation Paper No. 86, Department of the Interior; U. S. Geo- logical Survey, Washington, 1903. | —Co.Lui£, GEORGE Lucius. Ordovician Section Near Bellefonte, Pennsyl- vania, [Bulletin of the Geological Society of America, Vol. XIV, pp. 407-20; Rochester, October, 1903.] —DaLL, WM. HEALEY. Geological Results of the Study of the Tertiary Fauna of Florida, 1886-1903. [Extract from the Transactions of the Wagner Free Institute of Science, Vol. III, Part VI; Philadelphia, 1903.] —E.LLs, R. W. Notes on Some Interesting Rock-Contacts in the Kingston District, Ontario. [From the Transactions of the Royal Society of Canada, Second series, 1903-4, Vol. IX, Section 1V; J. Hope & Sons, Ottawa. | —FuLLER, M. L. Probable Pre-Kansan and Iowan Deposits of Long Island, New York. [From American Geologist, Vol. XXXII, November, 1903. | —Grasau, A. W. Paleozoic. Coal Reefs. [Bulletin of the Geological Society of America, Vol. XIV, pp. 337-52; Rochester, September, 1903. | —Kenp, J. F., AND KNIGHT, W.C. Leucite Hills of Wyoming. [Bulletin of the Geological Society of America, Vol. XIV, pp. 305-36; Rochester, September, 1903.| —Kunz, GEorGE F. The Production of Precious Stones in 1902. [Extract from Mineral Resources of the United States, calendar year 1902, U.S. Geological Survey; Washington, 1903. | —LAMBE, LAWRENCE M. The Lower Jaw of Dryptosaurus Incrassatus (Cope). [Reprinted from the Ottawa Naturalist, Vol. XVII, pp. 133-39, November, 1903. | : —NICKLES, JOHN M. The Richmond Group in Ohio and Indiana, and its Subdivisions; with a Note on the Genus Strophomena and its Type. —RoweE, JESSE P. Some Volcanic Ash Beds of Montana. [Bulletin of the University of Montana, No. 17; Missoula, Mont.] —RoweE, J. PERrky. Some Montana Coal Fields. [From American Geol- ogist, December, 1903. | 67 68 RECENT PUBLICA TIONS —RUuSSELL, I. C. Volcanic Eruptions on Martinique and St. Vincent. [From the Smithsonian Report for 1902, pp. 331-49; Washington, 1903. | —RussELL, I. C. Geology of Southwestern Idaho and Southeastern Oregon, [Bulletin No. 217, Department of the Interior, U. S. Geological Survey, Washington, 1903.| —SCHEI, P. Preliminary Report on the Geological Observations Made During the Second Norwegian Polar Expedition of the “Fram.” [Printed by William Clowes & Sons, London and Beccles. | —Schuchert, Charles. On New Siluric Cystoidea and a New Camarocrinus. [From American Geologist, October, 1903.] On the Faunal Provinces of the Middle Devonic of America and the Devonic Coral Sub-Provinces of Russia; with two Paleographic Maps. [From American Geologist, September, 1903. ] —SCHWARZ, G. FREDERICK. The Diminished Flow of the Rock River in Wisconsin and Illinois. [Bulletin No. 44, Bureau of Forestry, Depart- ment of Agriculture; U. S. Geological Survey, Washington, 1903.] —SHIMEK, B. The Loess and the Lansing Man. [From American Geol- ogist, December, 1903. ] —SIMONDS, FREDERIC W. The Minerals and Mineral Localities of Texas. [ Bulletin No. 5, University of Texas, Mineral Survey, December, 1902 | —SPENCER, J. W. Submarine Valleys off the American Coast and in the North Atlantic. [ Bull. Geol. Soc. Am., Vol. XIV, pp. 207-26, Rochester, July, 1903.] —STANTON, TimotTHy W. A New Fresh-Water Molluscan Faunule from the Cretaceous of Montana. [Reprinted from the Proceedings of the American Philosophical Society, Vol. XLII, No. 173.] —TayLor, F. B. The Planetary System. [Published by author; Fort Wayne, Ind., 1903.| —TicuT, W.G. Drainage Modifications in Southeastern Ohio and Adjacent Parts of West Virginia, and Kentucky. [ Professional Paper No. 13, U. S. Geol. Surv.; Washington, 1903. | —Topp, J. E. Concretions and Their Geological Effects. [Bulletin of the Geological Society of America, Vol. XIV, pp. 353-68; Rochester, Sep- tember, 1903. | —WHITNEY, MILTON, AND CAMERON, F. K. The Chemistry of the Soil as Related to Crop Production, [Bulletin No. 22, U. S. Department of Agriculture; U. S. Geological Survey, Washington, 1903.] —WINCHELL, H. V. Synthesis of Chalcocite and its Genesis at Butte. Mon- tana. |/é¢d., pp. 269-76; Rochester, July, 1903.| —Weodworth, J. B. Postglacial and Interglacial(?) Changes of Level at Cape Ann, Massachusetts. [Bulletin of the Museum of Comparative Zodlogy, Harvard College, Vol. XLII, September, 1903; Cambridge, Mass. | fourRNAlL OF GEOLOGY [VET OAK ENC AU(Clal, QOH Ane SPAN NV els Si Cal l@O NS Aa RAG AT ING THE WELLS.” Durinc the typhoid epidemic at Ithaca, N. Y., in 1903, acom- mittee of citizens began explorations for a source of artesian water to replace the surface supply then in use. This work was continued by the Ithaca water board, and the result was the sink- ing of thirteen wells in a limited area on the southern outskirts of the city. Prior to this an artesian well had been developed in the same area, yielding a daily flow of about 300,000 gallons from a series of Pleistocene gravels at a depth of about 280 feet. A majority of the new wells found water in what appear to be the same gravels; others failed to develop water. Besides these deep wells, there are a large number of shal- lower ones in the city of Ithaca which obtain artesian water in a gravel series found at depths usually from 50 to 100 feet. *Published by permission of the Director of the United States Geological Survey. I am indebted for valuable assistance in the preparation of this paper to the fol- lowing gentlemen: Mr. C. C. Vermeule, engineer in charge of the boring of the wells, for directing that samples be preserved for me; Mr. F. L. Getman, his assistant, for collecting the samples and for other valuable information; Mr. Lawrence Martin, of Cornell University, for aid in gathering information, and in consideration of the nature of the well sections; Dr. G. K. Gilbert, for placing at my disposal certain facts from his notebooks bearing upon the question of tilting of the land in central New York; Dr. William H. Dall, for identifying the mollusca; and Professor D. P. Penhallow, for identifying the plant remains. 2A more detailed statement of the bearing of this exploration on local water sup- ply will be published by the U. S. Geological Survey. Vol. XII, No. 2. 69 70 Nike. Sip, LLANE IMPORTANCE OF THESE WELLS. These wells have yielded three important geological results: (1) They have in two cases revealed the exact depth of filling by Pleistocene deposits, and have therefore given some additional facts concerning the form and depth of the Cayuga Valley. (2) Since samples were collected at frequent intervals in several of the wells, and records kept of all, they have revealed the struc- ture underlying the Ithaca delta down to the rock floor. (3) They have thrown some light on the occurrence of artesian water in Pleistocene deposits. DEPTH OF DEROSIGS. The wells are all located near the western margin of the delta on which the main portion of the city of Ithaca is built. The surface soil is clay and muck, and the region is evidently one reclaimed from Lake Cayuga by the same processes of lake fill- ing that are now at work on the outer edge of the delta on the north side of the city. While low and swampy throughout much of its area, this nearly level delta rises perceptibly toward the creeks that descend through gorges cut in the valley walls. These elevated sections are low, flat alluvial fans, raised above the general delta level by deposits brought down by the torren- tial streams that occupy the hillside gorges. The delta also rises gently toward the south, and at a dis- tance of about two miles south of Ithaca abruptly ends against the north face of the morainic complex which fills the valley thence to its present divide. It is evident that this moraine descends beneath the delta deposits. Two of the borings reached bed-rock, one (Fig. 1, C) at a depth of 260 feet, the other (Fig. 1, G) still further out in the valley, at a depth of 342 feet. A profile of the valley at this point is shown in Fig. 2. Farther north (Fig. 1, Ad and B) two wells, bored to the underlying salt, encountered rock at 430 and 401 feet respectively, the latter being 1,500 feet south of the former. These borings are not numerous enough to warrant any con- clusions further than that the maximum depth of the valley is at AL STAN WARE SIECTMONS: A Ty TILA GCA, N.Y. 71 least 430 feet below the delta at Ithaca, or about 25 feet below sea-level. Since soundings in Lake Cayuga reveal a depth of 435 feet in the deepest point, in which, of course, there is at least some filling, the borings at Ithaca do not add to the known depth of thenvalleysltusimot tol beinterred, however, that the deepest boring at Ithaca, although near the middle of the valley, really represents the deepest part of the valley in this region." The discovery of rock in these wells shows that the general slope of the lower valley walls is continued down with practically no change s((Hig.2)), vat least to, the depth reached in the artesian wells (Higa 1, Gand G). DESCRIPTION OF THE WELL SECTIONS. Both in the deeper artesian wells and in shallow ones in the city of Ithaca the upper layers are found to be a fine-grained, massive clay. In two cases it is reported as sandy. This clay layer is absent, or at least not continuous, near the eastern wall of the valley where alluvial fans have been built opposite the stream The depth of the clay stratum varies from approximately AO) to 60 feet: lusca and plant fragments, including pieces of reeds and wood, were mouths. Fragments of mol- found in several of the samples from ACU A HA Qa RS = VA \ \\ K - JG i} \ \\\ \ ; Ki : » C x ie ~ =. An ve PR SE | / aca Sheet” (U.S. Geological Sur- vey) to show the location of the artesian wells. 4, &, salt wells, rock at 430 and 4o1 feet respect- ively; C, Strang Well, No. 1, 286 feet, struck rock at 260 feet; D, Millard Well, No. 2, 259 feet; &, Old Clinton St. Well, 280 feet; F, South Well, 232 feet ; G, Strang Well, No. 2, 352 feet, struck rock at 342 feet; H, Trapp Well, 332 feet; Z, Holmes Well, 201 feet; /, Millard Well, No.1, 303 feet; A, several welis, close together, as follows: Illston Well, 286 feet; Strang Well, No. 3, 280 feet; Strang Well, No. 4, 279.9 feet; Strang Well, No. 5, 276 feet; and Strang Well, No. 6, 295 feet; Z, Millard Well, No. 3, 303 feet. this clay layer; and in two cases logs were encountered, one at 38-39 feet, another at 33 feet. *It is noteworthy, in this connection, that a well near the center of the Seneca Valley at Watkins had not reached the bottom of the drift deposits at a depth of 1,080 feet. 72 We sy LAK kx Beneath this clay layer, in every well of which there is a record, a series of coarser beds is found. These coarse beds vary greatly even in neighboring wells; but in most cases there are both sand and gravel layers. The bottom of the series of coarse sediments varies from 60 to approximately 120 feet, and the thickness in individual wells from 20 to about 70 feet. The coarser sands are clear and well washed; the gravels consist of well-rounded pebbles similar to those now brought down by the torrential creeks that enter the valley. In most of the samples preserved from these coarse layers plant fragments and mollusca were found. Seven logs were encountered, and the two logs found in the overlying clays were almost down to the level of the coarser series. Thus between the depth of 35 and 119 feet nine logs were encountered in boring thirteen six-inch wells. Since two of the wells passed through two logs each, logs were encountered in seven out of thirteen wells. The depth of the several logs is given in the following table: Wells Depth in Feet Material Sinan’ g icepccsecceun riences 56 -58 Gravel @©ldiGlintongS taste 63 Sand NT Fear Foe rencraey seta Note et Be) 2X0) Clay MME see as aoloa bane 118 -II19 Sand (BTAp Dapaneere sateen 48-50% Gravel and sand MTAD Disko homage aie eee 5514-56% Gravel and sand Millardst et ierasesssiiecines 50 -51% Sand and pure gravel SOuth perenne 33 Clay, somewhat sandy Slran gage aco errarcnenees te Ilo —II2 Probably at bottom of gravel In all the deep wells the coarser layers are underlain by a great thickness of clay, in which no molluscan remains were found, though in several samples small, indefinite plant fragments occur. In most of the wells the driller failed to preserve more than one sample, which he considered typical of the entire clay mass; but near the top and bottom of the series the material is occasionally reported as ‘‘clay and stone,” ‘clay and gravel,” or ‘clay and sand.’”’ In one well (the south well), however, samples were preserved every ten feet, and these samples show clearly the nature of the material. From top to bottom, that ARES S AUN WARE LETS (SEAGAMONS WAT PIA GA, IN. V7 73 is, from 70 to 200 feet, it is a fine-grained clay, at all depths lower than 100 feet containing small angular pebbles, in some cases scratched. These stones increase in number and size toward the bottom, and the proportion of sand increases to such an extent that below the depth of 135 feet the well-driller calls ita sandy clay. But down to the 200-foot level the stratum is unquestionably clay. Owing to the indefiniteness of the nomenclature used by well-drillers, and the failure in many cases, to preserve samples ite! sie Salo always easy to state exactly where the bottom of the clay series is. Using the Fic. 2.— Profile of the hill slope on the western side of the Cayuga Valley at Ithaca just west of the artesian well sites. This profile is continued down to the points ble, I place the base where rock was reached in the artesian wells. (Horizon- of the clay series in tal scale, 1 mile to the inch; vertical scale, 1,000 feet to the inch.) best judgment possi- the thirteen wells as follows: 220, 210-30, 202-70, 238-78, 230, 225-76, 200, 214-80, 202-22, 244-70, 240, 242-46, 234-70. The well which gives the 200-foot level for the bottom of the clay is the one from which most samples were obtained, and for that point may be accepted as correct. It is, however, the farthest south of all the wells, and it does not follow that the bottom at that point is the same level as the bottom at other points. On the contrary, all the evidence seems to indicate that the base of the clay series is decidedly irregular. As the base of the clay series is approached, and after it is certainly passed, a series of beds of marked irregularity is encountered. They are prevailingly coarse-textured, and in every well include some sand or gravel. In many of the wells SOUTH WELL STRANGZ2 MILLARPD2 STRANGS S7RANGS STRANG 4 STAANG 6 232 /7- 352. PT AD IHG 276 FT. 280F +r 2799/7 295 Fr ° @ ? ? ? ? Wa Hater LEGENA— Rees Se ?| Unknown Pree ]So7e Fic. 3.—Sections of seven artesian wells grouped approximately along a north- south line. South well, southernmost. Limits of the four series of deposits indicated No Mater. Fe in a general way by the lines connecting the different sections. STAANG/ OLD CLINTON ST MILLARDS TRAPP WELL HOLMES / MILLAPD/ 2E6 FT. ZEOF#T 29/1Fr Little Water Fic. 4.— Sections of six artesian wells grouped approximately along an east-west line. Strang 1, westernmost; Millard 1, easternmost. Limits of the four series of deposits indicated in a general way by the lines connecting the different sections. 70 Tien Se LAGI both sand and gravel are encountered. No two of the wells have the same sequence of layers, even though the wells are close together. Samples prove that some of the layers are water- washed sand and gravel, while others are unquestionably till, with scratched stones. Ina number of other places till is sus- pected, though the evidence is not sufficient to prove it. In this seriesof coarse deposits, water is found at varying depths in the different wells, and in differen: sediments. In some cases it is found ina sandy clay, called ‘‘quicksand,” in which there is so much water, under such pressure, that the sand is forced into the pipes in sufficient quantities to fill them and stop the water flow. Between this extreme and that in which the water is found in coarse gravel, there are several intermediate conditions. The largest flow is obtained from the coarse gravels. Beneath the unquestioned till, and in various places beneath the materials interpreted as probable till, is found a black sand in which from 50 to 75 per cent. of the material is quartz, the remainder being mainly dark shale fragments. In one of the wells that reached bed-rock this black sand rests on the rock, which was encountered at a depth of 342 feet. Neither here nor in the other well that reached rock, nor, in fact, in any of ‘the wells, was any older drift encountered. All the materials are such as might have been brought by the last ice advance, or deposited since the ice-sheet melted away. Whether deposits of earlier ice advances were never made here, or whether they were all swept away by the last ice advance, is not determined by the evidence. INTERPRETATION OF THE WELL SECTIONS. 7 Morainic lower series —The history of the accumulation of the 342 feet of sediment revealed by these well-borings is in most respects clear. That the bottom series of till, sand, and gravel is morainic seems proved by several facts: (1) the neigbor- hood of the massive moraine which rises above the delta two miles south of the wells; (2) the position of the coarse materials at the base of the series, from all other members of which they differ decidedly; (3) the apparently irregular outline of the VA SIEEES PAUN aN AEE 12 SIE CH LOLN Sy Ade TELA CA Ni ¥ WH. upper portion of this lower series of coarse materials; (4) the marked variety of materials composing the lower series, which vary from gravel and sand to till, thus closely resembling the moraine that rises to the surface farther south; (5) the large per- centage of water-washed material, again resembling the condition in the moraine farther south. This large amount of water- washed material would be expected where the ice-front stood in a deep lake, as was the case here. Glacial lake clay.—It is well established by the evidence of various overflow channels, by well-defined elevated deltas at various levels,* and by lake clays on the hill slopes, that the Cayuga Valley was occupied by a steadily expanding, ice-dammed lake, with its level frequently lowered as successively lower out- lets were discovered by the melting back of the receding ice- sheet. This lake condition lasted for a long time; in fact, until the Mohawk outflow was discovered. The length of this period of lake stage cannot be stated; but it was sufficient for the ice to have melted back at least forty miles. Of necessity a great amount of clay must have been deposited in this lake, not merely that supplied from the glacier, but that washed from the hill slopes, and that brought by the streams which descended the steep hill slopes. The great thickness of clay found in all the wells, usually between the 100 and 200-foot levels, is interpreted as lake clay deposited in this ice-dammed lake. The great thickness of the clay stratum, its uniform character, its occurrence in all the wells, the general absence of animal remains and the presence of only minute fragments of plants, the occurrence of scratched pebbles that might have been ice-borne, and the increasing coarseness of the clay toward the bottom, all harmonize with _ this interpretation. No opposing facts were discovered. Recent lake clays —The uppermost clay beds, between the o and 40-50-foot levels, with abundant organic remains and an almost complete absence of sand and pebbles, are interpreted as lake ™See papers by FAIRCHILD, Bulletin Geological Society of America, Vol. VI (1895), PP. 353-74; Vol. X (1899), pp. 27-68; and by Watson, WVew York State Museum Report 57 (1897), Part I, r55-r117. 78 R. S. TARR clays formed in the same manner as those now accumulating. They are believed to represent modern lake-filling similar to that now in progress at the head of the lake. When the number of torrential streams that descend the valley wall is considered, the amount of lake-filling which this stratum represents is not excessive. These streams have formed deep gorges, having not only removed much drift, but also cut deeply into the shale. The material thus removed seems ample to make this deposit. The upper series of coarse sediments— A much more difficult problem is presented by the sand and gravel series, sandwiched between the two clay beds, and covering an area whose length north and south is known to be over one and a quarter miles. Some widespread change in condition is here represented, indi- cating a stage during which clay deposit was interrupted, appar- ently after the ice-dammed lake ceased to offer opportunity for deposit of lake clay, and before conditions appeared which per- mitted the accumulation of the later lake clays. This period of interruption of lake clay deposit must have been one in which, at the site of the wells, the conditions were either those of a shallow lake or else of absence of lake water. Both the coarseness of the sediment and the abundance of organic remains -indicate this. It is inconceivable that at a distance of three-eighths of a mile (the distance from the outermost well to the nearest valley wall) gravel and sand could be deposited in a lake from 56 to 118 feet deep, especially at a point remote from the mouth of any stream. It is inconceivable, also, that gravel and sand could be deposited in such a lake over so wide an area. Nor does it seem probable that such a large number of logs would be accumulated in lake deposits. The chance discovery of seven logs in boring thirteen six-inch wells through this sand-gravel series indicates a large number of logs in the series. This point becomes all the more striking from the fact that only two logs were found in the recent lake clays, and both of these near the sand-gravel series, while no logs were encoun- tered in the ice-dammed lake clays. Professor Penhallow has determined the two specimens of wood preserved as follows: (1) South Well, 30-35 feet in clay, ARTESIAN WELE SECTIONS AT ITHACA, N. Y. 79 Pinus rigida ; (2) Millard Well No. 1, 50 feet, gravel, the com- mon tamarack, Larix americana. Dr. Dall reports that the mol- lusca belong to the following genera: Valvata, Planorbis, Amni- cola, and fragments of Pisidium or Sphcerium. These are all of thesame nature, namely fresh-water, and such as are found in the con- fervee or other fine-textured vegetable matter, such as grow in quiet places in the course of brooks, or in ponds or lakes at the mouth of brooks. They would hardly be found in the unsheltered waters of a large lake like Lake Erie. Correlation of coarse sediments with Iroquois stage. —The evidence seems conclusive that these sands and gravels were either shallow-water, lake-margin deposits or else stream-made land deposits, and that they were succeeded by lake conditions. In seeking for an explanation of these phenomena, land-tilting seems the only rational hypothesis. It is a well-known fact, as clearly shown by Dr. Gilbert, that the land has been tilted in this region since the deposit of the beaches of the Iroquois shore line. The Iroquois lake stage immediately succeeded the stage of ice-dammed lake in Cayuga Valley. Therefore these sands and gravels are in the right position for correlation with the Iroquois beach stage. In a letter to me Dr. Gilbert supplies the following informa- tion: Correlating the upper bar at Richland Junction (563 feet) with the lowest bar at Weedsport, there is a gradient of 2.9 feet per mile. Correlating the upper Richland bar with the lower cut terrace at Montezuma, there is a gradient of 2.6 feet per mile. Correlating the upper Richland bar with a bar at Cayuga, there is a gradient of 2.7 feet per mile. The lines on which these measurements are made do not, precisely corres- pond with the direction of greatest slope of the plane of deformation, but I judge that the line from Union Springs to Ithaca makes about the same angle with the direction of greatest slope,so that these figures might be applied without correction. A correction for direction would increase the estimates for gradient. From Union Springs, where the Iroquois beach disappears beneath Lake Cayuga, to the site of the wells is approximately twenty-nine miles; and, taking 2.7 feet per mile as a gradient, the Iroquois shore line might be expected to appear at Ithaca at a depth below present lake-level of about 78 feet. Several of 50 Tee Sr LEA ich the wells passed through coarse material at this depth, and in several of them the coarse materials reach a much greater depth than this. On the other hand, the records of four of the welis show that the sands and gravels were passed through before that depth was reached. In his letter Dr. Gilbert makes this further statement: From Richland northward to Adams Center the gradient is 4.3 feet per mile, and north of Adams Center it is still steeper. So it is possible that a correction might advantageously be applied to southward flattening of gra- dient. Some such correction, the exact amount of which is not clear, would seem to harmonize better with the distribution of the gravels as revealed by the well sections. Several facts, as follows, seem to warrant correlation of these deposits with the Iroquois stage: (1) the difficulty of otherwise explaining the sand-gravel series; (2) the evidence of coarseness of material, and of plant and animal remains, all of which point either to land or shallow-water conditions; (3) the position of the series, resting on deposits which appear to have been com- pleted just before the Iroquois stage; (4) and the fact that their position is approximately at the level to be expected on the theory of formation during the Iroquois stage. No facts oppos- ing this correlation are known, and no other rational explanation suggests itself. It is therefore proposed as an interpretation of the phenomena. SUMMARY OF EVENTS. On the basis of these interpretations, we have revealed the following postglacial history of the Ithaca delta: First moraine was formed while the ice-front stood in a deep lake, in which the morainic material was largely assorted. After this stage there was a long period of lake clay deposit in an ice-dammed lake whose area was expanded, while its level fell by successive drops to lower levels as the steady melting back of the ice-front discovered lower and lower outlets to the north. In this lake there was floating ice, but little, if any, animal life. When the Mohawk outlet was finally discovered the ice-dammed lake nearly, if not quite, disappeared from the site of the artesian wells. The PARES SHAUN | VUEIEL VS EALOINS) Ai ly LILA GAS IN.) Ve 81 land surface at the site of these wells was then from 60 to 120 feet below the present delta surface ; but, owing to the depression of the land in the north, the lake waters either could not then reach this far, or, if they did, produced only a shallow lake. At this stage trees grew and mollusca thrived, while a series of sand and gravel layers were laid down whose depth in the several wells varies from 20 to 70 feet. Elevation of the land in the north tilted the basin of Lake Cayuga faster than the deposit of sand and gravel was made, ultimately covering the coarse deposits with lake water. Some of the sand and gravel may be due to the work of the lake waves as the tilting of the land caused an encroachment of Lake Cayuga farther and farther south. The fact that the pres- ent surface of the delta contains no sand and gravel excepting near the stream mouths, may be explained as follows: (Ge) the levelness of the delta; (2) the recency of the delta—it is still so swampy at the well sites that it is flooded at least once each year; (3) the fact that the streams now bring less material, having already cut through the drift into the rock. SOURCE OF THE ARTESIAN WATER. It is believed that the source of the water in the upper gravels is the alluvial fans opposite the mouths of the streams that descend to the Ithaca delta. Into these fans much of the stream water sinks, in some cases entirely disappearing at all times excepting in periods of flood. It is not absolutely certain that the gravels of the alluvial fans are continuous with the sand- gravel series encountered in the shallower artesian wells; but this is to be expected, since the conditions which favor the deposit of coarse sediment must have existed continuously near the mouths of the streams that descend the hill slope. Some such source near at hand is indicated by several facts: (1) a reported vari- ability in volume; (2) the moderate pressure of the wells, which in some cases barely forces the water from the ground; (3) the composition of the water, which indicates a shorter underground journey than that of the water in the deep wells; (4) the marked difference in composition and purity of the water from various wells. 82 TR (Se, LAIR For the water found in the deeper sands and gravels the source is believed to be the moraine which occupies the Cayuga Valley from the divide nearly to the well sites, a distance of over eleven miles. Numerous streams descend to this moraine, supplying much more water for percolation than the mere rain- fall. The moraine is to a very large degree made of sand and gravel, offering the best of conditions for the entrance of water. The hardness of the water and the temperature (52° in August and December) both indicate a long underground journey ; and the great pressure, which forces from one of the wells a steady volume of 300,000 gallons a day, also indicates some fairly dis- tant source. To account for the pressure observed it is necessary to find a source much higher than the well sites. No such source is to be found to the north because the Jake occupies that region; to the east and the west rise high hills in which are nearly hor- izontal strata of shale and sandstone. This leaves the moraine to the south as the only possible source of the water; and this source is not only ample, but, if the above interpretation of the well sections is correct, there is a direct connection between the surface moraine and the buried moraine gravels which supply the water. Raa ARR IrHaca, N. Y. THE ROLE OF POSSIBLE EUTECTICS IN ROCK MAGMAS. THE appearance of Professional Paper No: 18, Zhe Chemical Composition of Lgneous Rocks, Expressed by Means of Diagrams, by Professor Iddings, completes a trilogy of works* which will leave a permanent impression upon the nomenclature and progress of petrography. No one who has not done a little work of the sort can begin to conceive of the amount of labor employed in the comparatively small paper of less than one hundred pages which has just appeared. And even so, it is probable that vastly more in the way of experiment in trying to group the facts in different arrangements has been done, of which no trace remains. We see the finished railroad survey, and can form some conception of the engineering, but we cannot know how many lines may have been run through the dense forests before settling upon the final track. It is easy also to notice possible improvements in the track when it is once completed; and if I make some suggestions along this line, it is only with the greatest respect for the distin- guished authors of the new chemical classification. Whatever modifications the future may have in store, I think that there will be much of their work which will endure, and that such terms as ‘“‘persalic” and ‘‘dofemic” will after a little become household words to the petrographers of all countries. I think the remark of Mr. Iddings, that we must look upon a rock as the chemist looks upon a solidified mass of mixed salts, that ‘‘we must think of the study of igneous rocks, their magmas and rela- tionships, as purely physico-chemical problems, involving the measurement and comparison of mass and force, and their definite quantitative expression,” as one of very great importance. When he goes on to say that there are no recognizable groupings of ' The Quantitative Classification of 1gneous Rocks, by CROSS, IDDINGS, PIRSSON, AND WASHINGTON, and Professional Paper No. 14 by WASHINGTON, being the two previous works. 83 84 ALFRED VC LANE, rocks or noticeable subdivisions of chemical series, that chemi- cally similar rocks occur in genetically different families, that it” follows that the subdivisions of all igneous rocks into groups for the purpose of classification must be on arbitrarily chosen lines, | and that it is no argument against a classification if a rock of great importance belongs on the boundary of a classificatory division, that any system of classification will be as natural as any other system, he is rather too pessimistic. Merely asa mass of mixed salts, there are certain common relationships which hold, and of which any system of classification, the new one included, must take account. He has pointed out a number ot such facts: (1) In the great majority of rocks alkalis do not exist in excess of that required to make feldspar or nephelite; (2) The commonest rocks are in composition like the average of all rocks which has the silica percentage of 58.7 to 59.77 and an alkali-silica ratio of 0.083 to 0.088. Iddings, in fact, suggests the possibility that all rocks have been derived from one com- mon magma by splitting. If so, this splitting must have gone on under chemical and physical laws which are to be traced and followed in the classification. There is, however, another possi- bility which I wish to suggest, namely, that in the process of splitting there is a tendency in one of the fractions toward that same common average.’ To this point we will return later. There are numerous other facts which Iddings points out, such as the inverse relation between silica and alkalis, and iron and magnesia, and various other notes on pp. Ig, 64, 65, 70-81, of which any classification may take account, as, in fact, the new one does, to a very large degree. Upon that depends its ser- viceability. The fact that there are exceptions does not invali- date the importance of this relation, provided they are not too many. Now, if one looks over the general diagram, Plate I, which Iddings has prepared to see that it is true that there is ‘‘no clus- tering of analyses’’ and ‘no natural subdivisions” (p. 17), I can- not agree with him in the conclusion he draws. In the first place, *This would fit especially well in a planetesimal theory; in fact, would be almost necessary in view of the wide prevalence of basaltic magmas. EUTECTICS IN ROCK MAGMAS 85 as he himself has pointed out, there are no more alkalis in most -rocks than alumina and silica to go with them. In the second place, there are very few rocks with high silica percentage down about to the point representing the average rock in which the ratio of alkali-silica is less than 0.04. As the amount of alkalis increases, the number of analyses becomes more numerous. Somewhere between the ratios 0.08 to 0.09 they are most abun- dant. After that they diminish somewhat and become less up to a ratio between 0.12 and 0.16, and then they become more numerous once more. If we suppose this not to be accidental, we must assume @ éendency on the part of a certain group of magmas toward an alkali-silica ratio between 0.08 and 0.09. I fixed on 0.083 from the diagram. This, I noted, was the ratio of the average rock. Expressing this in fractions, the alkali-silica ratio would be 1:12. This suggested to me at once a combination of ortho- clase or albite and quartz with equal molecular proportions of silica. This again reminded me of Rosenbusch’s spherulite-form- ing microfelsite, which he supposes to be of the composition K,O-Al,O,-4 SiO,, with x greater than 6. It is also the com- position of micropegmatite where the quartz and feldspar are equal. All these facts harmonize with the supposition that the combination (Na,O, K,O)- Al,O, - 6SiO,+6SiO, is the eutectic ratio of alkali and silica. If such magmas are composed mainly of alkalis, alumina, and silica, these will tend to crystal- lize out whichever of these components happens to be in excess, and the analyst will be likely in the long run to obtain more analyses near this ratio than either above or below. This suppo- sition is obviously compatible with the idea that the igneous rocks are derived from fusion either of pre-existing sedimentaries or of planetesimals. We have remarked that down to where silica equals 0.59 and 0.57 there seems to be a belt of analyses having about this ratio, with comparatively few as the ratio dropped. Now, of course, these rocks are not all simply feldspar and quartz. How do we explain this belt or line? Simply by supposing that this eutectic relation holds, even though other elements were combined with the silica, that the 6 silica may be more or 86 Ib) GID) (Cs JOINS, less in combination with other elements commonly ceccurring in rocks, and yet that the eutectic ratio of alkali:silica :: 1:12 would hold true. Possibly it must be in combination with H,O or some other oxide. The first element, of course, to be considered would be the lime. What composition of albite and anorthite would give the same eutectic ratio? A little simple algebra proves that it is the commonest labradorite (Ab, An,). Now, this feldspar is, as we all know, one of the commonest in rocks, and more easily fusible than either albite* or anorthite; so that 1t is strictly eutectic: “In other words, the diagram seems to show that there is probably a eutectic series from a micropegmatite with quartz about equal to albite or orthoclase down to labradorite. If this theory is true, then there would be a tendency in rocks where the alkali-silica ratio was less than 1:12 to have the silica crystallize out first in a porphyritic way as quartz; and, on the other hand, if the alkali ratio were a little greater, there might be a tendency to have the feldspar crystallize out early, so as to bring the residual magma down to the eutectic ratio. A large excess of alkali would bring it near another eutectic. Corresponding to these would be a more or less quartzose marginal zone where the crystallization first began. That this is liable to be modified more or less by irregular changes in temperature, pressure, and environment, I need hardly add. But what about the other elements, in particular the iron and magnesia? If silica were there in excess of the eutectic ratio, they would combine with it very easily and, such compounds being relatively insoluble in the magma, crystallize and separate out, in sharp form embedded in the cutectic compound. In har- mony with this, we find some of the few analyses which are markedly below the eutectic ratio containing considerable quan- tities of lime, magnesia, and iron. They may be aggregates of these earlier eliminations of the magma—divergent splits. Some of them are such, I judge from the description. Moreover, the minerals formed in such magmas must have high silica ratios, be augite or enstatite rather than olivine. * According to the latest researches of Doelter and others this is not so—for albite is 10° to 15° more fusible. HOTECTIES IN ROCK MAGMAS 87 We have remarked that the eutectic ratio continues in a line from micropegmatite to labradorite (Ab, An,). This feldspar _ has a silica percentage of 53 and an alkali-silica ratio of I: 12. This is so close to the ratios of the average rock that they are within the limits of the probable error; in fact, we may say that the average rock has the same silica percentage and alkali ratio as labradorite. Now, we will notice upon the chart that once the silica falls below this ratio there is marked change in the behavior of the analyses, and they seem to stream off toward the lower right-hand corner, the alkali ratio dropping with the silica ratio. We might infer, therefore, that for rocks less silicious the eutectic ratio above given did not hold; or, rather, it may hold, so far as the alkalis are concerned, that they still find the most fusible compound Na,O-: Al,O,:6Si0,, and 6 ROSiO,, but that in the presence of an excess of bases there is some other equally or more fusible compound ROSiQO,. This, remember, is a purely theoretic inference from laws of chemistry and Iddings’s diagram. A moment’s thought shows how amply it is confirmed by petrographic research. This more fusible mineral is augite. As the percentage of SiO, in augite -is about the same as in labradorite (between 55 and 44), and cannot anywhere in the pyroxene group get above that of enstatite (60), we see why the distribution of analyses turns a square cor- ner, and we find them quite frequently from 0.58 SiO, down with all kinds of alkali ratios. We may go on to ask if there is any eutectic balance between labradorite and augite. As the ratias of lime to silica and percentage of silica are practically the same in the augite and in the labradorite series, the question as to the predominance of the one or the other tendency will probably be a question of balance mainly between the femic and alkaline constituents. The femic constituents in a magma mainly of fel- sitic eutectic are only slightly soluble and tend to crystallize out early as magnetite, biotite, hornblende, etc., even though present in very small quantities; and even though they increase markedly in abundance, their solubility or fusibility is still small, except as they can be taken into the augite molecule. An excess of mag- netite or olivine crystallizes readily.‘ The metallurgists tell us tI am neglecting a lot of minor matters. 88 ALFRED C. LANE about this in their slag formule. Consequently, it is a priori probable that labradorite and augite magmas mix readily and without tendency to split; though, when they actually come to crystallize, petrographic observation teaches that the labradorite feldspar is the earlier and the augite later for the rocks on the right or lower silica side of the average rock. The reverse is true on the other side. We must remember that, so far as chemical affinities allow, the eutectic (most fusible, least solidi- fied) magma will contain a little of every element going; and therefore we are not surprised at the complexity of the composi- tion of the augites which are collected in the table accompany- ing the statement of the Quantitative System. The latest researches show that basaltic magmas are fusible at distinctly lower tempera- tures than either labradorite or augite. Nor are we surprised that, if we draw a line from the average rock-analysis point, or that for labradorite, in the average direc- tion in which the less silicious and less alkaline analyses diverge therefrom, it will strike the line of no alkalis at 43 per cent. SiO,, or somewhat less; for this is not only the lowest silica percentage which augite, diallage, and the feldspar series reaches (anorthite), but the maximum for olivine. So in the femic magmas there is a clear tendency away from extremely different percentages of iron, lime, or magnesia. Probably the ratio CaO: MeO Be® 3322: 1s not var irom the eutectic one. In the following group of analyses from Lighthouse Point, for instance, it is clear that at the center, where there was opportunity for differentiation and adjustment of the magma, and the eutectic would accumulate, there is more lime than at the quickly chilled and cooled contact. The magma had an excess of magnesia and iron for eutectic relations, and this is shown petrographically by an early generation of olivine and magnetite. ; Returning to the Iddings’s diagrams, it seems, from a study of the distribution of Class I, that a few rocks are included (anorthosites, canadases) which it goes against the grain to include —rocks with an unusual amount of lime, but really tribu- tary to the femic eutectic; that the real natural family is from EUTECTICS IN ROCK MAGMAS 89 ANALYSES OF STONE FROM LIGHTHOUSE POINT, UNDER DIRECTION OF E. Dy CAMPBELL, (BY Eo EB. WARE, JUNE 30, 1903). 616mm 4,115 mm 7,600mm Contract (2.2 feet) (13.5 feet) (24.9 feet) DISTANCES EROMPNIARG INH Il | eee eee | eee ee OR Mole- Mole- Mole- Mole- Noor cule No. 3 cule Now6 cule No. 8 cule SHO Gs 46.98 | 0.783 | 47.67 0.795 | 47.25 0.787 | 47.10 0.785 Al,O,;. 17.85 O.175\ || 17-55 0.172 | 18.00 0.176 | 17.47 0.172 Fe,0, 3.13 0.019 2.51 0.016 2.21 0.014 2.66 0.017 FeO .. 10.30 | 0.143 | 12.59 0.175 | 12.42 0.172 | 12.93 0.179 IM @ Reavy ectsitterey xovassicelemmisrelay eis (a@)7.10 | 0.177 5.65 O.141 6.35 0.159 | 6.88 0.172 CAO Re ree eee eeiaseal| (2) O54 7m pe OnLo2cuTO N75 0.192 | 11.45 0.204 | 10.27 0.183 Sodiumioxide serenely: 2.04 | 0.033 | 2.22 0.035 | 1.96 0.031 1.91 0.031 iPotassiumioxidemensesee seis 0.60 ! 0.006 | 0.65 0.007 | 0,66 0.007 | 0.59 0.006 H,O at above 800° C........ ()1.97 | 0.168 | 0.35 O.112 Sea | aesarenes ol Gente Eis Ovaterro as Greer etracrice (a)1.55 | 0.086 0.40 0.028 Besoin usteeeite Saceove eae (CO piigsenne cocnone samecenedcnd 0.20 | 0.005 | 0.18 0.004 Seopa ence Lata | Stats 12-(0)5 *aeee qooo dose cobU mono na ae 0.143] 0.001 | 0.169! 0.001 0.158 | 0.001 | 0.161 | 0.oor Sin 0.097} 0.003 0.183 0.0057} 0.086 0.0027] GC.III 0.003 Ce ere sentans fiaicie eaten che necs aie 0.07 | 0,002 | 0.05 0.0014] 0.02 0.0006} 0.09 0.0025 ING A(O) ee a reac cma oe 0.26 0.003 0.19 0.003 0.18 0.003 0.15 0,002 100,760 IOI. 102 100.744 100,322 Ratio, alkalis: SiO). specter 0.048 0.053 0.0483 0.0472 Ratio, pore space: solid space, i INV asOleNereeisiye cece | O84 73 0.0012 0.0032 0.0018 Spa Groinvasoleney.)--)-- scl = 2.83 3.02 3.01 3.02 (a) Checked later in another sample. (6) Determined on new sample, first method incorrect. labradorose as a limit up; and that those which occur with more than 30 per cent. of magma belonging to the femic eutectic are rare. Moreover, it will be noted that there is among the orders between orders 4 and 5 a new principle of classification intro- duced. This corresponds approximately to the extra alkaline eutectics. Similarly, the rockallases appear very isolated and strange in the salfemic family. I shall not, however, pretend to discuss the analcitic, melilitic, and other alkaline eutectic magmas, with which Iddings is much better acquainted than I. My object is rather to suggest that ultimately a more rational and natural, and therefore useful, classification of analyses might be attained if our main magmatic groups were defined by the eutectic within whose influence they come, so that in splitting one part will be nearer the eutectic and the other farther from it. I have Iddings’s diagram to thank for this suggestion. gO ALFRED CGC LANE I am not going to say all that might be said about such inter- growths as perthite and sigterite (nepheline-albite), and other pegmatitic intergrowths as keys to eutectic proportions, because I do not wish to write a treatise, but rather a review—I fear too long—pointing out what seems to me inferences to be drawn from Iddings’s diagrams. But the loss of mineralizers might change ithe eutectic, Porunstanee, iF theseutecticsbe, Na, Ox Al,O, + 6SiO, =-6H,O -Si@,, a loss of Hi) © might mean the replacement of the H,O by %(CaO-Al, O,) which might be a chemical quantitative change entirely worthy of recognition in a quantitative chemical classification, zf it proved of sufficient importance; as much so as any magmatic split. I think that the indications of Iddings’s diagrams are that it is not from a quantitative chemical standpoint of primary impor- tance, although his figures take no account of the water. It is not to be forgotton that in a holocrystalline rock the H,O of | the magma, while it may be in combination in biotite, analcite, etc., is very likely to be concentrated in drusic and microdrusic cavities, and not be noticed in the analyses at all, or, if at all, then in the water given off below 110°; but a maximum idea of this quantity may, however, be derived from the porosity, which in the dike mentioned above is for the more crystalline part less than a third of I per cent. While, therefore, Iddings’s diagrams do suggest a natural grouping and classification based on the various eutectics of various magmas, yet the time is not yet ripe for such a perma- nent arrangement. We do not know enough about the eutectics. In the meantime, the new system of pigeon holes has some advantages, and many, especially of the minor groups, may endure. Still we can see that the old divisions of rocks might be grouped around the average rock and given a more precise chemical meaning, as follows: 1. Acid, z.¢@., alkali-silica ratio 0.013 —, and silica percentage 0.58-++, or limited as in diagram, eutectic ratio toward which crystallization takes place, alkali: SiO, ::1: 12. Granites and diorites, and many syenites. ‘syIOI SNOAUSI IY} JO UOT}VOYISSeIS jeorurayo aatyjeyjuenb yeinyeu ve jo Ayyiqissod ay} pur ‘auty onoayna urew ay} Suoje Bursaysnyjo [wanyeu ay} Moys OF J ld s,SSuIpp] wo1 peylpow weisviq o s+ os $3 03 s9 on SL o8 $8 06 sé 001 Ht HH ; Ht HE Stages H H + t t t T pases’ z T t t I rrr a Ht SHEET See Hf i = t Ht + i¢ 1 - eeeaaee t f : t i ne . + tt T ~ 5 + t t t oy t tt 1 1 i H +H3 f a : ite : ! t ii ry ty t + iH o P iin HH H Ht t eas ttt | t HH iH Ht t ag : = : i ee Pe re Ht I t t oat i : tH + : : aoes seeiees tie a iH 1 se oh e Ht aeeseeeseceget H eesaea! H posses! Seeenanay xi S oz ae : a { i 4 4 | i tH L 1 ee H t t 7 +4 t oe ry a H+ 4 : +H t } t ¢ : $2 anes r tt +} 4 +H H+ a +} Hf Q2 ALT RED i Crwele AINE The general rules for computing the norm of a rock can be simplified if made applicable only to this division. 2. Basic. Silica percentage, 0.58—, bounded, as shown on diagram perhaps, or by some other line expressing the fact that the eutectic is not melilite, but augite. Eutectic ratio, Ca: (Migs He) SiOr a sie 1e27 Basalts, gabbros, peridotites, etc. Computing the place of such rocks in a quantitative system is quite simple. “All the alumina can be combined with potash, soda, and _ sufficient lime and counted salic, everything else femic, so that the ratio sal: fem is quickly found, and the ratio of soda to potash and alkalis to lime are found incidentally. 3. Alkaline. All other rocks, which obviously should also be subdivided, perhaps, as shown in the diagram, into a femic and salic group, and farther yet. What ought to be done is carefully to study the whole field, with due regard to magmatic splitting, watching the last crystal- lization, and determine as nearly as may be what are the eutectic ratios in the silicate magmas. Then the work should be experi- mentally verified, in the new geological laboratory that I hope we are to have. Then, finally, we shall no longer have to envy the metallographist,’ who measures the areas of micropegmatite (eutectic) copper and copper oxide, and areas of solid copper, and says: ‘‘So much eutectic with 3.45 per cent. Cu,O, and so much plain copper: there must be just so much copper and so much oxygen.” In such study it is the last-formed minerals, with a little of nearly everything in their molecule, which show by their compo- sition the eutectic proportions of the different constituents. Since writing the above I have noticed in the Bedlage Band XVII of the Mewes Jahrbuch, p. 516, and especially pp. 546--64, an article by Schweig, and in the Centralblatt, 1903, p. 605, a note by Linck on a series of experiments exactly along the line ‘HOFMAN, GREEN, AND YERXA, “ A Laboratory Study of the Stages in the Refin- ing of Copper,” Zransactions of the American Institute of Mining Engineers, October, 1903. EUTECTICS IN ROCK MAGMAS 93 suggested, and in part only confirmatory of the suggestions above. For instance, Linck (pp. 606, 607) found a magma with 49.05 SiO, and an alkali silica ratio of 0.115, after absorbing all the silica it could at 1300° C. change to a magma with 52.20 SiO, and an alkali silica ratio of 0.082, almost precisely on the supposed eutectic line. Another with 62.62 per cent. silica absorbed a mixture of ferrous and ferric iron until the silica per- centage dropped to 60.58 and the magma could be nearly expressed mds (Nal@) On Al OF 26 SiO7 =, CaO Al Ox 2 SiO, +3 (Ca Mg, FeO) SiO,. In Schweig’s extensive series of experiments, starting with a strongly alkaline magma (17.5 per cent. alkali molecules), then adding separately silica, alumina, iron, magnesia, and lime to saturation, and then later adding also silica to saturation with the other oxide, only in case silica is added to saturation do we find analyses which are comparable with any of the rock analyses plotted by Iddings. The inference is suggested that the natural igneous magmas of the alkaline group are always able to absorb all the silica they will take. When magnesia or iron oxide or alumina is added with the silica the alkali-silica ratio drops only to about 0.15. But the magma is much more capable of absorbing lime, and after satura- tion with lime and silica the alkali-silica ratio becomes 0.895 and the SiO, 70 per cent., bring it well into the eutectic belt, which apparently therefore is that of magmas saturated with lime and silica. All the glasses were cooled as quickly as possible, and no attempt was made to determine the temperatures of solidification. It would be interesting to repeat the experiments and note the latter. ALFRED C. LANE. LANSING, MICH. A PRACIURE VAAIIbR YS Vsil reve THE correspondence between the drainage system in the region north of the Yellowstone National Park and the system of fractures traversing the rocks is so striking that there can be little doubt of a causal relationship between them. The studies of de la Beche in England, of d’;Omalius and Daubrée in France, of Kjerulf and Brogger in Norway, and recently of Hobbs in Connecticut have demonstrated that in certainlocalities there is the closest correspondence between the valleys or drainage system and the fractures in the underlying rocks recognized as faults or joints. And Daubrée in his classic work Etudes synthé- tiques de géologie expérimentale, has shown how a more or less rect- angular system of nearly parallel joints may be produced by torsional stress as well as by compression. The almost universal application by geologists in former times of the idea of the controlling influence of fractures on drainage, and the exaggeration of the importance of faulting in this con- nection, coupled with the evident independence of many drainage courses from faulting, factors, led to a revulsion on the part of modern physiographers and their obvious dependence on other from the views of earlier geologists to such an extent that the influence of rock fractures on drainage courses has been mini- mized, if not altogether neglected, in recent times. One reason for the revulsion from the idea of fracture drain- age systems is to be found undoubtedly in the emphasis formerly laid on the conception of faulting as an essential element in the problem. This was introduced in the term “ fault valleys,” (vadlées de failles) of d’Omalius, and has been memorialized in the vignette on the title-page of the Annual Report of the United States Geologt- cal Survey. That such valleys occur is well known, but that they form a small fraction of the whole is beyond dispute. It is to be noted, however, that the demonstrations of Daubrée deal chiefly with joints or fractures (cassures), rather than with faults (fazlles), the two being but different phases of the same 94 A FRACTURE VALLEY SVSTEM 95 phenomenon, fracturing; one, a result of small or imperceptible displacement; the other, of profound, or at least notable, disloca- tion. The torsional experiments of Daubrée produced more or less rectangular systems of nearly parallel fractures, some of which might be called joints, and others faults. The studies of Becker,t Van Hise,? and Hoskins3 have established the laws and frequency of parallel and of conjugate joints (systemes conjugués of Daubrée) and the phenomena have become familiar to all geologists who have worked in regions of well exposed disturbed rocks. The phenomena are in part as follows: There are faults in more or less parallel lines through wide extents of territory inter- sected by faults at various angles often nearly go”. A dominant or major fault is frequently accompanied by par- allel fractures of minor degree, which are in some cases Close to the dominant fault, in other cases at considerable distances from it. Faulting is accompanied in parts of its course by crushing and brecciation of the wall rocks; in other parts the sides of the fault are closely pressed together without evidences of breccia- tion. Further, there are portions of a fissure plane where the walls are not closely compressed, but may be actually open. Planes of fracture and fissuring are often lines of special decomposition of the rock material, and are for this reason less resistant than the neighboring rocks. From these facts it is evident that the course of drainage, in following lines of least resistance, may cut its way along fault lines where they are in brecciated, loosely aggregated, or decom- posed rocks, but may leave the faults in places where unbroken rocks are closely compressed. Moreover, streams may find less obstacle in cutting unfractured softer rock than in removing fractured harder material. ™Gro. F. BECKER, Monograph III, U. S. Geological Survey (Washington 1882), pp. 156-87; Bulletin of the Geological Society of America, Vol. IV (1893), pp. 41-75; and Zransactions of the American Institute of Mining Engineers, Vol. XXIV (1894), pp. 130-38. 2C. R. VAN Hise, Szxteenth Annual Report, U. S. Geological Survey (Wash- ington, 1896), pp, 633-78. 3L. M. HoskINs, zé7@., pp. 845-74. 96 JOSEPTT PS IDDINGS Numerous parallel joints would aid erosion, if it were once located along a dominant fissure, by permitting the falling in of the sides of the drainage channels. Consequently, it is reason- able to expect that planes of fracture will in some cases become lines of drainage and pronounced erosion, while in other cases they may be disregarded by streams. It is also known to every field worker that dislocations in large areas of massive rocks are easily overlooked, and are often diffi- cult to determine when suspected; and further that minor fract- ures are generally neglected in geological field work. It follows from this that much evidence that may exist connecting the location of drainage with rock fractures has not been collected, and much that might be sought, by the very nature of the prob- lem, may not be found, because the bottom of a valley is usually filled with loose material that conceals the rocks through which the valley has been cut. A study of a mountainous region such as that lying north of the Yellowstone National Park must convince one that fracture systems have had more influence on the location of drainage than is ordinarily supposed, or than can be actually demonstrated, perhaps, by the evidence obtainable from the region in its present condition of rock exposure. With regard to the presentation which follows, it is to be remarked that the problem of a possible fracture drainage sys- tem was not in mind when the writer was in the field, and no special search was made for evidence bearing on the question. The argument offered is based on such observations on the struct- ure of the region as were made in the field by Mr. W. H. Weed and the writer, and on the drainage features of the map prepared by the topographers. The topographic map which is reproduced is a reduction of that of the Livingston Quadrangle, Folio 1 of the Geologic Atlas of the United States. The geology of the district may be found in the folio. In this connection the map of the Three Forks Quadrangle Folio 24, and those of the Yellowstone National Park, Folio 30, should be studied." 1U. S. Geological Survey, Geologic Atlas of the United States, Folio 1, Washington 1894; Folio 24 and Folio 30, 1896. HA APRN CIS S. OMI AINE SIESIS INL Q7 A simplified drainage map in which the drainage lines are some- what straightened is given on the thin sheet, which is intended to be placed over the topographic map. The purpose of straight- ening the drainage lines is to present a simpler expression of the system, rendering the persistency of some of the directions more evident, and permitting the relationships of the various directions to be more easily noted. It does not follow that drainage channels would be more in accord with fracture systems if they were straighter. Fractures and faults are not necessarily straight. They are more often curved or crooked, as is the case with most of those observed in the region under discussion. The fault lines and dikes are printed in blue on the thin sheet. It is probable that the irregularities of direction in the drainage, as it is drawn on the topographic map, are more in accord with the fractures than the straightened lines traced on the drainage map. The value of the tracing is in the simpler expression of the system, serving the same purpose as a simpli- fied statement of a highly complex set of relationships. It may be looked upon as a diagrammatic statement of the drainage system. The region embraced by the map lies between the meridians MOmandel tie vandilatitudes 45 - land 46... 1t is immediately, north of the Yellowstone National Park, and contains the Snowy Range, the eastern slopes of the northern portion of the Gallatin Range, part of the Bridger Range, and the southern slopes of the Crazy Mountains. The Yellowstone River traverses the quadrangle from the south and southwest to the northeast, its tributaries intersecting the country in all directions. A study of the topographic map reveals the angular char- acter of much of the drainage system, and the prevalence of certain parallel and sub-parallel lines which appear in various streams and occur in quite diverse portions of their channels. Along parallel lines different streams may be flowing in opposite directions; one stream may be near its source, another near its mouth, having other portions of their channels trending in other directions. The persistency of these lines becomes more striking when the geological structure of the region is taken into account and it is observed that certain drainage lines traverse rocks of 98 JOSE REG PS LD DILNG:S: such diverse nature as gneiss and schist, volcanic tuff breccia and solid lava, limestones, sandstones, and shales. The relation of some of these directions of drainage to known fracture planes will be pointed out. The dominant drainage lines in the southern three-quarters of the quadrangle trend about _northeast-southwest and northwest-southeast, more nearly N. 30° E. and W. 30° N., the angle between them being approxi- mately 90°. There are other systems of almost rectangular lines somewhat differently oriented, namely, north-south and east-west. These are well developed in the central eastern part of the quad- rangle. Let us consider these systems in some detail, commencing with the main drainage channel, that of the Yellowstone River. This enters the quadrangle on the south in a curiously zigzag channel, known as the Third Canyon, cut in crystalline schists. The longer zigzag lines run northwest, and the shorter are almost at right angles. The northwest direction is followed from Gar- diner to Reese Creek in Cretaceous strata, is exchanged for a more northerly direction as far as Yankee Jim, where the north- west direction is followed through a narrow gorge into the open valley, where a right-angled turn is made to the northeast. From Reese Creek to the open valley the river traverses crystalline schists. The dominant northwest line of the river just described is parallel to the most profound fault within the region; namely, that which threw the whole sedimentary series, including the Laramie Coal Measures, down below horizons of crystalline schists. It is a fault of more than 11,000 feet and probably is more than 16,000 feet. This fault appears to end abruptly at Cinnabar Creek, the principal throw being southeast of Reese Creek; It is interesting to observe that the Yellowstone River in no place exactly follows the fault line as it is located on the present surface. But it is evident from the topography of the country south of the river and east of Gardiner that parallel fractures must exist on both sides of the fault line, and these must control the northwest direction of the tributary streams and the longer A FRACTURE VALLEY SYSTEM 99 lines of the zigzag Third Canyon. These lines persist to the southeast within the boundary of the Yellowstone National Park, the fault itself being covered by younger volcanic lavas. As already remarked, the great fault appears to end at Cin- nabar Creek, but the Yellowstone River follows a line parallel to it through the gorge at Yankee Jim, and parallel lines of lesser drainage are plainly shown in the vicinity. Whether the fault actually ceases at Cinnabar Creek cannot be determined, because of the covering of volcanic rocks forming the surface of the mountainous country westward. Beyond these lavas, however, in a direct line a similarly profound fault is exposed in the Madison Range in the Three Forks Quadrangle. This was accom- panied by a similar throw of the sedimentary strata south of the crystalline schists, and is in a direct line with the fault just described. The total length of the combined fault lines is over sixty miles. Parallel to this great fault are several well defined ones of less extent, the largest lying north of Mill Creek Basin. Here there are two sub-parallel faults which unite in the head of the North Fork of Mill Creek. The dominant one of these has been traced from the Boulder Canyon westward to the Yellowstone Valley, a distance of twenty miles. The shorter fault has been traced for thirteen miles. They are not observed west of the Yellowstone Valley in the Gallatin Mountains because of the covering of volcanic lavas, but they appear again with the same characters in the Madison Range just west of the lavas, evidently passing beneath them. The western faults are exposed for twenty miles, and there can be no doubt of their persistence beneath the lavas. In this case the total length of the principal fault would be sixty-four miles. In the bare gneisses two miles north of the fault line at the head of the North Fork of Mill Creek joint planes or small faults are clearly visible, parallel to the main fault and having a hade to the south, indicating normal faulting. In other parts of the region there are faults parallel to those just described. A small one within the area of crystalline schists occurs in the southeast corner of the quadrangle, crossing 100 JOSEPH 2. LDDINGS Slough Creek. It is exposed for a short distance only, being covered at both ends by volcanic rocks. A still smaller one was observed six miles farther north. These indicate the presence of fractures parallel to the profounder faults. North of the body of crystalline schists a northwest fault exists in the valley of the East Boulder River ; another is situated south of Livingston Peak; and two smaller faults occur between this mountain and Livingston. Two more are located just west of the Yellowstone River in this vicinity; and there is a short one northwest of Mount Ellis, in the central western margin of the quadrangle. The presence of major fractures in a northwest-southeast direction being clearly established, the occurrence of parallel fractures of minor importance is rendered highly probable, some in fact, having been observed in the field. Corresponding to these northwest-southeast fractures are the channels of many of the tributaries of Slough Creek, Buffalo Creek, and Hell Roaring Creek; the minor branches and the main chan- nels of many of the streams flowing into the Yellowstone, as far as Livingston, notably those in the Gallatin Range. The most remarkable instance, which is a good example of a fault valley, is the valley of East Boulder River. Returning to the consideration of the course of the Yellow- stone River, it is seen that upon leaving Yankee Jim Canyon it flows at right angles through a broad valley in a northeasterly _ direction for forty miles. More strictly, the first twenty miles are about N. 30° E., the next sixteen miles being more northerly. From the mouth of Shields River beyond Livingston the course is south of east, nearly at right angles to its course at Livings- ton. This is followed by a right-angled bend at McAdows Canyon, after which the course is northeast to the edge of the quadrangle. The broad valley of the Yellowstone between Yankee Jim Canyon and the Lower Canyon, which valley continues south- west up Tom Minor Creek, lies in the direction of a scarp fault forming the western flank of the Snowy Mountains. The sum- mits of gneiss and schist at 11,000 and 10,000 feet from Emigrant ANT RA CHORE VALEE VCS VS EMM IOI Peak to Mount Delano rise abruptly on the east side of the val- ley, while on the west there are the Jong sloping spurs of volcanic lavas overlying sedimentary strata in the northern part of the Gallatin Range. The location of this fault was not discovered in the field, as it is undoubtedly obscured by the valley deposits. It dies out abruptly before reaching the Lower Canyon, near Livingston, and is probably most profound north of the cross fault at Mill Creek. This northeast direction is the same as that of two pronounced faults in the southwest corner of the quadrangle that enter it from the Yellowstone National Park. One lies in the drainage channel of Cinnabar Creek; the other is in the valley of Reese Creek. It is probable that a third fault parallel to these occurs in the valley of Gardiner River west of Mount Everts, but it has not been definitely located. These faults terminate in the great northwest fault in Yellowstone Valley south of Sheep Mountain. The throw of the Cinnabar Creek fault is to the west, but the extent of the displacement is not determinable. It has been traced for a distance of eighteen miles. The Reese Creek fault is clearly recognizable east of Electric Peak, where the throw is to the east and the displacement more than 6,000 feet. It con- tinues southward as a scarp fault along the east flank of the Gallatin Mountains in the. Yellowstone Park. It is known for twenty miles, and disappears under lava. The throw of the fault west of Mount Everts is to the west. A minor fault parallel to those just described occurs in the gneiss east of, and parallel to the channel of Hell Roaring Creek. Several northeast-southwest faults of slight extent have been noted north of the great body of crystalline schists. One is a short spur connected with the fault in the valley of the East Boulder River. Two are connected with the northwest-southeast faults southeast of Livingston, and another is south of Mount Ellis in the western part of the quadrangle. In each of these cases it is interesting to observe that there are conjugate faults at nearly right angles. In the direction of the northeast-southwest fractures, besides the Yellowstone Valley and Cinnabar and Reese Creeks, which 102 JOSEPH P. IDDINGS are definitely located on fault lines, there are the channels of Bear Gulch and Crevice Gulch, Hell Roaring and Slough Creeks, with tributaries to Boulder River and Mill Creek, and the upper portion of the West Boulder. Another conjugate system of fractures is indicated by the rect- angular drainage with lines almost north-south and east-west. The direction of the main Boulder River through its canyon is the same as that of several of its head branches, of Buffalo Creek to the south, and of several north-south drainages to the east. The east-west direction is found in tributaries of the Boulder River, in the North Fork of Mill Creek, and in a stream in line with this east of East Boulder plateau. In this case there is a fault nearly parallel to the North Fork of Mill Creek connected with the northwest-southeast system of faults. An east-west fault has also been observed by Mr. W. H. Emmons near Hay- Stacl:\Peak at the head’ of Boulder River. ~ here cam belittle doubt that in this portion of the quadrangle there is a system of fractures in an east-west and north-south direction. The sculpturing of the canyons of the Boulder and West Boulder, and of several to the east, seems to require the action of other agencies than those of ordinary erosion. These narrow canyons have been cut 3,000, 4,000, and 5,000 feet into gneiss and schist within a few miles of their heads, and their drainage basins seem quite inadequate to furnish sufficient water for so great and such deep erosion. It seemsas though the rocks must have been rendered more susceptible to abrasion by being fract- ured or jointed. A similarly placed system of drainage channels exists in the northwestern corner of the quadrangle west of Shields River and east and south of the Bridger Range, but no system of fract- ures or faults has been noted in this region. The northeastern corner of the quadrangle is intersected bya distinctly rectangular system of drainage. The channels of the Yellowstone and Boulder Rivers, with the tributaries of the Shields River, have a northern trend nearly at right angles to those of Shields River and several. creeks flowing southeast from the Crazy Mountains, and numerous small tributaries of the A FRACTURE VALLEY SYSTEM 103 Yellowstone and the Boulder Rivers. That these directions are parallel to existing fractures is shown by the trend of dikes of igneous rock that traverse this part of the country. A number of dikes occur in Crow Indian Reservation onthe Boulder River. They trend northwest almost exactly parallel to the small streams in this vicinity. Two longer dikes occur on Gage Creek, trend- ing in the same direction. In the valley of Shields River there are dikes having a northeast-southwest trend. The Crazy Mount- ains are filled with innumerable dikes radiating in all directions from the core of igneous rocks lying north of the Livingston quad- rangle. That portion of the mountains within the limits of the map is traversed by dikes trending north-south in the middle of the group, on the east side trending east of south, and on the west side trending west of south, but not parallel to the drainage channels in the foothills. Apparently these fractures are immedi- ately connected with the intrusive core of the mountains, and have not extended into the surrounding sedimentary strata for long dis- tances. The system of fractures indicated by the drainage is prob- ably more profound and was produced by deeper-seated forces. There are dikes and lines of intrusive rocks parallel to neigh- boring drainage channels in the vicinity of Haystack Peak, Emi- grant Peak, and at the west base of Sheep Mountain. Besides the drainage that may have been controlled or initi- ated by fracture lines, there are within the area covered by the map excellent examples of channels that have followed softer rocks, where the minor topographic features of the country con- form to the position and character of the sedimentary strata, and where fracture and faulting appear to have had no influence on the drainage. It is the larger valleys and dominant channels that exhibit the general relationship between drainage and facture: That the drainage system of this quadrangle is closely related to the structural features of the rock formations will not appear surprising when the geological history of the region since Lara- mie times is taken into account. The dislocation of the crystalline schists and sedimentary strata that followed the deposition of the Laramie resulted from 104 FO SEPTATE ALD INIG:S} torsional shearing stresses of great magnitude, as is shown by the rapid variation in the amount of displacement along the planes of faulting. These fractures have occurred in nearly parallel planes, and also in a nearly rectangular system, and resulted in faults and joints. The fractured and dislocated rocks were eroded to a very great extent at a very early period. Thus from a large area the entire sedimentary covering, 10,000 feet or more in thickness, was removed from the underlying crystalline schist before Eocene times, and the surface of the country presented an irregular topog- raphy not greatly different in character from that of the pres- ent day. Upon an irregular surface of crystalline schists in the present valley of the Yellowstone River, just south of the Living- ston quadrangle, there rest horizontal beds of Eocene volcanic tuff with ancient tree trunks in vertical position. Erosion has reduced the rocks in this vicinity to nearly the same surface at successive periods, as 1s shown by the occurrence of surface extru- sive lavas of very different ages in close proximity. The direc- tion of the drainage in this locality has undoubtedly shifted repeatedly during this long lapse of time. Since volcanic activity commenced, fracturing and faulting have taken place at intervals through the Tertiary period. The profoundest faulting and erosion took place before Eocene times and the extravasation of Eocene tuffs. This was followed by great accumulations of Miocene tuffs and lavas, which were fractured and dislocated along the faults in Cinnabar and Reese Creeks, and elsewhere. These were greatly eroded in late Ter- tiary time before the eruptions of Pliocene rhyolite which has been faulted in its time. In the faulting of the Electric Peak and Sepulchre Mountain blocks there was unquestionably a displacement along the line of the earliest northwest-southeast fault against which these blocks terminate proving a recurrence of fracturing and displacement along old lines of jointing and faulting. And while there is little or no evidence in this region that successive faulting has often taken place at widely remote geological periods along the same lines, it seems probable that profound jointing may establish Vp, | f. ig a a gees ay SN XY pie i BREEAM : me; = 74 Ss SS { Cea es vA ~ ye Tr : . Zg AZ Se ZL Oa ea Y al A ee yf. Coy I/y, is | ‘SS in as JOS ie Sass qe ( ame: ys ah che eM SE i, linet eben \iralondaney ‘ena; seaeste Vaogeanbarnnen BPs vivid f NY “Ai aad 7 ' 7 en errra \ { : 4 ; Fe L} qavracient ai aN Ya Sars i PATI 4 re ONT, Loco NOS Revere ae). eae “4 Fs egress me 5 i my, hi 1 sPb Adel rl rd mn Sear taionaianl pipiens © ue fone FM el oS Saar are aemce eT ear Sree Sm CNRS TEN de: a MONTANA LIVINGSTON SHEET ey = Pid oa S hee 0 Ws RS = DSWIS U.S.GEOLOGICAL SURVEY. yew POWELL. DIRECTOR Se fai My Z SANS Sy =, = y os ANT Be ANS ar | S i y Zw »») \ = Py: = tS 4. AAA WIG 15 : J aw KANE A } Ny A ety as f X aD ap) MMS ZZ = GA a, EN = See « y S y V6 = Y 4A, A ie = a0 eS fs oS 2 ~ Sh rat (ap == — yc S TC G PEN =< 7 an ) XY- ~< Se os 5 = i g = i? IK cS, UD) RD 4 N aa J. Olas b (ER i we : (INS AO SA ( "| 2 ee ~, > g s \ii XQ BEER CUI ) GIONS aN AAAS MEE CS A W'S =a \ Gee SS Sie S; eee eS, ee Contour Interval 200 feet bicker Pesta Ry SeeRem eT es i : | | U,S.GEOLOGICAL SURVEY. Wie JW POWELL, DIRECTOR os ESSHG\ \ AG = See SS Bo nS MONTANA LIVINGSTON SHEET =i a WOM AY y Cee, BS % ( of MES Ve dees f= ve : py man: cE é A PRA GCLORE VALLEY SVSTEM 105 planes of weakness along which subsequent movements of a minor character may take place affecting superimposed rocks, so that a system of Eocene, or even late Cretaceous, joints may be extended upward into overlying rocks in Miocene time provided similar dynamic action of a pronounced character occurs in both periods as it has in this region. JosEepH P. Ipp1ncs. CUSPATE FORELANDS ALONG THE BAY OF QUINTE. TABLE OF CONTENTS. INTRODUCTION—THE BAY OF QUINTE. MOVEMENTS OF THE WATERS OF THE Bay OF QUINTE. Currents. Waves. THE FORELANDS AND BARS. . Sand spit below Bogart’s dock. . Gravel bars on Picton Bay. . Terrace and bar near Allison’s wharf. . Prinyer Cove spit. . Pleasant Point spit. . Fish Point spit. . Amherst bar. . Calf Island loop bar. THE ORIGIN OF THE V-TERRACE AND V-BAR. CON Am FW NN CONCLUSION. INTRODUCTION—THE BAY OF QUINTE. THE flat-lying limestone regions immediately to the north and east of the east end of Lake Ontario are traversed by a number of deep valleys with graded side slopes on their lower courses. These valleys are probably of preglacial origin, and were carved at a time when the relative altitude of the several parts of the Ontario lowland was different from what it is at present. The partial submergence of a number of these valleys, tributary to one another, has formed the water body known as the Bay of Quinte. This bay extends from near Kingston, at the east end of Lake Ontario, toward the southwest for a distance of over fifty miles,"and nowhere has it a breadth exceeding two miles. A reference to the accompanying general map will show its remarkable zigzag course.*. For purposes of study it may be tFor a discussion as to the probable origin of this valley see “The Trent River System and the St. Lawrence Outlet,” Bulletin of the Geological Soctety of America, Vol. XV. 106 i aro) 0 | TO MOM Sali Ul 97090 ALNIQO 2° AVE av W lt ZA 1 —_F, @\ F ~Z WAN SW Zz WAZ \\ f Woy Wa ay : | - OD als CCN Zw a\\ S , ZA f& AAT —Zas UY MAS GAY 2 AC uojsbury 22UDd VN KX 108 ALFRED W.-G. WILSON divided into three parts: the Trenton-Desoronto section, trend- ing a little north of east; the Desoronto-Picton section, trending to the west of south; and the Picton-Kingston section, trending nearly northeast. The upper section is comparatively shallow; for the most part the shores are rocky; and no characteristic cuspate forelands have been noted along them. The middle section, sometimes known as the Nine Mile Reach, has much deeper water, and the valley sides are steep, often inaccessible cliffs of Trenton limestone. The maximum relief is about 185 feet. Much of the shore is rocky, but along the east side there are, in places, small amounts of modified drift lying between the water’s edge and the front of the adjacent escarpment. In one place, a short distance below Bogart’s dock, shore drift derived from this material has formed a small fore- land of fine sand which resembles the V terrace with the rimming bars which Gilbert describes as occurring on the shores of Lake Bonneville. On the west side of Picton Bay there are also two small spurs of shore drift which seem to be associated with talus cones from the face of the cliff. Along the third section of the bay there are four excellent examples of the cuspate foreland and one long flying spit. Some of these cuspate forelands have a remarkably close resemblance to the V terraces and V bars of Lake Bonneville. Parts of the shores of this section of the bay are also rocky, but the amount of drift, both till and stratified material, is greater than else- where. Off the west end of Amherst Island the water has its maximum depth of 230 feet. The valley reaches its maximum relief of 284 feet near Glenora. The south shore is bordered by the steep escarpment of a cuesta which rises about 200 feet above water level near Glenora. The height gradually decreases eastward, and in Amherst Island it is only about 50 feet. The north shore rises gently inland. On the south shore rock exposures are numerous; on the north shore glacial drift fre- quently occurs, bed-rock less often. Of the four cuspate fore- lands to be described, three occur on the south shore; the flying spit is located at the extreme eastern end of Amherst Island, also on the south side of the bay. CUSPATE FORELANDS AT BAY OF QUINTE 109 The material which forms the loose débris of the shore is in part derived from the wasting of the cliffs, in part from the gla- cial deposits. The material which forms the single spit which occurs on the east side of the middle section, and also that of the spit which occurs on the north side of the eastern section of the bay, seem to be wholly of glacial origin. The materials of the three forelands and the flying spit which occur along the south side of the eastern section of the bay are largely derived from the bed-rock where it outcrops along the shore, but there is a slight admixture of gravels derived from the glacial deposits. MOVEMENTS OF THE WATERS OF THE BAY OF QUINTE. Currents —Before describing each of the spits in detail, and discussing the question of their origin, it is considered advisable to say a word about the movements of the waters of the bay. As is well known, there are no appreciable tides on the Great Lakes; hence tidal currents do not enter as a factor in the dis- tribution of shore waste. The volume of water discharged by tributary streams into the upper part of this bay is considerable, but its ratio to the total amount of water in the bay is so small that no appreciable outflowing currents are set up. It is alto- gether doubtful that any portion of the bay water below Deso- ronto has a normal current from this cause of over a mile per day. The seiches of Lake Ontario periodically affect the height of the water of Kingston. Accurate data are not at hand to permit of any statement of their exact periodicity, but by calculation it should be about sixteen hours between wave-crests. The change at Kingston ordinarily does not exceed a foot and a half, except during and after exceptional storms, when it is much greater. The water that is backed into the bay at the time the crest of the seiche-wave is at Kingston must theoretically cause an oscillatory movement in the bay, as the crest and trough of the seiche- wave travel up the bay. At Napanee, at the head of the naviga- ble portion of the Napanee River, about seven miles above Desoronto, this seiche-wave often makes a difference in water level oteaboute 3 teets =) Hlere, however thes water is), backed I1O ALFRED W..G. WILSON into a narrow funnel-shaped opening. Out on the open bay very slight changes in level are occasionally noticeable, but no records of their amounts are available. It may, however, be stated that they are very slight, and at no time, except at the upper part of the Napanee estuary has the writer been able to determine the existence of any noticeable current due to this cause. It may be stated that the currents in the bay produced by this cause are not capable themselves of transporting any of the material which is moved along the bay shore. It is true that they may slightly accelerate or retard the currents which are concerned in the active transportation, but they are much too weak to be in any way considered as active and effective agents in transportation. Where they have been observed at their maximum the water is perfectly clear, although the bottom is covered with a fine mud which settles rapidly when stirred. Approximate estimates as to the strength and importance of the seiche currents can also be made at the Murray Canal. This canal is four miles in length and connects the upper end of the bay with Presqu’Isle Bay, this latter bay connecting directly with the open lake. The crest of the seiche reaches Presqu’- Isle Bay some hours before it reaches Kingston. Consequent on the rising of the waters at Presqu’Isle Bay a current sets in eastward through the canal to the head of the Bay of Quinte. Some hours later the crest of the wave advancing from the Kingston end of the bay, having had about 110 miles farther to travel, reaches the head of the bay, and occasionally may start a current through the canal in the opposite direction. Unfortu- nately, it has not been possible to carry on simultaneous observations at several points on the bay, nor at any one point continuously for a long enough period to establish the time rela- tions of these oscillations. The existence of the currents through the canal has been established. These currents in the canal are farther complicated by wind-action which generates surface cur- rents. From observations made during periods of calm weather the author would infer that the current to be attributed to the seiche alone rarely exceeds five miles per day. It must be noted that until careful quantitative observations are made there can be CUSPATE FORELANDS AT BAY OF QUINTE Ielar no definite statement, for even during calm weather the momen- tum of wind-generated currents causes them to continue for a considerable period, and it is difficult to distinguish positively between these residual currents and the true seiche current. In the absence of accurate observations of the time of oscil- lation of the seiche, we have no means of knowing whether the crest of one wave starts the currents through the canal from one direction; at {the * time’ when the trough of another is at the opposite end of the canal; in other words, we do not know whether the current periodically reaches its greatest possible maximum value. Assuming a mean depth in Lake Ontario of 65 feet, and making some allowance for retardation of the advance of the wave-crest up the narrow bay, a calculation of the time of oscillation of the seiche along the line of direction of the most prevalent storms suggests that the periodicity of coincidence of crest and hollow at opposite ends of the canal will not be the same as the period of the seiche. The mean depth of the canal is 11 feet; the breadth at the bottom, 80 feet; the breadth of the water surface, 125) feet. elit may be inferred from the small volume of water moved through the canal by the seiche current that in the much broader, deeper bay the actual currents generated by the seiche oscillation must be very slight. Waves.—The effective agents in the transportation of the shore débris are wind-waves, and the longshore currents which are associated with them. The size of the material transported and the rate at which it travels must necessarily depend upon the strength of the waves; these in turn depend upor wind velocity, and, in the Bay of Quinte, upon wind direction. Observations which have extended over a considerable period have shown that resultant effective transportation along the shores of the Great Lakes depends in part upon the direction of the most prevalent winds, in part upon the length of the stretch of open water across which the acting wind has come. The larger storms usually determine the resultant direction of transportation. Now, in the case of the Bay of Quinte, the steep sides of the valley in which the waters of the bay lie so guides and controls the winds I12 ALFRED W. G. WILSON that we find that the efficient wave and wave-current work in shore transportation is done by those winds whose direction con- forms nearly with the axial direction of the several sections of the bay. The narrowness of the bay, coupled with the depth of the valley, is such that even violent storms blowing across it can do less efficient work than is done by the much gentler local breezes blowing up or down the bay. In this locality the prevailing direction of the wind during the summer is from the southwest; but, in spite of this, it is found that, because of the considerations to be noted below, there is virtually no continuous. transportation eastward except along parts of the lower portion below the Upper Gap. There seems rather to bea constant oscillation to and fro. Because of the shape of the bay and its position the directions from which efficient winds and their accompanying waves can come are the northeast and the southwest. The material which forms the forelands varies from fine sand in one example to large rock plates weighing over four pounds each. All the spits but one are built of coarse and fine gravel or shingle. In most cases the material is almost all so coarse that its transportation must be attributed to the wave itself, rather than to the action of any longshore current during the intervals that the wave may have raised it off the bottom, though no doubt these currents assist in that transportation to a small extent. It is moved in part by rolling along the bottom, but even some of the largest fragments are frequently lifted clear of the bottom and carried along with the wave. The shape of the oblong or rhomboidal plates (rarely over an inch thick, and with an area on the flat side, varying from ten to thirty square inches) materially facilitates this mode of transportation. THE FORELANDS AND BARS. 1. Sand spit below Bogart’s dock.—This is a small spit which consists wholly of fine sands derived from the adjacent cliff cut in modified drift. The spit measures about 245 feet across the base and extends about 100 feet out from the shore line. The normal width of the beach between the cliff front and the water GOUSPATERAOREPEAND SCAT BAY OPLOCINTE Ios is about ten feet. In its present attitude the axis of the spit inclines toward the southwest or slightly down the bay. A ref- erence to the accompanying sketch plan will show the present existing conditions. There is a central triangular terrace at water level, marshy, but filling with sand which drifts in or is washed in by rains or waves. Bordering this are two distinct sand ridges rising about 2 feet above water-level. The outer ridge has imponded a small amount of water between itself and the inner ridge. A third ridge has been begun on the outside of these two. ; Referring to the general map, it will be seen that effective MT WN: Miki I Fic. 2.—Small sand spit below Bogart’s dock, June 1, 1903. transportation must always be by winds blowing nearly parallel to the axis of the middle section of the bay. The present shore line, both above and below, is certainly just as irregular as it is here—it would be described as slightly wavy. There is no stream discharging near here, and there is no evidence of a local landslip having modified the shore line in such a way as to cause the beginning of the building of the spit at this point. In the field it was at first very difficult to see in this case why it should have happened to be formed here and not at a half-dozen other apparently similar places. It happens, however, that there is a very slight, though noticeable, difference in the curvature of the shore line at this place, and it seems as if, under certain special conditions of wind-action from the east of north, the longshore wave and wave currents first started to build a terrace.and later. a bar outward from the slight salient in the shore line of this point, and that the same waves gradually turned the end of this free bar as it reached deeper water, giving it its curved form, and finally tying it on to the shore again. This bar was subsequently 114 ALFRED W. G. WILSON modified and its curves readjusted by waves coming up the bay. At a later period the second bar was built outside the first, under a similar succession of conditions, the waves most actively con- cerned in its construction coming from the southwest. The third portion was in part built during the summer of 1903, under the action of a series of storms from the northeast. During the process of its building the waves cut into the earlier bars on the north side, producing the concave curve in the shore line at this point, and depositing the eroded material nearer the apex of the spitvon the) far side) ofithe axis of the initial orm, pro- ducing the asymmetrical form shown in the plan. If their action continued long enough under the conditions existing at the time the observations were made, the bars would be extended in a very much larger loop and would inclose a very much larger lagoon. The rounding of the end of the spit and the shaping of the convex and concave curves on the south side were actually done by the same set of waves which brought the material to form the outer cap of the spit. In this, as in several other cases, even where the material was coarse gravel, the apex of the spit lies so far off-shore that waves curving obliquely toward it from either direction will only have their shoreward ends retarded as they advance obliquely on the shore. The off-shore portions advance in the deeper water virtually unretarded, and thus the wave front is rapidly curved around the end of the spit. Mate- rial moved along a side of the spit toward the end, when dis- charged at the apex, will often be carried around the end by the more vigorous unretarded portion of the same or the next following wave to that which accomplished its final discharge at the point. , This sand spit seems to be rather an evanescent thana perma- nent feature of the shore. The present spit is, from the charac- ter and size of the sedges growing in the lagoon area, inferred to be several years old, probably not more than five. 2. Grand bars on Picton Bay.— On the west side of Picton Bay, nearly opposite the west end of the third section of the Bay of Quinte, are two peculiar bars forming two distinct loops, convex outward, joining the shore by two short concave curves of adjust- CUSPATE FORELANDS AT BAY OF QUINTE I15 ment. The beach between the cliff-foot and the water is here quite narrow, usually less than 6 feet in width. Above and below the two-loop bars in question the shore line is slightly sinuous, but the beach is of very uniform width. Between the two bars there is a stretch of 78 feet where there is not enough beach gravel to cover the bed-rock, and the cliff rises directly from the COT u tl) 100' NNO N Amiri guy Lyi GZ FIG, 3. water, here about a foot in depth at the shore line. The south loop is 220 feet in length, and the north one 280 feet. The north loop holds a long, narrow little pond between it and the old shore. The low area between the south bar and the old shore was above present water level, and was nearly filled'’with gravel. The sudden departure from the normal conditions along this shore to form these bars is difficult of explanation. In the pres- Wy x yy ww AHA yinan™ ee OL E CO Fic. 4.— Foreland near Allison’s dock, May 22, 1903. ent instance it is possible that a small landslip from the cliff may have temporarily changed the shore line in such a way as to necessitate readjustment by the waves. On the other hand, they may have been formed under the action of the waves alone on the normal shore line, under conditions referred to below in a general discussion of the origin of the forms here described. In this latter case they represent initial stages of a form which reaches its perfection in the V terrace and V bar. 3. Terrace and bar near Allison’s wharf—On the north shore of the eastern section of the bay at Allison’s dock, there is a sea- 116 ALFRED W. G. WILSON’ cliff 25 feet in height, cut in till. To the east the cliff becomes much lower. About half a mile east of the wharf occurs one of the most perfect examples of the V-bars. The sea-cliff of till here has a height a little under 5 feet. There is a narrow beach about 20 feet in width. The front of the cliff behind the fore- land is more subdued than elsewhere; it is graded, and is covered with sod. The bars which inclose a triangular lagoon are built of gravel and sand. The material of the east arm is chiefly a coarse gravel; that of the west is gravel with a much larger per- centage of finer material and some sand. On the inner side there is a small amount of clayey soil which has gradually been blown or washed into the lagoon. The bars are of at least three periods of formation. The oldest rises 3.2 feet above water level the next oldest 4 feet, and the present one about 3 feet. The older beaches have been in- part cut off by the newer, as shown in the plan. The inclosed lagoon is triangular in outline, with rounded corners. The base on the old shore measures about 210 feet, the apical distance along the axis is about 135 feet. The depth of water is about 18 inches. It is more or less grown over with water plants and grasses. The east arm of the triangle measures 144 feet; the west, 165. The apex of the spit is rounded and the nearly straight sides join the shores with short concave curves of adjustment. The east arm of the bar is much higher and wider than the west arm, and its outer end has several times been truncated by stronger storms from the east. The present form of the spit is thought to be due to the activity of the waves, chiefly from the east. The western arm has been straightened and smoothed off at frequent intervals by the less violent, but more constant waves from the southwest. The bottom on which the terrace rests here slopes rapidly downward under the bay, the 100-foot contour lying less than a quarter of a mile off shore. A reference to the general map shows that this spit is located very near one of the most salient points of the north shore of this section of the bay. On the ground its actual location is about a quarter of a mile to the east of this point, and hence it is sheltered by the point from the storms which blow directly GOUSPAPEVORELANDS ATT BAYV OF OGLNILE. Ly, down the bay from the southwest. Waves which travel up the bay from the east would apparently have their maximum effect on the beach at this point. A little farther east there is another minor point, too small to show on the plan. Beyond this toward the large point (a drumlin) shown on the plan, about a mile and a quarter east of Allison’s wharf, the shore débris is very much coarser. Both to the west and east the rawness of the shore cliffs and the coarser beach débris show that there is much more active erosion going on there than in the immediate vicinity of the Fic. 5.— Foreland about half a mile west of Prinyer Cove, June 1, 1903. spit. The inference there seems to be that just at this locality we have a region of relatively quiet water and less activity, where material eroded by the waves acting alternately at different intervals tends to accumulate. 4. Prinyer Cove spit— About a mile west of Prinyer Cove there is a slight salient on the shore line which is tipped by a small V terrace and rimming bars inclosing a triangular lagoon. The axis of the spit lies nearly at right angles to the trend of the shore line. The spit is 275 feet in length and measures about 300 feet across the base. The sides are nearly symmetrical, and the inclosing bars are built of gravel. The inclosed lagoon is in part filled up with rank marsh vegetation; near the edges are some large trees. The apex of the spit shows the lines of succes- sive additions on alternate sides. Inside the present beach only one of the earlier beaches is well preserved. This has been in part 118 ALFRED W.G. WILSON cut into during the readjustment of curves when the present beach was built. The land behind the shore is overlaid by a thin sheet of till. It slopes gently bayward, and the inner margin of the lagoon gradually merges into the mainland. Both on the east and west there is a low cliff above the beach having a height of about 2 feet. The cliff and beach that must have existed behind the lagoon have long since disappeared. The gravel bars on the sides rise about 3 feet above water level. That.on the east is a little larger, and consists of coarser material than the one on the west. Almost all the gravel composing the bars is derived from the adjacent bed-rock—a nodular shaly limestone of Trenton age. 5. Pleasant Point spit. This is the largest and the most inter- esting of all the forelands on the bay. The general form of the foreland is shown by the accompanying plan. The material of which it is built is almost wholly gravel. The eastern side con- sists of very coarse shingle containing numerous flat plates of all sizes up to three or four pounds in weight. The.west arm, on the other hand, consists chiefly of smaller rounded pebbles, rarely over an inch in diameter, and there is also a certain amount of fine gravel and sand. To the west of the foreland there is a shore cliff about 20 feet in height, of which at least the upper 5 feet are glacial till. The base of the cliff is shaly limestone, and the width of the normal beach is between 6 and Io feet. Itis strewn with coarse cobbles, there being very little fine material such as is found on the arm of the spit a fewyards away. The old cliff runs behind the spit ; twice it changes its direction, recording significant changes in the growth of the spit. Its height at the base of the eastern arm is only about 5 feet. It continues as a low bluff for some distance to the southeast. The drift varies in thickness, but near the spit its thickness 1s about 2 feet. The original foreland so far as it can be traced, lay a little farther to the west than the present one, and was very similar in shape and size to that near Prinyer Cove. At the present time there are seven distinct beaches. Counting east from the inner triangular lagoon, the first three of the beach mounds or ridges each VGCUSPATE HORELANDS AT, BAY, OF QOUINTE TIQ rise only about a foot above present water level. They are nearly parallel, and between them we find two long, narrow ponds. The fourth beach, the largest and highest of the series, extends nearly the whole length of the spit. The next two are also of consider- , able height and breadth, and are best preserved near the outer end. In the readjustment of the curves during the formation of co i IMMUN Fic. 6.— Sketch plan of Pleasant Point Foreland, May 23, 1903. the seventh or modern beach the waves have cut through the sixth and fifth, and are now acting on the fourth near its shore end. On the west side traces of only one ancient beach could be found between the present modern beach and the triangular lagoon. It is assumed, in the counting, that this is the correlative of some one or more of the first six of the earlier beaches found on the east side. Both the beaches on the west side cut across the ends of the first three of the earlier beaches, and the modern one cuts across the ends of the other three as well. The fourth beach on the east, the highest and broadest of the series, rises 120 ALFRED W. G. WILSON about 6 feet above present water level, or at least 8 feet above the bottom of the lagoon. The beach on the west is only about 2 feet high, except near the apex of the spit. A reference to the general map will show that immediately to the east of the point we have a gap—the Upper Gap— in the side of the Bay of Quinte valley, through which storm waves from the open lake can have access to the bay. The waves which will have most effect on the shore are those coming froma little to the east of ‘south: although the waves of a storm from the east or south will also be capable of effective work. On the Fic. 7.— Sketch plan of about 100 feet of the other hand, the spit is apex of the Pleasant Point spit, May 23, 1903, showing the shifting beach ridges and terraces. exposed on the west only to waves traveling up the bay before a wind having a very limited distance in which to act. Hence we find that the larger waves from the open lake have been steadily carrying material around the point, and depositing it in the slack, but very deep, water behind. The point of the spit is now out as far as the 70-foot contour. The much larger size of these waves has been the important fac- tor in determining the coarseness of the material of the eastern part of the spit, in piling it so high, in determining the amount which has been brought here, and in causing the spit to travel slowly eastward. The material which forms the west arm is in part derived from that brought by the bigger waves to the east side and subsequently carried around the point, partly by the same system of waves which brought it, but chiefly by the waves coming up the bay from the northeast at other times. Some of it is brought from the shores to the west. One record of the changes which take place at the apex of the beach under the action of CUSPATE FORELANDS AT BAY OF QUINTE WAI different storms is shown in the accompanying sketch. Material is transported very rapidly along the eastern side of the beach, in spite of its coarseness. Along the west the travel seems to be much slower because of the relatively small size of the waves. This spit must be very old. Near the outer end of the fourth beach, the highest of the series, is an oak tree sixteen inches in diameter. This beach andthe earlier ones are covered with a thick growth of large cedars. 6. Fish Point spit.—tThis spit is not so large nor so well developed as the others. The reasons for this are twofold: first, on the east the source from which material may be drawn is only about half a mile of beach, and on the west the distance is not much over two miles; in the second place, there is almost no drift cover, and the rocks here seem to be a little less shaly than elsewhere, consequently the supply of gravel is not so abundant. The gravel which occurs on the beaches on either side of the point is i very coarse, many of the rounded pebbles Fic. 8.—Fish Point! exceeding two inches in the longest diameter, Foreland, May 24, 1903. and there are numerous large plates up to ten pounds in weight. The gravel at the spit is smaller than elsewhere, that on the east 22 ALFRED W.-G. WILSON side probably a little coarser than that on the west. The spit, as a whole, resembles a cap which has been built by the gravels on the end of a minor salient of the mainland by the waves when readjusting the shore curves. The main portion of the spit con- sists of a large irregular or wavy topped terrace of coarse gravel, built out in front of the mainland. For the most part the earlier beaches have lost their individual identity. At the outer margin several of the later ones are still persistent, inclosing shallow lagoons. The spit was particularly interesting as it exhibited several features, which are described in detail because it is thought that their mode of formation is an index of the way in which the large V bars and V terraces were built up. The eastern side of the spit at the water line had a serrate margin, there being ten dis- tinct, well-marked minor cusps, which for convenience in descrip- tion may be called cusplets. Each of these had a long, gently curving shore line on the side toward the advancing waves. The free end of the cusplet was joined to the main shore by a short, abrupt, concave curve. Sometimes the free end of the cusplet was drawn out into a sharp, well-developed point. The best- formed cusplets had a sharp median ridge extending down the axis, and often prolonged as an apical spine at the free end. The outer slope, toward the water, was very steep, at first almost a straight line, and then gradually curving around to the normal subaqueous beach curve. The inner slope was much flatter. The curve of the shore line of the individual cusplets was approxi- mately adjusted to the curve of advance of the front of the waves which were building and shaping them (see Fig. 9). The finer gravel lay on the longer back slopes, the coarser fragments, often small plates rather than rounded pebbles, were concentrated on the steeper frontal slopes. These serrations on the side of the spit seem to owe their origin to the attempt of the waves of a particular series of storms, coming from a nearly constant direction, to readjust the curvature of the shore line to the curvatures of their own fronts. Off shore the waters are very deep, and the shore line of the bay is yet in a very young stage of its development; consequently CUSPATE FORELANDS AT BAY OF QUINTE 123 the waves traveling obliquely toward the shore are. not symmet- rically and systematically retarded. The wave does not advance on the shore parallel to its front but comes up obliquely (see Fig. 9). The result is that the gravel was moved obliquely up the slope of the beach, and then obliquely downward with the return of the wave, but always with a resultant in a direction Fic. 9.—Showing the relation of the wave-fronts to the serrate margin of the east side of Fish Point Foreland. parallel to the shore. During the period of observation the débris moved along the long curve of the cusplets very rapidly, and then, when discharged into the deepest water at the free end, would either fall at once to the bottom, or might happen to reach the end just in time to be carried across the intervening space by the rush of the less retarded part of the wave which had not yet reached shore. Material would thus be rolled along the long slope by the breaking edge of the wave, but, when discharged at the free end, it was often bodily carried several feet past the 124 ALFRED W. G. WILSON spine of the cusplet and up to the main beach by the more pow- erful, less retarded portion of the waves—there to be rolled slowly or rapidly along the long slope of the next cusplet, where the process was repeated. The size of the cusplet in some cases seemed to be increasing, but several seemed to have reached a maximum stage. Given a constant material, the limit of size seems to depend upon the size of the waves and their periodicity. These little cusps are formed during the period of a single storm, or series of storms, when the waves advance in an oblique direction on a previously evenly curved shore. Their forma- tion and their symmetrical arrangement seem to be due to two factors. In the first place, very frequently the undertow is able to carry material down the slope of the beach a little farther than the front of the wave can move it up, within certain limits. Consequently, although some of the material moved up the slope by the front of the wave lodges, some of it moves down with the undertow, and a small percentage of this latter material may move out beyond the zone at which the next oncoming waves can move it up the beach. Hence there will be a slow but gradual accumulation just beyond this line, which in time will even modify the direction of the long shore currents. A second and more important factor in the production of these serrations along the shore is the development of nodal lines along which material tends to accumulate. Where the waves are advancing at an angle to the shore there will be a number of waves breaking at the same time at different points along the shore. As the spacing of the waves is nearly uniform, if the shore line were perfectly straight, these points of simultaneous wave-breaking would be equidistant from one another. On a curved shore the spacing will be systematic, but the distances between breaking points will not necessarily be equal. Now, the undertow which flows out from one wave as it breaks will inter- fere with the advance of the next following wave, if it meets that wave on that part of the shore where the orbital motion is nearly a straight line up the’ beach. This happens very frequently where part of a wave is retarded by a cusplet while the other CUSPATE, FORELANDS AT BAY OF QUINTE 125 part passes the free end with little retardation. The result will be a tendency for the material moving down the slope with the undertow, and up the slope with the advancing wave, to be dropped at a symmetrically arranged series of points. The obliquely moving waves also move débris along the shore in the resultant longshore direction of the wave advance. The result of the combined action of these different factors is that gradually a little bar is built out from the shore by which the waves attempt to readjust the curvature of the shore line to a curvature appropriate to their direction of advance. Because of the nearly uniform spacing of the waves, these bars will begin at a number of symmetrically arranged points. Because of the normal, uniform slope of the subaqueous floor, the maximum distance from shore at which the undertow can materially inter- fere with the advance of the next wave will be located at a nearly uniform distance off the initial shore line, and this will tend to limit the size of the individual cusplets. The size is also limited by the distance between the crests of the waves. The building of the cusplets further modifies the form of the shore line, the slore of the bottom, the direction of the advance of the waves, and the direction of the longshore currents; but with waves of constant size an equilibrium will be established, at which time the cusplets will have their maximum size. If the waves are irregular, cusps may not be formed at all. The same waves which had built the serrate margin along the eastern side of this foreland had built a small flying spit at the apex. Between the free end of this small flying spit and the main beach a very small A-shaped point was also gradually built up. The waves coming from the east in the direction indi- cated by the arrows (Fig. 10) swung around the point, giving it the form shown in the figure. The fronts of the waves assumed the form of a series of helicoidal curves as they swung around the point as if ona pivot. As many as eight waves could be counted swirling around the west end of the flying spit at the same time, the moving crests looking not unlike the spokes of a gigantic horizontally rotating wheel. The relative positions of the succes- sive wave-fronts are shown by the dotted lines in the figure. 126 ALFRED W. G. WILSON Material which had rounded the- extreme tip of the flying spit was actually carried across the narrow water space between the flying spit and the little conical point being deposited on the outside of X \ \ \—a \ \ AN \ NY \ ‘i \ NZ \ S SS \ \ S uh Ss Se \ NN Se XN ~s ~ sy ~ “ ~ ~~ Xx Ss RI — ~ Se v ~— x NS Te yy" DINU way ye LE TY DI ns uw SEV) 1H) yl es oo A\\\\\\ TT ae RO eae WER ee v eh WW" » Ss 34 > SS Se SS y R — s S ~e d Se y Tce a SSS Y == y Ss iS) J S=5) y Seen sine 3 x = a] / > Z-l = y, 2 ZH ez 2 Em x Wee = ENS Pe oS ay ot | = |S — a Vee ~ - Z “tLe Lel \ i =I at = Bel =| 54 5 / 1 Se =) / i] | = ~\ y j 15 = / ! rt eal / / ~ Be / =} \ ee / = at / = es / Si f = 7 = 4 = Ns =! es p= Al = =I a S a = 5 : E Zh ©) Be = a zy = = = 3 = nN” =) =} a = -) Kat S =) Fic. 10.—Sketch plan of the apex of Fish Point Foreland, May 24, 1903. CUSPATE FORELANDS AT BAY OF QUINTE M27, the cone. As each wave came in, the water in the small lagoon rose and fell. The outflowing current seemed to be the control which shaped the inner curves of the cone. A little farther to the west the same waves were increasing the size, rounding the ends, and otherwise modifying the two larger cusplets (Fig. 11), which, | EAR at SSA IRs RIE VIE TR EEE EEE NINE SY Fic. 11.— Two well developed cusplets in the foreground, the apex of the small loop spit appears in the background. North side of Fish Point Foreland. judging from their initial forms, had evidently been built some time before by a storm blowing from the west. 7. Amherst bar.—Waves rolling into the bay through the lower gap from Lake Ontario have built a long gravel bar off the east end of Amherst Island. This bar runs nearly north from the end of the island and is nearly two miles in length. Most of it is submerged, but near the island a portion rises as a sharp ridge several feet above water level. The eastern end of Amherst Island is low, and the shore is rocky. Most of the gravel forming the bar has been moved along the south shore of 128 ALFRED W. G. WILSON the island by southwest storms off Lake Ontario. The portion of the bar that is above water level has a peculiar curved form, due to the many complex modifications which such a bar may undergo under the influence of minor storms. Some of these are well shown near the free end of that portion of the bar which rises above water level. On the south side of the free end we find two large, well-developed, south- pointing cusps, bounded by curves which are concave lakeward. These cusps seem to owe this form to the action of waves advancing from _ the southeast and the southwest at different times. 8. Calf Island loop bar.—A\though not in any way associated with the Bay of Quinte, it seems desirable to include in these descriptions a reference to the loop bar off the east end of Calf Island. The island lies about four miles northwest of Stony Point, and half a mile to the west of Stony Island. Storm waves blowing down the lake naturally divide at the island and pass on either side of it. Coarse gravel derived from the limestone rock, by which the main island is underlaid, has been piled in two high ridges, one leading off from either side of the island. The two unite in a rather sharp_point about 350 yards from the east end of the rocky part of the island. The crests of the bars are about g feet above water level, and between them is a deep, narrow pond. The south bar is about 60 feet wide, and has equal slopes on either side; the north bar is a little wider and more irregular. Similar forms are to be looked for off the northeast ends of several of the other rocky islands in this part of the lake. Off the east end of Grenadier Island two long flying spits have formed, inclosing between them a shallow bay known as Basin Harbor. This bay is gradually filling up. The free ends of the two spits are curving toward each other, and, given time enough, we would expect them to unite. In the meantime, the inclosed basin will be partly filled by sand either washed in by the waves or blown in from the bars. The outer slope of the bars will still have the steep gradients of such forms; their height will depend upon the depth of the adjacent water. In time there will thus be formed off Grenadier Island a huge terrace, with running CUSPATE FORELANDS AT BAY OF QUINTE 129 bars, which in form will approximate in shape to the typical V- terrace and V-bar. THE ORIGIN OF THE V-TERRACE AND V-BAR. Four of the forms which have been described in the preceding paragraphs agree very closely, both in form and location, with Gilbert’s description of the type examples in Lake Bonneville.’ In his descriptions of the type examples Gilbert notes that: They are built against coasts of even outline, usually but not always, upon slight salients, and they occur most frequently in the long narrow arms of old lakes. In discussing the origin of the form he states: In some cases the two margins appear to have been determined by currents approaching the terrace (doubtless at different times) from opposite direc- tions ; and then the terrace maryins areconcave out- ward, and their confluence is prolonged in a more or less irregular point. In most cases, however, the shore drift appears to have been carried by one cur- rent from the mainland along one margin of the ter- race to the apex, and by another current along the remaining side of the terrace back to the mainland. The contours are then either straight or convex. The bars which border the terraces he attrib- utes to a later period during a slight deepen- ing of the waters of the lake, after the ter- races had attained their full size. While the lake stood at the higher level, the linear embankments were built at the outer mar- gins. N \ Sy | Fic. 12.—Sketch plan of about 500 yards at the apex of the portion of Amherst Bar above water level on May 25, 1903. Direction of wave advance shown by the arrow. The author’s studies of the forelands in the Bay of Quinte lead him to suggest the following hypothesis as to the mode of origin of the forms here described. In ‘the first place, it must be noted that the level of the water in the bay varies *U. S. Geological Survey, Fifth Annual Report, 1883-84, p. 98. 130 ALFRED W.G. WILSON considerably with the seasons, being a little higher in late spring orearly summer than at any other time. The level of Lake Ontario also changes considerably during a season. Both of these factors may have some bearing on the formation of the terraces and bars. The changes in level due to the larger seiche waves must occasionally be even greater than these seasonal changes. None of the forms show any evidence which could be interpreted as being due to these seasonal or periodical changes in level. In a previous paragraph a detailed description was given of the process by which small cusps were produced along a shore. Under the continued action of waves of moderate amplitude the dimensions of these small forms would gradually increase, and eventually they would reach a size which could easily control the shore currents and wave direction of even moderate storms. In the present instance the bay is completely frozen over from about the middle of November until the first of May. During the season of open water the only effective storms are those which chance to be blowing up or down the bay. To be effec- tive, they must have a constant direction, for a considerable interval of time. Hence, while moderate breezes which generate small waves are frequent, violent storms which can modify the work of all previous lighter winds and waves are rare. When they do come, their first work would be to readjust the shore curves developed during the previous interval. The chances that they would preserve a suitable direction long enough to efface the work of the previous, more or less contsant, but less energetic, storms are very slight. The construction of the small triangular terrace may in part be attributed to the leveling action of some such storms as these. In all observed cases, although the terrace under the triangular lagoon had a slight slope outward, its slope was not so great as that of the adjacent shore a little distance on either side of the sand spit; from which it is inferred that there had been some filling. Whether such a process could produce a very much larger terrace than those noted is uncertain. In other cases the portion of the terrace included between the bars may have been partly filled in by the CUSPATE FORELANDS AT BAY OF QUINTE 13a waves themselves after the formation of the bars. Such a ter- race is in course of construction off the east end of Grenadier Island. A similar process is causing a great deal of inconven- ience at several harbors along the north shore of Lake Ontario, where two artificially constructed bars in the shape of piers inclose a harbor which periodically fills with sand that has to be removed by dredging. In some cases the inner lagoon may have been filled after the bars were formed, by ordinary processes of transportation which tend to fill hollows and lessen the grade of steep slopes. The size of the terrace would also depend upon the size of the water body, and upon the character of the material. The tendency will always be for the waves bringing the supplies of material to heap this up in the form of a bar. In the later stages, when the accumulation has become considerable, the larger storms would not be able to efface these bars, though they will reshape them and pile the material higher on the outer mar- gin. On the outer side of a bar, below water level, the material has a gentle slope to below wave base. Beyond this the inclina- tion of the front slope will be the angle of repose for material of the kind. In the case of all the forms on the Bay of Quinte, where the water drained off it would be found that the forelands would have steep frontal slopes, with an elevation in several cases of about 60 feet. The top would be a nearly flat terrace, with gently curved edges, and rising above it at a little distance from the margin would be the sharply defined rimming bar. In the smaller examples the same waves which build the one side of the foreland carry material around the end of the spit and distribute it for a shorter or longer distance, according to their size, on the other side. On some occasions the same waves may shape both sides at the same time, but usually it is found that the adjacent sides are shaped alternately. In some cases the greater proportion of the material comes from one side, and its redistribution on the opposite side of the spit is effected by other waves from a different direction and at another time. In the case of Point Pleasant spit it seems to be slowly shifting east- ward, as material brought from the southeast accumulates on Teg2 ALFRED W. G. WILSON that side. At the same time less rapid erosion is taking place on the west side under the action of less violent waves. CONCLUSIONS. In conclusion, it may be stated that the forelands here described seem to have been built wholly by the action of waves acting either directly or indirectly in association with longshore cur- rents which were intimately associated with them. The location of the forelands is associated with some more or less salient feature of the coast which has influenced the direc- tion of wave advance and the course of longshore currents, and has localized the effective transporting action of both. Their formation is due to the control exercised on wind direction and on wave direction by the form of the bay. The form of the forelands is due to the peculiar character of the long, narrow water body on which they are situated, the conditions being such that only certain classes of storms can be effective agents in the shore transportation. The immature character, and consequent imperfect adjustments of sub-acqueous portions of the shore is an important control in wave-work. The V-terrace and the associated V-bar upon it, in the instances here studied, are regarded as products of the same agent, and do not necessarily imply a change in water level. The evidence from Point Pleasant spit implies that there has been no signifi- cant change in level during the long period of growth of the greater part of the spit. ALFRED W. G. WILSON. DEPARTMENT OF GEOLOGY, McGill University, Montreal. waCOMERIBULION LO Wik. SUD OF THE INTE R- GUACINE GORGE, PROBLEM: Topography of the Finger Lake region—The topography of the Finger Lake region is too well known to American geologists to require any detailed description here. The rocks, which are almost wholly Devonian, consist of great deposits of shale and sandstone, with a few thin beds of limestone. These rocks have never been greatly disturbed and lie nearly horizontal, with a slight southward dip. In the Cayuga Lake district there is a series of gentle folds which cross the lake valley in the east-west direction. A glance at the even sky-line presented by the hill- tops shows that the region is a great plateau. This plateau has been so deeply dissected that it resembles a mountainous coun- try, with the hills often rising several hundred feet above the valley bottoms. About fifteen miles south of the heads of the lakes is a dissected divide that Professor Tarr® has characterized as being ‘‘high and diverse in topography.” From the divide the plateau slopes northward and merges into a drift-filled region at the northern ends of the lakes. Here, doubtless, there was an escarpment in preglacial times, but it is now nearly obscured by drift. Cayuga Lake valley — From the divide at Spencer Summit the valley of Lake Cayuga extends northward a distance of nearly fifty miles, when it is lost beneath the drift. Professor Tarr3 has called the divide at Spencer Summit a divide of ‘destructional origin.” He considers the depth of the drift here to be slight; and from the steepness of the walls he infers that the divide must have been higher in preglacial times, ‘Shaving been lowered *This paper was originally written as a thesis for the master’s degree at Cornell University. Since the preparation of the original manuscript enough new informa- tion has been secured to warrant a slight revision, and therefore some changes have been made. The writer is indebted to Professor R. S. Tarr for many valuable sugges- tions concerning the field investigations and the preparation of the original paper. 2R.S. Tarr, Bulletin No. 5, Geological Society of America, p. 340. 3/bid., p. 341. 133 134 GEORGE €. MATSON by glacial erosion.”’ There is also good reason for believing that the divide has been lowered by stream erosion. The ice in its advance would close the outlet of the lake valley, causing a lake to be formed between the ice front and the divide. The drain- age of this lake across the divide would continue until the ice had advanced to the divide. In receding the ice would again | cause the formation of a lake in the valley, which would exist from the time the divide was uncovered until the ice retreated far enough to uncover a lower outlet. The drainage across the divide would naturally tend to lower it. The amount of erosion would vary with the length of the time. The presence in Cayuga Valley of the well-developed terminal moraine of the Wisconsin glacial epoch points to the existence of an ice-dam in the valley fora long time. This fact points to the probability of the divide having been considerably lowered by stream erosion. The effect of the drainage across the divide would be influenced by the rela- tive altitude of the northward- and southward-flowing streams. If the. southward-flowing streams had cut considerably below the level of those flowing northward, the water would fall into the deeper valleys, and the divide might be destroyed in a very short time by the recession of this waterfall. If, as seems more probable, the northward-flowing streams had reached the lower level, the removal of the divide would be much slower. On this point Professor T. L. Watson has said: _It can hardly be doubted that the Laurentian tributaries were the stronger streams, therefore encroaching upon the territory of the other system, and thereby causing the southward migration of the divide.* We should also bear in mind that there is a possibility of a differential uplift having rejuvenated the Laurentian streams just before the glacial period. Mr. M. L. Fuller? failed to find evi- dence of this uplift in the area covered by the ‘“Elkland-Tioga Folio: It has been frequently urged among geologists that the advent of the earliest Pleistocene ice-sheet was preceded by a general uplift of the north- ern half of the continent, affecting the surface throughout the northern por- r“Some Higher Levels in the Post Glacial Development of the Finger Lakes,” Report, N. Y. State Museum, Vol. I (1897), p. R. 68. 2M. L. FULLER, “ Elkland-Tioga Folio,” U. S. Geological Survey, p. 7. THE INTERGLACIAEL GORGE PROBLEM 135 tion of the United States. In western Pennsylvania, however, the presence of Pleistocene river gravels on rock terraces several hundred feet above the bottom of the present gorge of the upper Allegheny River indicates that the last stage of the active erosion did not begin there until after the first ice invasion, though the uplift and the inauguration of the erosion in the lower reaches of the river may have been somewhat earlier. The uplift recorded by the rock terraces immediately adjacent to the Susquehanna in the eastern portion of the state is of questionable date, but would appear to be of late Tertiary or early Pleistocene age. In the Elkland-Tioga region there appears to be a slight notching in the bottom of the old valley of Pine Creek and some of its tributaries, but it is believed that this was not produced until after the southward deflection of the lower portion of the creek through the gorge south of Ansonia. This diversion, as will be described more fully in the discussion of: the earliest glacial stage, was probably due in great measure to the accumulation and overflow of waters ponded in front of the advancing ice-sheet, and the con- sequent reduction of the divides and the cutting of a new channel in which the stream persisted even after the ice had disappeared. The notching of the bottom of Pine Valley and its branches was a result of the diversion through the new and lower channel, and affords no evidence of uplift. The Elkland-Tioga region is not far from the Finger Lake region. It is; however, on the south side of the divide, and includes some of the streams which are tributary to the head- waters of the Susquehanna River. If rejuvenation had effected the Laurentian drainage, this would tend to increase rather than diminish the advantage of the northward-flowing streams over those flowing southward. The hills, which rise steeply at the southern end of the lake valley, become lower and more gently sloping as you pass north- ward. The valley also widens rapidly toward the north. While a mature stream valley ought to become wider and the walls more rounded toward the mouth of the stream, the change here is so rapid as to suggest that there must be some other explana- tion to account for a part of the difference. One cause which has probably contributed to this end is the northward differential depression which occurred at the close of the glacial period. If this depression amounted to no more that two feet per mile, it would have made a difference of one hundred feet in the relative height of the land at the ends of Cayuga Valley. It has been suggested that the difference in the topography at the 136 GEORGE C. MATSON two ends of Cayuga Valley is due to a difference in rock texture.’ The southern part of Cayuga Valley is cut in the hard, durable Portage sandstone, while farther north are the soft Hamilton shales which break down much more easily when exposed to the agents of weathering. Another point of considerable, though undetermined, importance in this connection is the intense glacia- tion to which the northern end of the lake valley has been sub- jected. As suggested by Dr. G. K. Gilbert,’ the belt in which the northern end of the lake hes has been much more intensely scoured by the ice than the belt south of the lake. The width of the lake varies from about a mile near its southern end to about three miles near Aurora. The surface is 378 feet A. T., and the greatest: depth is: 432) teets, jon considerably more than half its length the bottom of Cayuga Lake is below sea level. The drainage of Cayuga Valley is northward into Lake Ontario. Direction of preglacial drainage.—While some of the earlier writers believed that the preglacial river which occupied Cayuga Valley flowed southward, all the later students are agreed that it drained northward into the Ontario basin. The hypothesis of a northward preglacial drainage is based on the following lines of evidence: the general northward slope of the land; the wider and more mature aspect of the valley as one passes northward from the divide; the increasing number of tributaries near the divide; and the fact that the bottoms of the mature tributaries become lower from the divide northward. While all these lines of evidence have been affected by the various changes accompany- ing glaciation, it has not been sufficient to destroy their value as evidence in this connection. Since there are no facts opposed to the theory of a preglacial drainage toward the north, we may regard the hypothesis as established. Postglacial gorges.— With few exceptions, all the tributary streams which occupy mature preglacial valleys enter the main valley through narrow, rock-walled, postglacial gorges contain- New Vork Geological Survey of the Fourth District (1843), p. 225; Monograph XLII, U.S. Geological Survey, p. 79. ? Paper read before the Geological Society of America, December, 1902. THE INTERGLACIAL GORGE PROBLEM 137 ing many rapids and waterfalls. The length of these postglacial gorges is usually less than three miles; and the amount of fall varies from about 100 feet to over 500 feet per mile. The descent is usually accomplished by means of a series of cascades, though Fic. 1.—Taghanic Falls. in a few cases there are cataracts of considerable height. Tag- hanic* Falls, the highest, measures 200 feet. Many of these falls have developed on account of an alteration of hard sandy layers with layers of soft, easily eroded shale. A few cataracts * This spelling is in accordance with that on the Genoa atlas sheet of U. S. Geo- logical Survey. 138 GEORGE C. MATSON also occur over the Tully limestone, which caps the soft Hamil- ton shales. In general, the amount of fall in the postglacial gorges diminishes toward the north, although there are some exceptions to this rule. The principal cause for this gorge condition, in the lower part of the tributary valleys, is the fact that the streams have been turned from their old channels and forced to cut new ones. Another cause which has contributed to the same end is the widening and deepening of the main valley by glacial erosion. In some cases the streams have possibly been turned aside by a moraine dam, but the most common obstructions are the deltas deposited in the ice-dammed lakes* which formed in front of the continental glacier. These lakes fell to successively lower levels when lower outlets were uncovered, and the streams continued to flow along the course occupied when the lakes fell. The new channels are mostly south of the old course, and Professor Tarr’? has attributed this to the effect of the prevailing north winds on the waters of the extinct lake. Coy Glen, being exposed to the south winds and protected from the north winds, lies north of its old channel. Interglacial gorges ——In every case where the streams have been turned aside in the manner just described there is a lower course in the same broad valley, which is itself a gorge. Pro- fessor Tarr has already called attention to these gorges in his Physical Geography of New York.3 Interglacial (?) gorges—In central New York there are numer- ous gorges which are broader than the postglacial valleys and partially obscured by glacial till, showing that they were formed either during preglacial or interglacial times. This class of val- ley is especially well illustrated in Six Mile Creek, where its rela- tion to the broad, mature preglacial valley is wellshown. In one case, near Taghanic Valley, lake beds containing fresh-water fossils have been found beneath the till. One naturally thinks of these gorges as being interglacial in 1H. L. FatrcuHILD, Bulletin V7, Geological Society of America (1895), pp. 353- 74; T. L. Watson, Report, New York State Museum (1897), pp. 55-117. 2 Physical Geography, N. Y., p. 177. 3Pp. 178, 179. THE INTERGLACIAL GORGE PROBLEM 139 origin, and this explanation seems, at present, the most probable; but all that can now be said with certainty is that they antedate the last advance of the ice. The question of these gorges has a very important bearing upon the whole subject of the drainage history of central and western New York. Were the gorges due to interglacial conditions or to an uplift of preglacial times ? Leverett refers to similar gorges in his monograph’ on The Gla- cial Formations and Drainage Features of the Erie and Ohio Basins. The valleys of this hilly country present marked differences in topography. In some valleys the slopes from top to bottom havea mature aspect, while in others the upper part of the slope is mature, but the lower part is gorge-like and youthful in appearance. The phenomena suggests at once that some val- leys have remained below the level of stream-cutting, while others have been undergoing a marked trenching. In these which have been deepened, the old valley bottoms are traceable along the brow of the rock gorges or canyon valleys, for the old valleys are generally broader than the new ones. Insome cases, however, the new valleys occupy the whole width of the bottoms of the old ones, and there is only the change in the angle of the slope of the valley bluff to mark the depth of the old valley. - There is, in some valleys a series of complex terraces or rock shelves, of which one set or system stands at the brow or border of the canyon valley, and the othersat higheraltitudes. There are also, insome cases, rock shelves inside the trenches of the canyon val- leys. The set of trenches standing at the brow of the canyon valley is, how- ever, a far more persistent feature than any of the others, and it is this set which receives chief attention in the ensuing discussion of drainage systems. It seems to mark a true gradation plain, formed when the stream was in con- dition between degrading and aggrading its bed. All the preglacial tributaries of Cayuga Lake Valley which have been examined have gorges cut in their bottoms, and these gorges are wider, and in many cases deeper, than the postglacial gorges. The approximate width of the drift-flled gorges can usually be ascertained without much difficulty. On account of the drift-filling, the depth is not readily determined ; however, Evans, who studied Taghanic carefully, mapped the old valley bottom as continuous above the level of Lake Cayuga.’ Ten Mile Creek.—Ten Mile Creek, despite its name, is only about six miles in length. It rises near the village of Danby and flows a little west of north, entering the Inlet about two miles * Monograph XL/, U.S. Geological Survey, p. 80. 2 Rk. M. Evans, Thesis on Taghanic (1897), Map. II. 140 GEORGE C. MATSON south of Ithaca. The stream occupies a broad, mature valley with gently sloping sides. This valley divides at Danby, the branches becoming narrower and the walls steeper for some dis- tance beyond the village. While Ten Mile Creek now rises at Danby, there is evidence that the preglacial divide was farther south. Danby Creek, which rises near the east branch of Ten Mile Creek, flows southward through a valley which gradually narrows for a distance of three miles and then widens again. Michigan Creek, which heads near the west branch, flows through a similar valley; but in this case the narrowest part of the val- ley is only about two miles from the source of the stream. The narrowing of these mature valleys cannot be explained on the ground of rock texture, for the rocks here are of uniform charac- ter. The narrowing may be explained by supposing that the narrowest part represents the divide between these streams at the time the broad valleys were being formed. The divide hypothe- sis is strengthened by the fact that the highest hills are on either side of the narrowest parts of the valleys; and by the additional fact that the present divides are clearly of constructional origin, being composed of low drift ridges. The drift ridges are within the limits of the region covered by the last ice-sheet ; and the present location of the divides was determined by the material deposited at the ice-front." About a mile from the Inlet, Ten Mile Creek enters a narrow postglacial gorge ; and in less than a mile it falls, by a series of rapids and cascades, through a vertical distance of about 420 feet. The highest falls are the two cascades near the edge of the inlet, which together measure 190 feet. Running nearly parallel to this gorge is a drift-filled gorge, which is much broader and deeper than the postglacial gorge, as shown by sections I and 2. The buried gorge is now occupied by a small stream which has removed the drift down to 520 feet A. T., but has not reached the rock bottom. The width of this gorge is 250 yards ; its depth is not known, but the north wall rides 156 feet above the drift which forms the bed of the present stream. The lower end of this stream is through a small postglacial gorge, cut in the north «For location of the divides, see Map I. THE INTERGLACIAL GORGE PROBLEM I4I wall of the filled gorge. There is a rock terracet on the south side of the buried gorge at 620 feet A. T., which probably repre- resents a former level of the stream, for there is a tributary gorge on the north side near the cemetery which enters the:main gorge at about the level of this terrace. The tributary shows as a gap in the wall of the main gorge, just below the cemetery and it is crossed by a small stream just above the cemetery. A short dis- ; ily pw l _ —— iI iH i i ‘itydyay! il ih Section I. Section 2. V. Scale obs Bonrene Cie tae Portage Alluvium Ile ey Sandstones Clay and Shales IDK, Ay tance from the inlet the present stream crosses a small buried gorge, which must have been another tributary, for the rock is practically continuous at 720 feet A. T. between the old gorge and the postglacial gorge. If we look upstream for the continuation of the large buried gorge, we find that it appears just above the delta which was built by the stream in Glacial Lake Ithaca.? Above this delta the gorge can be traced nearly a mile to where it is finally obscured by the drift. In that distance the stream has cut two small postglacial gorges around drift obstructions which block tThe writer has begun a study of these rock-shelves which he hopes will lead to an explanation of the direction of flow, and at the same time elucidate some other points. The photograph taken in Six Mile Creek shows two terraces. 2T. L. Watson, Report, New York State Museum (1897), pp. r55-r117. 142 GEORGE C. MATSON the main gorge. One of these postglacial gorges is shown on Map II. Map II comprises an area 1,000 yards long, and from 250 to 600 yards wide. The lowest part of the map is at the city reser- voir, 880 feet A. T. The contours were mapped on either side of the stream up to 1,000 feet A. T. Above the 1,000-feet con- tour’ the surface rises very gently to the divides between Ten Mile Creek and the neighboring streams. There are three ridges on the map which extend nearly east and west, and rise to a height of 100 feet above the stream. Each ridge is composed partly of drift and partly of rock covered with drift. Channels of Map I1— The stream enters the map through a broad drift floored channel, and, after crossing the rock in the southern ridge through a narrow postglacial gorge, it enters another drift-floored channel, through which it flows to the city reservoir. Below the reservoir the stream flows through the narrow postglacial gorge described earlier in this paper. In addition to the channels now occupied by the stream, there are four other channels within the area of Map II. Two of these channels pass beneath the southern ridge, one east of the present stream (see map, A), and one west of the present stream, and just ‘south of the house (see map, 2), (Where is, aachannel beneath the middle ridge west of the rock outcrop (see map, 2), and one beneath the northern ridge, east of the rock outcrop (see map, C). Evidence of the existence of channels—TYhe general evidence of the existence of these channels is of two kinds: (1) the trend of the rock outcrops; (2) the existence of well-defined indenta- tions on the upstream side of each of the ridges. These inden- tations were produced by the stream swinging against the soft drift ridges. The drift was not entirely removed from the chan- nel, because the rock in the ridges prevented the formation of a broad meander within these channels. “In the ‘case of the channel (D), west of the postglacial gorge, the drift ridge has been so badly eroded that the rock-walls show above it, and near the southern end of this channel the wall has been exposed down to the bottom of the channel (Plate V). The hypothesis of a THE INTERGLACIAL GORGE PROBLEM 143 channel (A), east of the postglacial gorge, is strengthened by the fact that the channel west of the postglacial gorge is too narrow and too shallow to be the continuation of the broad chan- “nel south of the ridge. The existence of a channel (4) beneath Fic. 3.— Terraces in the drift-filled gorge of Six Mile Creek. the middle ridge is very apparent in the field; and is fairly well shown by Plate VI. A boring made on the top of the ridge to a depth of ten feet did not strike rock, although the bottom of this hole is more than twenty feet below the highest rock in this ridge. The channel between the middle and north ridges seems rather narrow to be the continuation of the broad channel on the east side of the map, and the well-defined indentation (see Plate 144 GEORGE C. MATSON VI and Map II, C) on the south side of the north ridge, points strongly toward the existence of a channel beneath the ridge. Nature and size of channels—The fact that the rock rises per- pendicularly above each of these channels indicates that they are gorges, and the rock in each of the ridges forms a rock island. The postglacial gorge is 25 yards wide, 50 yards long, and go feet deep. The gorge south of the southern ridge is over 250 yards wide and 100 feet deep. The drift has been removed down to gio feet A. T. without exposing the rock bottom. The gorge (D) just south of the house is 35 yards wide and 90 feet deepyi lit has its bottom at oro feet Avia Mhisiiseat themsame time level as the bottom of the postglacial gorge. The gorge running east from the house is 125 yards wide and 110 feet deep. Its bottom, as indicated by borings near the house and between the rock in the two ridges, is below 890 feet A. T. The gorge just south of the reservoir is more than 175 yards wide and it has been cut to 880 feet A. T. without encountering rock. This would give it a depth of 120 feet. The gorge between the mid- dle and northern rock islands, is 125 yards wide and has a drift bottom goo feet A. T. The outcrop along the stream at the east end of this gorge, is an extension of the rock in the northern ridge and it was probably originally at the same height, having been lowered by the swinging of the stream against it. It now stands 905 feet A. 1. The arrangement of the sections of gorges into a series of continuous gorges depends upon size, depth, and position. Gorge No. 1* passes beneath the southern ridge, east of the middle ridge, and east of the rock in the northern ridge. From the slope of the rock-floor of the broad valley, as indicated by well records and rock outcrops, this channel must lie approximately in the axis of the broad valley of Ten Mile Creek, already described. Gorge No. 2 passes beneath the southern ridge, westward past the house, beneath the middle ridge, and west of the northern ridge. That this gorge could not pass between the middle and northern ridges is shown by the fact that its bottom *The numbering is for convenience only, and does not indicate the supposed chronological order of formation. THE INTERGLACIAL GORGE PROBLEM 145 is lower than the low rock outcrop which extends along the eastern edge of this channel. We still have left Gorge No. 3 and Gorge No. 4. These gorges may possibly belong together, though that is by no means certain. By taking the low ridge of rock east of the stream as one edge of Gorge No. 4, we can reduce its width to 75 yards, but it is still nearly three times as Fic. 4.— A view in the drift-filled gorge of Ten Mile Creek. wide as Gorge No. 3, and it is also more than 10 feet deeper. If, however, we assume that there was a fall somewhere between the two gorges, it would account for the difference in depth; but this assumption does not explain the difference in width. Downstream extension of these gorges——A\l\ these gorges must have entered the inlet through the broad gorge below the reservoir, which lies north of the postglacial gorge, for the rock- wall of the Inlet Valley is continuous for two miles on either side of the drift-filled gorge, except where there are postglacial trenches. It is possiblethat there is a local divergence of the small 146 GEORGE C. MATSON drift-filled gorge. The stream which cut this small gorge may have turned aside and entered the broad drift-filled gorge through one of the channels which were mentioned earlier in this paper as being tributary to this gorge. Origin of the rock islands.—As already shown, there are three rock islands on Map II. In considering the origin of the rock islands, the first question is: Could they have been formed during the normal stream development? A stream abandons its old course and takes a new one when ox-blow lakes are formed; but this happens when the stream is flowing on a flood plain, where the work of cutting across a spur is comparatively easy. In the case which we have to consider the rock is hard enough to offer considerable resistance to stream erosion. Moreover, the shape of the rock island is a serious objection to this theory. If they had been formed by a stream meandering, they should be rounded on the upstream side. Reference to the map will show that all of the rock islands are cut off squarely on the upstream side. A second way in which rock islands may be formed is by lateral swinging of two streams until they cut through the divide which separates them. If the rock islands had been formed in this way by the uniting of the main stream and a tributary, they should taper to a point on the upstream and downstream sides. None of the rock islands have this form. In the case of the southern rock island, this hypothesis would meet with another objection from the fact that the rock-wall of the old gorge is continuous for one-half mile above the area of Map II. | From what has just been said it seems impossible to attribute the rock islands to normal stream action. Another possible hypothesis is that the rock islands have been formed as a result of glaciation. In this connection, two possibilities arise. Some of these gorges may have been overflow channels of glacial lakes, or the stream may have been forced to cut new channels around glacial deposits which obstructed its former course. Can any of the gorges be the overflow channel of the glacial lakes? The gorges cannot be the overflow channels of a glacial lake, for there is no place where a lake could form with an out- let at this point. Any lake formed in front of the ice in the THE INTERGLACIAL GORGE PROBLEM 147 inlet would drain over the divide at Spencer Summit, and the drainage of a small lake in Ten Mile Creek valley would natu- rally have to pass over the divides, not along the axis of the valley where these channels are located. For this reason, any gorges which would be formed as overflow channels ought to show on the divides, where a careful search has revealed none. Fic. 5.—Looking down stream. Postglacial gorge of Map No. II on the right. Gorge D on the left. Could the gorges have been formed interglacially? After glaciation, streams naturally begin to flow along the lowest courses. If the drift deposit is great enough to obscure much of the preglacial topography, the streams may take very different courses from those occupied preglacially; but if the amount of drift-filling is slight, they will naturally follow preglacial drain- age lines. There may be partial abandonment of old drainage lines, as in the Genessee River," or the stream may follow *An enumeration of the cases of reversions of drainage which have been described by various writers would be too large an undertaking for the scope of this 148 GEORGE C. MATSON: approximately its preglacial course. It is possible, then, that these may be interglacial channels, but it would be just as impossible to have them all formed in one interglacial period as to have them all formed during normal stream development. The hypothesis that some of the gorges were cut because of drift obstructions requires no different conditions from those existing in scores of places near here. Ten Mile Creek has been forced to cut three short postglacial gorges, because of such obstructions. The acceptation of the above hypothesis calls for a greater number of epochs of deglaciation than has heretofore been recognized in this region; but when we consider the great complexity of drift deposits, which has been recognized else- where, we do not feel that it is necessarily an objection. This hypothesis has no facts opposed to it, while all the others are open to objections which arise from conditions in the field ; therefore we may regard it as established. Age of the gorges—A considerable lowering of the divide between the Cayuga River and the northward-flowing streams might have produced a rejuvenation of the streams tributary to Cayuga Valley. There is, however, little doubt that the Lauren- tian streams were stronger than the Susequehanna; consequently such a rejuvenation at the time of the first glacial invasion is improbable. What may have been the exact conditions govern- ing the erosion at the divide during the remainder of the glacial period is, as yet, unknown. Ten Mile Creek was a small tributary near the source of the preglacial Cayuga River, and would not feel the effect of an uplift until long after the rejuvenation had begun in the lower reaches of the stream. It is also questionable whether the uplift occurred long enough before the glacial period to permit the rejuvenation of the entire Laurentian drainage before its lower reaches were obstructed by ice. Nevertheless, the possibility of such rejuvenation must not be ignored. paper. Among the best known and most extensive are those of upper Ohio, described by CHAMBERLAIN AND LEVERETT, American Journal of Science, Third Series, Vol. XLVII, No. 280, pp. 247-82. For descriptions of local reversions see Physical Geography of New York (1902). THE INTERGLACIAL GORGE PROBLEM 149 While we cannot eliminate the possibility of a preglacial reju- venation, we can limit the effect in Ten Mile Creek valley. The bottom of the broad valley in Ten Mile Creek is about 720 feet A. T., and there isa shelf in the drift-filled gorge at 620 feet A. T. A differential uplift increasing at the rate of three feet per mile (which would be a fair estimate) to the northeast might cause trenching to 620 feet A. T. without producing a rejuvenation of the streams of the ‘“Elkland-Tioga Folio.” To permit erosion to the lowest known point in this gorge would call for an uplift twice as great, while to reach the lowest known depth of any of Fic. 6.—A view of the middle ridge looking down stream. the gorges would require an uplift which would effect the streams south of the divide. Therefore, it seems safe to conclude that the main trenching of the broad valley was interglacial, and that in the case of Ten Mile Creek the preglacial uplift did not cause the stream to cut below 620 feet A. T. There is considerable evidence that the broad valley of Six Mile Creek contains two drift-filled gorges with their bottoms below 540 feet A. T. This fact would seem to indicate at least two periods of deep trench- ing after the first glacial invasion. The length of the deglaciation intervals—In considering the length of the intervals of deglaciation, we are confronted by a great many difficulties. One of the most troublesome points is our lack of knowledge of the depth of the various gorges. Another difficulty, which is almost equally troublesome, is the uncertainty as to their chronological order. From the fact that I50 GEORGE C. MATSON the buried gorge (2) south of the house has its bottom at g1o feet A. T:, while all the others have been excavated to a still greater depth, we may conclude that excavation of the gorges to their greatest known depths would require a large amount of erosion since the glacial period. If all the gorges are inter- glacial, the ice must at one time have retreated far enough to permit gorges to be cut down to, if not below, the present lake level. This would probably require a northward drainage. A comparison of the relative widths and depths of the various gorges shown on Map II brings out the fact that the amount of erosion required to excavate the buried gorges was con- siderably greater than the amount accomplished in postglacial times. This subject is complicated by the fact that the drainage basin of Ten Mile Creek has been reduced in postglacial times; but even making a large allowance for this reduction in drainage area, we may safely say that the shortest interval of deglacia- tion was probably fully as long as the postglacial time, while the others were considerably longer—in at least one case, several times as long, Cause of the gorge-cutting—In the formation of the broad valleys the streams must: have reached base level. The only thing that could bring about rejuvenation of the streams, result- ing in gorge-formation, is an uplift of the land. This uplift must have been of long duration, even though the streams may have found some of their work of degradation, below the old valley floors, accomplished by glacial erosion. The amount of this uplift cannot be determined; but from the fact that the buried gorge of Six Mile Creek has been cleared of drift down to 420 feet A. T., without reaching rock bottom, it follows that the uplift must have been sufficient to permit the stream to cut below that depth. The buried gorge of Butternut Creek, on the west side of the Inlet Valley, is over 250 feet deep; hence the amount of uplift must have been at least 250 feet. Since Butter- nut Creek is at present flowing 250 feet below the old valley floor, we may also assume that the postglacial altitude is probably fully 250 feet greater than the altitude when the broad valleys were formed. THE INTERGLACIAL GORGE PROBLEM I51 Summary.— Briefly summarized, this paper is intended to show the existence of a series of complex gorges which are considered interglacial. The minimum number of epochs of deglaciation is two; the maximum number, four. The amount of ice recession was probably sufficient, in at least one case, to permit a north- ward drainage. The length of the epochs of deglaciation can be only roughly estimated, but the shortest was probably as long as postglacial time, and the others were doubtless much longer. The main trenching of the broad valleys was interglacial, and the minimum amount of interglacial elevation is placed at 250 feet. The land is probably more than 250 feet higher at the present time that it was during the time of the formation of the broad valleys of Cayuga Lake region. GrorGE C. Matson. CHAMPAIGN, ILL. AN INTERGEACIAL WALEEY- IN EELNOTS: In this paper it is our purpose to trace the complex history of the present Embarras River valley, in southeastern Illinois, which lies partly on the newer Wisconsin drift and partly on the older Illinoian drift, crossing the terminal moraine of the former about midway of its.course. North of the moraine the Embarras Valley is entirely postglacial, with little relation to any former valley. South of the moraine the youthful postglacial valley, as far as it extends, lies within a wider and more mature interglacial valley, which, in turn, usually lies within a still more mature preglacial rock valley. There is evidence in Illinois and vicinity of several advances and retreats of the ice within the glacial period. During each retreat the melting of the ice freed rock waste, which had been carried along beneath the ice or frozen into the glacier, and deposited this waste as a sheet of till over all the land uncovered by the melting away of the ice. One of the oldest of these till- sheets is the Illinoian, now exposed over most of the southern and western parts of the state, but concealed by subsequent sheets in the northeast, the most extensive of which is the Wisconsin. In the southern two-thirds of the state the I]linoian and Wisconsin sheets are apparently the only ones represented, and their com- mon boundary is the great terminal moraine, built during a pro- longed halt of the ice-front at its farthest advance in the Wisconsin stage, and called here the Shelbyville moraine. The time between these two ice invasions, represented by the two drift layers, corresponds with two or more stages represented elsewhere by differentiable till-sheets, but known here only by erosion and weathering in the older drift. In other words, in southern Illinois, we have an old till-sheet extending far south, the Illinoian; a much younger till-sheet extending southward about to Paris, Mattoon, and Shelbyville, the early Wisconsin; and a strong terminal moraine lying along their common boun- dary. Fig. 1 shows the distribution of these features. For 152 AN INTERGLACIAL VALLEY IN JLLINOIS 153 convenience the time between the two glacial stages mentioned may be called an interglacial epoch, remembering that in other places it may be broken up into several stages of advance with shorter interglacial stages. ig tenes hh Terminal Moraine Wisconsin Drift. 1 Illinoian Drift. Fic. 1.—Cartogram showing distribution of Early Wisconsin and Illinoian Drift and the Shelbyville moraine and others, also the position of the Embarras River with reference to these areas. Initial letters for towns and cities mentioned intext. Adapted from Leverett, Il. Glac. Lobe. Monog. 38, U.S. G.S. Since the general slope of the state is from the north to the south, most of the streams flow in that direction, and of those taking their sources on the new drift several necessarily break 154 GEORGE D. HUBBARD through the terminal moraine and proceed across the old drift to the Wabash, Ohio, or Mississippi Rivers. Among these streams which rise within the moraine loop and find their way out through it is the Embarras, to which the study now turns Rising in several more or less swampy regions, or from tile drains in Champaign county, and receiving several small streams en route, the Embarras proceeds southward across the nearly level tracts of the Wisconsin drift-sheet, and through several minor terminal moraines, until it finally breaks through the large Shelbyville moraine south of Charleston in a clear-cut youthful gorge a hundred feet or more in depth. From this point it continues a more quiet, meandering stream across the plains of the Illinoian drift, to its debouchure into the Wabash River, near Vincennes, Indiana. Through the early part of its course the Embarras makes its way, leisurely in dry weather, but rapidly in rainy weather, along open, stream-made ditches, or in semi-canals opened by dredging- shovels, in a channel not large enough to carry the flood waters. The stream can hardly be said to have a valley, only a channel, in this part of its course, but after twenty-five to thirty miles have been covered the ditch has become enlarged and a valley formed, which gradually becomes wider and deeper until the moraine is reached. During all this distance extreme youth is a chief characteristic; the valley walls are very steep—so steep, in fact, that landslides are common, and the side streams are short. The main valley attains a depth of seventy-five or eighty feet before it reaches the moraine and is cut in the drift; but occasionally bed-rock is reached, and at such places the valley often narrows to a few feet, and the stream goes through a little rock-floored gorge with rapids or tiny cascades. In the moraine the valley is perceptibly deeper, and near the southern margin of it attains a depth of over one hundred feet, while -at several points within the ten or twelve miles through the moraine the water runs on a rock floor, being definitely constricted within a rock gorge at two points. As the stream comes out upon the older drift, the valley decreases markedly in depth, and increases slightly in width. ANVTINDIERGLA CIAL VALLEY IN [LLINOTS 155 The decrease in depth is due, of course, to the absence of the newer drift-sheet and its thickened border, the moraine; the slight increase in width may doubtless be ascribed to the same cause. The valley continues to widen as the stream increases in size, but does not deepen much in the fifty-mile remainder of its course. Its fall is about that of the general slope of the land. Very rarely does the stream touch bed-rock, and at no point is the valley perceptibly constricted because of encountering the more resistant layers. In low water the stream quietly meanders along to its mouth. At anumber of points within the first ten miles south of the moraine opposite valley bluffs are apparently of different heights. This was discovered to be due to the fact FIG, 2.— - - -, supposed boundary of preglacial valley; 4, interglacial valley filled with; 4, Wisconsin overwash material; C, present valley — postglacial; D, present stream bed; “, remnant of terrace on east side; /, present flood plain; G, second bottoms; Z, Illinoian drift sheet. Vertical exaggeration about 25. that the present gorge-like valley is within an older, more mature * valley, whose sides are reached by the inner valley only on the east, and not continuously there. Hence along the west side there are practically two continuous bluffs, the inner one being from a few rods to three-fourths of a mile within the outer one. But on the east side the stream has pushed its inner bluff much nearer the outer one, reaching it at a number of points and blending with it in one continuous descent from the uplands to the present flood-plain. Where this stage has not been reached there are two bluffs on the east side, as on the west, but they are rarely far apart. When the bluffs have coalesced on one side and not on the other, opposite bluffs seem to stand at discordant levels. Fig. 2 illustrates the case. The second or outer bluffs often have a scarp nearly as steep as will stand, say 20-25°, but are usually of gentler slope. 156 GEORGE D. HUBBARD In the latitude of Bradbury, some three miles south of the outer margin of the moraine, the second bluff attains nearly the same height above the first as the first does above the stream. Farther north the first bluff rises higher with reference to the stream and the outer bluff, until, near the moraine, the second becomes almost imperceptible, while farther south the first gradu- ally decreases in height until it becomes practically mz. Fig. 2 is a cross-section, somewhat idealized, about opposite Bradbury, at a point where the two eastern bluffs coalesce. Fig. 3 is a bird’s-eye view much conventionalized, of the conditions along the west side of the stream, as discussed above. What is the relation of all this to glaciation? Evidently there are two periods of erosion represented. Is one postglacial ? Where, in time, is the other? The nature of the materials in the two sets of bluffs and their relation to surrounding features alone can solve the problem. Certainly there must be a preglacial valley somewhere near this present stream, unless the whole plan of the preglacial drain- age has been upset. The Ohio, Mississippi, Illinois, and lower Wabash in this latitude are, in general, about where they were in preglacial time; hence the general direction of other large streams must be similar to that of their earlier courses, but they are not necessarily in exactly the same beds. The older drift, according to the records of borings and coal shafts, and the exposed sections © in stream and railroad cuts, is from twenty to fifty feet. thick, rarely more, often less. Between the outer or second bluffs which continue down the stream is a valley, increasing from a mile to three miles in width, and in the deeper portion of its cross-section about fifty feet deep, rarely revealing bed-rock in its floor or sides. Both are practically of continuous drift, correspond- ingin composition, physical characteristics, and state of weathering with the general drift-sheet on either side. The valley may be followed for fifty to sixty miles with no material change of char- acter, so far as these features are concerned. There are three time-periods to be considered in the solution of the origin of the valley in question. First, is it a preglacial valley? Second, was it formed since the retreat Of the last 1¢e- at AUN ND RGIEA CIAE VALEENA TIN TLETNOLS: 157 sheet? Third, is it interglacial, formed after the advance of the ice that left the Illinoian and before that which left the Wiscon- sin drift? If it can be shown that the valley is neither preglacial nor postglacial, but is subsequent to the deposition of the [lli- noian drift and prior to that of the Wisconsin, its position will be established. The three possibilities will be taken up in the order mentioned. Since this valley, walled by the second bluffs, is cut in the Illinoian drift-sheet, it must have been formed subsequent to the laying down of that till-sheet. The character of the walls also precludes the possibility of its being due to rejuvenation Fic. 3.— Looking N. W. from above east bluff at south end of the Wisconsin over- wash in the interglacial valley. R—=Embarras River, F = present flood plain, S = “second bottoms ’’— the terrace top, U—upland surface. Vertical scale very greatly exaggerated. just prior to the first advance of the ice, as may be the case with those gorges cut in rock, but more or less filled with drift of a late age, such as have been reported in New York. Again, since the valley is so rarely cut down to bed-rock, when the latter is usually very near the surface outside the valley, it is probably in a preglacial, well-matured valley, whose limits are reached only when the present valley encounters ledges. Furthermore, since the valley in the [llinoian drift apparently lies within the pregla- cial valley, the latter must have been nearly, but not completely, filled and obliterated by the drift-sheet of this early ice invasion. Had it been completely filled, how could the stream now be so continuously withinit? It is also true that no large percentage of the drift-filling has been subsequently removed, else the walls of the preglacial rock valley would have been more frequently dis- . covered. Likewise, it is reasonable, if not imperative, to suppose 1538 GEORGE D. HUBBARD that when the valley in the Illinoian drift was first formed it extended to a greater depth than the present stream has cut, because the latter, in the few miles of double gorge, is not yet running on Illinoian drift, not having succeeded in cutting through a subsequent filling of the valley in the Illinoian. This leads to a consideration of the nature of the inner, lower bluffs which accompany the present stream some miles beyond the terminal moraine, and are represented by the second bottom terrace front in Fig. 3, S.. They are of gravels and sands and, farther south, clayey, fine saads, and partially sorted and strati- fied drift material corresponding in composition and stage of weathering with that of the earlier Wisconsin, being similar also to deposits on the uplands spread out as an apron in places along the southern margin of the moraine. The second bottoms between the inner and outer bluffs on each side of the stream, where they are not blended into one bluff, are made of the same materials. Wells, on these bottoms, show a section of the same materials to a depth of fifteen to thirty feet. No deeper wells were found in it. Furthermore, the relation of this terrace material, thick near the moraine and feathering out to nothing southward, points to the same conclusion—namely, that the materials of the walls of the inner gorge are layers of overwash material carried out beyond the terminal moraine by the over- loaded stream flowing from the melting ice, and deposited by that stream in order to aggrade its course, and thus increase its carrying power to a strength equal to its remaining load. Since this material is of Wisconsin age and is laid down within a valley already formed, that valley cannot be postglacial. The third and only remaining alternative is that the valley represented by the outer bluffs is one formed between the deposition of the Ill1- noian and that of the Wisconsin drift-sheets, and hence may be called an interglacial valley. Its width and general maturity differentiate it from the known postglacial valleys, while its relative lack of greater maturity, even though in unconsolidated materials, declares it not to belong to the preglacial drainage. The same facts speak for an interglacial time period consider- ably longer than the postglacial period has been. AWN TINDERGEA CHAT VALLEVAIN TEEINOTS 159 Subsequent to the formation of the overwash plain in the interglacial valley and the associated retreat of the ice, the habit of the stream has been changed, because its load has been greatly reduced. Probably its volume has also been somewhat cut down, but not to correspond with the reduction in burden; hence, whereas the stream was aggrading its course during the Wisconsin stage, it has been degrading during the postglacial stage, and has so far succeeded in removing the previously deposited material that the stream now flows in a gorge, a half mile or more in width. The stream is still widening this valley, and the second bottoms(Fig. 2, G) still remain to be removed. The second bottoms constitute terraces of better agricultural soil than that of the vicinal upland. They correspond in orign with those described by early writers in many New England valleys, and especially those in the Connecticut and tribu- taries so thoroughly studied recently by Professor W. M. Davis. They may be correlated with many terraces in Illinois valleys, because nearly all streams heading on the new drift and flowing out through the moraine are terrace-bordered. The conditions differ to some extent from those reported by Professor Davis in that the postglacial excavation has not proceeded sufficiently as yet to encounter the walls and floors of the old valley. There- fore there are no defending ledges, no series of terrace steps, and very few cusps. The stream has cut back as far, or a little farther, each time than on the previous swing, and hence there is one long step from the second bottoms or terrace-top to the present flood-plain. It is unfortunate that more attention has not been given to the terraces of the Illinois, Iowa, and Wiscon- sin streams. There are many such terraces, and no doubt a careful study of them would yield valuable returns. Summary.—The history of the present Embarras River valley is a complex one. North of the moraine it is entirely postglacial, and probably bears little relation to its precursor of interglacial time, still less to the preglacial valley. Very likely near the moraine and in it where the valley is deepest the stream has cu its recent channel mostly in the filling of a former valley, but the occurrence of occasional rock-ledges witnesses to the errors of 160 GEORGE D. HUBBARD the stream in attaining its former course. Farther back from the moraine, even to the stream’s sources, the drift is so deep and the channel is so shallow that the absence of bed-rock in the bottom of the valley proves nothing as to the position of the earlier stream and its valley. South of the moraine the till is much thinner, and a valley with such depths as those of the Embarras would often, if not almost constantly, reach bed-rock, were it not within an earlier valley. Hence there was, first, a broad, well matured preglacial valley similar to those at present in nearby carboniferous rock in unglaciated parts of Indiana, Kentucky, and southern Illinois. Second, this valley was nearly filled with drift of Illinoian age. Third, after the retreat of the ice of the Illinoian stage and before the Wisconsin invasion, a valley was excavated in the above-mentioned filling. This probably extended some distance north of the present terminal moraine, possibly as far as the present sources of the Embarras. Fourth, this interglacial valley was partly filled with overwash material (entirely filled, north of the moraine, with drift) of Wis- consin age. And, fifth, the present youthful postglacial channel has been cut in the filling of the interglacial gorge, but the process has not gone far enough to reach bed-rock where the stream is fairly within the earlier valleys. GEORGE D. HUBBARD. CORNELL UNIVERSITY, Lthaca yin. Ys REVIEWS. SUMMARIES OF PRE-CAMBRIAN LITERATURE FOR !902-1903. II. [Continued fron p. 62.] Cake Vern: A. P. COLEMAN AND A. B. WILLMOTT. “The Michipicoten Iron Ranges.” “Geological Series,” University of Toronto Studies, 1902, pp. 39-83. See also Report of the Bureau of Mines, Ontario, 1902, pp. 128-51. Coleman and Willmott describe and map the Michipicoten iron ranges. The rocks are classified as follows : (@lGaunentiante. acs is4 ac Gneisses and granites { Basic eruptives Upper Huronian...... ee eruptives AC HS aaah co eeest Doré conglomerate Z ( Eleanor slates Helen iron formation | Wawa tuffs | | Gros Cap greenstones Lower Huronian...... KG The Gros Cap greenstones are basic eruptives with ellipsoidal structure corre- sponding in position and character to the Ely greenstones of the basement complex of the Vermilion district of Minnesota. They are in part basal to the other rocks of the district, but in part also they are interbedded with the rocks of the Helen iron formation. The Wawa tuffs are acid schists having the composition of quartz- porphyry or felsite, usually in the form of tuff, ash, or breccia, and sometimes show stratification, taken to indicate deposition by water. Slates of distinctly sedimentary origin, occurring in thin bands near Eleanor Lake and called the Eleanor slates, are referred to the Lower Huronian. Their relations to the Helen iron formation are not known. The Helen iron formation, 500 feet thick, comprises banded granular silica with more or less iron ore, black slate, siderite with varying amounts of silica, and griinerite schist. All are found well developed at the Helen mine, and all but the griinerite schist have been found in the Lake Eleanor iron range also, while granular silica and siderite occur in large quantities in every important part of the range, though small outcrops sometimes show the silica alone. All of the rocks of the iron formation contain con- siderable amounts of iron pyrites. The grained silica and the granular silica is similar in certain respects to the jaspers and ferruginous cherts of the United States, and their origin is believed to be the same. ‘They differ in being often soft, pulverulent, and brecciated. The black, graphitic slate, forming a thin sheet just under the iron range proper west of the Helen mine and at other points in the region, seems closely related to the granular silica, being composed of the same material with a large admixture of carbon which smears the fingers. Iron ore is mined in the Helen mine, and this mine is described in detail. The 161 162 REVIEWS ore body is located at the east end of the deep Sayers Lake basin, partly above and partly below the old water level. The lake has now been drained, and the ores appear in a great amphitheater opening out to the west. The rocks immediately associated with the hematite are siliceous ore, ferruginous cherts, or grained silica rocks. These are mapped as immediately surrounding the iron ore, and also as forming for the most part the north wall of the amphitheater. The east wall of the amphitheater is composed of iron carbonate which shows gradations into siliceous ore and into hematite ore. ‘The south wall is composed of Wawa tuffs. The ores are believed to have resulted from the secondary alteration of an original iron formation consisting mainly of iron carbonate, grained silica, and limestone, in part interbedded with the Wawa tuffs, but mainly deposited above them. The iron formation and the tuffs were folded up together, with the result that the tuffs were formed into a trough underlying the iron formation, and the iron formation within this trough was folded and brecciated. Percolating waters then altered the iron carbonates. Probably the chief solvent of the carbonates was acid ferric sulphate or sulphuric acid resulting from the oxidation of the iron pyrites, which are found in considerable quantity throughout the iron formation. The ore body has resulted directly from the alteration of iron carbonate, the oxidation of the iron sulphide having yielded but little ore. The oxidation of the iron took place where solutions of iron carbonate came into contact with waters bearing oxygen. The principal areas of iron formation possibly bearing iron ore at Gros Cap, Sayers, and Boyer Lakes, just east of the Helen mine, around Brooks Lake, south of Long Lake, just east of Goetz Lake, in Parks Lake, and between Parks and Kimball Lakes. The Upper Huronian rocks are represented principally by the Doré conglomerate, occurring typically at the mouth of the Doré River and thence eastward beyond Michipicoten Harbor, and to a less extent in other parts of the district. This con- glomerate is unconformably above the Lower [luronian rocks of the district. It con- tains pebbles of granite, felsite, conglomerate, granular silica of the iron formation and breccia. The Doré conglomerate is cut by acid intrusives in dikes and bosses. These are the latest rocks of the region. é The Laurentian granites and gneisses have not been studied in detail in the Michipicoten district, but their associations with both Lower and Upper Huronian prove them to be post-Huronian eruptive masses. Comment.— As noted in the above paper, there is very close similarity in lithology and succession between the rocks of the Michipicoten district and the rocks of the Vermilion iron-bearing district of Minnesota, although described under different names. The rocks above called Lower Huronian and Upper Huronian are called respectively Archean and Lower Huronian by the United States geologists. There is substantial agreement in the matter of the origin of the ores in the two districts.» The Wawa tuffs are also similar to the Palmer gneiss of the Archean of the Marquette district. The granites and gneisses described as Laurentian for the Michipicoten district are similar in character and relations to granites in United States districts referred to the lower Huronian. ‘J. MORGAN CLEMENTS, “The Vermilion Iron-Bearing District of Minnesota, Monograph XLVI, U.S. Geological Survey, 1903; C. R. VAN Hisk, “ The Iron-Ore Deposits of the Lake Superior Region,” Tzverty-first Annual Report of the U. S. Geological Survey, Part ILI, pp. 305-434. REVIEWS 163 A. P. COLEMAN. ‘Rock Basins of Helen Mine, Michipicoten, Canada.” Bulletin of the Geological Society of America, Vol. XLIII (1902), pp. 293-304. Coleman discusses the origin of the rock basins of Boyer and Sayer’s Lakes of the Michipicoten district of Canada, the former containing the Helen iron-ore body. He holds the lake basins to have resulted from the solution of the iron-bearing rocks long before glacial time. A. B. Wit~motTr. ‘The Nomenclature of the Lake Superior Formations.” JOURNAL OF GEOLOGY, Vol. X (1902), pp. 67-76. Willmott discusses the nomenclature of the Lake Superior formations, this being practically a consideration of Van Hise’s ‘‘Iron-Ore Deposits of the Lake Superior Region.” ? He argues principally against the correlation of the Animikie series with the Upper Huronian of the original Huronian area. He states that there can be no doubt that Logan in 1863 included within his Huronian two series —the one typically represented by the banded jaspers, the other by the slate conglomerate and the jasper conglomerate. This has been uniformly followed from that time forward by all Canadian geologists, and by many American, the vertical green schists and their interbedded banded jaspers being considered Lower Huronian. Professor Willmott doubts the advisability of attempting the separation of the green volcanics and sedi- ments, except in limited areas of economic value. Here each would be given forma- tional names, just as Van Hise has done with the Ely greenstone and the Soudan iron formation. In other places the volcanics and eruptives will take the name of the sediment with which they are associated. The lowest sedimentary series of the Lake Superior region is the Lower Huronian. These sediments were included in the areas mapped.as Huronian by Logan in 1863, and, although not actually found in place by him, were recognized from their fragments, and to him should be given the credit. As so used, the term ‘“‘ Lower Huronian” is nearly equivalent to the term “Archzean” as used by Van Hise, and the term “Upper Huronian” is equivalent to Van Hise’s ‘Lower Huronian.” Accordingly, the Animikie, or the Upper Huronian of Van Hise, is younger than the original Huronian series. That the Animikie is later than the true Upper Huronian or original Huronian may be shown in the following ways: I. Stratigraphically it is the third series of sediments upwards from the bottom of the geological column in the Lake Superior region; the Upper Huronian is the second. 2. Lithologically, the two series are quite different, and so presumably are of dif- ferent age. There is very little conglomerate at the base of the Animikie; in the ‘Huronian the quartzites, slate conglomerates, and jasper conglomerates are of great thickness. The oolitic jaspers found in the Animikie are quite absent from the Huronian. The shales, so important in the Animikie, are almost unknown in the Huronian. The laccolitic sills of the Animikie are lacking in the Original Huronian. 3. Structurally, the two series are usually said to be alike in that both lie flat and undisturbed. While this is quite true of the Animikie, it is only partially true of the Huronian north of the Georgian Bay, and is untrue of the Upper Huronian about Batchawana and Michipicoten. Coleman? and Murray3 have described cases of tC. R. VAN Hise, “The Iron Ore Deposits of the Lake Superior Region,” Twenty-first Annual Report of the U. S. Geological Survey, Part II], pp. 305-434. ? Bureau of Mines, Ontario, 1901, p. 189. 3 Geological Survey of Canada, 1858, p. 95. [64 REVIEWS vertical dip within the so-called Original Huronian, and others have been observed by myself. These seem to occur around the outer portion of the Huronian basin, and more gentle dips obtain in the central part. Evidently the Huronian has been sub- jected to forces which the later Animikie has escaped. 4. Assuming that the large areas of eruptive granite-geisses in the Lake Superior region are of the same age, we find that the Upper Huronian has in many cases been pierced by them, but that the Animikie always overlies them. A. P. COLEMAN. ‘The Huronian Question.” American Geologist, Vol. XXIX (1902), pp. 325-34. Coleman discusses the Huronian question, his argument being mainly against the correlation of the Animikie series of the Lake Superior region with the Upper Huronian series. Evidence that the Animikie is unconformable above his Upper Huronian series is summarized, and emphasis 1s placed on the points that both the Upper Huronian and the Lower Huronian differ lithologically from the Animikie ; they are metamorphosed and schistose as compared with the Animikie; and they are much folded aad highly tilted, in marked contrast to the Animikie. Comment.—For the most part the terms “Upper Huronian” and “ Lower Huronian,” as applied by Professors Willmott and Coleman to rocks owd¢s¢de of the part of the Original Huronian area of Logan on the north shore of Lake Huron, are to be correlated respectively with the “Archean” and “Lower Huronian” of the United States geologists, and thus Van Hise’s “Upper Huronian” or “ Animikie” comes above their ‘Upper Huronian.” For such areas, therefore, there is no marked difference of opinion as to the number and succession of series, but only difference in names. However, when it comes to the correlations of these series with the rocks of the Huronian series on the north shore of Lake Huron the difference is fundamental. Coleman and Willmott, in common with other Canadian geologists, apply the term “ Upper Huronian” to the entire series north of Lake Huron mapped as ‘“‘ Huronian”’ to underlying greenstones, green 9 by Logan, and apply the term “Lower Huronian schists, and jaspers (as typically developed in the Michipicoten district). This Lower Huronian series, with the addition of certain “Laurentian” granites, corresponds approximately to what Van Hise, following the terminology of the U. S. Geological in this and other parts of the Lake Superior region. ? Survey, has called the “Archzean’ But the sediments which Logan mapped as ‘ Huronian,” and which are classed as “Upper Huronian”’ by Willmott and Coleman, have been divided on the north shore of Lake Huron by Van Hise and Pumpelly, following Alexander Winchell, into the “Lower Huronian” and “Upper Huronian”’ series, the break being placed at the base of the Upper slate conglomerate, It is with these divisions of the Original Huronian series that the correlation of the Upper Huronian and the Lower Huronian series of the rest of the Lake Superior region has been made by Van Hise. Field work done on the north shore of Lake Huron during 1902 by Professors Van Hise, Seaman, and the writer presents further evidence of the correctness of this correla- tion. A full discussion of the evidence is not possible here, but it will be presented shortly in a general monograph on Lake Superior geology now in preparation. ANDREW C. Lawson. ‘‘The Eparchean Interval: A Criticism of the Use of the Term Algonkian.” 2udletin of the Department of Geology, Uni- versity of California, Vol. III, pp. 61-52. Lawson criticises the use of the term ‘‘ Algonkian.” Heemphasizes the impor tance of the interval, which he calls the Eparchzean interval, between the ‘‘ Huronian’ REVIEWS 165 and Animikie series, that is, between the Lower Huronian and Upper Huronian, of the United States Geological Survey, and argues that no one term such as “ Algon- kian” should include a break of this importance. It is proposed to restrict “Algon kian”’ to the Animikie and Keweenawan rocks, and to retain Dana’s term “‘Archzean’’- for all rocks below the Animikie, z.e., below the Eparcheean interval, and also to retain the terms “ Laurentian” and “ Huronian,” as subdivisions of the Archzean, with the significance originally given them by Logan. The correlations of the Animikie with the Keewatin of Minnesota and with the Upper Huronian of Lake Huron is regarded ? as an error, because of dissimilarity in lithology, stratigraphy, and in relations to intrusives. Following Willmott and others, itis believed that the Animikie is younger — than the (Upper) Huronian series of Lake Huron, and thus later than the Eparcheean interval. Emphasis is placed on the marked lithological similarity in the sedimentary series below the Eparchzean break in the Lake Superior and Lake Huron region, and the probable correlation of these rocks with the Original Huronian series. Summar- ized in tabular form, the correlation proposed is as follows: ( Cambrian (Upper division or Potsdam only). Unconformity. Paleozoic, + { Keweenawan, | Algonkian. { Unconformity. l | Animikie = Penokee = Upper Marquette. Eparchean Interval. { Huronian =Upper Keewatin = Lower Marquette, etc. Unconformity. Laurentian, so called, granite gneisses, etc. (intrusive in the Ontarian) and the Carlton anorthosites, 4 [err tes Huronian =Crystalline schists of south shore | Onan d invaded by granite gneisses. Unconformity. [ [ Coutchiching, Archean Comment.— Concerning the correlation of the Animikie with the Upper Huron- ian the comment on Professors Willmott and Coleman’s articles, summarized on a preceding page, is pertinent. Dr. Lawson implies that the Animikie has been corre- lated by the U.S. Geological Survey with the entire sedimentary Huronian series of the north shore of Lake Huron, while it has been correlated only with the portion of this series above the limestone; and against such a correlation his argument loses some of its force. If the Keweenawan and Animikie series are not Cambrian, as they are held to be by Lawson, but are pre-Cambrian, as held by the U.S. Geological Survey, then Lawson’s objection to the term ‘“‘ Algonkian,” as replacing in part the old term ‘“Archzean,’’ would prevent its application even to the Animikie and Keweenawan rocks to which he restricts it. As already noted, the nomenclature of Lake Superior and Lake Huron series is being fully discussed in a general monograph on Lake Superior geology now in preparation. Arguments for the adoption and retention of the term “ Algonkian’ ? will be summarized, together with new arguments developed in recent field work. C. R. VAN Hise. “Geological Work in the Lake Superior Region.” Pvo- ceedings of the Lake Superior Mining Institute, Vol. VII (1902), pp. 62-69. Van Hise briefly sketches the history of geological mapping in the Lake Superior region, calling attention to the difficulty of preparing accurate maps, and concludes that the maps which have been published from time to time since the earliest map of Foster 166 REVIEWS and Whitney represent reasonably close approximations to the facts as then known, and that, notwithstanding their many imperfections, they have been of service at the time of publication. J. E. Spurr. “The Original Source of the Lake Superior Iron Ores.” American Geologist, Vol. XXIX (1902), pp. 335-49. Spurr discusses the origin of the pre-Cambrian iron ores of the Lake Superior regions. He repeats his conclusion that the Mesabi ores have resulted from the altera- tion of a green ferrous silicate uf the class of glauconite, and further states that his conclusion in reference to the Mesabi iron formation may be ‘ most of the other Lake Superior iron ores.” ‘probably applied to Comment. — This paper is practically a reply to a brief abstract published in the Engineering and Mining Journal of an informal talk given by the writer before the Geological Society of Washington. While fully agreeing with Mr. Spurr’s major conclusion that the Mesabi ores have resulted from the alteration of ferrous silicate granules, the writer has emphasized certain facts which seem to prevent the applica- tion of the name “glauconite”’ to this silicate.* As to the statement that conclusions applicable to the Mesabi ores apply to most of the other Lake Superior iron ores, presumably this is based on certain similarities in granules and concretionary forms to be observed in the iron-bearing rocks of the Mesabi and Gogebic districts. This similarity the writer has discussed elsewhere,’ and believes it will afford no support for Dr. Spurr’s somewhat sweeping statement. C. K. Lerru. ‘A Comparison of the Origin and Development of the Iron Ores of the Mesabi and Gogebic Iron Ranges.” Proceedings of the Lake Superior Mining Institute, Vol. VII (1902), pp. 75-81. Leith compares the origin and development of the Gogebic and Mesabi iron ores. The ores of the two districts occur in the same geological horizon; they result from the alteration, under weathering conditions, of a ferrous compound of iron, through the agency of percolating waters, and are localized in channels of vigorous circulation of water. But the differences in the development of the ores of the two districts are important. The original ferrous compound of iron is mainly iron silicate in the Mesabi district, and iron carbonate in the Gogebic district, although both substances appear in each district. The localization of the ores in the Gogebic dis- trict during their concentration has been within clear-cut pitching troughs with defi- nite shapes, while in the Mesabi district the very gentle folding of the iron formation, its fracturing, and the absence of intrusives combine to make the channels of vigor- ous flow within the iron formation most devious, resulting in the curious and exceed- ingly irregular shapes now to be observed in the Mesabi ore deposits. The original ferrous silicate from which the ores develop in the Mesabi district is in minute homogenous granules, the form of which remains even after the substance is changed. Associated with these granules are undoubted concretions of iron oxide and chert with concentric structure. In the Gogebic district there appear numerous =C. K. Lerru, “The Mesabi Iron-Bearing District of Minnesota,” Monograph XLITI, U.S. Geological Survey, 1903. 2C. K. Leiru, “ A Comparison of the Origin and Development of the Iron Ores of the Mesabi and Gogebic Iron Ranges.” Proceedings of the Lake Superior Mining Institute, 1902, pp. 75-81. (See summary below.) REVIEWS 167 concretions with concentric structure, which Van Hise has shown to develop during the alteration of iron carbonate; andassociated with these are rare granules of iron oxide and chert in varying proportions, which may represent altered ferrous silicate granules similar to those of the Mesabi district. Evidences of the existence of original ferrous silicate granules in the Gogebic district are not sufficiently numerous to warrant modi- fication of Van Hise’s conclusion that the ores have developed from the alteration of iron carbonate. C. K. Letra. ‘The Mesabi Iron-Bearing District of Minnesota.’ Mono- graph XLIII, U. S. Geological Survey, 1903. Leith describes and maps the geology of the Mesabi district of Minnesota. The district is two to ten miles in width, extending from near Grand Rapids on the Mis- sissippi River to Birch Lake, a distance of approximately one hundred miles. The main topographic feature is a ridge known as the Giant’s (or Mesabi) Range, which’ extends the length of the district. The geologic formations represented in the district belong, in ascending succession to the Archean, Lower Huronian, Upper Huronian Keweenawan, Cretaceous, and Pleistocene. They are all separated by unconformities, The core of the Giant’s Range is formed by Archzan and Lower Huronian rocks; except for the portion in ranges 12 and 13, where Keweenawan granite forms the core. On the south flank rest the Upper Huronian rocks, containing the iron-bearing forma- tion, with gentle southerly dips. The Keweenawan gabbro lies diagonally across the east end of the district. The Archean rocks consist principally of green rocks of great variety, including dolerites, metadolerites, basalts, metabasalts, diorites, and hornblendic, micaceous, and chloritic schists. The more massive rocks frequently have an ellipsoidal structure, which is characteristic of the green igneous rocks of other parts of the Lake Superior region. In addition to the green basic rocks, there are present small areas of granite and porphyritic rhyolite. The Lower Huronian series consists of sediments and granite. The sediments are graywackes, slates, and conglomerates, all metamorphosed, with bedding and schistosity practically vertical. They may be as thick as 10,000 feet, but it is thought more probable that the thickness does not exceed 5,000 feet. The Lower Huronian sediments rest unconformably upon the Archean rocks, as shown by basal conglomerates containing fragments of all the varieties of rocks found in the Archean. The Lower Huronian granite forms the main mass of the Giant’s Range westward from a point near the east line of Range 14 W. It is intrusive into both the Archzan rocks and the Lower Huronian sediments, and has produced strong exomorphic effects in both. The Upper Huronian or Animikie consists of three formations —the Pokegama quartzite at the base, above this the Biwabik formation (iron-bearing), and above this the Virginia slate. The Pokegama quartzite comprises vitreous quartzite, micaceous quartz-slate, and conglomerate. The thickness ranges from 0 to 500 feet, averaging about 200 feet: The conglomerate at the base indicates unconformable relations of the Pokegama formation to the Archean and Lower Huronian rocks. The Biwabik formation, the iron-bearing formation, comprises ferruginous, amphibolitic, sideritic, and calcareous cherts, siliceous, ferruginous, and amphibolitic slates, paint rocks, “‘greenalite”’ rocks, sideritic and calcareous rocks, conglomerates -and quartzites, and iron ores. Cherts make up the bulk of the formation. The 168 ; REVIEWS original rock of the formation is shown to consist largely of minute granules of green ferrous silicate, thus confirming Spurr’s conclusion. ‘The material was called “ glau- conite”” by Spurr, but is here determined to be a hydrous ferrous silicate entirely lack- ing potash, and thus not glauconite. It is named “greenalite” for convenience in discussion. The cherts and iron ores are shown to develop mainly from the alteration of the greenalite granules. The slates are in thin layers interbedded with the other phases of the iron formation. The paint rocks result from the alteration of the slates. The conglomerates and quartzites form athin layer from a few inches to perhaps 15 feet or more in thickness at the base of the formation. They pass upward into fer- ruginous cherts of the iron formation rather abruptly, though usually at the contact the chert and quartzite are interleaved for a few feet. The conglomerate of the iron formation rests upon Pokegama quartzite, indicating a slight erosion interval between the Biwabik and Pokegama formations, although the interval is not shown by dis- cordance in bedding, which is parallel in both. Heretofore the quartzite and con- glomerate in the iron formation have not been discriminated from the rocks of the Pokegama formation. In the eastern portion of the range the iron formation is in contact with the Keweenawan gabbro and granite, and near this contact has suffered profound metamorphism. The characteristic rocks of this area are amphibole- magnetite-cherts. The thickness of the formation may vary from 200 to 2,000 feet. The average may be 1,000 feet. The Virginia slate is essentially a soft slate or shale formation, but it contains graywacke phases, near its base a little limestone, and near its contact with the gabbro is metamorphosed into a cordierite-hornfels. The normal slate phases of the formation may be distinguished with difficulty in isolated occurrences from the slate layers in the Biwabik formation. The separation of the two is of importance to the explorer, and hence an attempt is made to determine criteria for their discrimination. The thickness of the Virginia formation cannot be measured within the district, but from analogy with the Penokee-Gogebic district and the extent of the low, flat-lying area south of the Mesabi range supposed to be occupied by the slate, the formation is believed to have a very considerable thickness. The slate grades, both vertically and laterally, into the Biwabik formation. The entire Upper Huronian series is well bedded, conformable in structure (although having a thin conglomerate between the Biwabik and Pokegama forma- tions), and dips in southerly directions at angles varying from 5° to 20°, and excep- tionally at higher or lower angles. The series is gently cross-folded, and the axes of the cross-folds pitch in southerly directions. Accompanying the folding is consid- erable jointing, especially in the brittle Pokegama and Biwabik formations. Indeed, in these two formations the folding is brought about mainly through relatively minute displacements along joints, while in the Virginia formation the folding has taken place mainly by the actual bending of the strata. : The thickness of the Upper Huronian series within the limits of the district mapped may average about 1,500 feet ; but if the total thickness of the slate formation outside the limits of the district be taken into account, the total thickness of the Upper Huronian series is probably several times this figure. The relations of the Upper Huronian series to the subjacent formations are those of unconformity, as evidenced by basal conglomerates, discordance in dip, difference in amount of deformation and metamorphism, distribution of the series, and relations to intrusives. oo REVIEWS 169 The Keweenawan rocks consist of gabbro, diabase, and granite, all of which are intrusive into the rocks with which they come into contact. The north edge of the gabbro runs diagonally across the east end of the district from southwest to northeast, resting upon the edges of each of the members of the Upper Huronian series, and at Birch Lake against the Lower Huronian granite. North of the gabbro margin, in Range 12, are isolated exposures of diabase which may represent sills associated with gabbro intrusion. The granite forms the crest of the Giant’s Range through Ranges 12 and 13. This granite has not heretofore been discriminated from the Lower Huronian granite. The exomorphic effect of the gabbro and the granite upon the Upper Huronian series has been profound. J. MorGAN CLEMENTS. “The Vermilion Iron-Bearing District of Minne- sota.” Monograph XLV, U.S. Geological Survey, 1903. Clements describes the geology of the Vermilion iron-bearing district of Minne- sota. Elaborate general and detailed maps, accompanying this report, are based on field work by Clements, Van Hise, Bayley, Merriam, and Leith. The district ranges from two to eighteen miles in width, and extends from a little west of Lake Vermilion in a direction a little north of east to Gunflint Lake on the international boundary, a distance of about one hundred miles. The rocks of the district are described under the headings “Archean,” ‘ Lower Huronian,” and ‘‘ Upper Huronian,” representing series separated by marked uncon- formities. The Archzan of the Vermilion district is*divided into three formations, as follows, given from the base up: the Ely greenstone, the iron-bearing Soudan formation, and the granites of Vermilion, Trout, Burntside, Basswood, and Saganaga Lakes. The Ely greenstones consist of basic and intermediate igneous rocks widely dis- tributed in anticlinal areas, as shown by the distribution of the overlying sediments. They were originally rocks corresponding in character to intermediate andesites and basic basalts. They have been extremely altered, but retain in many cases in striking perfection the original structures, such as ellipsoidal parting, and spherulitic and amygdaloidal structures. A study of their various textures and structures shows that these greenstones are unquestionably of igneous origin, and are largely of volcanic character. Many of them have been rendered schistose by pressure. The greenstones have also been strongly affected by the contact metamorphism due to the intrusion of great granite masses. Asa result of this intrusion, there have been produced from the greenstones amphibole-schists, which form a marginal facies of the greenstones, lying between them and the adjacent granites. The greenstones have also been met- amorphosed by the Duluth gabbro of Keweenawan age, and granular rocks have thus been produced which in most cases show the originai textures of the greenstones, but contain also a development of fresh biotite, hypersthene, brown-green hornblende, and magnetite. The Soudan iron formation is widely distributed in the western part of the dis- trict, but is practically wanting in the eastern half. It is found mostly in narrow belts, which consist largely of greenstone so intimately associated with the iron formation that it has been impossible to separate them on the map. The formation consists of (1) a very subordinate fragmental portion made up of some conglomerate, clearly rec- ognizable as having been derived from the underlying greenstones, grading up into sediments of finer character; and (2) lying above this fragmental portion, the iron- 170 REVIEWS bearing formation proper, which consists of siliceous rocks, largely white cherts — though varying in color from white, green, yellow, and purplish to black —with red jasper and carbonate-bearing chert, griinerite-magnetite-schist, hematite, magnetite, and small quantities of pyrite. These iron-bearing rocks are clearly of sedimentary origin. They do not now present their original characters, but are presumed to have been derived from rocks that were largely carbonate-bearing, ferruginous cherts. The relation of the iron formation to the adjacent greenstones is clearly that of a sediment- ary overlying an igneous series. The few basal conglomerates of the iron formation that have been found consist of pebbles derived from the underlying greenstone, showing conclusively their relationship. This relationship is obscured, however, in most places, by the absence of the conglomerates, and by the fact that the iron formation has been very closely infolded in the greenstone. In consequence of the extreme folding and of the impossibility of determining different horizons in the iron formation, it has been impracticable to ascertain its thickness. The iron-ore deposits of the Vermilion district show a striking analogy with those of the Marquette district. Like them, they may occur in two positions with respect to the iron-bearing formation. They are found, first, at the bottom of this formation, and, second, within it, the ores in both cases being the same in character. The Ely deposits are typical of the deposits occurring at the base of the formation. They are found at the bottom of a closely compressed syncline of the iron formation where it lies in the relatively imper- vious greenstone. The source of the iron was, in the first instance, the Ely greenstone. From this it was removed through the action of water and collected in the Archzan sea to form the sedimentary deposits of the Soudan formation. After the folding of the formation this disseminated iron was carried by downward percolating waters into places favorable for its accumulation, such as the bottom of this synclinal trough, where it was precipitated by oxygen-bearing waters coming more directly from the surface. Pari passu with this precipitation silica was removed, affording space for the accumulation of the iron to form the ore deposits as now known. The Tower and Soudan deposits differ only in detail from the Ely deposit. Granites, intrusive into the Archean, occupy a wide area, and are named from the topographic features with which they are conspicuously associated. That these intrusives are older than the Ogishke conglomerate (Lower Huronian), which succeeds in age the Soudan formation, is shown conclusively by the fact that pebbles derived from them occur in this conglomerate. The general period of intrusion of all of these acid igneous rocks is placed between the time of the deposition of the latest sediments of the Archzean and that of the deposition of the earliest sediments of the Lower Huronian series. The Lower Huronian occurs in two detached areas, one of which, known as the Vermilion Lake area, extends from the western limit of the area mapped, in the vicin- ity of Tower, to within about eleven miles of Ely on the east, and the second of which, known as the Knife Lake area, begins about seven miles west of Ely, and extends eastward to the eastern limit of the area mapped. At the base of the series there lies a great conglomerate, known as the Ogishke conglomerate, containing pebbles and finer detritus from all of the rocks of the Archean. Above this conglomerate, in the eastern portion of the district, there are found in a few localities small masses of the iron- bearing Agawa formation. This formation is petrographically the same as the Soudan formation. Init, however, there is in places adevelopment of the carbonate-bearing facies. No iron ores have been found in it. Overlying the Ogishke conglomerate, in the REVIEWS 171 western portion of the district, and the intervening iron-bearing Agawa formation where present in the eastern portion of the district, there occurs a thick series of slates of varying character, to which the name “ Knife Lake slates” has been given. These slates have been very closely folded, and have been more metamorphosed where intruded by granites of Giant’s Range, Snowbank Lake, and Cacaquabic Lake, and by the Duluth gabbro. These igneous rocks occupy a considerable area, and their intrusive relation to the Lower Huronian are unquestionable. The Lower Huronian sediments now stand nearly vertical. The Upper Huronian or Animikie series is found in the extreme eastern portion of the district, where it is continuous with the Animikie of the Mesabi district to the west and Thunder Bay to the east. At the bottom of the series occurs an iron-bearing formation known as the Gunflint formation. Above this occurs a great slate-gray- wacke formation, to which the name “Rove slate” has been given. The Gunflint formation is correlated with the Biwabik formation of the Mesabi district. It has a very limited development in the Vermilion district, and its most interesting phases are especially well developed in the vicinity of Akeley Lake. In general the formation has a monociinal dip to the south-southeast at a low angle. It has been extremely metamorphosed by the Duluth gabbro, and where most metamorphosed the rocks are composed of coarsely crystalline bands of quartz, of varying width, alternating with coarsely crystalline bands of magnetite ore reported to vary from one inch up to ten or twelve feet in thickness, and of bands of dark-green, brown, or black rocks that consist of combinations of quartz, augite, hypersthene, hornblende, olivine, and mag- netite as the principal minerals, but associated occasionally with some ferruginous carbonate, actinolite and griinerite. The Duluth gabbro and the Logan sills, referred to the Keweenawan, occur in the eastern portion of the district. The gabbro is found to metamorphose all of the sediments already enumerated, and is thus shown to be one of the youngest rocks of the district. It is also found to be intrusive in the Keweenawan volcanics. A number of facts are enumerated to show that the gabbro and the Logan sills are of essentially the same petrographic character, although they exhibit minor differences that are readily explicable when one considers the relative amounts of the two rocks. After a consideration of these facts, and of the stratigraphic relationship of the rocks, the conclusion is reached that the gabbro and the sills are of essentially the same compo- sition and age, having been derived from the same parent: mass of magma. In certain localities in the Duluth gabbro there are found masses of titaniferous magnetite of varying but small size with some associated minerals. These masses grade into the surrounding gabbro, and were formed as the result of processes of segregation. Cutting the Duluth gabbro are acid dikes and dikes of basalt and diorite. The entire district has been much folded and metamorphosed, resulting in a marked north of east and south of west trend of the Archean and Lower Huronian formations, marked principally by schistosity. Comment on the Vermilion and Mesabi Reports.— Detailed work in these districts has developed a number of points bearing on the general stratigraphy and correlation of the rocks of the Lake Superior and Lake Huron districts. In the Vermilion district the rocks now called Lower Huronian had previously been referred to the Upper Huronian by the U. S. Geological Survey, and the sedi- mentary Soudan iron formation, now mapped as Archean, had previously been called Lower Huronian, and separated from the greenstones and granites supposed alone to 172 REVIEWS represent the Archean. The Lower Huronian and Archzean of the present report correspond approximately with the Upper and Lower Keewatin of the Minnesota Sur- vey, although there are minor differences in the reference of the several geological units to these divisions. In the Mesabi district the rocks underlying the Animikie of the Upper Huronian had previously been lumped together as Archzean by the U. S. Geological Survey. They are now shown to be divisible into (1) a sedimentary formation, referred, with its associated intrusives, to the lower Huronian, showing remarkable similarity in lith- ology and structure to the Lower Huronian of the Vermilion district; and (2) an igneous series, referred to the Archzean, with marked similarity to the igneous rocks of the Archean of the Vermilion district. The Lower Huronian and Archean thus cor- respond roughly to the Upper and Lower Keewatin of the Minnesota Survey. This division of the Keewatin was not made in the Mesabi district by the Minnesota Geo- logical Survey, although Dr. Grant noted the occurrence of rocks characteristic of the two divisions, and suggestéd the possibility of their separation. The correlation of the Animikie series of the Vermilion and Mesabi districts with the Upper Huronian series of the north shore of Lake Huron is the same as in previous reports of the U. S. Geological Survey, and is the feature of the correlation which has been severely criticised by Canadian and other geologists, including Coleman, Willmott, Winchell, and Lawson, who hold the Animikie to be unconformably above the origi- nal Huronian series of the north shore of Lake Huron, from which the term “ Huron- ian” comes. Comments on their arguments are made in connection with summaries of their articles on a preceding page. The reference of the sedimentary Soudan iron formation to the Archzan, instead of including it in the Huronian and thus making a threefold division of the Huronian, as is now possible in the Marquette district, has also been criticised. The defense of such a use of the term ‘‘ Archzean ” involves a discussion of the principles of pre-Cam- brian nomenclature not here warranted. Such a discussion will be made in a final monograph on Lake Superior geology now in preparation by the U. S. Geological Survey. N. H. WINCHELL. Some Results of the Late Minnesota Geological Survey. American Geologist, Vol. XXXII (1903), pp. 246-53. Winchell summarizes some results of the work of the late Minnesota Geological Survey. Those referring to the pre-Cambrian are as follows (the numbers are Profes- sor Winchell’s) : 5. The discrimination of two iron-bearing formations in northern Minnesota, thus separating the Mesabi range stratigraphically from the Vermilion. ‘This observation was continued into Wisconsin and Michigan by a visit to those states, and the same dualty was pointed out in the iron regions of those states, and was announced for the first time in the Minnesota report for 1888. It has since been discovered that there is still a third iron horizon in northeastern Minnesota, not mentioning the titanic iron ore of the gabbro. It is the upper Keewatin, the others being in the Lower Keewatin and the Taconic. 6. The separation of the Archzan of Minnesota into two non-conformable parts, viz., the Upper and Lower Keewatin, with a great basal conglomerate between them. 7. The determination of the oldest known rock of the Lake Superior region, a greenstone called Kawishiwin, the bottom rock of the Keewatin, the supposed earliest crust of the globe. REVIEWS WAS 8. The great quartzite formation, which cuts quite a figure in the geology ot Wis- consin and Minnesota, is nonconformable upon the Animikie, and is a member of the fragmented beds of the Keweenawan. This has been named “Sioux quartzite,” “Bar- aboo quartzite,” and “New Ulmquartzite.” It is that which contains the red pipestone (catlinite) in southwestern Minnesota. It is the western representative of the Pots- dam sandstone, of Potsdam, N. Y. This quartzite seems to be the representative of the Middle Cambrian, as the Beckmantown is of the upper Cambrian. g. The origin of the Mesabi iron ore is referred to a greensand, which has been altered, affording iron ore by concentration of the iron in certain favorable positions. Cotemporary with this alteration was a concentration of silica, and this was increased by oceanic precipitation. The original greensand was found to become pebbles, and to increase into angular masses that were neither sand nor pebbles, but rather breccia. These breccia masses have at first an amorphous crystalline texture and grade into a form of the iron-bearing rock which was named “ taconyte,”’ and the whole was referred to volcanic action, being different forms of suddenly cooled volcanic glass and rhyo- lite, broken and distributed by beach action. While this volcanic débris was under- going this transformation, great quantities of silicia were set free from the glass; but this silicia immediatly saturated the débris, producing spotted jasperoid, taconyte, and sedimentary jaspilyte. Having reached this result on the Mesabi Range, it opened the door to the under- standing of the iron ores of the Vermilion Range, and at once the rhyolitic forms and all the igneous associations of those ares with basic igneous rocks were elucidated, thus confirming Wadsworth’s idea of the igneous origin of the jaspilytes of the Marquette region — rather the igneous origin of the rock which later was changed into jaspilyte- Io and 11. After prolonged field examination, the Minnesota Survey reached the conclusion that the granites of the Archean grade into gneiss, the gneiss into mica- ceous gneiss and mica schist, and finally into less and less metamorphic rocks that show a plain fragmental structure and sedimentary origin. There was found no exception among the Archean granites. The granites are of two dates of formation—one at the close of the Lower Keewatin, and one at the close, or after the close, of the Upper Keewatin. A later granite, associated with the gabbro, and grading into it, is of the Keweenawan, and another did not spring from a deep source, but is a surface prod- uct of metamorphism carried to the extreme of fusion, on clastic materials that were later than the basal.greenstones. Adventitiously they form intrusions in some of the later (and especially into the clastic) greenstones, but they are not known to penetrate the oldest greenstones. Tentatively the alkaline and the acid siliceous elements in these early sediments were supposed to have been derived from the atmosphere, as the basal crust could not have afforded them. In the same manner the gabbro, which becomes acid and grades into syenite, was derivdd from the metamorphism and fusion of the greenstone with their clastic varia- tions. Diabase was found to pass insensibly into gabbro; but, on the other hand, it is also certain that it was the original form of all igneous greenstones, and that it must have had, and still has, a deep-seated source. These belts of intensest metamorphism, whether productive of granite or of gabbro, have a parallelism with each other, and with the northwestern rim of the great syn- clinorium of the basin of Lake Superior, marking successive continental folds, in har- mony with a system which continued through Archean and Taconic time, and even into the Upper Cambrian. 174 REVIEWS Comment.— Conclusions 5 and 6 are essentially in accord with conclusions reached by United States geologists who have worked in this area, although differing in nomen- clature and minor points. The same may be said of Conclusion 7 in the matter of a greenstone being the oldest rock in the state, although Professor Winchell is alone in calling it the earliest crust of the globe. From Conclusions 8, 9, 10, and 11 the United States geologists dissent 2 ¢ofo. Adequate discussion of these conclusions would involve covering the entire range of Minnesota geology. ‘The reader is referred to Monographs XLITI, and XLV,? and to pp. 305-4343 of Part III of the 7wezzty- first Annual Report of the U.S. Geological Survey for such a discussion. N. H. WINCHELL. ‘Sketch of the Iron Ores of Minnesota,” American Geologist, Vol. XXIX, pp. 154-62. Winchell briefly describes the iron ores of Minnesota, and incidentally sketches their geological relations. No new points are added to those previously presented. ROBERT BELL. ‘Report on the Geology of the Basin of Nottaway River.” Annual Report of the Geological Survey of Canada for rgoo, Vol. XIII, Part K, Igo02. Bell describes and maps the geology of the basin of the Nottaway River. Granites and gneisses referred to the Laurentian occupy the larger portion of the area. They are for the most part intrusive into the crystalline schists referred to the Huronian. Huronian rocks occur principally in a large area that is near the center of the region, and in small areas north of the center of the region and south of Lake Mistassini in the eastern part. The large tract of Huronian rocks forms a part of the great belt of Huronian rocks extending continuously from the eastern side of Lake Superior to Lake Mistassini, a distance of seven hundred miles. The Huronian may be grouped in three classes, namely: (1) crystalline schists, together with some other rocks forming a comparatively small proportion of the same series ; (2) massive green- stones; and (3) granites. The schists embrace a considerable variety, but the greater part of them are dark green and hornblendic or dioritic, and they often pass into more or less massive greenstones, so that it becomes difficult to map the two varieties sepa- rately. Dolomite, quartzite, arkose, conglomerate, and agglomerate are exceptional occurrences. J. BuRR TYRRELL AND D. B. Dow tinG. ‘Reports on the Northeastern Portion of the District of Saskatchewan and Adjacent parts of the Dis- tricts of Athabasca and Keewatin,” Annual Report of the Geological Survey of Canada for rgoo, Vol. XIII, Parts F and FF, 1902. With map. Tyrrell and Dowling report on the northeastern portion of the district of Sas- katchewan, and adjacent parts of the districts of Athabasca and Keewatin, comprising an area adjacentto the north end of Lake Winnipeg. The east, northeast, and northern *C. K. Leiru, ‘“‘The Mesabi Iron-Bearing District of Minnesota,” Monograph XLITTI, U.S. Geological Survey, 1903. 2J. MoRGAN CLEMENTS, “The Vermilion Iron-Bearing District of Minnesota,” Monograph XLV, U.S. Geological Survey, 1903. 3C. R. VAN HisE, The Iron Ore Deposits of the Lake Superior Region, 7zwenty- first Annual Report of the U. S. Geological Survey, Part II], pp. 305-434. REVIEWS 175 portions of the area mapped are occupied by Laurentian and Huronian rocks, of which the Laurentian rocks are in the larger areas. They consist of granites and gneisses, some of which are intrusive into the Huronian, and some of which are probably basal to it. Huronian rocks are found in small areas at Cross Lake, at Pipe Lake, and in the large area extending from Wekusko. Lake to Athapapuskow Lake. They consist of conglomerates, quartzites, basic eruptives, and greenstones, and altered schists, similar to rocks of Lawson’s Keewatin and Couchiching series. A. P. Low. “ Report onthe Geology and Physical Character of the Nasta- poka Islands, Hudson Bay,” Annual Report of the Geological Survey of Canada for rgoo, Vol. XIII, Part DD, 1903. Low describes the geology of the Nastapoka Islands, Hudson Bay. ‘The rocks forming the islands are in descending order as follows: Feet 1. Rusty-weathering, dark gray siliceous rock containing ankerite (carbonate of iron and mag- nesia, and magnetite - - - - 65s - 6 he - 5 2 E = - 20-100 2. Dark gray siliceous rock containing magnetite, with small quantities of ankerite - - 50-250 BaeRedtaspilytesrichuinghematiterorewi.4) i pee =) a aw een Sit Wee oan Eero =Too 4. Red jaspilyte poorinhematiteore - = : : - : - - 2 O 4 - 5-20 5. Purple, or greenish-weathering, dark green, graywacke shales - - - - - - - 10-70 6. Red jaspilyte poor in hematite ore - 2 - < 2 é 6 5 2 5 s 5 o-s5 7. Light greenish-gray sandstone and shale - 2 = = = : : oe : - 10-300 8. Fine-grained dolomite - - - - - - ° - = é = a 2 2 - 0-50 There isa general dip toward the westward, or toward the sea, of from 5° to 15°, There are north-and-south faults, the upthrow being almost on the west side, with the result that the rocks appear in north-and-south ridges. The displacement is small and rarely exceeds one hundred feet. Another system of faults lies transverse to the first system. Large areas of similar unaltered sedimentary rocks occur throughout the peninsula of Labrador, and are probably the equivalents of certain of the iron-bearing series about Lake Superior and of those to the westward of Hudson Bay, hand specimens from these localities being undistinguishable. On former maps of portions of the peninsula of Labrador, the areas of rocks belonging to this formation have been colored as belonging to the Cambrian formation, and in the earlier .reports on this region the rocks were thought to be a part of that system, owing to their unaltered condition, in contrast with all the other rocks of that vast area that were either crystalline granites and other irrupted rocks, or crystalline schists and gneisses, so completely metamorphosed as to have lost all trace of their original sedimentary nature, if any were sediments. These highly crystalline rocks were classed as Laurentian or Huronian, and were considered to be much older than the unaltered rocks of the so-called Cambrian areas. More extended and closer study of both the unaltered and crystalline rocks, and of their relations to one another, has changed the views of the writer; and he now considers the unaltered, so-called Cam- brian rocks to be the equivalents of many of the gneisses and schists classed as Lau- rentian (Grenville series), and the Huronian areas of the Labrador peninsula to represent a portion of the unaltered rocks and their associated basic eruptives (traps, trap-ash, etc.), altered by the irruption of granite and rendered schistose by pressure The granites which have been classed as typical Laurentian, always cut and alter the bedded rocks wherever seen in direct contact with them, and are consequently newer than the latter. 176 REVIEWS During the past season very thin layers of carbon with some resemblance to organic forms were found in the sandstones of Cotter Island; these have the appearance of lowly organized plant life, lower than the known fossils from the lowest beds of the Cambrian; and consequently this formation is older thanthe Cambrian. It is proposed, therefore to class these so-called Cambrian unaltered rocks as Laurentian, as they represent the oldest known sedimentary rocks in the northeast of America, and probably in the world. West Virginia Geological Survey. Vol. 1, Oil and Gas; Vol. II, Coal; and map showing the occurrence of coal, oil, and gas in West Virginia, By I. C. WuirTE, State Geologist. Proressor I. C. WHITE, state geologist of West Virginia, has just issued a map showing the distribution of coal, oil, and gas areas in that state. The base of the map is topographic, with contours of 1,000 feet, and is, all in all, the most accurate map of the state which has ever been published. The map shows both the coal areas and the coal mines of the state. Of the former, the Pittsburg, the Allegheny- Kanawha, and the New River-Pocahontas are differentiated. In the aggregate, the coal areas cover nearly one-half of the state. The areas of natural gas and oil, though more restricted, are still extensive. The map, just published, is a welcome supplement to the excel- lent volumes on Oil and Gas (Vol. I, issued in 1899), and Coal (Vol. II, issued in 1903). No state geological survey has issued economic reports of greater worth. While in the case of both volumes the treatment is primarily economic, the general structural relations of the Mississippian, Pennsylvanian, and Permian series, as developed in West Virginia, are clearly set forth. Re DS: Geographic Influences in American Fiistory. By ALBERT PERRY Brigham. The Chautauqua Press, 1903. Pp. x-- 366; 61 illustrations. American History and its Geographic Conditions. By ELLEN CuuRCHILL SEMPLE. Boston: Houghton, Mifflin & Co., 1903. Pp.466;, 16 maps: Tue above books are pioneers in a most interesting and important field, too long neglected. American history has been profoundly influ- enced by geological and geographical conditions. ‘To ignore these controls is to make history very largely empirical. To recognize them is to go a long way toward making history a rational science. To REVIEWS 177 appreciate, for example, the geographic conditions which controlled every move ofthe contending armies of the Civil War in Virginia is to make intelligible a chapter of American history from which otherwise one gets but a confused, meaningless impression. The scope of Professor Brigham’s book is indicated by the chapter headings: “‘The Eastern Gateway of the United States ;” “‘Shore-Line and Hilltop in New England;” “The Appalachian Barrier;” “The Great Lakes and American Commerce;” ‘“‘The Prairie Country;’’ ZCowonmricesand» Canes.) lhe, Civils War) no Wheres inittle: Rain Falls;”’ ‘Mountain, Mine, and Forest.” As these topics suggest, the author’s view-point is always that of the geographer. The treatment of the subject is simple and somewhat popular, the book being designed primarily for a non-professional class of readers. Miss Semple’s book is a distinct contribution. The geographic influences which have shaped the trend of American history, from the discovery of the continent to the present, are treated in a scholarly and judicious manner. The inclusive character of the book is shown by the titles of the chapters: ‘“‘The Atlantic States of Europe the Discov- erers and Colonizers of America;’’ ‘‘The Rivers of North America in Early Exploration and Settlement;” ‘The Influence of the Appa- lachian Barrier upon Colonial History ;” “The Westward Movement in Relation to the Physiographic Features of the Appalachian System ;”’ “Geographical Environment of the Early Trans-Allegheny Settle- ments;” “The Louisiana Purchase in the Light of Geographic Con- ditions;”’ “Geography of the Atlantic Coast in Relation to the Development of American Sea Power ;” ‘“ Geography of Sea and Land Operations in the War of 1812;” “Spread of Population in the Mis- sissippi Valley as Affected by Geographic Conditions ;” “Geographic Control of Expansion into the Far West: the Southern Routes;” “Expansion into the Far West by the Northern Trails;” ‘‘Growth of the United States to a Continental Power Geographically Determined ;”’ “The Geography of the Inland Waterways;” ‘The Geography of the Civil War;” ‘Geographical Distribution of Immigration;” “Geo- graphical Distribution of Cities and Industries ;” ‘‘Geographical Dis- tribution of Railroads;” ‘‘The United States in Relation to the American Mediterranean ;”’ “The United States as a Pacific Ocean Power.” Miss Semple has been skilful in the selection of material from the great mass of scattered data. Irrelevant matters are invariably excluded, and the conclusions reached are generally fundamental. New light is 178 REVIEWS thrown on many topics. English success and French failure in North America are shown to have been largely due to geographic conditions. At the north the French followed every stream into the interior in quest of furs. ‘They spread themselves thin over an enormous area,” and therefore failed. The Appalachian Barrier confined the English to the coast, and the many resulting advantages contributed to their success. It is popularly supposed that our possession of the Louisiana territory is due to a series of fortunate circumstances in European poli- tics. Miss Semple shows that, once having passed the Appalachians, geographic conditions made it inevitable that the Americans should control the interior at least as far west as the Rockies. The purchase of Louisiana was the occasion, not the cause, of the acquisi- tion of the trans-Mississippi country. That must have come sooner or later. Even if the French had established themselves in Louisiana, they could not long have resisted the operation of geographic factors and the enterprising spirit of the western people, itself in part a product of environment. The trans- Mississippi region, hopelessly arid beyond the one-hundredth meridian, could never have supported a large enough population to resist the Ameri- cans, with whom the common navigation of the Mississippi would soon have brought them to blows..... Had England conquered Louisiana from France—the chance which Napoleon feared —even her superior colonizing methods could not have made the country support a population large enough to cope with the thickly planted American settlements in the wide, rich, well- watered regions to the east. In a conflict between a cis-Mississippi and a trans- Mississippi power, the former had every geographical condition in its favor —coast-line, rivers, climate, soil, and habitable area. The Americans were destined to hold the West. The purchase hastened and facilitated the process. The excellent discussion of the War of 1812 throws much light on a period which, to be understood, must be approached from the geo- graphic side. Geographic conditions made this a frontier war and controlled all operations. The author does not overestimate the importance of the geographic view-point when she says: The sea-fights of this war, if studied merely in their chronological sequence as presented in the ordinary school histories, leave only a confused impres- sion, of which the student, young or old, retains little at all and less that is valuable. But an analysis of the geographical distribution of these engage- ments reveals a wide underlying system which explains their purpose and brings order out of an apparent chaos. The Gadsden Purchase has been almost universally condemned as a purchase involving the payment of an enormous price for a small REVIEWS 179 tract of worthless land. Miss Semple maintains that because of the great strategic importance of the Gila River depression as a passway to the coast, money was never better spent. Miss Semple believes the most potent factor in American expansion to have been the abundance of free land. The exhaustion of the sup- ply has led to a recent exodus of westerners into Canada, over 50,000 going in the three years following 1899. It is pointed out that we must look to the recently initiated national system of irrigation in the arid West for the checking of this migration. The arrangement of the matter in the book is not always the best, and a very few important topics are slighted. For instance, the dis- covery of gold in California does not receive due emphasis as a factor in American expansion. Such shortcomings are few, however, and the book is to be heartily commended to all students of geography and history. lalelalesieh RECENT PUBLICATIONS. —BONNEY, T. G., AND PARKINSON, J. Detrification in Glassy Igneous Rocks. [Quarterly Journal of the Geological Society, Vol. LIX, 1903, pp. 429-43.] —Butts, CHARLES. Fossil Faunas of the Olean Quadrangle. [Reprinted from Report of State Paleontologist, 1902; University of State of New York, Albany, 1903.] —CLARKE, JOHN M. Construction of the Olean Rock Section. [Reprinted from Report of State Paleontologist, 1902; University of State of New York, Albany, 1903. | —COLE, LEON J. The Delta of the St. Clair River. [Geological Survey of Michigan, Lansing, 1903.] —Committee on Underground Temperature, Reports of the. Fifteenth, Sixteenth, Seventeenth, Eighteenth, Nineteenth, and Twenty-second Reports. [Neues Jahrbuch fiir Mineralogie, etc. ] -—CUMINGS, EDGAR RoscoE. Development of Some Paleozoic Bryozoa. [American Journal of Science, Vol. XVII, January, 1904. | —Day, ARTHUR L. AND VAN ORSTRAND, C. E. The Black Body and the Measurement of Extreme Temperatures. [Reprinted from Astrophysical Journal, Vol. XIX, No.1; Chicago, January, 1904.]| —FIsHEeR, O. On Deflexions of the Plumb-Line in India. [From Philo- sophical Magazine, January, 1904. | —GIRARDIN, M. PAUL. Rapport sur les observations glaciaires en Haute- Maurienne dans les Grandes-Rousses et 1l’Oisans dans 1’été de Igo02. [Commission Frangaise des Glaciers, Paris, 1903. | —GLENN, L.C. Devonic and Carbonic Formations of Southwestern New York. [Reprinted from Report of State Paleontologist, 1g02; Univer- sity of State of New York, Albany, 1903.]| —HUNTINGTON, ELLSWORTH, AND GOLDTHWAITE, JAMES WALTER. The Hurricane Fault in the Toqueville District of Utah. [Bulletin of the Museum of Comparative Zodlogy, Harvard College, Vol. XLII; Cam- bridge, Mass., February, 1g04.] —JEFFERSON, MARK, S. W. Wind Effects. [Reprinted from Journal of Geography, Vol. III, No. 1; Chicago, January, 1904. | —-JULIEN, ALEXIS A. Genesis of the Amphibole Schists and Serpentines of Manhattan Island, New York. | Bulletin of the Geological Society of American, Vol. XIV, pp. 421-94; Rochester, December, 1903.] 180 (OOM OF CEOLOGY APRIL-MAY, 1904 ICE-RETREAT IN GLACIAL LAKE NEPONSET AND IN SOUTHEASTERN MASSACHUSETTS. INTRODUCTION. LAKE NEPONSET is the name applied to the body of water which, during the final retreat of the Wisconsin ice-sheet, occupied the upper portions of the valley of the present Neponset River a few miles south of Boston. The existence of this lake was first pointed out by Professor W. O. Crosby, who regarded it as one of a series of more or less open lakes, the waters of which had gathered between the general ice-margin and the higher lands bordering the north- ward-sloping valleys in the region to the southwest, south and south- east of Boston during the period of retreat. ‘The more important of the water bodies beginning at the west are designated by Pro- fessor Crosby as Lakes Sudbury, Charles, Neponset, and Bouvé. The deposits and history of Lake Bouvé have been discussed in detail by Dr. A. W. Grabau, while the Sudbury, Charles, and Neponset lakes have been defined and discussed in a more general way by Professor Crosby and Mr. F. G. Clapp. In the writings of Crosby, Grabau, and Clapp,’ the view, though not definitely stated, seems to have been tacitly accepted that, although there were doubtless many minor irregularities of the ice-front, the margin as a whole preserved a rather definite and regular terminal tFor Mr. CLApp’s present views see paper on “Relations of Gravel Deposits in the Northern Part of Glacial Lake Charles, Massachusetts,” pp. 198-215 of the present number of the JOURNAL OF GEOLOGY. Vol. XII, No. 3 181 182 M. L. FULLER facing, such as characterizes living glaciers. It was apparently con- ceived by the writers named that, as the face drew back to the north, a line of glacial lakelets came into existence at the heads of the northward-leading valleys. ‘These, as the ice retreated, were con- sidered to have grown in size and to have coalesced until the Sud- bury, Charles, and Neponset lakes, and possibly also Lake Bouvé, became united into a single lake many miles in width and length, and of considerable area. The studies of the present writer in Lake Neponset have led to the conclusion that the ice in that region, instead of retreating with a definite and somewhat regular front, had become absolutely stag- nant before the history of the lake began, and that its disappear- ance was characterized by marked irregularities along lobes, deep re-entrants and detached blocks being the rule rather than the excep- tion. Moreover, the marginal distribution of the deposits makes it seem probable, if not certain, that there was no general body of water such as was postulated for the Sudbury-Charles-Neponset stage, or even for the simple Neponset stage itself, but that the depos- its, generally considered as marking the lake level or levels, were laid down in a series of small and more or less independent lakelets existing along the margins of the residual valley lobes or about entirely detached masses of ice. In urging the improbability of the existence of large lakes with definite levels in this region during the earlier stages, however, the writer does not wish to be considered as denying the existence of considerable bodies of water in the lower portions of the valleys during the closing stages of the lakes when the ice-lobes and blocks had practically disappeared. In the following discussion of Lake Neponset the terms “lake” and ‘‘bay” are used to designate those portions of the basin of the Neponset River and its tributaries in which glacial sediments were laid down in standing water irrespec- tive of time, elevation, or of the character of the water bodies in which the deposition took place. Stoughton Bay is simply a portion of the Neponset basin, lying in the vicinity of the town of the same name. In this bay the con- ditions which the writer believes to have characterized the ice-retreat in the region under discussion are recorded very definitely in the ICE-RETREAT IN GLACIAL LAKE NEPONSET 183 distribution and topographic expression of the deposits; and, as the history of the retreat in this locality is believed to be essentially the same as in the other glacial lakes of the region, the bay has, for the purpose of discussion, been selected as a type. STOUGHTON BAY AREA AND ITS DEPOSITS. Topography and drainage.—In a broad way it may be said that Stoughton Bay occupied a somewhat oval area surrounded by an interrupted belt of hills which, starting northwest of North Stoughton, extends southward about four miles, and then curves first to the west and then to the northwest, finally terminating in the range of hills southwest of Canton. ‘The breadth of this basin is about four miles, and the length a trifle greater. The crests of the hills con- stituting the boundary vary from 200 up to 420 feet in altitude, while the intermediate cols vary from 190 to 250 feet in elevation. Notwithstanding these relatively low gaps in the southern rim of the basin, it seems probable that the preglacial drainage, like that of the present period, was by way of the Neponset River to the north. Both the lower level of the rock-floors of the channels and the greater width of the valleys at the northern end of the area bear out this supposition. The possibility of a deflection of the drainage to the southwest through the valley now occupied by Massapoag Pond, has been considered, but although the valley is relatively broad, the frequent projection of rock ‘‘islands”’ through the out- wash drift deposits, with which it is filled, appears to indicate that the rock bottom is much higher than in the present valley at Canton Village. The Canton valley, however, is not a broad one, the rock outcropping at relatively short distances from the stream on both sides, and probably underlying it at no great depth. The high elevation of the bottom of this channel, as compared with the low elevation, which probably characterizes the near-by Neponset valley bottom, would appear to indicate, either that the former is not the main preglacial drainage channel, or that the Neponset valley has been materially deepened by glacial erosion. The latter appears to be the more probable. In fact, the drainage and topography of even the bed-rock areas of this portion of Massachusetts show very few of the characteristics of normally developed drainage systems. 184 M. L. FULLER Relatively little is known as to the minor details of the rock topog- raphy, within the basin, as the floor is deeply covered, except perhaps in the center, with extensive planes of stratified drift, which rises in places to nearly 150 feet above the marshy tracks along the main streams, which in turn are of unknown, though probably not very great, elevation above the rock-floors beneath them. The wells in the region are generally shallow, and in most instances afford little information”of value. The details of the plains and of the cols in the rock-rim, many of which served as channels for the overflowing waters, can be considered to the best advantage in the discussion of the various stages in the history of the bed. STAGES OF DEPOSITION. The history of the Stoughton Bay area of Lake Neponset may be said to have begun when the first body of water came into existence between the ice and the retaining walls of the basin. Such a lake- let must of necessity have been of small size at the beginning, but soon became enlarged through the melting back of the ice. The outlets may have been over low points in the rock-rim, or along the edges of the ice. As the melting progressed, the expansion of the lakelets continued until the lower outlets of the waters were uncov- ered. The different levels of the waters are indicated approximately by the altitude of the stratified deposits, which were laid down during the different stages. In Stoughton Bay these deposits are of two general levels, the higher standing at 250 feet, and the lower at about 190 or 200 feet above the level of the sea. The higher stage is named from the town of Stoughton, the greater part of which is located on the high-level deposits, while the lower stage is named from the village of Springdale, near which the lower planes are strongly devel- oped. The distribution of the materials show that not all the deposits, even of a single stage, were laid down in the same body of water, but accumulated, on the contrary, in more or less separate lakelets. The principal water bodies of the higher stage were the Rattlesnake Hill, East Sharon, and Stoughton lakelets. while the leading bodies of the lower stage were the Flm and Springdale lakelets. Soe iii 4 ki ; f ‘ | : i rh 0 ET fi Toe Aseer7, G.I A (; : x ) “NAS : % No} 5 hc ~ My a . "SN (/ I} i i < ntsseret 1.—Retreatal Stages in Stoughton Bay of Glacial Lake Neponset. j Rock and till, generally higher than “high-level” plains. _ High-level plains and outwash of East Sharon and Stoughton lakelets. Gradation deposits between high- and low-level plains. — Low-level plains and outwash of Elm Street and West Stoughton lakelets Gradation deposits below low-level plains. Ice-masses at close of low-level stage of lakelets. Position of ice-margin at opening of high-level stage. Position of ice-margin at close of high-level stage. RUZ oe uM Ue| Position of ice-margin at close of low-level stage. 186 M. L. FULLER EARLIER OR STOUGHTON STAGES. Raiilesnake Hill lakelet—This is the smallest of the recognized lakelets of the Stoughton Bay area, being less than half a mile in length and only about one-eighth of a mile in width. It was formed in a re-entrant just north of the col between Rattlesnake Hill, rising to 420 feet on the northwest, and another granitic hill, about 360 feet in height, on the southeast. The rock-floor of the col is not now exposed, but as there was free drainage to the south, the deposits though constituting a somewhat broad and flat sheet, are probably of slight thickness at the crest. The surface of these deposits stands at 250 A. T. The deposits of the locality may be divided into three classes: (z) the outwash deposits in the channel leading southward from the Fic. 2.—Diagrammaiic north-south section through Rattlesnake Hill divide, show- ing outlet and broken delta deposits. (4, B, etc., show successive stages of ice-front during its recession.) col; (2) the flat top deposits at and just north of the crest; and (3) the irregular and broken delta deposits on the north (Fig. 2). The outwash gravels, which constitute a gentle sloping deposit extending down the valley to the southward, were evidently formed not later than the period when the ice-margin rested at the point represented by A, as the unfilled kettles and other depressions between the imperfect deltas to the north indicate that in the later stages little or no material was being carried into the lakelet. The deposits at and just north of the crest consist of sands and fine gravels, and evidently represent the perfected lake deposits. They present an almost perfectly flat surface, one-eighth of a mile in width, and perhaps twice as long, which is so ill drained that in the wetter seasons the water stands over several acres, though of a depth of only a few inches. It is now the site of a cranberry bog. From this flat the rock hills rise with a sharp line of demarkation abruptly on each side. ee ICE-RETREAT IN GLACIAL LAKE NEPONSET 187 At the time the deposits were laid down the ice probably presented a fairly definite face to the lake, the position being not far from the crest of the pass possibly near A. Later, however, the margin became broken up into detached blocks at its edge, the location of these blocks being represented by the imperfect planes and kettles lying just north of the crest. Continuing to the north, the deposits of stratified material end rather abruptly, the lower portions of the northward-leading valley, through which the ice drew back, being practically free from them. ‘This would seem to indicate, either that the glacial streams had been diverted at some point farther north, or that they no-longer carried any material quantities of sediments. Otherwise the deposits would have continued to accumulate in the ~ lakelet which, in the later stages, must have reached some distance to the north of the deposits previously noted. The waters passing out from the lake to the south, being no longer overloaded with materials carried in suspension, began the work of eroding out the deposits laid down in the earlier stages, with the result that a thick- ness of ten feet of sand and gravel was removed throughout nearly the entire width of the valley leading south from the pass, the original level being represented only by an occasional terrace remnant stand- ing about ten feet above the present valley floor. Imbedded in the silt, which here constitutes the floor, is a granitic bowlder, nearly fifteen feet in diameter, probably stranded on the rock or till sur- face underlying the silts during the melting of the ice and surrounded by subsequent deposits of sand. East Sharon lakelet—This name is applied to the body of water which lay between the granitic hills, one and one-half miles south of East Sharon, and the ice-front after the latter had shrunk back from the valley sides. The lakelet had a total length of approximately two and one-half miles. The greatest extent of open water was south of East Sharon (southwest of West Stoughton), where the lake meas- ured a mile or more across. A mile southwest of West Stoughton the lake became contracted into what must have been simply a rather broad and sluggish lateral stream,which, however, opened up again to the southward into a marginal body, one-quarter of a mile or more in width, which continued to the north base of Rattlesnake Hill. In this compound lakelet the deposits of the East Sharon stage 188 M. L. FULLER were laid down by the superglacial streams from the surrounding ice. The sand and gravel plains are most perfectly developed along the hillsides, as the lakelet was there shallowest and soonest free from ice. In such situations the plains were frequently built up to a hori- zontal upper surface, coinciding approximately with the water level, and are generally free from kettles. As the ice receded, lower por- tions of the hillsides were uncovered, and the water became deeper. The ice, how- ever, melted back most Fic. 4. rapidly along the sur- Fics. 3 AND 4.—Showing mode of formation’ of face of the lakelet, gradational deposits: By marginal deposits before Teces- leaving p r0j ecting sion of ice; 4, plain and gradation deposits after recession edges beneath the water, which became covered with sands and gravels. On the further melting of the ice and its disappearance from beneath these gravels, the materials were let down into irregular accumulations along the sloping valley sides, constituting the gradational deposits between the upper and lower plains, and between the lower plains and the present valley floor in the Stoughton Bay area. When the ice-wedge was very thin, gentle and fairly regular slopes resulted when the materials were let down, but when thicker, steeper and kettle-pitted slopes resulted. Dry Pond, in the southern part of East Sharon, is an example of such a kettle. This is the explanation of the change from a flat to a gently sloping plain, and finally to the irregular hummocky slopes which characterize the plains at many points in this area. ‘The steeper ice-contact slopes, which are especially well developed in portions of the east side of the East Sharon plain, were formed where the marginal ice-wedge was only slightly developed. ‘That the slopes of the plains cannot be regarded as purely depositional is shown by the fact that the inclina- tion of the surface is opposite that exhibited by delta plains. There can be no question that the materials of the plains came from the ice. This is attested by the rounding of the pebbles, the of ice. ICE-RETREAT IN GLACIAL LAKE NEPONSET 189 lack of agreement in composition between the gravels and the adjacent rocks or tills from which they must otherwise have been derived, and the absence of even the most local postglacial deposits of similar character in the region. The lake-level was regulated by the altitude of its outlet, which was through the notch southeast of Rattlesnake Hill, as in the case of the eariler lakelet of that name. ‘The outlet stream was marginal as far south as the northeast base of the hill, but there the waters passed onto the ice, on which they continued until the notch was reached, as shown by the absence of erosional or depositional features in the intervening area. That there was no outlet over the divide south of Massapoag Pond, which lies some two miles west of Rattlesnake Hill, is shown by the fact that the rock-floor is from ten to twenty feet or more lower than at the Rattlesnake Hill outlet, and as much below the level of the East Sharon and Stoughton Plains, the highest of which it could not, therefore, have controlled. The field evidence, moreover shows that the Massapoag valley was occupied by ice until a late stage in the history of the Stoughton Bay region. Stoughton lakelet—This is by far the largest of the lakelets in the Stoughton Bay region, and receives its name from the town of Stoughton which is located in the middle of its area. Like the East Sharon lakelet, it was compound, being composed of a larger water body at the north and a smaller one at the south, the two being apparently connected by a narrow channel along the eastern margin in the vicinity of Stoughton. Like the East Sharon lake, it was also located, at least so far as the northern or main portion is concerned, at the termination of a rock hill projecting between two converging valleys in which the ice still remained. The breadth of the northern portion of the lake from north to south was about one and one-half miles, and from east to west about one and one-quarter miles. ‘The southern portion of the lake was more irregular. From east to west it measured somewhat over a mile in length, but it was less than half a mile in width. In the northern half of the lakelet the broad Stoughton plains were deposited. While in a general way these plains present a uni- form upper limit, the plain was in many places never built up to a gO M. L. FUELER perfectly smooth surface, but is characterized by broad, gentle undu- lations, merging into kettles about the margins. Some parts, however, are as flat as a floor. In the southern half of the lakelet the plains are generally quite broken, though occasional flat-topped areas exist. The margins of the plains nearest the valleys present in both areas the same gradational type as the East Sharon plains. The natural outlet of the Stoughton lakelet would at first sight appear to have been southward along the valley followed by the rail- road, or along the similar valley half a mile farther east. The alti- tude of these outlets in each instance, however, is less than 200 feet, while the flat tops of the sand plains, which probably indicate the approximate level of the water of the lake, is 250 feet, or the same as that of the Rattlesnake Hill and East Sharon deposits. It seems clear, therefore, that the southward valleys mentioned were still blocked by the ice. The elevations of the Stoughton plains, corre- sponding as they do with that of the Rattlesnake Hill outlet, suggest that it was through this notch that the water passed on to the south. To do this it must have crossed the tongue of ice still occupying the valley north of Ames Pond. An examination of the locality seems to show beyond reasonable doubt that this was in fact the case, the waters crossing just north of the road leading eastward from near Dry Pond. The waters in passing deposited much material on the ice, which, on the melting of the latter, was left as an irregular belt of sands, gravels, etc., across the valley at the point indicated, and practically connecting the Stoughton with the East Sharon plains. LATER OR SPRINGDALE STAGES. Elm Street lakelet—This lakelet came into existence after the ice-margin had melted back to a point a mile or more to the west of the position it occupied during the deposition of the Stoughton plains. The lakelet proper was about three-quarters of a mile wide from north to south, and a mile from east to west. In it was deposited the typical flat-topped plains traversed by Elm and Water Streets about a mile southwest of Stoughton. Its upper surface stands at an elevation of 210 feet, and the plain is bounded on the northeast, north, and west by sharp ice-contact slopes, while on the east and south it slopes off into rolling deposits, apparently quite distinct from ICE-RETREAT IN GLACIAL LAKE. NEPONSET IgI ordinary fore-set slopes, and probably representing deposits along the subaqueous ice-wedge, afterward let down as irregular gradational masses, as explained in the case of the East Sharon plains. The altitudes of the upper limit of the flat-top deposits indicates that the Rattlesnake Hill outlet had been abandoned, and that a newer and lower outlet to the south had been opened up through the valley occupied by Ames Pond at a level of about 210 feet. The ice had not entirely left the valley, a narrow tongue still remaining in the center, along both sides of which, as indicated by terraces with ice-contact slopes toward the valley, outflows took place. On the west side the water followed straight through the valley, but on the east side a portion passed off to the eastward by sub-outlets through notches located respectively near where the highway crosses the pond and just above the south end of the pond. West Stoughton lakelet.—The West Stoughton lakelet was a body of water one and one-half miles long and one-half a mile wide, extend- ing from Springdale on the north to a point about a mile west of Stoughton on the south. The plain which now marks its position, though fairly flat in places, is much more broken than most of the other plains described, presenting in many places rolling and kettle topography characteristic of deposition open and around more or less detached sheets or blocks of ice. Fine ice-contact slopes extend all the way around the edge from West Stoughton to Springdale, marking the margin of the ice, still occupying the center of the valley. The elevation of the surface of the West Stoughton plains is from 190 to 200 feet. It seems to have been formed at a slightly later period than the Elm Street plain. The lower altitude of its surface is probably due to the fact that the ice in the depression now occupied by Ames Pond (artificial) had finally disappeared, opening up an outlet a few feet lower than that existing at the time the Elm Street plain was formed. The surplus waters probably escaped through the main and two sub-outlets in the vicinity of Ames Pond as in the case of the Elm Street plain. Closing stages.—Within the limits of the Stoughton Bay area there are no records of Lake Neponset later than the stage marked by the deposition of the low level plains just described, unless certain unimportant irregular deposits occurring in the center of the valley 192 M. L. FULLER are regarded as marking the final passing away of the ice. The ice already reduced to a mere line along the center of the valley could have remained but a short time, and, as the later outlets appear to have been to the north, no further stages of the lake are recorded. RETREATAL CONDITIONS. Absence of ice-movement.—That ice-movement had ceased before the opening of the lake history of the Stoughton Bay area is clear from the character and distribution of the deposits laid down. Nowhere in the region are there any deposits of the nature of moraines, such as would have been formed at the front of a living ice-sheet. This applies both to the uplands and to the valleys, which are alike free from till accumulations formed either along a gen- eral ice-front or at the terminations of valley ice-tongues. The form and structure of the stratified deposits also fail to show any forward movement of the ice-bodies, along the margins of which they were laid down. Folding and faulting of a nature indicating thrusts are absent in all exposures which have been seen, and the ice-contact slopes are practically free from till accumulations, such as would have been formed if the ice had possessed sufficient motion to bring fresh materials to the front. The topographic forms are always those which would result from a receding ice-margin, never from an advancing one. The gradational deposits described on pp. 185, 186 are especially significant in this connection. The strongest evidence, however, is from the distribution of the deposits, which, as has been described, form relatively narrow and successive strips along the sides of the valleys, showing conclusively, when examined as a whole, that the ice-shrinkage was not from south to north, or any other fixed direction, but always away from the valley sides. At the latest recorded stage, as shown by the sur- rounding deposits, only a narrow strip of ice remained along the center (Fig. 1). Neither this nor the earlier and wider masses pos- sessed the straight or gently curving outlines characteristic of living ice-tongues, but were marked by all the sinuosities and irregularities of an irregularly melting stagnant ice-margin. Drainage oj the ice-sheet—The highest stratified deposits of Stoughton Bay are the Rattlesnake Hill, East Sharon, and Stoughton ICE-RETREAT IN GLACIAL LAKE NEPONSET 193 plains, which were deposited in standing water at an altitude of 250 feet. These plains have an aggregate area of at least five square miles, and a probable average thickness of 75 feet. Certain reasons have already been given (pp. 183, 184) why these deposits cannot be regarded as derived from the surrounding uplands, but the most conclusive argument is their immense bulk, considered in connec- tion with the absence of erosion features on the surrounding uplands. That the material came from the ice, therefore, there can be no doubt. The question then is: Were they deposited by subglacial or by superglacial streams ? Half a mile southeast of East Sharon the plains, which here con- sist of sands and of gravels with pebbles up to several inches in diameter, stand at an altitude of 245 feet, or over 100 feet above the valleys on the east, north, and west, across which the materials must have been borne. If the gravels were transported by sub- glacial streams, they must have been suddenly lifted a distance of 1oo feet. To move a pebble 2 inches in diameter along a level bottom requires a current with a velocity of 2.8 feet per second; to lift it upward requires a velocity of about 3 feet per second. This would correspond to an upward pressure along the assumed ice- tunnel across the valley to the north of 43 pounds to the square inch, and to indicate a head of 2 feet per mile in a straight conduit like that of a circular pipe 6 feet in diameter. In a conduit with the many irregularities of a glacial tunnel the head would probably have to be ro per cent. greater, or approximately 31% feet per mile, in order to give the observed velocity. Slightly lower figures would be given by larger conduits. Such a head could be found at a dis- tance of many miles to the northwest. In living glaciers subglacial streams may possibly exist for considerable periods of time, but it seems beyond the ground of probability that a subglacial stream, with the pressure indicated, could continue over hill and valley for long distances beneath the stagnant ice-sheet without finding a pas- sage for the more ready escape of its waters. The temperature of a subglacial stream is always, according to observations, slightly above the freezing-point. The corrosive action of a stream, with the velocity and pressure indicated, would rapidly eat the roof of its tunnel, until during a somewhat extended period, 194 M. L. FULLER such as would be required for the upbuilding of extensive plains like those of the Stoughton Bay area, a base-level was reached. In crossing a valley the roof of the tunnel would be at least as high as the median height of the valley walls on the up-stream and down- stream sides. The motion of the water would be mainly super- ficial, and the lower or slack-water portion of the tunnel in the valley or depression would become filled to grade with deposits, which would eventually be left as an immense esker. No deposits of such a character occur in connection with the plains in this region. In case of the plains of the Stoughton Bay area, there can be no doubt that the streams supplying the material were englacial or superglacial. Of the two, superglacial streams are the more prob- able. It is clear that there was a connection across the ice between the Stoughton and East Sharon plains, while the waters from both flowed a short distance over the ice in their passage through the Rattlesnake Hill outlet. If superglacial drainage existed at one point, it is likely to have existed at several, and, in the absence of all evidences of subglacial drainage, can reasonably be accepted as the predominant type. It is not assumed that subglacial drain- age was not common in other regions, nor that it did not exist to some extent in the region now under discussion; but it played little or no part in the upbuilding of the plains. Inception oj lakelets—In an earlier paragraph it has been pointed out that there is every reason to believe that the ice had become stagnant before the first of the lakelets of the Stoughton Bay area came into existence, and it is probable that all movement had ceased while the entire surface of the region was still covered by the ice. As the melting of this stagnant ice-field went on, the hilltops began to appear, and the superglacial drainage was obliged to seek the notches for outlets. Seemingly, the first pass to be uncovered was the one immediately west of Rattlesnake Hill, at an altitude of approximately 325 feet; but it appears to have been so situated that it was not available to any of the superglacial streams, for an examination of the ground shows that there was no outflow of water through this pass. Following this, the next to be laid bare as the melting pro- gressed must have been the notch immediately east of Rattlesnake Hill, at an altitude of 250 feet. This is the outlet which is known as the Rattlesnake Hill outlet. ICE-RETREAT IN GLACIAL LAKE NEPONSET 195 Lakelets oj the earlier stages—The position of the ice-front at the beginning of the flow through the outlet was very near the crest of the pass, as indicated by the distribution of the plains. Later, how- ever, the ice drew back slightly, while farther north it receded at the same time from the valley walls sufficiently to bring into existence the East Sharon lakelet on the west side of the valley and the Stoughton lakelet on the east. Ice-masses still occupied the Massapoag and Ames Pond valleys, and the two southward-leading valleys southeast of Stoughton, leaving the Rattlesnake Hill notch as the only available passage to the south. The outlets of both lakeléts were through this pass, to reach which the drainage of both, as has been explained, passed over the ice itself for a short distance. O pening oj the Ames Pond outlet.—The earlier or high-level stages of the Stoughton Bay lakelets were brought to a close when, by the shrinking away of the ice from the retaining hillsides, lower outlets were opened through the valley of Ames Pond. This valley is marked on both sides by terraces exhibiting ice-contact faces on the sides nearest the pond, constituting, in fact, true kame or morainal ter- races—a feature not often recognized in eastern Massachusetts. From these it is evident that an ice-tongue still remained in the mid- dle of the valley. This was at first possibly connected with the larger remnant of the ice farther north in the valley, but the melting of the ice, which had previously proceeded slowly by ordinary surface ablation, now probably went on with much greater rapidity, owing to the action of running water on both sides of the valley mass. This rapidly cut away the ice, already greatly reduced by the stream cros- sing from the Stoughton to the East Sharon lakelet, with the result that the Ames Pond mass was early separated into an independent block. Lakelets oj the later stages—The Elm Street lakelet was the first of the lower-level bodies to be formed, and the waters found outlet along both sides of the Ames Pond block, forming the kame terrace deposits mentioned. ‘The waters in part passed out from the valley by a bowlder-strewn gorge through granite ledges at the head of the bay just southeast of the highway across the pond, in part by the wider, sand-filled valley between rock hills just above the lower end of the pond, and in part directly southward along the main valley. 196 M. L. FULLER Following the Elm Street, the West Stoughton lakelet was formed. By this time the Ames Pond block had become much contracted, or had entirely disappeared, affording an outlet, from 10 to 20 feet lower than the earlier, through the valley now occupied by the pond. ‘These later waters carried little sand, and are represented, therefore, not by deposits, but by their erosive action. The faces of the terraces of the valleys were shaped in places by these waters, but in the main they are of the nature of ice-contact slopes. CONCLUSIONS. The results of the study of the Stoughton Bay area, in so far as they apply to retreatal conditions, prove: (1) that in this area the ice had become stagnant before the inception of the lakelets, and remained so throughout their history; (2) that the ice, previous to the emergence of the hills through its surface, was reduced almost entirely by superficial ablation; (3) that the subsequent melting was most rapid along the margins of the projecting land masses, because of the radiation of heat from them and the concentration of drainage along their borders; and (4) that the shrinkage continued to be out- ward in all directions from the exposed land masses, until the ice was reduced to narrow lines along the deeper valleys, and finally dis- appeared essentially simultaneously from all portions of the area. APPLICATIONS. Extent of glacial lakes.—Hitherto, in the discussion of glacial lakes in eastern Massachusetts, as was pointed out in the introduc- tion, it has been assumed that the ice retreated with a rather definite and regular margin, in front of which extensive bodies of water accumulated in the northward-sloping valleys. While the observa- tions of the writer have not been sufficiently extended to discuss in detail the conditions in other lakes, it has appeared almost certain from reconnaissance studies that the size of the open water bodies in the various lakes at any given time were much smaller than has usually been supposed, and that the marginal lakelets were very numerous, and definite and regular margins relatively rare, the ice having disappeared, in many instances, in the same manner as in the Stoughton area. Which type of retreat prevailed can be determined ICE-RETREAT IN GLACIAL LAKE NEPONSET 197 only by careful field work, though the facts now at hand seem to indicate that the mode of retreat discussed in the present paper is the most common. Conditions in southeastern Massachusetts—A study of the numer- ous, and often very large, kettle depressions in the stratified drift plains south of Middleboro, and at various other points in Plymouth county in southeastern Massachusetts, shows that throughout large areas in this part of the state extensive masses of stagnant ice existed during the final disappearance of the ice. How broad a belt of such ice existed at any one time or place cannot be readily determined, but in general the belt of no motion was probably not over ten or twenty miles in width, as more or less indefinite morainal bands indicating active ice-movements are found at intervals across the area, while occasionally more pronounced deposits occur. NM. FULLER: U. S. GEOLOGICAL SURVEY, Washington, D. C. RELATIONS. OF GRAVEL DEPOSITS IN: THE NORTH ERN PART OF GLACIAL LAKE CHARLES, MASSA- CHUSETTS. INTRODUCTION. AmonG the most impressive features of the glacial deposits in eastern Massachusetts is the great abundance of sandplains, which occur by scores in all the large valleys. The characteristic flat sur- faces, lobate fronts, steep or kettled ice-contact slopes, and tributary eskers testify to the formation of the plains as deltas in glacial lakes, through transportation of the sand and gravel from the ice in which it was incorporated by powerful glacial streams flowing either upon or beneath the ice, and deposition in standing water along the glacier front. In association with the true sandplains are areas of kames, eskers, and other irregular deposits varying from mere patches of gravel to tracts several square miles in extent. At intervals during a period of two or three years the writer has had opportunity to study in detail the sands and gravels within a certain limited area, and takes this opportunity to give a few of his results, hoping they may throw some additional light upon the glacial-lacustrine history of the region. LAKE CHARLES. Lake Charles is the name given to the lake, or series of lakelets, which occupied the valley of the Charles River during the decay of the latest or Wisconsin ice-sheet.* | The region drained by the river embraces portions of the counties of Suffolk, Norfolk, Worcester, and Middlesex, and is covered by the Boston, Framingham, Dedham, Franklin, and Blackstone topographic sheets. All portions of the valley are crowded by thick deposits of sand and gravel, among which flat sandplains are conspicuous, wave-cut shore lines have been occa- sionally noted, and on the east and south sides of the basin a num- ™W. O. CrosBy AND A. W. GRABAU, American Geologist, Vol. XVII (1896), No. 2, pp. 128-30; F. G. Capp, “Geological History of the Charles River,” Techno- logical Quarterly, Vol. XIV (1901), No. 3, pp. 171-201; No. 4, pp. 255-69; and Ameri- can Geologist, Vol. XXIX (1902), pp. 218-33. 198 RELATIONS OF GRAVEL DEPOSITS 199 ber of well-defined lake outlets are known, each of which, as a rule, corresponds in level with the upper surfaces of a particular series of plains. Moreover, a study of the individual plains shows that a num- ber of them often reach approximately to a common elevation, and, with few exceptions, they may be classified into one of several groups distinguished by their different levels. The elevations of the most abundant of these—not including those near sea-level—have been found to be approximately 150, 170, 200, and 270 feet in elevation, although local groups and isolated plains occasionally occur at inter- mediate levels. A fact thought to be of considerable significance is that plains having a common elevation are distributed over very wide areas. Deltas of the 170-foot series not only are known north and south of the Dover highlands and in the adjacent Neponset valley, but are very abundant across the divide of the Sudbury River in the vicinity of Cochituate, and even far down the valleys of the Sudbury and Concord Rivers. The widespread distribution of plains of the same general elevation at first appears to indicate the presence in the region of a large open body of water, which with suc- cessive retreats of the ice-margin, opening outlets for the lake at con- secutive lower levels, allowed the water to fall repeatedly to the next stage below. Recent studies by Mr. M. L. Fuller in the Neponset basin indicate, however, that in that region there was no such simple sequence of events; and the present writer, working independently, finds evidence in Lake Charles indicating that here too the open lake hypothesis can be accepted only in part. DESCRIPTION OF THE REGION. The location of the main region under consideration in the pres- sent paper is shown in Fig. 1. In a general way, it includes that portion of the Charles River basin lying south of the Boston & Albany Railroad in Newton and north of the Dedham and Dover highlands. More definitely, it is that part of the drainage basin of the present Charles River lying within that of the preglacial Charles. The portions of the valley lying north of the Boston & Albany Rail- road, and those east of the Brookline and West Roxbury highlands, are not included, being within the limits of what has been designated Lake Shawmut—a glacial lake corresponding with later stages of the ice-retreat. 200 FREDERICK G. CLAPP Topography.—aA glance at the topographic map shows that the region may conveniently be divided into the following subdivisions: the Western highlands, the Dover and Dedham highlands, the Brook- line and West Roxbury highlands, the Needham-Waltham valley, MAP OF THE 3 CHARLES RIVER BASIN i AND VICNITY F.G. CLAPP. Mikes: TGs eh. and the regions intermediate between uplands and lowlands, the last division consisting of the greater part of the towns of Newton, Needham, and Wellesley. Another feature of importance is the broad area of modified drift extending westward from Wellesley to the Sudbury River, and overlying the site of a broad preglacial RELATIONS OF GRAVEL DEPOSITS 201 valley too feet or more below the present surface. This valley— the ancient outlet for the Sudbury valley drainage, is supposed to continue beneath Wellesley to the valley of Rosemary Brook, and thence northward to the present Charles River at Riverside. The broad marshy area east of Needham, in which the bed-rock surface is known to lie at a considerable depth below sea-level, was probably occupied by a tributary of the Charles which entered the main river just north of Highlandville. The valley between South Natick and Dedham contained no single preglacial stream. The largest unbroken upland region is the broad band extending westward from the Nepon- set River across Dedham, Westwood, Dover, and Sherborn, which in preglacial times separated the lowlands on the north and south into two distinct drainage systems, since united along the course of the Charles River. DESCRIPTION OF THE DEPOSITS. In a provisional classification we may recognize the following types of deposits as occurring in this portion of Lake Charles: 1. Flat-topped delta-fans—typical gravel plains—the level of which varies but a few feet in a single plain. These deltas almost invariably have a beautiful ice-contact slope, but the common lobate front may or may not be developed. For this class of deposits the common name of ‘‘sandplain” is used, although the plains considered consist largely of gravel. 2. Kettle plains, or deposits with a definite maximum elevation, but so broken up by kettle-holes as to have little semblance to plains. In origin, they differ from true sandplains only in having been formed among crowded ice-masses. 3. Eskers, or deposits of glacial streams. These are often con- spicuously tributary to deltas. 4. Irregular deposits of gravel along the sides of valleys, usually consisting of great numbers of small hills or kames densely crowded together, for which the name ‘“‘moraine-terrace”’ is sometimes used. 5. Kames—the name comprising all small, irregular hills of sand and gravel not included under (3) or (4). 6. Thin coatings or undulating deposits of gravel, apparently having no definite relations. 202 FREDERICK G. CLAPP In the present study particular attention is given to the first three types. SANDPLAINS. Although in ascertaining the elevations of the plains the writer has not had opportunity to do any leveling, yet from the excellent con- touring on the new Boston sheet, and in some instances by the aid of the older topographic maps, the approximate elevations have been interpreted in the field. Within the immediate area under consid- eration twenty-five typical plains have been studied, which are here named, with their approximate heights above tide. The distribution of the plains and their associated eskers is shown in Fig. 2. 1. Newtonville plain, 150 feet. 14. Dedham plain, 120 feet. 2. Newton Highlands plain, 150 feet. 315. Wigwam plain, 120 feet. 3. Newton Center plain, 150 feet. 16. Islington plain, 120 feet. 4. Winchester Hill plain, 150 feet. 17. Pegan plain, 160 feet. 5. Nahantan plain, 150 feet. 18. Noanet plain, 150 feet. 6. Cow Island plain, r4o feet. 1g. Cedar Hill plain, 270 feet. 7. West Roxbury plain, 140 feet. 20. South Natick plain, 120 feet. 8. Dedham Island plain, 150 feet. 21. Trout Brook plain, 100 feet. 9. Lower Falls plain, 150 feet. 22. Greendale plain, 100 feet. to. Waban plain, 150 feet. 23. Riverside plain, 60 feet. 11. Wellesley plain, 150 feet. 24. Auburndale plain, 70 feet. 12. Needham plain, 170 feet. ; 25. Newtonville plain No. 2, 80 feet. 13. Birds Hill plain, 190 feet. Where the plains slope appreciably away from the ice-contact the elevation of the outer edge is taken, as representing most closely the true water-level. Most of them slope slightly, the difference in elevation sometimes amounting to as much as 20 feet, which makes it evident that plains having elevations given as 150 feet and others as 170 feet might have been formed as deltas in the same body of water. That such was probably not the case will be shown below. In addition to the plains already enumerated, several exceptionally fine developments outside the Needham area have been visited in this connection. ‘The most important of these are as follows: 26. North Wellesley plain, 170 feet. 29. South Framingham plain, 170 feet. 27. Pickerel Pond plain, 170 feet. 30. Medfield Junction plain, 170 feet. 28. Cochituate plain, 170 feet. A comparison of the elevations of the thirty deltas shows that twelve of the number have an elevation of 140-50 feet, six of them of 203 The remaining All below 120 feet were deposited during the Lake Shawmut stages. RELATIONS OF GRAVEL DEPOSITS twenty-five will be described somewhat in detail. 170 feet, four of 120 feet, three of 60-80 feet, two of 100 feet, one of 160 feet, one of 190 feet, and one of 270 feet. *€ ‘Sty yo Arepunoq 9y} sMoys our] ysep oy yj, ‘surejdpues pure sioyso [ediourd jo uornnquysiq—'z ‘org ake’ seas ope? PAGO SuOlg DANA youucoiddo 2a Ssaquinu 49440 PI0MASALA ‘susoyd /0 SLOQUNU Of 43/91 $39 20g Bul 7S Be df ~ “yn ul pasojjua SuaqUiny apiu aug ; A Esty Bl) 200T EX {0 Sarwopunog —~— O51 FAL SAIYS H. 79h ae) m surord puogs 3: Eyes f M900 1849 we SJ2DIUDI orf 02) g } o¢i|s VPRYY RNS mvyduiuoiy Yinos A Fe ATy att Sarat EF) ay aaeaed Bree . a rs Gx bial ' Biwi a P, pil | puitiy Ciel Lp tt 02 MAA/ s nv y 1127 IMO 7 UDgMaA/ ' 2UAUIMaAY (= 0B= SYSTEM. ESKER NEWTONVILLE-WEST ROXBURY SANDPLAIN This series of deposits is situated entirely on the east side of the Charles River, and consists of a continuous series of alternating plains and eskers extending from the northern end of the Newtonville 204 FREDERICK G. CLAPP esker, one-half mile south of the Boston & Albany Railroad, across Newton to the frontal lobes of the Dedham Island and West Rox- bury sandplains—a total distance of about six miles. Newtonville esker and plain.—The classical delta known as the Newtonville sand plain, which has been described by Davis and Gulliver,’ lies fully a mile south of the railroad, about half-way between Newtonville and Newton Center. It has an elevation of 150 feet, and averages fully three-fourths of a mile long and half a mile wide, occupying nearly the whole breadth of the valley of Laun- dry Brook. Several preglacial knobs rise through the plain, and to the east it rests against a low rock hill capped by a drumlin rising to an elevation of over 300 feet; on the north and west sides are charac- teristic ice-contact slopes; while a short stretch on the south forms good frontal lobes. Viewed from any standpoint, the deposit is a typical sand-delta, and well merits even more careful study than it has received. A large excavation on Commonwealth Avenue gives an unexcelled opportunity for viewing its internal structure. The feeding stream of the delta is represented by the Newtonville esker, which first appears as it rises from beneath the lower and more recent plain at Newtonville, about three-fourths of a mile north of its junction with the sand plain. On both sides are large areas of densely crowded kames and kettles, with which the esker in places becomes somewhat confused. Southwest of the plain there is also a small kame area. Unlike the southeast side, which has good frontal slopes, the southwest corner immediately south of Commonwealth Avenue breaks up into kames and kettles, which are composed of very coarse gravel, abound- ing in bowlders up to several feet in diameter. Within four hundred feet a further change takes place, by which the irregular hills become shaped into a typical esker, bounded on both sides by kames, and running in the direction of the Newton Highlands plain. Before reaching this there is a short break, due to a swamp, but the parallel- ism of the esker to the rock hill on the east, the close similarity between the levels of the two plains, and the apparently perfect *W. M. Davis, “The Subglacial Origin of Certain Eskers,” Proc. B. S. N. H. Vol. XXV (1892), pp. 477-99; F. P. GuLtiiver, “The Newtonville Sand Plain,’ JouRNAL OF GEOLOGY, Vol. I (1893), pp. 803-12. RELATIONS OF GRAVEL DEPOSITS 205 topographic relationship indicate strongly that the esker stream was the feeder of the Newton Highlands delta. Newton Highlands plain—This plain, somewhat irregular in shape, extends a little over a mile in an east-and-west direction, and half a mile in its greatest breadth north and south. Like the Newton- ville plain, it is bounded on the east by a rock hill, has ice-contact slopes on the north and northwest, and frontal lobes on the south; but the ice-contact in this case differs in character from the other by being extremely irregular, and in places much broken up by kettles. Onthe southwest the fore-sets run against the rock hill south of Eliot Station; on the north and south it is bounded mainly by low, swampy areas. It is for the most part a typical sandplain. Cook Street esker—The most interesting feature of the Newton Highlands plain is on the south side, near the Cook Street junction of the New York, New Haven & Hartford with the Boston & Albany Railroad. At this point a small knob of conglomerate rises above the glacial drift, and towards the north and east the fore-sets are broken and irregular, indicating the presence of ice here at the time of their deposition. Directly east of the conglomerate knob and south of Cook Street is a low esker running southward out of the sandplain. Beyond the first road it rises to a height of 30 feet or more above the swamp, and farther south widens to a breadth of at least 500 feet; but within 2,000 feet ends abruptly at a small brook. As this some- what peculiar esker has been extensively excavated, its internal structure can be easily studied. The south end being in itself some- thing of a delta, the main stream which formed it was probably not directly tributary to any plain. Winchester Hill eskers and plain.—A few hundred feet to the west, however, and holding very definite relations to the small plain west of Winchester Hill, is another esker. Starting half-way up the side of the rock hill south of Elot Station, it descends fully 4o feet. becoming at once a good-sized ridge, crosses the swamp to the kame field west of Highland Avenue, with which it soon becomes confused, and joins the sandplain at its northwest corner. The Winchester Hill plain is comparatively small, hardly one- eighth the size of the Newtonville plain, although topographically its relations are somewhat similar; it is bounded on the east by a drumlin, 206 FREDERICK G. CLAPP. the remaining three sides being free. Unlike the deltas already described, this plain has no frontal lobes, being surrounded on the north, west, and south by ice-contacts. Both to the south and west are extensive areas of kames and moraine-terrace. Nahantan plain.—Lying a mile and a half southeast of Winchester Hill, in the midst of a large kame field, is a plain having the same approximate maximum elevation as the surrounding kames and of the associated esker. In extent of isolation it goes a step beyond the Winchester plain, as its surface does not rest at any point against rock or till, and it is entirely surrounded by ice-contacts and kames. As the slope of the Brookline highlands is not over 1,000 feet distant, it is likely that rock exists only a short distance below the surface, and the plain occupies as truly a marginal position as the others men- tioned. In the area between this and the Winchester plain are two eskers, which, while they have not been observed to run directly into elther plain, are supposed from their positions to mark the streams which carried the water southward between the two lakelets. The esker running out of the Nahantan plain is better defined in its relations than the ridges similarly situated with reference to the preceding deposits. It quickly attains its individuality, and borders the kame area for nearly half a mile before finally disappearing below the Charles River meadows. One feature well worthy of special note occurs just south of the plain, where there is an enlargement of the esker barely 200 feet across, yet having all the characteristic structural features of the larger plains. South of the terminus of this esker the relations of the eskers of the series are not well defined. Short ridges appear-above the flood-plain west and north of Cow Island, and continue from near Spring Street southeastward, crossing Washington Street on the boundary line between Boston and Dedham, beyond which no attempt has been made to trace them. At the south- ern point of Cow Island is a good junction of the main esker with a ridge which crosses the river from the south. Cow Island, West Roxbury, and Dedham Island plains —The elevations of the first two of these are indicated by the map as about ro feet lower than the other plains of the system. This discrepancy may be real, but is more probably due to difficulty of interpreting the contours on a map of the published scale. In either case, it is believed RELATIONS OF GRAVEL DEPOSITS 207 that these three plains are of the same height, and were controlled by the level of a single body of water. They are all bounded on the west by steep and irregular ice-contacts, and the West Roxbury and Dedham Island plains rest visibly against preglacial knobs of land. With the Cow Island plain, on the contrary, no till nor rock can be found associated, and several wells drilled at the pumping station on the southeast side of Cow Island prove that the drift is here very thick. In one instance go feet of gravel was penetrated before reaching bed-rock. Conclusions and discussion.—From the foregoing description sev- eral characteristics common to the deposits may be formulated: 1. The plains are nearly always marginal—deposited about knobs of rock or till. Conversely, most of the preglacial prominences ris- ing above the valley floor have glacial deposits associated with them. This is probably partly due to the natural sinking of the decaying ice masses into the valleys, and in part to the principle that the heat reflected from rock masses accelerates the melting of the ice imme- diately surrounding them, forming hundreds of small marginal lake- lets, throughout the decaying zone of the ice-sheet. 2. The plains without exception have a definite ice-contact slope, which in several instances entirely surrounds them, except on the sides which are bounded by higher land. This feature indicates that some of the lakelets became entirely filled by gravel, forcing further deposits to accumulate in quantity along the channels of glacial streams or in superglacial lakelets. 3. Where typically developed, the deltas are associated with feeding eskers on the north, and with effluent eskers on the south. Such a relation can only mean that the glacial lakes, like bodies of water on the present land surface, had both inflowing and outflow- ing streams, the latter carrying off the surplus water and providing a transporting and depositing agent for the excess of gravel. This is in accordance with what has been observed on the Malaspina glacier.* The controversy in regard to the superglacial or subglacial origin of eskers will not be entered here, but the writer believes there is‘evidence that eskers of both classes exist in the region, and that tI. C. RUSSELL, “‘ Expedition to Mt. St. Elias, Alaska,” National Geographical Magazine, Vol. III, pp. 106-8. 208 FREDERICK (G, CLAPP. by careful study of the individual cases their differences in origin can be detected. 4. The elevation of all the plains of the system is very nearly or quite uniform, at about 150 feet above tide. Thus, notwithstanding the supposition that there can have been no large open lake in the region, there was certainly some connection between the lakelets adequate to maintain the water at a common level. The remnants of the decaying ice-sheet consisted of stagnant valley-blocks much broken up, the whole mass saturated with water, which was main- tained at a uniform level consequent on some factor outside the Need- ham area. Into this basin were discharged torrents of water, which carried the gravels and deposited them upon, beneath, and between the ice-masses. This view is satisfied both by the superglacial and subglacial theories of eskers, and further corroborates the supposi- tion that in this area there are examples of both classes. As the difference in elevation between the different plains of the series can not be more than a few feet, it is evident that in the waters of the lakelets themselves or below their level there can have been but little current. Notwithstanding this, it is known that water- worn bowlders of considerable size are abundant in many of the eskers and associated deposits, indicating a very powerful current. The most probable explanation, already alluded to, is that to the main superglacial or subglacial rivers were tributary many side streams which discharged into them from higher levels on the ice or surrounding land. The extensive kames and moraine-terraces are best explained, as shown by Fuller in the case of Lake Nepon- set, by supposing them to have been deposited upon masses of ice sometimes of considerable thickness, having a front sloping out- ward beneath the lacustrine waters. The probable conditions existing during this stage are indicated in Fig. 3, in which the lakelets are represented by the dotted por- tions and the stagnant remnants of the ice-sheet by shading. The . regions covered by a combination of these patterns usually corre- spond in position with large areas of moraine-terrace, and were at this time ice-tongues overlain by considerable depths of water, in which the superglacial and subglacial gravels were deposited. Gla- cial streams, so far as traceable, are shown by heavy black lines. RELATIONS OF GRAVEL DEPOSITS 209 Their most striking feature is that they are conspicuously divided into two series, one following in a general way each side of the main valley and known to have connected with the other only at a single GE: Ey > Beater ann ice z= ONE MILE Fic. 3.—Probable conditions during the 140-170-foot stage of the glacial lakelets. (Numbers show locations of plains, and correspond with list given in the text.) point. It is probable, however, that there were other connections which have not been preserved in the deposits. A second noticeable feature in the eastern series is the confluence at the Winchester Hill lakelet of two rivers, both of which have been traced for some dis- 210 FREDERICK G. CLAPP tance outside the map, and were apparently the only main streams draining the triangular area which they include. THE NEEDHAM-FRAMINGHAM SYSTEM OF PLAINS. For convenience of discussion this designation is applied to the group of plains lying between Needham and Framingham, having, in addition to a common elevation of approximately 170 feet, most of the characteristics of the preceding group. ‘These deposits cover an area beside which the Newtonville-West Roxbury area appears insignificant; yet, as the writer has had little opportunity to study them in detail, they will be dismissed with a few remarks. Belong- ing to the group are at least six plains—the Framingham, Cochit- uate, Pickerel Pond, North Wellesley, and Needham plains. All have good ice-contact slopes, most of them on certain sides only; but one plain, that lying east of Pickerel Pond between Wellesley and Cochituate, is entirely surrounded by characteristic steep and kettled slopes. This plain, like that forming Cow Island, is the only member of its group not resting against the rock sides of the valley, and, like Cow Island, is supposed to overlie a buried pre- glacial gorge. The most conspicuous topographic characteristic of this region is the abundance of kettle-holes, often occupied by ponds. Lake Cochituate, Dudley, Pickerel, Mud, Jennings’, and Morse’s ponds and Lake Waban are all kettle-ponds, overlying the course of the preglacial Charles River. Between, around, and upon the ice- blocks gravels were deposited, their mode of origin being demon- strated by their kettled character. The proportionate area of kames to kettles is much larger than in the deposits farther east, and the kettles are more isolated. Plains having in general the same eleva- tion extend far down the valleys of the Sudbury and Concord Rivers; thus emphasizing the suggestion that in the Sudbury valley there were at that time fairly open bodies of water. Needham plain.—This is the largest and most typical plain in the group. Hemmed in on east and west by rock hills, it is bounded on the north by a deep preglacial valley, in which the ice lingered and gave rise to deep kettles, extensive moraine-terraces, kames, and eskers. ‘Three or more eskers, with relations suggesting a super- glacial mode of origin, enter the plain from the north. Its southern RELATIONS OF GRAVEL DEPOSITS AAA side shows a beautiful lobate front, which, together with the absence of kettles and the scarcity of modified drift immediately to the south, indicates that in that direction was an open lakelet three to four miles in extent. The overflowing water, reinforced by the drainage from an extensive area to the west, found its outlet eastward through a large superglacial or marginal stream. Lower Falls, Wellesley Hills, and Waban plains.—These are all about 150 feet in elevation, and the first two originally formed a single plain. Extending in an irregular line from near Rice Crossing north of Wellesley Hills to Newton Lower Falls, and thence to the high morainic hills east of Riverside, is a kettled and kamey ice-contact nearly three miles in length. Large kettles south and east of the Lower Falls indicate that the glacier to the north was almost con- nected with the decaying ice-blocks on the southeast side of the plain; but that it was not quite continuous is indicated by consideration of the fact that the slopes on both sides of the river are erosion slopes, in which artificial excavations show nearly horizontal stratification, probably once continuous across the valley. ‘The Waban plain, lying closely adjacent to the Lower Falls plain on the east, was formed by the Auburndale esker stream.’ It is bounded on the north and south by ice-contacts, is much broken up by kettle-holes, and might properly have been classified with the Newtonville-West Roxbury group, with which it is closely connected. Other plains oj the series.—In the valleys east and west of Dover are two or more small plains at approximately the same elevation, which are shown by ice-contacts on their valley sides to be deposits of small marginal embayments. Plains of the Newtonville-West Roxbury type occur in the Medfield-Medway basin of the Charles River, and also in the Neponset valley. The uniformity of elevation over such a wide area strongly suggests dependence upon some lake- outlet east of the Neponset valley. No systematic search for such an eastern outlet has been made in the field, but it is known that in the southern watersheds of the Charles and Neponset Rivers there is no pass lower than 200 feet above tide. tJ. B. WoopwortH, “‘ Some Typical Eskers of Southern New England,”’ Proceed- ings of the Boston Society of Natural History, Vol. XXVI, pp. 197-220; W. M. Davis, “‘ The Subglacia! Origin of Certain Eskers,” ibid., Vol. XXV, pp. 477-99. 212 FREDERICK G. CLAPP The perfect ice-contact slope along the Boston & Albany Rail- road east of Wellesley Hills, together with the absence of any flat- topped deposits rising higher than 90 feet, throughout the area to the northward, and likewise from the entire region east of the Brookline and West Roxbury highlands, proves that the northern part of the Boston basin was then occupied by the still little-decayed ice-sheet, an arm of which extended up the Neponset valley as far as Norwood. PLAINS OF OTHER LEVELS. Cedar Hill lakelet—There still remain to be mentioned several important plains at levels different from any heretofore described. One of these plains occurs on the divide between Noanet and Mine Brooks in Dover. It lies at an elevation of 270 feet, and has a tribu- tary esker on the north and frontal slopes on the south. Beginning a short distance south of the plain, and extending to beyond Walpole, a distance of over four miles, is one of the finest eskers in the region, supposed to mark the course of the stream through which this lakelet discharged. Dedham plains.—In the towns of Dedham and Westwood are several plains having an elevation of about 120 feet. These are flat- topped, are either entirely or nearly surrounded by ice-contacts, have abundant kettles, and are not known to correspond with plains of any other group. Along the valley of the Charles River, between Dedham and South Natick, are a number of plains at elevations of 100 to 120 feet, which may belong to the same stage. Below this elevation all the plains are confined to the Lake Shawmut area, the western limit of which is marked by flat go-foot deposits north of the ice-contact at Riverside and Newton Lower Falls, and which extend eastward throughout the Boston basin, where they are recognized at all eleva- tions down to sea-level. HISTORY OF THE LAKELETS. The highest known plain in the region, the Cedar Hill plain, has an elevation of 100 feet above the next lower level, and must have been formed at the time of the earliest uncovering of the pass in the Dover highlands, which left a lake irregular in shape covering an area of nearly one square mile, walled on the east and west by rocky hills, and on the north and south largely by ice. The eskers in the RELATIONS OF GRAVEL DEPOSITS 2a valleys of Noanet and Mine Brooks suggest that this is another instance of a glacial lake having both inflowing and outflowing streams. The occurrence of so numerous and widely scattered sandplains having a uniform elevation, which, notwithstanding their wide distri- bution, were evidently formed in merely local lakelets, often resting in part on the ice itself, is best explained by the assumption that the ice was in such a state of decay as to allow a connection between the individual lakelets sufficient to maintain the water at a common level. Such being the case, the general level of the waters at the different stages must have been determined by one of two causes; either, first, by deposition in an arm of the sea, or, second, as was more probable, by topographic features, probably passes between the highlands on the east side of the lower Neponset valley, through which the water escaped to Lake Bouvé,* and thence to the sea. The probable explanation for the lack of plains between the 270- and 170-foot levels is the absence of any pass intermediate between these elevations on the northern side of the Blue Hill range. In the southern portion of Lakes Charles and Neponset, where passes of other levels occur, intermediate plains also abound. ‘The relations of ice, land, and water during the 170-150 foot stage are shown in Fig. 2. This stage of the lakelets existed during the entire time while the outlets remained at this elevation. When the decay of the ice had advanced sufficiently to open a 120-foot outlet north of the Blue Hills, the ponded waters fell to that level. There is no indication that during this time there was any considerable disappearance of the ice in the upper and central portions of the Charles River basin, for in these portions no plains are known below 160 feet. The absence of synchronous plains in the Lake Shawmut area indicates that this was still for the most part ice. By far the greater portion of the 120- foot plains is situated in the town of Dedham, their location being probably due to their favorable position at the head of the Charles River valley glacier. Another view of their origin is that they repre- sent merely a local stage, while the greater portion of the waters stood at a lower level. 1 A. W. GRABAU, “Lake Bouvé,” Occ. Papers, Boston Society of Natural History, Vol. IV, part ITI, pp. 564-600. 2A FREDERICK G. CLAPP CONCLUSIONS. The most important generalizations to be drawn from this study are as follows: t. The decay of the ice zm situ for many miles back from the ice- front—the decaying glacier consisting of a mass of stagnant ice over- lain and buried by sheets of water and by extensive deposits of sand and gravel. 2. The more rapid disappearance of the ice on the west than farther east, causing a nearly open lake in parts of the Sudbury valley, while as yet the ice in the lower portions of the Boston basin had not decayed sufficiently to allow the formation of a single sand plain. 3. The interdependence of the water levels of the individual lake- lets belonging to each stage, and their correspondence with some common outlet toward the sea. FREDERICK G. CLAPP. WASHINGTON, D. C. Dt ee bOPA DIE (QUARTZ PORPHYRY) OF NORTE CAROLINA.” INTRODUCTORY STATEMENT. WHILE engaged, during the past summer, in.a study of the gran- ites of North Carolina for the State Survey, opportunity offered for examination in the field of the well-known and interesting rock called ‘‘leopardite,’ county. Knowledge of the occurrence of this rock in the state dates back many years, and brief descriptions of it have been pub- lished from time to time by different writers, as noted below in the appended references. In 1853 Dr. Hunter’ briefly described, megascopically, the general appearance, including locality, of the leopardite found near Char- lotte, Mecklenburg county, North Carolina. He says: “It is noticed by Professor Shepard, under the head of feldspar, as the leopard stone of Charlotte, North Carolina.” Professor Shepard regarded it as composed of compact feldspar and quartz spotted by the oxides of iron and manganese. Hunter suggested the propriety of retain- ing the name “leopardite,” for the reason that it is quite charac- teristic of a rather unique rock. In the same paper the author refers to a second locality in Lincoln county, North Carolina, where leopardite had recently been found. Concerning the character of the rock in Lincoln county, he says: ‘‘The pervading stripes are, however, generally finer; and when broken diagonally, it presents a handsome arborescent appearance.”’ In 1862 Dr. F. A. Genth3 described the leopardite occurring near Charlotte as a true porphyry, and gave some general results of a microscopical examination of thin sections of the rock, including a chemical analysis. Still a third locality in North Carolina where ’ which occurs near Charlotte in Mecklenburg * Published by permission of the state geologist of North Carolina. 2C. L. Hunter, “Notices of the Rarer Minerals and New Localities in Western North Carolina,’ American Journal of Science., Vol. XV (1853, 2d ser.), p- 377- 3 F. A. GENTH, “Contributions to Mineralogy,” zbid., Vol. XX XIII (1862, 2d ser.), Pp: 197, 198. 215 216 THOMAS L. WATSON leopardite is reported to be found is referred to by Genth, namely, near the Steele mine in Montgomery county. More recently the leopardite occurring near Charlotte has been noted by Merrill" and Lewis’. After briefly describing the general appearance of the rock, Professor Merrill makes further statement of its economic value. In connection with his work on the building- stones of North Carolina, Lewis visited the locality to the east of Charlotte, where the leopardite is exposed, and, so far as contained in published accounts of the rock known to me, he was the first to note its true geological occurrence. Quartz porphyries in association with other closely similar acid volcanic rocks are developed, in places, over the central and the northwestern parts of the state. So far as known at present, the areas of acid volcanic rocks are confined to the volcanic belt which skirts the western margin of the Triassic sandstone in the eastern Piedmont region,’ and to several of the extreme northwest counties* of the state. These rocks show no essential differences, so far as they have been studied, from certain areas of similar ones which occur and are traced at irregular intervals northward along the Atlan- tic border region of North America as far as Newfoundland. Of those occurrences in North Carolina, the quartz porphyry found near Charlotte is the only one visited by me that shows the characteristic spotted appearance so suggestive of the name “‘leop- ardite.”’ Except for the mottled or spotted appearance produced by rounded black areas of metallic oxides, the Charlotte rock differs but slightly, if at all, in essential characters from quartz porphyries described from other localities. (See table of analyses on p. 223). «GEORGE P. MERRILL, Stones for Building and Decoration (New York, 1897), 2d ed., pp. 272, 273: 2J. V. Lewis, Notes on Building and Ornamental Stone, First Biennial Report of the State Geologist, N. C. Geological Survey, 1893, p. 102. 3 GrorGE H. WitttaMs, “The Distribution of Ancient Volcanic Rocks Along the Eastern Border of North America,” JouRNAL OF GEOLOGY, Vol. II (1894), pp. 1-32; J. S. Ditrer, “Origin of Paleotrochis,” American Journal of Science, Vol. VII (1899, 4th ser.), pp. 337-42. 4A. Kerry, Bulletin No. 768, U. S. Geological Survey, p. 52; Geologic Atlas of the United States, “‘ North Carolina-Tennessee, Cranberry Folio,” 1903. THE LEOPARDITE OF NORTH CAROLINA 21 “I LOCATION AND OCCURRENCE. The leopardite is exposed in a number of small outcrops at Bel- mont Springs, about one and a half miles east of Charlotte. Begin- ning on top of the hill, several hundred yards above the spring the rock is traced in outcrops over the surface for a distance of a quarter to a half mile in a north 30° east direction. It forms a true dike, intersecting a medium textured and colored, sheared and crushed, Fic. 1.—View showing the spotted appearance of the rock on a surface broken at right angles to the longer direction of the pencils. Photographed from a hand specimen. (One-half natural size.) biotite granite; and, so far, as it was possible to determine, the dike nowhere exceeds twenty-five feet in width, with a smaller average cross-section. A small opening in one of the outcrops from which some of the rock has been blasted reveals a sharp contact between the quartz porphyry and the inclosing granite. MEGASCOPIC DESCRIPTION. The fresh rock is nearly white, tinged the faintest greenish in places, and penetrated by long parallel streaks or pencils of a dead 218 THOMAS L. WATSON black color. When broken at an angle to the direction of the pencils, the rock surface appears spotted with rounded, irregular black points, ranging in size up to a half inch in diameter. At times the roundish points are somewhat irregular and only partially devel- oped, as shown in the lower left half of Fig. 1. These may be crowded close together over the surface, as seen in the figure, or or they may be entirely absent from some areas and irregularly distributed at wide intervals over others, as indicated in Fig. 3. Indeed, the black points are reported to fail entirely in the rock as the dike is traced northward, when the rock assumes a uniformly light color throughout. However, every outcrop and specimen of the rock seen by me contained them. A section cut parallel to the direction of the pencils presents a surface streaked with long, somewhat irregular, though roughly parallel, black lines, more or less perfect dendritic or fern-like forms (Figs. 2 and 3). I was shown recently a large slab of the rock col- lected from one of the outcrops since my examination in the summer of 1903, which, for perfection and delicacy of tracery in fern-like forms, was beautiful beyond description. The black streaks or pen- cils which characterize the rock are composed of a staining of the oxides of manganese and iron. The rock is cryptocrystalline in texture, breaking with a con- chodial fracture, and is intensely hard and tough. Minute quartz crystals of doubly terminated pyramidal faces are distributed through the rock at irregular wide intervals. These are nowhere abundant in the rock, but they are always present to some extent, and consist both of the light-colored and the dark, smoky, vitreous quartz. Indeed, unless carefully examined, the rock would ordinarily be pronounced non-porphyritic in texture, so small and scattering are the porphy- ritically developed quartzes. Megascopically, porphyritic texture is nowhere particularly emphasized in the rock, but its slight develop- ment is best seen on a weathered surface of the stone, where the unaltered quartz crystals, though few in number and widely scat- tered, contrast more strongly with the weathered surface and appear more conspicuous than in the fresh rock. Feldspars are also porphy- ritically developed, as described below, though the phenocrysts are difficult of differentiation in hand specimens of the rock. THE LEOPARDITE OF NORTH CAROLINA 219 MICROSCOPICAL DESCRIPTION. In thin sections the rock consists of a holocrystalline groundmass and scattered small porphyritic crystals. Flow-structure is not exhibited in the groundmass, and the phenocrysts indicate no orientation with respect to each other. The groundmass is micro-granitic in structure, though some sections show much of the micro-granophyric structure, with an irregular radial, spherulitic, structure developed in greater or Fic. 2.—View showing approximately parallel black streaks and pencils on rock surface broken parallel to the direction of the pencils. Photographed from hand specimen. (One-half natural size.) less proportion in all of the sections studied. When they form com- plete spheres, which is rarely the case, they usually exhibit somewhat irregular ragged peripheries, and further show usually between cross nicols a very indefinite black cross. The form of the grains in the typical micro-granitic areas of the groundmass is sharp and allotrio- morphic to partially idiomorphic. The principal groundmass minerals are feldspar and quartz, with much light-colored mica, and an occa- sional inclusion of prismatic apatite and zircon. Irregular minute grains of iron oxide are scattered through the sections, and stained 220 THOMAS L. WATSON areas from manganese and iron oxides, forming the dark spots and pencils in the hand specimens, occur. The thin sections are char- acterized by the complete absence of ferro-magnesian minerals. Feldspar is apparently in largest quantity, and is composed of both potash and plagioclase species. Occasional grains of micro- cline are recognized which show the characteristic microcline twin- ning. The unstriated feldspar grains so strongly resemble quartz that it is impossible in many cases to distinguish-them without the application of optical tests. Optical tests show the plagioclase to be albite—a circumstance entirely confirmed by the chemical analysis of the rock given below in the table of analyses under I, in which only the barest trace of lime is indicated, with soda in large amount and in excess of the potash. Some of the plagioclase exhibits polysynthetic twinning according to the albite law, and at times assumes lath- shaped forms. ‘The feldspar substance is generally fresh, but the individual grains are usually rendered dark by, abundant, closely crowded, minute, dark, dust-like particles, the identity of which could not be made out. Quartz is of the usual kind and presents no noteworthy features, further than its occurrence in small mosaics of interlocking grains, which occupy at times distinct areas in some of the thin sections. Light-colored mica, tinged a faint yellow, is very generally distrib- uted through the sections, in the form of irregular minute shreds, groups, and aggregated masses, the folia of which are at times imper- fectly arranged radially about a common center. A part, at least, of the mica is clearly secondary, while some of it is yet doubtful as to origin, whether primary or secondary. Its general appearance and association in the sections might very well indicate secondary formation for all of it. Phenocrysts of both quartz and feldspar occur in well-developed idiomorphic forms, usually in rectangular and squarish cross-sections. In the thin sections studied phenocrysts of feldspar are more abundant than quartz; and while the porphyritic texture is poorly developed in the hand specimens, it is very pronounced in the thin sections. The quartz phenocrysts show irregular fractures free from impurities, strained shadows, and occasionally inclose grains of feldspar. The porphyritic feldspars show in part broadly twinned bands of plagio- THE LEOPARDITE OF NORTH CAROLINA 22m clase, and untwinned orthoclase. These are frequently rendered nearly opaque from innumerable, closely crowded, dark inclusions not identifiable, along with minute spangles of colorless mica. Zonal structure is rarely observed, and cleavage is usually wanting. Around the borders of several of the feldspar phenocrysts slight embayments, produced by incipient resorption, are noticeable. Fic. 3.—View showing partially spotted and partially streaked rock, with tend- ency toward arborescent form manifested near the middle of the picture. Surface broken at an angle intermediate between that of Figs. 1 and 2. Photographed from hand specimen. One-half natural size. Several of the sections were so cut as to include areas of the black pencils which characterize the rock, megascopically. These are dis- tinguished, microscopically, from the white portions of the groundmass only by a distinct medium-to-dark yellowish-brown staining, some- what resembling that of limonite stain frequently observed discoloring tiny areas of the rock, derived from the partial leaching of any iron- bearing constituent in igneous rocks. No definite source of the staining was entirely indicated in any of the sections, but the areas clearly represent percolation of solutions of manganese and iron 222 THOMAS L. WATSON salts through the rock. Why the definite arrangement into long pencils and dendritic forms manifested megascopically, evidence is again lacking, for the textural relations of the mincrals in the dis- colored areas are precisely the same microscopically, as for other portions of the rock. The character of the staining suggests that the spotted and streaked appearance of the rock is a superficial phenome- non, and perhaps does not extend to any very great depth. Fic. 4.—View showing weathered surface of the rock. Partial leaching of the dark spots is emphasized in the upper portion of the picture. Photographed from hand specimen. (One-half natural size.) CHEMICAL. COMPOSITION. The chemical composition of the rock is given in analysis I of the table of analyses. The analysis of leopardite was made by Dr. F. A. Genth from the freshest fragments of the groundmass obtain- able. The most striking features of the analysis are (1) the very acid character of the rock, as manifested in the high SiO, content; (2) the nearly complete absence of CaO and MgO; and (3) the increased Na,O which is in excess of the K,O. The analysis, however, har- THE LEOPARDITE OF NORTH CAROLINA 223 monizes closely with the microscopic study of thin sections of the rock, for the absence of ferromagnesian minerals accounts for the very slight amount of MgO present, while the practical absence of CaO and the large percentage of Na,O prove the plagioclase to be albite, as. indicated above by the microscope. This analysis is compared in the table with a recent, more detailed one (II), of a quartz porphyry occurring in the northwestern part of the state, and with a spherulitic rhyolite (III) found east of the Charlotte locality in Montgomery county; and with analyses IV and V of well-known quartz porphyries occurring in other parts of the United States. A perusal of the figures given in the table will make clear the general similarity of the rocks, notwithstanding the rather striking differences indicated in some of the constituents. TABLE OF ANALYSES. I II II IV V SIO, ----------------- 75-92 79-75 79-57 73-12 72.85 AM,On soveesbaaonoecde 14.47 10.47 II.41 14.27 13.78 ING Oia seiner aes RE 0.88 0.64 0.20 Ouse 1.87 IO) Bienes alte a enre een 0.92 0.70 0.26 0.36 NIG OM cies eecoesieee 0.09 0.13 a little 0.24 0.42 CAO Te atae neil 0.02 ORES O27 TLO 0.87 IND Olt Gamer eens 4.98 1.36 3.46 3-43 4.14 IK 5 Oss ern bee ea eee 4.01 6.01 Bag 4.90 4.49 Jel Oita sa aeseaaae aie 0.08 0.18 0.68 0.22 ISL, Oa Pinel e, aoaeoos 0.64 0.60 0.61 0.73 0.54 TKO) posse sia ea eee aoe 0.15 Onn 0.08 0.44 5 OR ircgrarers is ee a= trace trace 0.03 Ooi TAO) fy Feiler ee eS 0.05 eeeie Fes te IM GNX OL Ae ee eee trace 0.06 0.06 SIRO) aia ees rere shsteu trace ee trace Ba Oe are eeeare sie sere en 0.06 0.05 trace a Oman cea ee trace trace IN @ pics ther ek pee See, eee anes ee 0.20 COs oem anEeeaEs eevee Bee See 0.77 A Otalliidcteccesicss Miartvetss= T00.OL 100.37 100.02 100.18 99.87 I. Quartz porphyry (leopardite), one and a half miles east of Charlotte, Meck- lenburg county, North Carolina. American Journal of Science, Vol. XXXIIT (1862, 2d ser.), p. 198. F. A. Genth, analyst. II. Quartz porphyry—two and a half miles northwest of Blowing Rock, Watauga county, North Carolina. Petrographic data by Arthur Keith. Contains quartz and orthoclase, with subordinate sericite, chlorite, and biotite. W. F. Hillebrand, analyst. Bulletin No. 168, U.S. Geological Survey, p. 52. 224 THOMAS L. WATSON IIL. Spherulitic rhyolite—Sam Christian gold mine, Montgomery county, North Carolina. Described by J. S. Diller, American Journal of Science, Vol. VII (1899, 4th ser.), p. 341. W. F. Hillebrand, analyst. Bulletin No. 168, U. S. Geological Survey, p. 53. j IV. Quartz porphyry.—Yogo Rock, sheet at head of Belt and Running Wolf Creeks, Little Belt Mountains, Montana. Described by Weed and Pirrson. Twentieth Annual Report, Part III, U. S. Geological Survey, pp. 520 ff. W. F. Hillebrand, analyst. Bulletin No. 168, U. S. Geological Survey, p. 125. V. Quartz porphyry.—Six miles east of Ironton, Missouri. Described by E. Haworth, Annual Report, Missouri Geological Survey, Vol. VIII, 1894, p. 181. Melville, analyst. WEATHERING. In some exposures of the leopardite the weathered surface of the rock, which is still hard and firm, presents a lusterless, dead, chalk- like whiteness, the black spots of which are more or less bleached, changed from black to a reddish-brown in color. ‘This alteration is brought out fairly well in Fig. 4, which is a photograph of a hand specimen of the partially weathered rock, reproduced one-half natural size. Bleaching of the spots is more emphasized along the top of the specimen, shown in the figure (4) in the contrasted lighter color of these spots to others in the same figure. When Fig. 4 is com- pared with those of the fresh rock, Figs. 1, 2, and 3, it is noticeable that all the spots in it have undergone some leaching, as indicated in their color being less intense or deep than for those in the fresh specimens of the rock. Tuomas L. WaTSON. GEOLOGICAL LABORATORY OF DENISON UNIVERSITY, Granville, Ohio. QUARTZ-FELDSPAR-PORPHYRY (GRANIPHYRO LIPAROSE-ALASKOSE) FROM LLANO, TEXAS. THERE occurs in the vicinity of Llano, Tex., a porphyry which is very interesting petrographically, and may prove equally so commer- cially. It forms a large body whose shape and geological occurrence have not yet been described. It is said to be quite uniform in char- acter. The material submitted by Dr. William B. Phillips, Director of the University of Texas Mineral Survey, for petrographical study is a gray porphyry with abundant phenocrysts of red feldspar and blue quartz, the matrix or groundmass being aphanitic to phanerocrystal- line. It appears to have a crystalline texture, but the individual grains are not distinctly visible without a microscope. ‘The rock is therefore mottled red and gray, with light blue spots of opalescent quartz. The phenocrysts vary in size, the largest feldspars being 10™™ in diameter, the largest quartzes 5"™. The quartzes exhibit a beautiful blue color, which is light sky-blue in the central part of the crystal and dark at the margin. The crystals are not all colored to the same degree; some are lighter than others. The color does not change perceptibly with a change in the angle of incidence, or in the position of observation, except that in certain positions there is a brilliant light blue luster. The feldspars are rather uniformly colored light Indian-red, the larger crystals being mottled with gray. The proportion of phenocrysts and groundmass estimated from the surface of the specimen and from three thin sections is: { quartz 5 7 ; pee Oa Phenocrysts | feldspar - a E fe = ADS Groundmass 5 j : 5 oae 100.00 Under the microscope the groundmass is seen to be holocrystal- line and microcrystalline, and is composed of feldspar and quartz in nearly equal proportions, together with a small amount of brownish- green mica, and still less fluorite, magnetite, apatite, and zircon. 225 226 JOSEPH P. IDDINGS The proportions in which these occur was determined by microscopi- cal measurement to be approximately, in 62.8 per cent. of ground- mass: Total Quartz, DIREC) - - - : - - 34.6 Feldspar, - 2922 - - . - - SSS A7, Biotite, 8.6 - - - - - = BeO Fluorite, HO. - - - - . 1.0 Apatite, nk - - - - - = Ont? 62.83 100.03 The fabric of the groundmass is uniformly heterogeneous, being a mixture of automorphic granular and micrographic. It consists of anhedrons of quartz, very free from inclusions, except some minute gas cavities, with similarly shaped anhedrons of microcline slightly clouded with alteration products, besides anhedrons of twinned albite with an approach to automorphism. ‘These anhedrons vary in size from o.1 to o.or™™ in diameter. Throughout the whole are scattered at short intervals granular clusters of graphic intergrowth of quartz and feldspar. The crystallization of the graphic parts was almost contemporaneous with that of the anhedrons, as these are developed in continuous orientation with the graphic clusters. The mica is xenomorphic in great part, and is in about the same sized anhedrons as the quartz and feldspar. It appears to have been almost contemporaneous in crystallization with these minerals. Its color is green to brownish-green. Fluorite occurs in irregularly shaped anhedrons, xenomorphic in form. It is colorless in thin sections, exhibits distinct cleavage, and is characterized by its low refraction and isotropic behavior. It is quite uniformly scattered through the groundmass. Apatite occurs in colorless microscopic prisms. Magnetite and zircon both occur in anhedrons in such small quantities that they were not measured. ‘They appear to constitute a small fraction of I per cent. of the rock. A careful study of the feldspars in the groundmass showed that microcline and albite are present in nearly equal proportions, and that they form separate and distinct crystals not perthitically inter- grown. | QUARTZ-FELDSPAR-PORPHVRY FROM LLANO, TEX. 227 The feldspar phenocrysts are microcline, with extremely minute and regular multiple twinning in two directions. The delicacy of the twinning suggests a possible soda content in the potash feldspar approaching soda microcline. ‘There is also a perthitic inclusion of albite in irregularly shaped shreds, and also a slight clouding due to alteration, which is probably kaolin with hydrous oxide of iron which gives color to the feldspar. The quartz phenocrysts contain multitudes of minute inclusions, rather evenly distributed through each crystal, except for a margin of © nearly pure quartz in some cases. ‘The inclusions are of two kinds, generally intermingled: one consists of extremely thin, colorless prisms, sometimes passing into lines of minute grains, like broken prisms; the other kind is in equally thin tabular crystals with six sides and trigonal shapes, and a light brown color. The colorless prisms have higher refraction than quartz, but the double refraction is not recognizable. They resemble apatite rather than rutile, having lower refraction than rutile and not being so long as rutile needles often are. The width of these prisms varies from 0.000800™™ to much less; that is, it is mostly a fraction of a wave-length of light. The brownish tabular crystals are equally thin, and range in diame- ter from o.co4™™ to much less. Studied by incident sunlight, they exhibit metallic reflections of a bluish-white and also of other colors. They have the crystal form and color of ilmenite. These inclusions lie at all angles within the quartz crystals, but there appear to be sets of parallel directions intersecting at various angles, so that in some positions many tabular microlites reflect light in one direction. The same is true of the colorless needles. They lie in parallel lines crossing at various angles, whose orientation with respect to the inclosing quartz does not appear to be definite. The sky-blue opalescent color of the quartz phenocrysts is undoubt- edly due to reflection of blue light-waves from the minute colorless prisms, whose width is a fraction of the length of light-waves. It is similar to the blue color of the sky. It is probable, however, that there is also blue light produced by interference of the light reflected from both sides of the minute tabular crystals, whose thickness is also of the order of a fraction of a light wave-length; so that both kinds of phenomena occur within these quartzes. 228 JOSEPH P. IDDINGS From the microscopical measurements of the minerals and the optical characters of the feldspars it is possible to estimate approxi- mately the chemical composition of the rock. The feldspars appear to be albite and orthoclase (potash microcline) in almost equal pro- portions in the groundmass, and the phenocrysts appear to have these molecules in nearly the same proportions. In assuming a chemical compositon for the brownish-green mica, the analysis of that in the soda granite (grano-liparose) of Cape Ann, Mass.,* was chosen. On this basis the chemical composition of the rock was calculated to be that given in column I. This was done before chemical analyses of the rock were made, and the result is of great interest as showing how far this method of estimation may be relied on in favorable cases. If the microscopical measurements had been made to include the mag- netite and zircon, the result would have been still more elaborate. Subsequently Analysis II was made by Mr. S. H. Worrell, of the University of Texas Mineral Survey, on a sample of the rock from the land of Mr. H. C. Harned, near Llano, Tex. As the alkalies were not separately determined in this analysis, Dr. H. S. Washing- ton very generously undertook to analyze material from the specimen studied microscopically. The result is given in Analysis IIT. . From I | II Ill IV SLO eee ae er nae Hae | 2 FAG 75.90 74.80 BAD © Sica earth oe ey 11.58 ek 12.07 II.44 Hes Olce terete seer 0.69 1.6 TOL TOY Be Ouages Sobetiaek Mee ween 2.61 TS) 1.45 1.62 IN AO) ae rei eees ir eeeepre THOM CEM eee ee oe 0.22 0.28 (Gri Ok oer as mole eet Sins 0.82 0.2 0.65 0.80 IN Os. aexeeuensascocdes 3.40 8.5 3.08 Buse ICO a oBeusn sun cnoHonGe 5.46 tr Roe 5-52 io ee ee 0.36 0.3 ee 0.23 TROMceni sae otee enone 0.20 0.5 0.38 0.40 DAN © Ye acu mee ea ee Sa OnOS edn ak ghee 0.15 ©.05 IFA Fe be Ok ie op oyaa te nen ep ae ene ce OAOra al oosecteks n. d. 0.49 COB OS ete Se po Adlbog Dies ah aw ea | Prelleny eaeraneei te NONE! |) sees Min @ nee Sacco eee 0.02 1.9 n.d 0.18 100.29 100.4 100.70 100.20 less 0.2 less 0.21 100.08 99 -99 ™See Table XIV in Quantitative Classification of Igneous Rocks (Chicago, 1903), mica analysis e. QUARTZ-FELDSPAR-PORPHYVRY FROM LLANO, TEX. 229 these analyses it will be seen how close the microscopically estimated chemical composition is to that determined by chemical analysis. The higher silica in III shows that the quartz in the rock was underestimated by 1.5 per cent., or that the piece analyzed by Dr. Washington was slightly richer in quartz phenocrysts. The following data were determined in the laboratory of the Mineral Survey of the University of Texas: Specific gravity, 2.64; corrected, 2.67. One cubic foot of the rock absorbs 9.47 ounces of water. Crushing strength, 15,300 pounds per square inch of sur- face. The alkalies in I and III are remarkably concordant, proving that the determination of the feldspars by optical means was correct; The lime determined in III corresponds to that estimated optically in fluorite and apatite. Fluorine appears only in Analysis I, and is very nearly correct, probably as much so as if determined by chemical means. The correspondence between the two oxides of iron in both chemi- cal analyses, II and III, the discrepancy in Anlaysis I, and the pres- ence of a small amount of magnesia in III show that the mica analysis chosen from the Cape Ann rock is not the proper composition for the mica in the porphyry under investigation. The probable composi- tion of this mica may be found by subtracting from Analysis II the chemical constituents of the known minerals—quartz, feldspar, fluorite, and apatite—and reckoning the remainder as mica and the extra quartz already mentioned. The result is as follows: extra quartz, 1.37 per cent.; mica, 8.6 per cent., having the composition (a). (a) (6) (a) (6) Si@seaeae sae Be) oil 35.26 IR Ons sameness 6.8 g.20 Mls Onssoaacsoos 19.2 10.24 ESO eee oe canS 270 ING Oe sae rere To 12.47 Oye ee eee ee 4-4 4.68 GO} Ghee oe aime 16.8 18.84 Mn @ geese hacnste 2.14 IMtO) Pesan eeeor Dats 3-24 (CHO r a aestesees 0.8 0.05 99.8 99 - 34 INE OLA eaGomer le tines eee 0.60 This is approximately the composition of a lepidomelane like that in the nephelite-syenite (grano-nordmarkose) of Litchfield, Me.," * Loc. cit., mica analysis }. 230 JOSEPH P..IDDINGS with a slight difference in the oxides of iron, and a notable amount of titanium oxide. It closely resembles the analysis (6) of lepido- melane from nephelite-syenite from the neighborhood of Lange- sundfjord, Norway, by Scheerer.* If this analysis of mica is used in the calculation of the chemical composition of the rock, the result is that given under IV. The mica is clearly a lepidomelane rich in iron and alumina and poor in magnesia. When a greater variety of micas has been separated from igneous rocks and carefully described and analyzed, it will be possible to estimate the chemical composition of a rock from a microscopical investigation with greater accuracy. The rock from near Llano, Tex., may be called a quartz-feld- spar-porphyry having the composition of a granite. In the “Quan- titative System of Classification” it is a graniphyro-liparose-alaskose. The norm calculated from Dr. Washington’s analysis, II, with the addition of fluorine determined microscopically from fluorite is given below under (1). The norm calculated from the mode by means of estimated Analysis IV is given under (2): (1) Norm (2) aes Mode (OMEN ARS ene SB Se mea 36.90 33.30 34.6 @rthoclases tees ee oe 31.14 32.80 27.8 MIMS eGooulosae os osoes 26.20 AG alo] 27.0 Conindumaeeseisee See ere Tage Ve un eat | aera seu Beats Ely persthen Caemmerc ster e r.69 2.42 Biotite 8.6 Mialome tite ine terre iyer-ie 165 340) 1.62 tr. lilmieniteyas emer sas ser 0.76 0.76 tr. AXP Atite voserercierst rey evsrerel eres 0.34 Tes 0.13 IMO SSoSsEsRe sao oUeeeE 0.70 1.00 1.00 100.45 99 - 67 100.03 The rock is a persalane with about 5 per cent. of femic com- ponents. Following the norm from Dr. Washington’s analysis, the quartz is so abundant that it is quarfelic, columbare, near the quar- dofelic order britannare. It is therefore a britannare-columbare. It is peralkalic of the most extreme kind, having no anorthite feldspar, the lime being entirely femic, in fluorite and apatite. It is alaskase near liparase, a liparase-alaskase. And with respect to ™See W. C. BROGGER, Zeitschr. Kryst. Min., Vol. XVI (1890), p. ‘191. QUARTZ-FELDSPAR-PORPHYVRY FROM LLANO, TEX. 231 alkalies it is sodipotassic, and hence an alaskose near liparose: lipa- rose-alaskose.. If the norm derived from the mode were made a basis of classification, the rock would be quardofelic, a britannare, “near columbare. From this it is evident that the rock is interme- — diate between these orders, and may be a liparose-alaskose, or an alaskose-liparose. JosEepH P. IpDINGs. CRYSTOSPHENES OR BURIED SHEETS OF ICE IN THE TUNDRA OF NORTHERN AMERICA. SHEETS or layers of clear ice have often been recorded as occur- ring in the alluvial deposits or in the sphagnum swamps of Arctic or sub-Arctic America, and most of the travelers who have made even short visits to the far north have noticed the occurrence of these ice-sheets in small escarpments on the edge of the tundra. I myself observed them in a number of places along the southern edge of the Barren Lands, in the country between the Mackenzie River and Hudson Bay, and drew attention to the fact that some of them, at least, were moving slowly down the gentle slopes on which they were lying. Since coming to the Yukon Territory I have had many opportunities of seeing and examining them in the frozen bogs which cover the bottoms of most of our valleys. As the mode of formation and growth of these ice-sheets for a long time appeared to be rather difficult to explain, the following remarks with regard to them may be of interest. ; The Klondyke gold-bearing district, to which my observations have lately been confined, and in which the deductions here set down were drawn, is a part of a great unglaciated belt or tract of country lying near the middle of the Yukon Territory in Canada, between the glaciated region which extends on both sides of the ‘“Chilcat” or Coast Range of mountains to the south and southwest, and the also glaciated region of the Ogilvie or Rocky Mountain range to the north and northeast. It is a country of high, well-rounded hills and deep, though flaring, valleys, in the bottoms of which flow streams with regularly decreasing grades. On one or both sides of these streams are everywhere deposits of alluvial material, varying from ten to a hundred feet in depth, consisting below of coarse sand and gravel, above which are fine sands with peaty and vegetable mate- rial, the uppermost layer, locally known as “muck,” usually con- sisting almost exclusively of sphagnum swamp. ‘The streams flow on beds of the coarser alluvial gravel or sand, seldom touching the underlying rocky floor, and are at present confined in relatively 232 CYRSTOSPHENES OR BURIED SHEETS OF ICE Zee shallow channels, the sides of which consist of the peaty and finer alluvial material. Ponds or lakes are conspicuously absent. The surface of the whole country, whether composed of “muck,” gravel, or rock in place, is almost everywhere permanently frozen, and while as yet comparatively few shafts have been sunk through this frozen layer, the evidence at hand would seem to show that it Fic. 1.—Ice formed by spring water in winter. has a thickness varying from forty or fifty feet on the higher, uncoy- ered parts of the hills, to two hundred feet in the moss-covered bot- toms of the valleys. Here and there, however, there are unfrozen channels in the otherwise frozen layer, through which springs issue from the sides of the hills, carrying water from the deeper saturated, and unfrozen ground through the frozen layer to the surface. Most of the known springs issue from the rock above the sur- face of the alluvial deposits in the bottoms of the valleys, but some have been exposed by mine workings beneath the ordinary surface levels. They all discharge more or less water throughout the year, 234 J. B. TYRRELL their flow being but slightly, if at all, affected by the conditions of the weather, or even by the most extreme seasonal changes of tem- perature. In summer those that issue from the rock above the alluvial deposits discharge over the surface into the nearest brooks or rivers, and, except where used as local supplies of clear cold water for household purposes, are rarely noticed; but in winter, when the thermometer occasionally falls as low as —6o0° F’., the water flowing out into the cold air freezes within a comparatively short distance, and by the close of the winter it may have formed a mass of ice many feet in thickness. ‘These ice-masses are locally known as “olaciers,’”’ and where they form along the lines of roads are often serious obstructions to travel. But, in addition to these masses of ice formed on the surface every winter, and which regularly melt away during the following sum- mer, other masses are formed beneath the surface in such positions that they are protected from the action of the sun and atmospheric agencies; and thus it is possible for them to increase from year to year to very considerable dimensions. ‘These underground masses of clear ice are also locally known in the Klondyke country as “gla- ciers,’ but the name “crystosphene” («puvatadXos “ice”; opny, wedge’’) is here suggested for them, as indicating a mass or sheet of ice developed by a wedging growth between beds of other material, while the name “‘crystocrene” («pyvs, fountain), is suggested for the surface masses of ice formed each winter by the overflow of springs. Crystosphenes are formed by springs which issue from the rock under the alluvial deposits that cover the bottoms of the valleys. As a rule, they occur as more or less horizontal sheets of clear ice, from six inches to three feet or more in thickness, lying between layers of “muck” or fine alluvium, usually where the “muck” is divided horizontally by a thin bed of silt or sand; and most of them, as far as my observation goes, are from two to four feet below the surface, though some are deeper. In area they differ greatly. One observed by the writer on the shore of Daly Lake, near the southern edge of the Barren Lands west of Hudson Bay, seemed as if it might underlie a square mile or more, while many of those in the bottom lands of the gold-bearing creeks of the Klondyke district vary in length from a hundred to a thousand feet, and in width from fifty CYRSTOSPHENES OR BURIED SHEETS -OF ICE 235 to one or two hundred feet, as shown by shafts sunk through them at various places. Speaking generally, these ice-sheets are of very even and regular thickness throughout, though they are not strictly horizontal, but approximate closely to the slope of the surface under which they lie. . For instance, the city of Dawson is built on an alluvial bottom land declining gently from the base of a steep hill to the banks of the Yukon and Klondyke Rivers, and a crystosphene which here underlies the surface at a few feet beneath it seems to have about the same slope. In another case a crystosphene was encountered on a mining claim on Hunker Creek three feet below the surface, and it was traced for five or six hundred feet down the valley, being everywhere at practically the same depth, while the surface itself had a slope of about one in a hundred, so that this apparently level sheet .of clear ice was five or six feet higher at its upper end than at its lower. Examples of this kind could be multiplied almost indefinitely, showing plainly that these ice-sheets do not partake of the character and attitude of frozen ponds or lakes. While these crystosphenes, or so-called ‘‘glaciers,” are usually of the nature of nearly horizontal sheets, occasionally they occur as veins or dikes of ice rising through the bed-rock into the overlying gravel. ‘Two such veins of ice were very well exposed in the under- ground workings on mining claim No. 39, below Discovery on Hunter Creek, where they evidently represented the former course of a spring, which had changed its point of discharge. More or less ver- tical masses of ice are also sometimes met with in the gravels them- selves, indicating the positions of former water channels from the bed-rock toward the surface. In the majority of cases crystosphenes are in the vicinity of springs that can be plainly seen issuing from the bases of the neighboring hills, but in other cases no such springs are apparent. In these latter cases, however, wherever the gravel has been removed, and the underlying rock has been exposed, springs have been found. While studying the origin of the crystosphene 600 feet long, already mentioned as occurring on Hunker Creek, no springs were apparent in the immediate vicinity, and at first it seemed as if the ice must have been formed from water flowing from a spring three or four hundred yards farther up the valley; but finally a little trickling 236 J. B. TYRRELL stream was found issuing from the rock several feet below the level of the top of the alluvial deposits. This was the source of the water that had formed the ice. The mode of formation of these underground sheets of ice is therefore somewhat as follows: Water, issuing from the rock beneath a layer of alluvial material, rises through the alluvium, and in summer spreads out on the sur- face, tending to keep it constantly wet over a considerable area. In winter, if the flow of water is large, and the surface consists of inco- herent gravel, the water will still rise to the surface, and there form a mound of ice. If, on the contrary, the flow from the spring is not large, and the ground is covered with a coherent mass of vegetable material, such as is formed by a sphagnum bog, the spring water, already at a temperature of 32° F., rises till it comes within the influence of the low temperature of the atmosphere above, and freezes. This process goes on, the ice continuing to form downward as the cold of the winter increases, until, a few feet below the surface, but still within the influence of the low external temperature, a plane of weak- ness is reached in the stratified and frozen vegetable or alluvial deposit, such planes of weakness being generally determined by the presence of thin bands of silt or fine sand. As any outlet to the top is now permanently blocked, the water is forced along this plane of weakness, and there freezes; and thus the horizontal extension of the sheet of ice is begun. While thus increasing in extent, the ice also increases in thickness by additions from beneath, until it has attained a sufficient thickness so that its bottom plane is beyond the reach of the low atmospheric temperature above; after which it continues to increase in extent, but not in thick- ness or depth. With the advent of the warm weather of summer the growth of the crystosphene ceases, but the cold spring water which continues to rise up beneath it has very little power to melt it, and its covering of moss or muck, being an excellent non-conductor of heat, protects it from the sun and wind, and prevents it from thawing and dis- appearing. Thus at the advent of another winter it is ready for still greater growth. Jo Bo AyRRELE: DAWSON, YUKON TERRITORY, CANADA, February 13, 1904. A COAL-MEASURE FOREST NEAR SOCORRO, NEW MEXICO. WE have grown so accustomed to consider that in the Rocky Mountain region the period represented in the eastern states by the deposition of coal was inimical to terrestrial life, or otherwise so different from the corresponding eastern time in its conditions that it is useless to search for a coal flora, that considerable interest attaches to any area, however restricted, in which a genuine Coal-Measure flora is present. Such an area became known to the writer a number of years ago, but it long proved impossible to study the locality in person. A small suite of fossils reported to have been collected in the fire-clay beds east of Socorro was placed in the writer’s hands by Mr. George Thwaites, and, after several fruitless attempts to identify the locality, the occurrence was reported, and, later, figures were given of four species of lepidodendrids which could be distinguished as distinct. There figures occur on Plate VII of the Bulletin oj the University oj New Mexico for 1900. No descriptions were given, and considera- ble uncertainty still existed as to the age of the formation from which they were derived. More recently the writer has not only been able to locate the place and add to the original collection-of specimens, but he has carried out a considerable amount of field work in the immediate locality which will suffice to settle the principal questions of stratigraphy so long left open. Whenever the nature of the contact between the granite and the superjacent rocks shall be studied carefully throughout the territory, many interesting points will be brought out. Whether the granite itself is a homogeneous element or represents various periods must be left an open question. We have every reason to conclude that it represents the metamorphosed sediments of an early geological time, and the writer has already reported instances where limestone beds within the granite have been found to contain what greatly resemble altered organic remains. It is also known that the gran- 237 238. C. L. HERRICK ites, in common with the immediately overlying strata, have suffered great erosion in the early part of the so-called Permian. This period of elevation, disturbance, and oscillation does not, however, seem to mark any great lost interval, so far as the record shows, while the interval represented by the granite contact is undoubtedly a consid- erable and variable one. ‘The disturbance which lifted the granites above the reach of the sea may not have been very extreme, and it is not likely that the present extreme of metamorphism was reached till long after, though it was certainly before the time in which the breccias and sandy beds of the Permian were formed, for these strata contain granite fragments in abundance. At any rate, when the sea began to return by subsidence of the granite, this encroachment was gradual, and from the south toward the north and northwest. This is proved by the fact that the expos- ures in the southeastern counties reveal vast deposits between the granite and the Carboniferous in which we have identified strata as old as the Burlington, superposed upon others yet older, but not identified as yet. In the southwestern part of New Mexico there are still earlier strata, some of which have been referred to the Devonian (Hamilton). The writer hopes to enter upon this subject more in detail in another place; at present it will be sufficient to indicate that as far north as in Socorro county the stratified rocks overlying the granite have revealed no remains indicating an age earlier than the Carboniferous, and the writer knows of no positive datum repre- senting anything older than the Coal Measures. It is true that there are reports of Subcarboniferous crinoids from the Graphic-Kelly beds lying upon the crystallines near Magdalena, but inasmuch as this 100 feet or more of crystalline limestone, which is so interestingly developed in the lead district, is not found in the locality under dis- cussion east of the Rio Grande, nor yet farther north in the Rio Grande valley, it may be said roughly that the boundary between the lower formations and the Coal Measures passes through Socorro county. Corresponding to the change just described is a still more pro- nounced geographical change in the character of the Permian and “red bed” deposits, which have an entirely different facies and suc- cession in the northwestern counties from the typical Texas sequence exhibited in the southeast. A COAL-MEASURE FOREST 239 The writer has described the Rio Grande valley as a region of early uplift—in effect an anticline, though the anticlinal axis is not exposed to view—and accordingly the dip of the strata is prevailingly to the west on the west side and to the east on the eastern side of the valley. The granite core of this anticline is largely removed, but is exposed in the bases of the Sandia, Manzana, San Andres, and Organ mountain ranges on the east side of the valley, as well as in the core of the great Sierra Ladrone uplift and the base of the Limitar Mount- ain on the west. The interval between these remaining buttresses of the great eroded mass is now, for the most part, filled with late strata composed of Tertiary (Santa Fé marls) and Pleistocene deposits, mostly nearly horizontal, except in proximity to the areas of late volcanic disturb- ance. There are also perforating necks and cones of basalt, as well as interstrata flows and dikes of the same material, but of an older period. In the American Geologist of recent date’ the writer announced the discovery of granite masses in the valley in the midst of the Tertiary beds, and gave a section extending from the Rio Grande east of Socorro through these granitic bosses. This section is so taken as to pass through the region here under discussion, though it does not cut the exact locality in which the fossil coal flora was found. These granite ribs (Fig. 1) are two in number, and are separated by an interval of perhaps half a mile, which is partly covered by the Tertiary beds, but, wherever cut by arroyos or canyons, the lower or Sandia formation of the Coal Measures is exposed in basin-like relation upon the granite, forming a shallow syncline (really a modi- fied monocline). The western rib or elongated boss of granite rises not over 100 feet above the canyon level, and presents a sharp slickensided fault to the west, with a course some twelve degrees east of north where measured. In some places the Tertiary wholly covers the boss, and to the north and south the granite passes out of sight beneath this formation, the entire length being perhaps two miles, as exposed. ‘The eastern exposure presents the overlying but unconformable Coal-Measure formation consisting of shales, silicious shale, quartzite, and ferruginous conglomerate near the contact, t “Laws of Formation of New Mexico Mountain Ranges,” May, 1904. 240 CE PEE RARNG KE followed by a large series (over 200 feet) of alternating shales, quartz- ites, and earthy or sandy limestones, the latter being quite fossilif- erous. Of the Graphic-Kelly formation which in the Magdalena Mount- ains, a few miles to the west, separates the Sandia series from the granite, there is no trace. The Coal-Measure rocks soon become horizontal, and at the east- ern side of the basin, where they abut upon the eastern boss or rib of granite, they dip to the west and he in juxtaposition to a fault exactly similar to that on the western side of the first-mentioned or western rib. In fact, the eastern rib is to all intents a repetition of the first, and the Sandia series is repeated on the eastern aspect in all essential respects as it occurs to the east of the western granite rib. Locally, however, there is a third fault which serves to repeat the granite. ‘This fault is a few hundred yards east of fault No. 2, and in the interval there are preserved a few of the Sandia strata, though in many places they are entirely removed by erosion. It is in the V formed by these two faults that the best example occurs of the fire-clay beds with the inclosed plant remains. In other localities the fire clay also occurs beyond the influence of this local fault, but these points are as yet little explored, and, in several localities visited, the granite contact has only a silicious shale in the place of the fire clay. ' The dip of the strata as they repose on the granite may be as high as 70 degrees, but more often about 45 degrees. Passing eastward, one crosses in succession a number of low hills, each capped by a bed of quartzite, the intervals being filled with shale and limestone, the superior resistance of the quartzites having thus impressed itself upon E. A COAL-MEASURE FOREST 241 the topography. When the top of the Sandia formation is reached, the topography changes, and the country rises in abrupt escarp- ments of massive to earthy limestone of the characteristic Upper ‘Coal-Measure habitus. At the base, this formation is locally very fossiliferous with the well-known Coal-Measure assemblage, a partial list of which will be given below. The first escarpment may rise some 100 to_150 feet, and is followed by others of the same character, indicating a probable aggregate thickness of 300 to 500 feet. The continuity of the section is here broken by what may be called the ‘‘Cane Spring monocline”’ or fault zone. ‘This will be more particularly described later on, and it will here be noted that this break forms in effect a rough delimitation of the Carboniferous trom the so-called Permian. ‘To the east of this the hills are mostly composed of earthy limestones, with shales of prevailingly reddish color, with gypsum beds and calcium anhydrite deposits to the base of the great escarpment forming the sky-line to the spectator from the west side of the river at Socorro. This ridge, which is the southern continuation of the Cibolo (Seboyo) Mountains, consists of from 500 to 600 feet of red beds, with gypsum and anhy- drite inclusions, capped by 150 to 175 feet of yellowish sandstone and 100 to 200 feet of gray, massive, non-fossiliferous limestone, with at least one small band of quartzite. Still to the eastward are hills containing the typical upper red-bed formations. To the south this whole series disappears beneath the Carthage coal fields of Cre- taceous age. To return to the plant beds, it will be noted that the fire clay, with its attendant shales, reposes directly upon the granite, or with a thin layer of quartzite between. The clay is of good quality, but very irregular, and has formed the base of the fire brick manufactured by the Socorro Fire Clay Co. In the clay, and especially in the shales overlying, are the remains of lepidodendrids and other coal plants referred to. A thin seam of coal is apparently present, though at the time of our inspection there was no opportunity to discover it in place. Report affirms that a boring one and one-half miles south encountered two small coal seams within 200 feet. | The thickness of the clay and attendant shales is not over 30 feet, of which half may be roughly estimated as clay. Then follows 10 242 CE EEE RICKS feet of silicious shale capped with quartzite, and 15 feet of similar shales also capped with about 5 feet of quartzite. This is all that remains of the great Sandia beds at this place, but to the east, on the other side of the third granite mass, the series is complete. This occurrence may be termed the “‘Incarnacion”’ fire clay, and the same name will apply to the granite exposure, it being the name of the mining district in which these exposures occur. There would seem to be no reason for separating the fire clay from the Sandia formation, it being but a local variation, and, even where the clay is not found, obscure plant remains attest the similarity of conditions to some extent." The Cane S pring fault area.—In this connection it may be desira- ble to call attention to this profound dislocation, the exact nature of which varies from point to point, but which is apparently found at a distance of a mile or more to the west of the main escarpment of the Cibola ridge. It lies nearly midway between the Incarnacion gran- ite ribs, with their faults, and the main range just mentioned. The typical exposure is, however, at ‘‘Ojo de las cafias,” or Cane Spring, a permanent water in the large arroyo joining the Rio Grande imme- diately south of Socorro and owing its existence to the faulted area. A section extending from the fault to the Cibola ridge is given, the data being approximate, but sufficient to give some idea of the condi- tions. : The disturbance is profound and has the appearance of being a double anticline with an interval separating the two members of some eighth of a mile. At Cane Spring, where the disturbance is most easily seen, about 700 feet of red, green, and yellow sandstone, shale, and marl are tilted up at a high angle, with the appearance of an appressed fold, the dip being to the west on the west side and to the east on the opposite side of the axis. The western limb is concealed by a covering of Tertiary, but on the east is a sharp fault, against which the nearly horizontal strata abut with little inclination. Thence eastward there are exposed by the arroyo about roo feet of red and yellow sand and marl to the second break, which is very similar to the «Since writing the above, similar plant remains (Lepidodendron sp., same as Fig. 9 below) have been found in the lower part of the Sandia series on the west side of the Ladrones Mountains, where the Sandia quartzite reposes directly upon the metamorphic rocks (schists and granites). A COAL-MEASURE FOREST to 4 Oo one just described. The section figured (Fig. 2) is a mile or more north of the spring, but has a similar double character. It is impossible to be certain that the disturbance is due to a raat ‘3 EH double appressed fold separated by faults, but fk p uy fly such is the appearance. At this locality the iy 9 : es 4 western side of the break is also exposed, and aie N ; nes ; ; ONS is a distinct fault, with much alteration and es && . : : 4, Ze excretion of quartz. A curious quartz breccia > & 2 By is a feature of the fault on the western side. [>. $k x SAe “se CK Ps > The upper Coal Measures or Permo-Carbonifer- ,7> ,°3,, ** ous appears on the west, and about half a mile c west of the fault we found fossils including Phillipsia major and other types known to be well up in the series. as Unfortunately, the paleontology of the Sandia — +s ww #97 && series is too little known at present to enable Pee P28e RN us to locate it definitely with respect to the divisions now recognized in the Middle West. We have determined in it, immediately above the fire clay beds, Productus costatus, S pirifer re cameratus, Ambocoelia planoconvexa, Seminula Ww argentea (small form), Derbia crassa (?), pa bryozoa, etc. In the lower part of the division = immediately following the Sandia beds there is a large fauna, of which the following may be mentioned: Hupachycrinus verrucosus, Reticu- laria lineatus, A mbocoelia planoconvexus, S pirijer 3 Interval gy Mive as n . 0-0 8 striatus, S pirifer cameratus, Pro- J as — ductus cora, Productus nebracen- « “3 a. . ° ° 6 SY 3 sis, P. semireticulatus, Seminula » ; ee : 8 ee nee argentea (large form), Derbia © & ©. 38 : RS awe crassa, Chonetes granulijera, © 8 ae : fs i Pas Ss FAN hs Dielasma bovidens (rare at this ‘ Nias Ae vu OK horizon, abundant higher up), Streblopieria (?). It may be noted that Eupachycrinus is known in the Kansas section from the Earlton to the Severly, Chonetes granulijera commences in the 244 CE ERR Cre Leucompton beds and extends upward, Streblopteria begins in the Earlton. So far as can be determined at this writing, the position is near the horizon indicated by Iola limestone, or within the Pottawatomie formation of Haworth. This is above the base of the Upper Coal Measures. Fic. 3.—Lepidodendron. thwaitest, sp. n. (One-half natural size.) is also common to find im- Below are given in tabulated form, for comparison, some of the more complete sections across the Permo-Carboniferous. It will be found that there is always a well- marked: change in lithological character of the rock at the horizon of the Manzano quartzite. Here is evidence of an unconformity by overlap. The lime breccias or small strata of lime with angular frag- ments of earlier lime and granite are quite characteristic of this general belt, and the Manzano beds them- selves usually contain vast quanti- ties of red granitic débris. Often vein quartz is present, pressions of tree trunks. All of these phe- nomena indicate proximity of land and a period of disturbance. The Sandia formation is easily distinguished, but it is more difficult to decide upon divisions of the Upper Coal Measures above that section. In general, there is a change of fauna, though numerous species pass unchanged throughout. Productus punctatus, Dielasma bovidens, and Meekelia are j Fic. 4.—Lepidoden- characteristic of the upper portions. The large 2%” ‘watlest, sp. n. Fusulina is found at a higher level than the common F. cylindrica. (Single bolster, en- larged.) A COAL-MEASURE FOREST 245 SECTION EAST OF SOCORRO. Massive lime - - - A it. Gray lime : = - 25- 50 ft. Shaly lime - = 3 6 Red quartzite 2 s Io Massive lime - : = 4 Gray lime - < - 65 Shales and shaly lime —- 15 Whitish to yellow sandstone 150 Top Sandia formation ft.). Red marly beds with lime, P / ion (235 ft.) etc. - - - IOO-I50 Red quartzite - Sy ee 4 Earthy lime - z 3 Io Red shale = - B TO Anhydrite - 2 Byso=) 75) ft. Quartzite - - = = Sandy shales, shales, and Shale - = = - IO nee : 2 : x6) Sandy lime (fossiliferous) - 6 Gypsum - s SS z 5 Quartzite - ie 8 Sandy shales, marls, and Shale 2 = Saat = sie) linen a a 5. itadie?. Limestone (fossilferous) 3 Red beds, as above - - Indef. Shale 5 S = 6 Lime (hard greenish, earthy) Io Limestone — - : - 4 Yellow sandstone -~ - 20 Shale pn wae 6 -Red sandstone —- 2 = 30 Quartzite- - : 5 A Silicious shales phale Seat te a Alternating shales (quartz- Limestone — - 2 - ites, green limestone, red Shale i ; : 5 0 flags) a - 300-500 Red coarse sandstone - 20 Red granitic conglomerate Green fissile shale - 2 25 quartzite. Earthy limestone - a 5 Massive limestone (fossil- Sandy shales and quartzite 50 iferous) - - - 25 Coarse quartzite - - 8 Manzano (massive) quartz- Clay and shale (Lepidoden- ite = = - - 4o- 50 dri ds) = - - = 35 Lime breccia. Contact quartzite - - Indef. Granite. Top of Upper Carboniferous formation (¢ 130). SECTION NEAR COYOTE SPRINGS, SOUTH- Massive limestones, shales 60- 75 ft. Se aa ee os Calcareous conglomerate - 5 (Permian not exposed; see below.) Limestone - - - 50 Massive limestone — - - ro ft. Sandy conglomeraie (4, 147) 5 Nodular limestone - - 3 Limestone - - - 5 Massive limestone - - 4 Silicious shale - - 2 Shaly limestone - — - 28 Earthy limestone = - 35 Granular limestone - - B Massive and shaly lime - iKele) Micaceous shales. Yellow sandstone (a, 66 ft.) 15— 20 Coyote yellow sandstone (b) 2-30 Shaly lime - - - 6 Massive limestone - 85 Massive lime - - - 3 Fusulina limestone - - 23 Shaly lime - - - 8 Shale (bryozoa) —- - 2 246 C. L. HERRICK = Massive limestone — - - 48 ft. Sandy lime - - - 17 tte Shaly limestone — - - 47 Massive limestone and San- Massive limestone - = 15 'dyshalylime - - 28 Shales (Flint Ridge beds) Be) Fusulina cylindrica beds Green quartzite - - 2 Shales. - - - - 28 Massive limestone (produc- Top of Sandia formation (1109 {t.). tus) - - - . 12 Sandy shale, with bands Nodular shaly lime — - 2 of limestone, alternating Massive limestone - - 30 with quartzite - - 45 Black shaly lime - - 48 Silicious fossil lime . 2 Massive lime - - 3 Sand and shales . - 22 Lime and lime shales - 40 Basal quartzite —- - 50 Sandstone (a) - - 2-4 Granite or gneiss. Near the locality in which the Coyote section was measured a section was taken which is more complete toward the top and serves to supplement the above section. That part of it above the Coyote sandstone horizon is here given: Manzano quartzite, red and yellow granitic sand, and conglomerate with silicified wood and granitic pebbles - - ee Se isicoatts Top of Upper Carboniferous. Earthy lime with lime breccia, etc. - - - = ae - 50 Earthy lime with green bands (bryozoa) — - - - - - - 25 Earthy cherty lime (Productus punctatus) — - - - - - - 16 Massive lime (Productus cora) - - - - - - - - 25 Black shaly lime (Dielasma bovidens) _ - ee - - - Be at] Bryozoa beds. Gray lime (Fusulina robusta?) — - - : - - - - - 6 Massive lime - - - - - - - - - - - 22 Coyote sandstone horizon. Below this point the section corresponds essentially with the sec- tion given above. ‘This corresponds toa total thickness of the Coal Measures of 750 feet, of which about 130 feet is formed by the Sandia beds. The section east of Socorro contains about 600 feet, of which 235 feet is contributed by the Sandia beds. In order to indicate the amount of variation in the same region, a section is here given corresponding with that last above recorded and taken only a short distance east of it in the Coyote district: Massive earthy lime with Bellerophon, Productus punctatus, etc. - =) gant: Allorisma and numerous fossils. | Sandstone - - - - - - - - - - - = A@ A COAL-MEASURE FOREST 247 Earthy fossiliferous lime - - - = = 2 e 3 A 17 ft. Sandy shale and granular lime - - - 2 é 2 : 2 5 Earthy limestone (Dielasma bovidens, Discina, etc.) - - = : II Calcareous sandstone - - - - - : 2 e 5 eee a Green sandy shale - - - - = : = E : Z 6 Reddish yellow sandstone (Coyote) : - : : 3 2 ik Ao Massive earthy lime. It appears that this last section falls a little short of reaching the Manzano quartzite and develops a great deal more of the sandy Fic. 5.—Lepidodendron Fic. 6.—Lepidodendron socorroense, sp. Nn. ‘thwaitest, var. striolatum, (One-half natural size.) var. n. (One-half natural size.) materials that are found farther west, though the Coyote sandstone has been observed to be 4o feet thick in one place, and entirely absent or reduced to a few inches within the distance of half a mile in con- tinuous exposure. For comparison, the following incomplete section may be given from the escarpment west of the Rio Grande in about Township 5 N., Range 4 W., of New Mexico P. M. This locality is northwest of the Ladrones Mountains. Lava flow (basalt). Red flags and shales (position uncertain) - - - - - 200-300 ft. Fault. Limestone with (Permian ?) fossils - - - - - - - 15 248 C. L. HERRICK Limestone and shale - - - - - - - L : = Oo Ite Sandstone shale with lime bands - - = : : : = Io Gray limestone - - - - - = = z = 2 ETCO White sandstone - - : - - = : 2 : z 40 Gray (gypsiferous) shales - - - - = = S : ONSET ys Yellowish sandstone - - - - : é : s s 100 Yellowish shale - - - - - : - : E = z 5 Reddish gypsiferous shale - - - a 2 : z 175 Anhydrite and limestone resting on fissile shale — - - - - - 100 Gray shales” - - - - = = - - ; Unexposed. Red flags and shales (several hundred feet). Carboniferous limestone (Coal Measures). At least 400 ft. The Carboniferous is separated from the detailed section by a valley perhaps a mile wide filled with low hills of the red beds, the dip being in all cases to the east. It is therefore very difficult to determine the thickness of the red beds, nor is it certain that there is no dislocation. There is the same difficulty in determining the extent of the fault at the top of the detailed section, but it is presumed that the throw is small, though there is Fig. es pi. an area of vertical strata intervening. dodendron socor- It is not difficult to conclude that the detailed roense. (Enlarged section here given corresponds in a general way with bolster of differ- ele that east of Socorro above the sandy shales, etc., marked as “indefinite.” In fact, it would appear that the thick anhydrite bed is a very convenient and rather con- stant bench mark, however variable it may be in thickness. References may here be made to the section given in the writer’s article on the ‘““White Sands.”* This section is from a locality com- paratively rich in fossils of a decidedly Permian habitus. PLANTS OF THE INCARNACION CLAYS. The following descriptions are given as a matter of convenience, in spite of the fact that it has not been possible to compare our speci- mens with eastern types. The forms seem to be new, and may serve to assist in calling attention to similar occurrences in the South- west. t Bulletin of the University of New Mexico, Vol. II, Fasc. 3, p. ro. A COAL-MEASURE FOREST 249 FIG. Io. Fic, 8.—Lepidodendron keyesi, sp. n. (One-half natural size.) Fic. 9.—Lepidodendron sp. ? (One-half natural size.) Fic. 10.—‘‘ Stigmarvia.’’ Root of one of the above ? CSE SEE RRNG KK bo on Oo Lepidodendron thwaitesi, sp. n. (Figs. 3 and 4.) (Cf. Bulletin of the University of New Mexico, Vol. II, Plate VII, Figs. 2 and ( ?) 4.) Trunks of large size. Bolster somewhat oval, kite-shaped to rhomboidal, about two and one-half times as long as wide, rather thickly set, but separated by high striated ridges; length of bolster, 0.9 in.; width, 0.4 in.; 1.2 in., and 0.45 in., 1.3 in. and o.5 in., in three cases; seven bolsters in the space of 4.5 in., measured obliquely along the rows in branch over 6 in. in diameter, seven in 5 in. in a larger specimen. Leaf scars (cicatrix) small, transversely oval, about one-fourth as long as entire bolster. Middle of bolster below the cicatrix marked with a ridge or depressed cauda, crossed by three very distinct frets. ‘Transpiring vents well marked on either side below the cicatrix, with which they are connected; vascular trace punctiform, often absent. Ligular scar near apex of cicatrix, triangular, often apparently notching apex of cicatrix. An escutcheon-like impression between transpiratory vents. Leaves sharply acuminate with slen- der tips, midrib strong, 3 in. wide and perhaps 8 in. long. Lepidodendron thwaitesi, var. striolatum, var.n. (Fig. 5.) Greatly resembling L. thwaitesi, and perhaps a variety of that species, but represented by smaller specimens in which the preserva- tion is not very perfect. Bolsters uniformly rhomboidal oval, surface flat, about five in the space of 2.4 in. measured diagonally in the rows. Length, 0.72 in., width, 0.35 in. Space below the cicatrix marked with numerous (5~7) irregular frets. Cicatrix appar- ently as in L. thwaitest. Lepidodendron socorroense, sp. n. (Figs. 6 and 7.) (Bulletin of the University of New Mexico, Vol. II, Plate VII, Fig. 1 ?) Trunks of moderate size, leaves slender. Bolsters rhomboidal to rhombic oval, acuminate at the ends, nearly symmetrical; lower portion, from the leaf scar downward, kite-shaped to rhomboidal, upper margin curved, the long axis marked by a well-defined keel, very prominent near the lower angle, where is an elevation; folar cicatrix about one-fourth the entire length of bolster; transpiratory append- ages obscure; vascular pore on the most prominent part of cicatrix; a deep pit below the cicatrix on median line; ligule large, well marked. Bolsters closely approximate, in well-defined oblique series; 0.7 in. long, 0.3 in. wide. In a trunk 3.2 in. in diameter g bolsters in space of 3.5 in., measured diagonally in the rows; in larger trunks g bolsters in 4.3 in. This species is well represented, and is variable in size, but the specimens leave much to be desired in detailed structure of the leaf scars. Lepidodendron keyesi, sp.n. (Fig. 8.) Trunk of large size. Bolsters large, obovate, flat, closely approximate, in well-defined diagonal rows. Eight bolsters in space of 4 in., measured obliquely A COAL-MEASURE FOREST 251 in the rows. Length of bolster, 0.75-0.8 in.; width, 0.5 in. Cicatrix nearly reaching middle of length of bolster: ligule very large, often causing the upper part of cicatrix to appear emarginate. ‘Transpiratory appendages large, oval, discrete. A groove extends from cicatrix to the lower angle of the bolster, and there are traces of frets. State of preservation imperfect. Named in honor of Dr. Charles R. Keyes, of the New Mexico School of Mines. Lepidodendron, sp. ?. (Fig. 9.) (Bulletin of the University of New Mexico, Vol. II, Plate VII, Fig. 3.) In form of bolsters this form is like L. brittst or L. aculeatum. All our specimens seem to be decorticated, so that it seems unsafe to attempt a description. It is even possible that these specimens belong to one of the above species, although in appearance they differ widely. : There are five bolsters in a space of 3.7 in., measured diagonally in the rows. Length of bolster, 1.6 in.; width, o.4 in. Vascular scar above the middle. Sur- face of (decorticated) specimen striolate. Coy SaERRICK: Socorro. N. M. THE VARIATIONS OF GLACIERS. IX." THE International Committee on Glaciers met in Vienna last summer, and the retiring president, Professor S. Finsterwalder, presented a report to the International Congress of Geologists, of which the following is a brief summary :? Although the committee has been at work only nine years, a time too short for very general results, still we can say that the thirty- five years’ period which Briickner found in the variations of the Alpine glaciers applies also to glaciers in other parts of the world. It is also probable that there are longer climatic periods than this whose course is very complicated; moreover, the individual character of the variations of special glaciers is, without doubt, dependent upon the topography of their basins. We therefore have to do with variations of a very complicated character, as they depend both upon climatic changes and upon individual characteristics. But we can say, in general, that the dominating tendency of glaciers at the present time is to retreat. There are many exceptions to this rule, the reasons for which are not understood. The Mont Blanc group showed some tendency to advance in the eighties, and within the last twenty years this advance has been transferred to the most easterly Alps, though not advancing regularly over the intervening regions. The Vernagtferner presents the most remarkable phenomena. This glacier increased 4oo™ in length between 1897 and 1902, which is not so extraordinary for it; but, in addition, the velocity of flow in a certain profile near the end increased from 17™ to over 250™ per year, and then suddenly within one year sank back to 80". The Vernagtferner is unique in other ways, and the careful study of its characteristics will add greatly to our knowledge of glaciers. Professors Forel and Richter have offered explanations of the * The earlier reports appeared in this JoURNAL, Vol. III, pp. 278-88; Vol. V, pp. 378-83; Vol. VI- pp: 473-70; Vol. Valo pp257—2550v Ol. VL spp set54=5 05 Ol exe pp. 250-54; Vol. X, pp. 313-17; Vol. XI, pp. 285-88. ? Comptes rendus du Congrés géologique international de Vienne 1903, pp. 161-69. 252 THE VARIATIONS OF GLACIERS 253 apparent irregularities of glacier variations. The increase in snow- fall in the reservoir results in a thickening of the upper part of the glacier, which thickening then progresses down the glacier, some- what like a wave, more rapidly than the ice itself. The ice thus advances faster than it melts, and the glacier is pushed farther down its valley.. When the supply in the reservoir diminishes, the pres- sure from behind also diminishes, and the ice moves less rapidly and melts back. Professor Finsterwalder then takes up the study of an ideal gla- cier, subject to periodic variations of thickness at the névé line, and subject to uniform melting along its whole length; and finds a mathematical equation which represents the changes in thickness and length which the glacier undergoes. Although these condi- tions are not exactly those which a glacier experiences, still they are near enough to give results which correspond fairly well with what is actually observed. The solution of the equation shows that the glacier in advancing takes a steep slope in front and moves with considerable velocity, and during the retreat takes a very gentle slope in front and its velocity is much diminished. The variation in thickness at the névé line produces a much greater relative varia- tion in the length of the glacier, and the change at the end occurs later than the change at the névé line. Under these conditions the glacier advances slowly and retreats rapidly. This is in contradic- tion to experience, as glaciers usually advance rapidly and retreat slowly. This contradiction probably comes from the fact that we have assumed a uniform variation at the névé line, which, however, is probably not at all regular, but makes a rapid increase and a slow diminution. If now we introduce a variable rate of melting, and assume that it has the opposite phase to the variation of thickness at the névé line, we find the variation in length increased and the state of minimum existing for a longer time, which Forel has described as characteristic of glaciers. It thus appears that the Forel-Richter theory is, in general, upheld by mathematical analysis; but there are many peculiarities which are still unexplained. In the obser- vations of the Vernagtferner glacier very great variations of velocity were found without corresponding increases in thickness. Lately Hess has shown that ice yields more rapidly to a given 254 HARRY FIELDING REID force as the time of application of the force increases. This might explain some of the difficulties, but many difficult questions of glacier physics arise in this connection. The International Committee serves as a natural point of union for all investigators of glacier phenomena, and is doing good work in encouraging glacier studies." The following is a summary of the eighth annual report of the International Committee on glaciers: REPORT OF GLACIERS FOR 1902. Swiss Alps.—Of the ninety-five glaciers which are being observed, seventy-eight were measured in 1902. The great majority of them are in a state of recession, and it is probable that the 680 other Swiss glaciers are receding also. ‘The recession is therefore general, though there is a slight tendency this year to advance, shown especially among the glaciers in the southwest Bernese Alps. The little glacier of Boveyre, which has been advancing for ten years as a result of a great increase in material due to an avalanche, has begun to retreat.3 Eastern Alps.—On the south side of the Ortler, two glaciers have retreated 3.5-11™; one has probably made a slight advance. In the Silvretta group three glaciers show a retreat of 4oo—500™ since the last maximum in 1850-60. In the Oetzthal the Vernagtferner, which last year showed an advance of 50™ and a remarkable increase m in velocity of from 210™ to 240™, has suddenly decreased its speed to not more than one-third of its highest value; it has, however, advanced 20™, and has swollen in its lower part. Its neighbor, the Guslarferner, has been stationary for many years. * Professor Harry Fielding Reid, of Baltimore, was elected president of the com- mittee for the ensuing three years, and M. E. Muret, of Lausanne, was re-elected as secretary. Professor Nathorst retired from the committee and was elected a corre- sponding member. Baron G. de Geer was elected to succeed him as representing the Arctic regions. Colonel J. von Schokalsky had already been elected to succeed the late Professor J. Mouschketow, who took such an active part in the work of the com- mittee. ‘The following additional corresponding members were elected: Professor Dr. A. Bliimcke, of Nuremberg;; Professor Dr. Hans Hess, of Ansbach; Professor Dr, A. Penck, of Vienna; and Mr. George Vaux, of Philadelphia. * Archives des sciences physiques et naturelles, Vol. XV, pp. 661-77; Vol. XVI, pp. 86-104. 3 Report of Professor Forel and M. Muret. THE VARIATIONS OF GLACIERS 255 Borings have been made in the Hintereisferner to determine its thickness. It has been’ completely pierced at a distance of 1,860™ from the end, where a thickness of 152.8" was found. This glacier has retreated 94™ in the last eight years, and its velocity of flow has been reduced 25-30 per cent. Hochjochferner, near by, has retreated only 1™, but has become somewhat thinner. Diemferner deserves special attention; it has advanced 30™ in the last two years, and 144™ since 1893. It has suffered a considerable change in form, and in the upper regions its sides very nearly reach the top of the old moraines. In the same general neighborhood seven glaciers have retreated 6-20™ since 1891; and six others have retreated 21-9o0™ since 1899; two of these were advancing a few years ago. One glacier observed in the Stubai group is in continued retreat. In the Zillerthal the Schwarzenstein glacier is retreating, but more slowly than last year. The Horn Glacier has ceased to advance and has retreated 4™, while the Waxegg has advanced 14™ on the average, and at one point 38™. In the Venediger region eight glaciers show recessions of from 2-17". The Krimmlerkees, which shows the greatest retreat this year, was advancing last year. In the Glockner group two glaciers have retreated 7~11.5™, while the Pasterze is stationary. In the Sonnblick group three glaciers show recessions of 5—20™ in two years. The Krummelkees, which advanced 7™ from 1899 to t901, has been stationary since then. In the Ankogl group two glaciers are retreating, but only half so fast as last year. On the other hand, the great Elendkees has changed its slight decrease of last year to a slight increase this year. On the whole, the retreat is more rapid this year than in pre- ceding years; nevertheless, a few glaciers are still advancing. We must mention that the Gepatschferner has been steadily retreat- ing since 1886, when exact measures began.* Italian Alps.—The Marmolada Glacier has retreated very con- siderably during the last forty years, though we cannot say what it is doing at the moment. The upper regions show signs of dimin- ishing, but the growth before 1883 was made. evident by the ice * Report of Professor Finsterwalder. 256 HARRY FIELDING REID closing up a grotto used as a sleeping-place by the Italian Alpine Club.. This grotto remained closed from 1884 to 1900. The Cris- tal, Sorapiss, and Kellerwand Glaciers have all retreated from a fraction of a meter to 3™. The snow-fields of Monte Cavallo show a considerable increase in size. The Lys Glacier, Monte Rosa, has retreated about 25™, shows a marked change in its form, and reveals newly deposited moraines. In general, the retreat of the Italian glaciers has continued, but the snowfall has increased.* French Alps.—Many glaciers have been’.observed under the direction of the French Committee on Glaciers, with the following results: On Mont Blanc the Bossons has greatly diminished; and the Mer de Glace, stationary for some time, now shows a marked recession. In the Maurienne eleven glaciers are retreating; one has been stationary since 1892. ‘The Glacier des Sources de |’Arc has retreated 1,250™ horizontally and 300™ in altitude since 1873. In the Grandes Rousses the Glacier des Quirhes has retreated about 25™ since 1899, and the Grand Sablat about 35™. The Glacier de la- Selle has retreated 600-800™ in the last thirty years. The Glacier des Etan- cons has retreated greatly, probably 1oo™ in the last fifteen years. Of its two tributaries which are now separated, one is advancing and the other is retreating. The Pvrenees.—The advance of the end of the nineteenth cen- tury has affected the Glacier de Vignemale, which has inrceased notably in thickness, though it has not advanced. The glaciers, in general, are distinctly retreating. Of the twenty observed, fifteen are retreating, and the others are either stationary or possibly growing, although none show any real advance from rgo1 to 1902.* Scandinavian Alps: Norway.—The summer of I901 was par- ticularly warm in Norway, so that the glaciers melted rapidly, and the snow-fields diminished to an extent never before seen. Many glaciers were retreating, and the glacial streams were much higher than usual. On the other hand, during the summer of 1902 the snow remained very late. Several glaciers of Galdhotind advanced 15-20™, and several receded, perhaps as much. On August 11, the glacier lake of Mjélkedalsvand was suddenly emptied and caused an inundation. Two glaciers near Olden are retreating. * Report of Professor Porro. ? Report of Professor Kilian. THE VARIATIONS OF GLACIERS 2 In the neighborhood of Folgefon several glaciers were retreating rapidly in the summer of 1901, whereas in 1902 they were advancing. Sweden.—The Mika Glacier has advanced 5™ since 1901, and the Solta 20™ since 1900. The Skuova has apparently been station- ary since 1897." Polar regions.—An expedition was sent to Greenland to study the inland ice between latitudes 68° 30’ and 69° 20’. Surveys of the Jakobshavn Fjord and the border of the inland ice to the south were made on a large scale. Photographs were also taken from marked points which may be used for future comparison. All the glaciers of the Jakobshavn Fjord are notably retreating; the rocks for 5.5™ above the present surface of the large glacier are entirely free of lichens, and the tongue of the glacier is 4*™ shorter than in 1883. The small glaciers flowing from the tributary fjords of the Jakobshavn show a similar, but smaller, retreat. Farther south, near Orpigsuit, the edge of the inland ice seems to be retreating, as the rock immediately above it is free of lichens and is covered with fresh striz. Photographs of the nunataks were taken which will serve to determine future variations. The Swedish expedition to the South Pole visited Royal Bay in the Island of South Georgia. Ross Glacier, which had retreated, according to the German South Polar Expedition, 800-go0o™ between 1882 and 1883, has since then advanced to the point where it stood 1M TOS25- Himalaya.—The Taschiny Glacier in Kashmir was retreating in 1875 and advancing in 1886. Some small glaciers in the Panjal Range were retreating before 1884 and advancing slightly some years later. In the Nun Kun the glaciers were advancing in 1902. In the Karakorum Range several glaciers in Schigar Valley advanced for eight or ten years before 1895. In the Saser-Nubra Mountains a slight advance took place in 1896.3 Caucasus.—The four glaciers on Mount Kazbek which have been retreating for some time have become stationary, and great accumu- lations of snow seem to indicate the beginning of a new advance. The summit of Kazbek, which in 1900 was almost free of snow, is now covered to a considerable depth. In the valley of the Guisel- * Report of Dr. Oyen. ? Report of Dr. Steenstrup. 3 Report of M. Chas. Rabot. 258 HARRY FIELDING REID Don the Djimara Glacier is still retreating, as are also two glaciers in the valley of the Ouroukh. The Mayl Glacier, on the northern slope of Mount Kazbek, was the scene of two terrible outbreaks which destroyed the baths of Kermadon in July, 1902. T’wo ava- lanches, originating in seven large snow-fields, came down a lateral gorge, and then passed over the surface of the glacier, following a course six miles long. The slopes of the mountains are not steep enough to cause this catastrophe, and it is probable that it was induced by an earthquake. Siberia—A number of small glaciers exist near the sources of the Oka River in the mountains southwest of Lake Baikal. Some of these were described by M. Radde in 1885. They are at present much smaller than they were then, and one of them seems to be on the point of disappearing. The Alatau or Kuznezk Mountains, 200 or 300 miles north of the Altai, do not contain any glaciers at present, but they show traces of former glaciation in the smoothed rocks and large moraines. In the Alatau of Sungaria there are many glaciers; those on the northern slope being generally larger than those on the southern. Two of the former have been sur- veyed. The mean height of the peaks in this portion of the range is about 13,000 feet. There are many glaciers in the Tyan Shan mountains which are rarely visited. The peaks of this chain have altitudes of from 16,500 to 17,500 feet; a new determination of the height of Mount Khantengri makes it 22,600 feet. Some of these glaciers are in a marked state of retreat; others do not indicate any definite variations. It seems in general that, if the glaciers are decreasing, there must be shorter intervening periods of growth, at least for some of the glaciers.’ REPORT ON THE GLACIERS OF THE UNITED STATES FOR 1903.” In May, 1903, the Muir Glacier was visited for the first time since the earthquake of 1899. Mr. C. L. Andrews, deputy collec- tor of customs at Skagway, and Mr. Case went from Skagway to the Muir Glacier in an open boat, photographed the end of the ice, * Report of M. Schokalsky. 2 A synopsis of this report will appear in the Ninth Annual Report of the Inter- national Committee. The report on the glaciers of the United States for 1902 was given in this JOURNAL, Vol. XI, pp. 287, 288. THE VARIATIONS OF GLACIERS 259 and showed by a map the changes which have taken place there. They found that the ice-front had retreated to a distance of from 3 to 34 miles from its former position, and almost the whole inlet was covered with floating ice very closely packed together. The ice-front now passes from the base of Mount Case northwesterly to the two small nunataks opposite, which have become united into one by the lowering of the ice. It then passes southwesterly to the corner of the large nunatak which separates Morse Glacier from the Muir. An area of 42 square miles has thus been taken from the glacier and added to the inlet. An area of 9 square miles of the inlet is closely covered with floating ice. The amount of ice which has been broken off from the glacier—if we assume an average thickness of 700 feet, which is probably not far wrong—amounts to about 91,000 million cubic feet. This is about fourteen times the amount formerly discharged annually into the inlet. The ice forms a terrace along the eastern side of the mountains, and Dirt Glacier ends as an independent tide-water glacier. Morse Glacier, which had already become an alpine glacier, separated from the Muir, no longer has its valley closed up by the latter’s ice.’ The new ice-front is composed of two parts. ‘The eastern part from Mount Case to the nunatak consists of ice which is practically stationary, whereas the western part receives all the active flow of the glacier. This portion does not differ materially in breadth from the old ice-front, and receives practically all the ice which was for- merly discharged into the inlet. It stands up as a vertical wall prob- ably about 200 feet above the water. In 1890 the surface of the ice where the glacier now ends was 500-600 feet above sea-level, so that this surface has been lowered 300-400 feet. The fact that the ice stands up as a vertical wall makes it probable that the water is fairly deep at this point, though probably not as deep as at the old ice-front. If this is so, the velocity of the ice near the present end is probably a little greater than near the old end, as the section is some- what less. This causes a tendency to advance, but the position of the end will probably remain almost stationary for some time; for if it advances materially beyond its present position, it will find no t “Studies of Muir Glacier,” National Geographic Magazine, Vol. IV (1892), p. 51- 2C. L. ANDREWS, “Muir Glacier,” zbid., Vol. XIV (1893), pp. 441-45. 260 HARRY FIELDING REID support on the east; it will broaden out, and will offer a longer line for breakage. If, on the other hand, there should be any material retreat, the ice-front would again become longer, resulting in still more rapid retreat, until, perhaps, the glacier withdraws above tide- level. There are eight glaciers of considerable size within easy reach of Skagway. These glaciers have been under observation by Mr. Andrews, who expects to continue to observe them. They are all retreating rapidly. Denver Glacier melted back 40 feet in two months in the summer of 1903. The “S” Glacier and the Upper Glacier have been retreating at the rate of 30 or 4o feet a year since 1898 (Andrews). The Mendenhall Glacier near Juneau retreated at the rate of 40 or 50 feet annually between 1892 and 1901. Immediately bor- dering its sides and end the ground is free of vegetation, but the shrubs and trees gradually increase in extent and size as we go far- ther from the glacier, either down its valley or up the mountain side. This increase in the age of the trees indicates the rate at which the glacier has retreated. It is very remarkable how rapidly trees have grown in this region, attaining a thickness of nine inches in twenty-five years and of nearly two feet in one hundred years’ Mr. Fernow has also noted the remarkably rapid growth of trees at the entrance of Glacier Bay, where trees only forty or fifty years old were 36 inches in diameter and 80 feet high.’ Throughout Oregon and Washington the last three years have been marked by excessive precipitations, and the snowfall of last year seemed to be the greatest of the three; but there is no evidence that this has yet resulted in the advance of the glaciers. Mount Baker is an interesting mountain, but it has received very little attention. It was ascended last summer by Mr. C. E. Rusk, who writes me that there are about ten glaciers on the mountain. ‘Two _of these, which he had the opportunity to examine, showed signs of marked retreat in recent years. The Eliot Glacier on Mount Hood has retreated slightly since *MARSDEN MAnson, “Forest Advance over Glaciated Areas in Alaska and British Columbia,” Forestry Quarterly, Vol. I (1903), pp. 94-96. 2 See Harriman Expedition, Vol. Il, pp. 249-52. THE VARIATIONS OF GLACIERS 261 tgo1. (Langille.) ‘The glaciers on the south side of Mount Hood also show evidence of retreat during the same interval. (Mont- gomery.) There are two or three glaciers on Mount Jefferson, but we have no evidence as to their variations. The peaks of the Three Sisters surround a large amphitheater five miles wide, which opens toward the east and formerly held a large glacier, but this has shrunk so much that it has now broken up into four small glaciers. There are three others on the outer slopes of the mountain. The moraines of the glaciers in the amphitheater stand up 30 or 4o feet above their surface. They are still fresh and free of vegetation, showing that this diminution of the ice has taken place in comparatively recent years. The reduction in thickness of the ice seems to be more marked than the reduction in length. Ice still remains under the moraines, which is a further indication of the short time since they were formed. The double crests which some.of these moraines present are ascribed to the melting of the ice under them. One of these glaciers showed in its Bergschrund projecting layers of ice separated by layers of dirt similar, on a small scale, to the projecting layers of ice found at the end of Greenland glaciers.' In the present instance, however, these projections are not ascribed to shear, but to differential melting; for where the snow is shaded from the sun the projections do not exist. (Russell.) Comparison of photographs of Lyell Glacier in California taken in 1883 by I. C. Russell and in 1903 by G. K. Gilbert show only very slight recession, whereas the McClure Glacier, close by, has suffered a marked retreat during the same interval. (Gilbert.) This difference may be due in part to the shapes of the two glaciers, Lyell being much broader than it is long, whereas the McClure presents a definite tongue. Professor LeConte has found a new glacier just below the east- ern precipice of Mount Jordan, in northern California, and thinks that there are other small glaciers along the eastern slope of the Sierras in this neighborhood. There has been less snow in the Sierras *I. C. Russet, “Glacier Cornices,’ JOURNAL OF GEOLOGY, Vol. XI (1903), pp- 783-85. 262 HARRY FIELDING REID of California this year than for many years past, and probably all the glaciers are retreating. (LeConie.) The Chaney and Sperry Glaciers in Montana show a marked retreat. The former, though a small glacier, has retreated 200 yards or more in the last eight years. (Chaney.) 3 The snowfall on the Arapahoe Glacier in Colorado was unusually small in 1902. In 1903, however, it was unusually large, and seems to have produced a noticeable effect on this little glacier. The ice is somewhat thicker and the front slope of the glacier steeper, but there is no apparent change in length, except at two points where streams have effected a slight recession. From September, 1901-2, the precipitation was below normal and the temperature above nor- mal at the Weather Bureau stations nearest to this glacier, whereas it was just the reverse from 1902-3. Silt was found on the moraines similar to that found last year, and as it is impossible to sup- pose that the glacier has advanced over the moraine and retreated again within a year, the former explanation of this silt, which required rather violent fluctuations of the glacier, must be abandoned. It is probable that the silt is due to dust blown from the mountains upon the snow and left on the moraine when the snow melted. ‘This is a more satisfactory explanation, but it shows nothing with regard to the glacier changes.* HARRY FIELDING REID. GEOLOGICAL LABORATORY, John Hopkins University, March 17, 1904. Nore.—Since the above report was written, the third volume of the Harriman Alaska Expedition, on Glaciers and Glaciation, has appeared. It is written by Mr. G. K. Gilbert, and is a valuable contribution to our knowledge of the Alaskan coast glaciers. Mr. Gilbert collates all information regarding the variations of these glaciers up to 1899, and adds the observations made by himself and by other members of the expedition. The positions of the ends of many glaciers are shown by pictures and maps. The glaciers discussed are too numerous to be named here, but we must mention the general fact that the glaciers of Glacier Bay and of Disenchantment Bay show very great recessions during the last hundred ‘Junius HENDERSON, “Arapahoe Glacier in 1903,’”? JOURNAL OF GEOLOGY, Vol. XII (1904), pp. 30-33. THE VARIATIONS OF GLACIERS 263 years, whereas those of the southwestern slopes of the Fairweather Range and those of Prince William Sound have quite lately been as extensive as they have probably been for several centuries. Mr. Gilbert makes some general sugges- tions of a theory to explain such curious anomalies. He also presents clearly the very striking evidence which the topography of the Alaskan coast offers in favor of the power of glaciers to erode their channels. The volume concludes with a theoretical discussion of the influence of the sea on the pressure which tide-water glaciers exert on their beds below sea-level. FAREWELL LECTURE BY PROFESSOR EDUARD SUESS ON RESIGNING HIS PROFESSORSHIP:? In the last lecture we occupied ourselves with the structure of South America. We saw that the earlier volcanic occurrences are restricted entirely to the Cordillera of the Andes, but that in the course of their appearance there are long interruptions. We have therefore arrived at the close of our hasty survey of the earth’s entire surface, and today we will review the events which have been set forth during the last two semesters. The present lecture, moreover, also closes my active life as a professor, and I stand at the end of a career of teaching at this university, which I have been permitted to enjoy for eighty-eight semesters. Before I take up the short summary mentioned, I believe it suitable to say a few words in regard to the changes which our science has under- gone during this long period. My collegiate work as lecturer on general paleontology was begun October 7, 1857—two years before the appearance of Darwin’s book, The Origin oj Species. It is well known that in the eighteenth century prominent thinkers, as Leibnitz, Herder, and others, properly recognized the connection and unity of all organic life. But, at the beginning of the nineteenth century, Cuvier, essentially by means of the fossils of the chalk of Montmartre, was able to present the surprising evidence that there had lived on the earth genera of animals which today are wholly extinct, and that similar changes have again and again occurred in the animal kingdom. He thus concluded that there had been repeated revolutions. In this he was followed by the great majority of inquir- ers, and at that time—the year 1857—everyone was completely under the influence of Cuvier’s views. Personally, a paper by Edward Forbes, on the influence of the glacial period on migrations, had a great effect on me; the article merits reading even to this day. ™ Given July 13, 1901, in the Geological Lecture Hall, Vienna University; taken stenographically by Mr. H. Beck. For the original lecture, in German, see Mitth. Pal. u. Geol. Inst. , Universitat Wien, 1902, pp. 1-8. Translator, CHARLES SCHUCHERT. 264 FAREWELL LECTURE : 265 After the appearance of Darwin’s book, there occurred a great and general change of view in all branches of biology. In fact, outside the great discoveries of Copernicus and Galileo, there can- not be cited another example having so deep an influence on the general opinions of naturalists. Darwin was not the first to con- ceive and pronounce upon the unity of all life; but that he was able to produce stronger proofs and to direct the trend of thought con- stitutes his undying fame. In the field of paleontology the consummation of this change did not, of course, go on so simply and, at least with us, so entirely in accordance with the views of Darwin as one is apt to imagine. The Darwinian theory of the variability of species was essentially based on selection and related appearances. Paleontology, however, teaches otherwise. It teaches that the terminology for single divisions of the stratified terranes, characterized by their fossil remains, finds appli- cation over the entire earth. Therefore from time to time there must have occurred, in some way, general changes affecting the entire physical condition of the world. Nor is there seen a per- petual and continuous changing of organic beings, as would be the case through the constant influence of selection. On the contrary, there are entire groups of animals appearing ‘and disappearing. Darwin sought to explain this by means of gaps in our knowledge, but today it is known that these supposed gaps possess too great a horizontal extension. There now arises the thought that the changes in the outer con- ditions of life have a controlling influence. I may here state that on this question there was some correspondence between Darwin and our widely mourned Neumayr, and that Darwin in no wise took a dissenting stand against the objections. In this connection it is most remarkable that the great and general knowledge of pale- ontology which I have just indicated, should apparently have made upon so great a mind as Darwin’s less of an impression than those small lines of variation noticed in certain fossil fresh-water snails, as, for example, in Valvata or Paludina. Here and there conditions are combined which permit somewhat closer analyses of the relations of this subject. This for instance, is the case in the superposition of the Tertiary land faunas of Europe, 266 PROFESSOR EDUARD SUESS more particularly of Vienna. Here one recognizes the following: Living beings are dependent, on the one hand, on certain outer physical circumstances, as climate, moisture, etc.; on the other hand, they also are mutually and socially dependent upon one another. Every living province—or, as it is usually expressed, every zoélogi- cal province—forms, as it were, an economical unity in which for so many flesh-eaters there must be so many plant-feeding food animals; for so many plant-feeders, so many food plants; honey- sucking Lepidoptera presuppose flowers; for insect-feeding song- birds a certain number of small insects are necessary, etc. The disturbance of one member of this unity can possibly destroy the balance of the whole. According to all appearances, such disturbances have occurred from time to time in land faunas, and they may have been of very diverse kinds. Then again an entire fauna is seen to vanish over all Europe, or over a still greater region, and a new fauna comes in to take its place. This new fauna nevertheless always has a more or less strongly vicarious relationship to its predecessor; it is clearly a variation of the former, probably in the main a resulting adap- tation to changed conditions; and even if the sequence of strata were completely unknown, one could readily discern which was the first, the second, or the third fauna. Besides this, the numerous phylogenetic lines which unite nearly all the great groups of fossil animals; or the unity in the developmental nature of single organs, as the extremities; or the general super- position of gills and lungs; or the rows of striking harmonies that exist between the development of certain groups of animals, and of single individuals of these groups—all indicate with certainty the correctness of the Darwinian basal idea, namely, the unity of life. Stratigraphic geology and paleontology show that the evolution of organic life was probably never completely interrupted, but that it did not go on ina uniform manner. Disturbances have occurred. The struggle for existence continues; yet it is only of secondary importance. Single very old types, as Hatteria (Sphenodon), have continued to maintain themselves to our day with but slight changes. Allow me now to speak of a few tectonic questions. When I began my collegiate work, there prevailed, especially in FAREWELL ADDRESS 267 Germany, the idea that mountain chains were built symmetrically; one group of oldest rocks formed the lifted or lifting axis, and upon each side were arranged younger rocks in parallel zones. Thus, you will still find in my own writing on the substructure of Vienna, in the year 1862, a presentation of the Alps as symmetrical mountains. Of course, this idea did not prevail without objections. At nearly every gathering of German naturalists at that time, the old Bergrath Diicker arose to protest against it. No one listened to him. With Schimper it was the same. The authority of Leopold von Buch, who expressed himself for symmetrical construction, remained unshaken. Then Leopold von Buch died. Upon this primary question of modern geology you will find no explanation for the origin of mountains in the leading text-books of that time, as, for instance, Lyell’s justly celebrated Principles oj Geology. For the investigation of this problem no part of Europe was more advantageously situated than Austria. There the land is arrayed before us in unusual variety. Hardly anywhere in Europe are tec- tonic contrasts so plainly presented—contrasts between the Bohemian Mass and the Alps, between the portion of Russian table-land beneath the Galician plain and the Carpathians, the peculiar connection of Alps and Carpathians, the continuance of the Turkestan depression over the Aral Sea into the depression of the Danube and to Vienna, and much besides.. In the year 1857 the idea was still often main- tained that the deposits found in the eastern Alps did not occur at all outside of the Alps, so great were the difficulties which the applica- tion of the accepted stratigraphic divisions of England and south Germany bore to the strange occurrences in the Alps themselves. Soon, however, it was recognized that in the Bohemian Mass the stratigraphic sequence was far less complete than in the adjoining regions of the Alps, and that in Bohemia particularly there is an extraor- dinary interruption of marine deposits extending upward into the Middle Cretaceous, whereas in the Alps all these great epochs are represented by marine strata. This same transgression of the Middle and Upper Cretaceous shows again in Galicia, then far into Russia, on the other side of the French Central Plateau, on the Spanish Meseta, in large parts of the Sahara, in the valley of the Mississippi, and northward over this region to the vicinity of the Arctic Sea, in 268 PROFESSOR EDUARD SUESS Brazil, finally on the shores of central and southern Africa, in east India; and, in fact, over such extraordinarily vast regions that it became impossible longer to explain such transgressions of the sea, according to the older views of Lyell, by means of the elevation and depression of continents. Through this and similar observations the newer idea has recently come into prominence that some general change must have occurred either in the shape of the hydrosphere or in its entire volume. It was seen that by the forming of a new oceanic depth, due to sinking, a certain amount of the hydrosphere was drawn off into the new depres- sion, and that at the same time there appeared to be a general land elevation, or, more correctly, there must have resulted a general sinking of the beach lines. The older view of the numerous oscilla- tions of the continents has also given way more and more to the teachings of marine transgressions, and through the denudation of continents, a more. exact examination into the actual mountain move- ments has become possible. If one were to assert that the Alps are folded, but that the Bohe- mian Mass is not, and that because of this there has resulted a dam- ming up, then this assertion would not be exact. The Bohemian Mass is also folded, and there is at present no known portion of the earth’s surface of which at least the archaic base is not folded. The difference, however, consists in this, that the folding has ended early at certain places; at others it has continued into a later or very late time, and possibly has also continued with a change in the ground- plan. In this respect central Europe shows a quite peculiar arrangement. The oldest folding is seen in the gneiss of the western Hebrides. Younger and of pre-Devonic age are the folds of the Caledonians, which can be traced down to Ireland. On these, farther south, are ranged the Armoricanian and Varischian folds, which embrace south- western England, Normandy and Brittany, the Central Plateau, the mountains of the Rhine, and the Bohemian Mass, inclusive of the Sudetes. Its principal folding was accomplished before the close of Carboniferous time, but minor movements of various kinds have followed. The Alps and Carpathians even underwent decided folding In the Miocene. Each part has moved northward toward the pre- FAREWELL LECTURE 269 ceding, or toward the horsts, in which the earlier member was dis- solved by sinking, and thus Europe has resulted through a succession of younger and younger folds. Meanwhile, more and more light came regarding the strange development which certain Mesozoic deposits, particularly the Trias- sic of the Alps, show when compared to the north-lying lands, as Wiirtemberg or Franconia. The observations in Asiatic highlands, especially in the Himalayas, taught that this type of Triassic develop- ment has a very wide distribution toward the east; and it even became possible to prove that directly across present Asia, from the existing European Mediterranean to the Sunda Islands, there once extended a continuous sea. This sea has, as you know, received the name Tethys. The old continent along its southern side was named Gondwana Land, and that on its northern side, Angara Land. The present Mediterranean is a remnant of Tethys. This Mediterranean, however, consists of a series of areas of diverse construction, and we have had opportunity to convince our- selves that, since Middle Tertiary time, first a portion was separated, as, for instance, the Danube plain, then a portion was added, as the Aigean Sea. The progress of geological research during the last ten years, however, has been so extremely great that a far more extensive knowl- edge of the seas has become possible. ‘They are of different kinds. We examine a world-map, and thereby, in accordance with oft- repeated warning, seek to guard against the deception which the dis- tortion of Mercator’s projection so easily produces. We see that, with the exception of the two Chinese rivers, Yang-tse-kiang and Hoang-ho, hardly another great stream finds its way to the Pacific Ocean. All waters of the continents flow toward the Atlantic or Indian Ocean. Many years ago the Russian General von Tillo drew on a little map the watershed of the earth, and showed how surprisingly small an amount of fresh water the Pacific receives. These two oceanic areas differ also in a feature of far greater importance. At the beginning of these lectures I noted the remarka- ble fact that from the mouth of the Ganges eastward to Cape Horn the continents are bounded ocean-ward by long arcuate mountain ranges, all of which appear to be moving toward the Pacific Ocean. 270 PROFESSOR EDUARD SUESS When, however, one follows the coast from the mouth of the Ganges westward, and again to Cape Horn, totally different conditions are met. Disregarding the bending of the mountains at Gibraltar and vicinity, which the American Cordilleras in the Antilles also show— at both places, as you know, folded mountain chains do approach the Atlantic area, but they bend backward as if held back by some secret force—one sees encircling the Atlantic and the Indian Oceans only similar amorphous coast lines, namely, such as are in no wise predicated by the structure of the lands. Therefore we have distin- guished a Pacific and an Atlantic type of coast. We can go still farther. In whatever direction one proceeds from the land to the Pacific, an unfolding sequence of marine series is seen. If one goes from the wide Archean areas of South America, on which lie horizontal Paleozoic sediments, toward the west, in the Andes are found marine beds of the Jura, the Lower Cretaceous, also the Middle and Upper Cretaceous. It is the same if one goes from the old Laurentian Mass in Canada westward toward the sea. This is also the case in Japan, etc. From the foregoing we may conclude that the Pacific is of very ancient origin, and that it has existed for an extraordinarily long time. With the other oceans it is different. When one nears the Indian Ocean, horizontally disposed marine beds are met with, not folded strata as in the Pacific area. ‘These, however, do not begin with the Trias, but in east Africa as in western Australia start with the Middle Jura, and in Madagascar with the Middle Lias. Similarly, on the shores of the Atlantic Ocean horizontal non-folded strata are found, and these, in west Africa as in North America and Brazil, begin with the Middle and Upper Cretaceous. From this we conclude that the Pacific Ocean is older, the Indian Ocean younger, and the Atlantic Ocean essentially still younger. I have mentioned yet another ocean, Tethys, which in Mesozoic times lay across present Asia, and whose remnants constitute our Mediterranean. ‘The entire area of Tethys is laid in folds, and from the Pacific Ocean to the Caucasus throughout these folds are also moving southward; their margins in the south are overthrusted; the entire province of the sea is crushed from the north, and even rem- nants of the old southern foreland—the Gondwana Land, or the FAREWELL LECTURE Rafik Indian peninsula—are included within this folding. You have heard that Kinchinjinga and its neighbors, the highest peaks of the earth, though within the folds of the Himalayas, still have, so far as known from their foothills, the stratigraphic sequence of Gondwana Land. We will now take a glance at the distribution of the lines of folding on the earth’s surface. In the region of Lake Baikal lies an extensive, somewhat crescentically arranged mass of very ancient Archean rocks. It is folded, with a nearly northeast strike in the east and a northwest strike in the west, and the folds are of pre-Cambrian age. This old strike locus or vertex embraces Sabaikalia, northern Mon- golia, and the East Sajan. Farther northwest there is developed another, younger vertex, or a second center of folding—the Altai. From this second younger locus proceeds an extraordinarily great system of bow-shaped folds, which, in an almost incomprehensible manner, embraces the entire Northern Hemisphere. The Altai encircle the old vertex, and its bows repeat themselves in the east from Japan and Kamschatka to the Bonin Islands. ‘Toward the west they form the broad ranges of the Tian-shan and Bei-shan. Their southeastern branches appear in the bows of Burmah. In front of them to the south lie the marginal bows of the Himalayas— the Iranic; and farther along, the Tauric-Dinaric bows. ‘They press over the Caucasus to Europe, and form here the two previously men- tioned chains of folds. These two chains of folds are themselves preserved in different ways. ‘The one, older, embracing the Varischian and Armoricanian folds, is first discernible in Mahren. It reaches the Atlantic Ocean in southwestern Ireland and Brittany, and disappears as a Rias coast. Years ago, however, Marcel Bertrand called attention to the fact that such a broad and mighty mountain system—on the Atlantic coast it is as broad as the bows of the Himalayas—could not possibly suddenly end here, but that in all probability it is continued to the other side of the ocean in the Rias coast of Newfoundland. As you have heard, Marcel Bertrand accordingly continued the Armoricanian primary lines directly across the ocean to the Appalachians. Of the Appalachians, however, it has been learned in recent years that they are far longer than was formerly believed. They form a bow which is not, as in the Asiatic and European chains, 272 PROFESSOR EDUARD SUESS folded toward the convex side, but toward the concave side, first westerly, then northerly, and continues west of the Mississippi into the Washita Mountains. The second or younger type, the Altai, strikes with decided flexing, narrowed through older horsts, from the Balkans to the Carpathi- ans and the Alps, and at Gibraltar the latter join those bows of the western Mediterranean that are completely reversed. Let us return once more to. North America. As we have heard, the American term as Laurentia the wide Archean: area which embraces the region of the Hudson Bay, middle Canada, and the central part of the United States. The Appalachians to the east and south of this mass, as we have seen, have a concave strike, are folded toward Laurentia, and vanish in the Washita hills. West of Laurentia, also, it is similar. It could have been shown that the Cordillera, whose connection with northern Asia has of course not yet been established, is, on its eastern side, in Canada, also folded toward Laurentia. It, too, bends toward the south with a more and more concave strike; continuing through Mexico, it is folded to the northeast, and then part of its folds finally turn toward Cuba and in the direction of the Antilles. Thus on both sides is North America encircled by concave- striking chains of folds. It is as if the folds extended away from Asia and toward Laurentia. This entire grand phenomenon may be illustrated by a comparison. By the eruption of Krakatoa the oceans were moved; long waves proceeded from the place of erup- tion, traveled around the entire earth, and met themselves on the other side of the sphere. This is merely a comparison, not an explanation. In the Southern Hemisphere the state of things is wholly differ- ent. For some time it has been known that in East India and South Africa, during Permian and Trias time, there flourished identical land floras—the Gondwana floras. Accordingly, it is concluded that these two continents were once united, and the area was named Gondwana Land. Later such floras were also found in Australia; then in the Argentine Republic. Thus it spread around the south. But the conclusion drawn from this as to the continuity of so great a continent was shattered by the circumstance that not only the FAREWELL LECTURE 273 characterizing plants of Lower Gondwana, but, in addition, the South African occurrences of associated animals, were also found in the Permian deposits of Perm in north Russia. What then results is an exceedingly similar distribution of land plants and land animals of that time, and a great continent in the south; yet immediate proof of its continuity is lacking. In fact, only on the Pacific margins of this supposed or actually united continent is it found that folding has taken place, and, indeed, in the east of Australia and the west of South America; while the intermediate Atlantic and Indian coasts are without younger folds. It is true that more recently, folding of pre-Carboniferous time has been described in South Africa, but in general the entire area between the western South American Cordilleras and the eastern Australian Cordilleras appears dead and unmovable. This is in contradis- tinction to the great diversity in movements of the Northern Hem- isphere. In general, these are the chains which we have sought to follow in detail in the course of these two semesters. The attempt toward a geometric arrangement of the mountain chains, which recently has been undertaken by distinguished specialists, finds, I fear, but little confirmation in actual occurrences. The tectonic lines that are met in nature tend generally at most to follow straight lines only in fissures or faults. The foldings, however, maintain them- selves more like long waves, and they give way to the older horsts. This is seen more clearly in the youngest Alps, or that branch of the Altai trending toward Europe; the bows of the Banda Islands are similar. I should now like to’say a little about the conditions of life upon the earth. We have already spoken of the wide distribution of the land faunas and land floras of Lower Gondwana. Earlier types of Carboniferous land floras had spread themselves from the Arctic region to South Africa. The Culm flora is known in Europe, Mon- golia, and Australia. Still more noteworthy is the fact that in the basalt streams of western Greenland there are interbedded plant layers of Lower and Middle Cretaceous, as well as of Tertiary times, and that during all this period there lived in this Arctic region first ferns and then leaf-bearing trees. In a word, in west Greenland 274 PROFESSOR EDUARD SUESS are seen occurrences of different times which throughout cannot be brought into harmony with the climatic conditions of the glacial period nor with those of the present; thus this entire younger epoch appears as an exception. One gets the impression that not at all times did there exist the present diversity of climate, and also that the diversity of life was not at all times a varied one. The great Indian land fauna of today, with its tigers and elephants, can be con- sidered as an independent unity, but here and there it 1s accom- panied by older Malayan remnants which increase the diversity. Gentlemen, as you see by this attempted survey, I can point out only some of the various directions in which our studies may be continued, and there exist so many hundreds and hundreds of questions that all, even the keenest ambition, will find the portals open and may hope for satisfaction. New discoveries are in pros- pect for all conscientious inquirers. In the course of the years I have seen and experienced much. In the beginning a man has honestly to endeavor with zeal, and with certain restrictions upon himself, to learn the detail; and some- times the hair whitens before he is in a position to obtain a general view and to risk a first synthetic attempt. This first step to syn- thesis is, however, the deciding step in the life of the inquirer. Soon he notes that his judgment obtains more consideration among his colaborers; he becomes more careful and conservative with the same; and finally the hour arrives in which his soul is filled with the highest satisfaction, because he has been able to add to human knowledge some new view or a new fact—a feeling over against which everything naturally vanishes that the outer world is able to offer in acknowledgment. Bulwer Lytton says in his novel: ‘‘When a man of great age is surrounded by children, he then sees at the end of his days, not a period, but only a comma.” This applies in equal measure to the inquirer and to his students. This is my good fortune, which today becomes my portion. Many have departed from us. The dumb tablets in our collec- tion halls give their names, and it is our duty today to remember them gratefully. Stolizcka found his end on Kara-Korum, Lend on Kilima-ndjara, Foullon on Gaudalcanar; Rodler brought his death FAREWELL LECTURE 275 germ from the Bachtyari Hills; we all think of Oscar Baumann with admiration. I rejoice today with all my heart that I am enabled to greet, not a series of students, but generations of students, from the renowned gray-haired members of the Royal Academy to the young fellows with sharp eyes. To the young ones among you I should at this moment like to say another word. ‘The old ones know it already. In the course of these forty-four years much has occurred on the earth, but nothing at this time so penetrating, nothing so decisive for the entire culture of humanity, as progress in the natural sciences. Into all departments of human life and doings it has entered; it influences and changes our social conditions, our philosophical conclusions, our political economy, the strength of states, everything. He who will look closer, however, can perceive that, besides the natural sciences, the naturalist himself is coming more and more to the front, that his social significance is being recognized, and that the worth of his studies is being more valued. Accordingly, the growing generation of inquirers has an increased duty, which consists in this, that the ethics of their personal life shall become more precise, so that, by the increasing influence of naturalists on all social and state life, the naturalist will also feel himself more worthy to take part in the guidance of intellectual humanity. And now I have reached the comma. When I became a teacher, I did not cease to be a student; and now that I cease to be a teacher, I shall not cease to be a student so long as my eyes see, my ears hear, and my hands can grasp. With this wish, I therefore do not step out, but take up my former position. And now I thank you all from the depths of my heart for your presence, and beg of you to retain for me a friendly remembrance. TE DI EORDAD THE recent publication by the United States Geological Sur- vey of monographs on the Mesabi, Vermilion, and Menominee iron- bearing districts marks the approximate completion of a prolonged and systematic study of the great Lake Superior ore-bearing series and of the associated pre-Cambrian formations. ‘There is yet to appear a supplementary and final volume on the geology of the Lake Superior region as a whole, which will bring together and correlate the general conclusions and maps of the district monographs. It is understood that this will be submitted for publication in 1905. This notable series of reports was formally inaugurated in the late eighties by Dr. R. D. Irving, but. was prefaced by his work on the Wisconsin survey in the seventies and early eighties. Dr. Irving’s death in 1888, however, permitted him to see only the beginning of the monographic work in the Penokee-Gogebic district. The further execution of the plan fell to his associate, Dr. C. R. Van Hise, who, during the past sixteen years, has carried it forward to its present advanced stage. The first of the monographs to appear was the Penokee-Gogebic monograph, No. XIX, published in 1892. The field work for this was done jointly by Irving and Van Hise, but the writing of the monograph fell largely to the latter. In 1895 appeared the Marquette monograph, No. XXVIII, by Van Hise in collabora- tion with Professor W. S. Bayley. This was followed in 1899 by the appearance of the Crystal Falls monograph, No. XXXVI, by Pro- fessors J. Morgan Clements, H. L. Smyth, W. S. Bayley, and C. R. Van Hise; in 1903, by the Mesabi monograph, No. XLIII, by Dr. C. K. Leith; and finally, during the past winter, by the Vermilion mon- ograph, No. XLV, by Clements and the Menominee monograph, No. XLVI, by Bayley, the latter having just come from the press. Monograph V, on the copper-bearing rocks of Lake Superior, by Irving, should also be mentioned, although it was published before (1883) the inauguration of the plan for the comprehensive investiga- tion of the iron-bearing districts. 276 EDITORIAL 2747 The total expenditure of the survey for this great series has been less than $150,000. Considering the magnitude of the iron-mining industry, and the intricacy and importance of the geological forma- tions involved, the expenditure is very conservative. Larger sums than this, we are reliably informed, are spent by single mining com- panies in the Lake Superior region in the exploration of very limited areas; indeed, it is in a large measure due to the co-operation of such companies that this series of monographs has been prepared within the amount named. The work of the survey is highly appreciated in the Lake Superior region, where a good geological and structural map has come to be regarded as an absolute prerequisite to intelligent underground ex- ploration and mining development. The working maxim has been firmly established that before expensive underground exploration is attempted it is economy to spend whatever is necessary—which is usually a comparatively small amount—to ascertain all that can be learned at the surface. A proper geological map of the ore-bearing district saves many thousands of dollars in sub-surface exploration by limiting the areas which it is worth while so to explore. Professor Irving and his colleagues mapped in detail the iron-bearing formation of the Penokee-Gogebic district in the later seventies and early eighties, and up to the present time no ore-bearing areas have been found outside the narrow limits they laid down. While elsewhere, from time to time, exploration has enlarged the boundaries of the ore- bearing formations mapped by the earlier surveys, it is within limits to say that a vast amount of subsurface exploration which might otherwise have been done in barren areas has been localized in more promising fields by means of the geological maps. This series of monographs constitutes a valuable contribution to the theory of ore deposition and to the intricate geology of the best known pre-Cambrian formations. Not only have they an intensive value to local mining men and to pre-Cambrian geologists, but they subserve a broader and scarcely a less important function in the dis- semination, through the industrial and scientific world, of information relative to the sources, the extent, and the modes of origin of products which are the dominating factor in the most important metallic indus- try of America, if not of the world. ACE. REVIEWS. Grundziige der Geologie des unteren Amazonasgebietes (des Staates Paré in Brasilien). Von Dr. FRIEDRICH KATZER. Leipzig, 1903. Pp. 302, royal 8vo; illustrations and one geologic map of the state of Para. Tue author of this work has brought together and published in a volume of convenient size the chief matters of interest in regard to the geology and geography of the lower Amazonas, and especially of the state of Para. The volume opens with a geographic sketch of the region. ‘This is fol- lowed by a brief history of the work done on the geology, and biographic notices of the men by whom the work has been done. It is an interesting fact that the bulk of our knowledge of the geology of Brazil dates from a visit made to that county in 1864 by Louis Agassiz. ‘The work of Agassiz him- self upon the geology was of no great importance, but his inspiration was far-reaching. C. F. Hartt, one of his assistants, returned to Brazil several times to continue work on the geology, and eventually died in that country. He took with him several assistants—Derby, Rathbun, H. H. Smith, and others—who have continued the work. After Hartt’s death, in 1877, Derby remained in Brazil and has devoted his life entirely to the study of Brazilian geology. It was through his influence that descriptions were finally published of the rich Silurian, Devonian, Carboniferous, and Cretaceous faunas of Brazil. Dr. Katzer himself was formerly geological assistant in the museum of Natural History at Para, and in that capacity he traveled extensively through the Amazonas country. His interest in the geology of the region has led him to publish this volume even after he has left Brazil. The biographic part is followed by descriptions of the different groups of rocks, and plates are given of the most important fossils from the fos- siliferous horizons. The volume is one of much value to those who wish to obtain a general knowledge of the geology of the Lower Amazonas region without having to seek it through a large number of papers published at widely different times and places. J. C. BRANNER. REVIEWS 279 The Correlation oj Geological Faunas: A Contribution to Devonian Paleontology. By HENRY SHALER WILLIAMS. [Bulletin of the United States Geological Survey, No. 210.] Washington, 1903. In the investigation of geologic problems concerned with correlation, two fundamentally different concepts must be kept continually in mind. The first of these has to do with rock strata, the media in which fossil organ- isms are preserved, and the classification of formations; the second has to do with fossil faunas or assemblages of organisms preserved in the rocks, and the classification of time periods. In the broad correlation of geologic formations the data furnished by the faunas are of prime importance, and too much cannot be said of the value of exhaustive researches upon fossil faunas as faunas. The paper by Professor Williams on The Correlation of Geological Faunas is essentially a treatise upon the methods of investigation of fossil faunas, in which the Middle Devonian fauna of the New York province, characterized by Tvopidoleptus carinatus, is especially used for illustration. In the first two chapters of the work, ‘‘The Principles of Correlation”’ and the ‘‘ Geological Expression of Faunal Migrations” are discussed in a manner applicable to any problem involving the study of fossil faunas. Chapter 3 is devoted to an application of the principles discussed in the pre- ceding chapters, to an investigation of the history of the 7’vopidoleptus cari- natus fauna. In treating of the ‘‘Shifting of Faunas” in chap. 4, illustra- tions are again drawn from the Devonian faunas of the New York province. The principles involved, and the effect of the shifting as expressed in the faunas themselves, are fully discussed. In considering the ‘“‘equivalency”’ of formations in chap. 5, examples are taken from the correlation of the Devonian formations of New York and Ohio. The sixth and last chapter of the treatise is devoted to the “‘ Bionic Value of Fossils.”” The application of the data furnished by species, genera, etc., of organisms, for the classifi- cation, not of rock formations, but of time periods, is discussed, and the chapter closes with the statement of a proposed “‘bionic time scale.” In this paper Professor Williams has assembled the more important results, both material and philosophical, which he has secured in the course of his long-continued investigations of the Middle and Upper Devonian faunas of New York. Many of these results have been previously pub- lished in various shorter papers, but here they are for the first time brought together in compact form. The paper is full of suggestions and should be studied by every student of fossil faunas. STUART WELLER. 280 EDITORIAL The Evolution oj Earth Structure. New York: Longmans, Green & Co., 1903. By T. MELLARD READE. THE author concludes that the continents and ocean basins arise from differences in the specific gravities of large sections of the earth. ‘These specific gravities are not stable, but are subject to slow changes consequent upon changes of temperature. A rise of temperature and local increase of volume create protuberances which may be of continental extent. A fall of temperature and decrease of volume lead to depressions; which may cul- minate in the formation of deeps. ‘“‘Thus it follows that these departures from the regular spheroidal forms are not original and permanent; nor are they features which have been growing from the dawn of geological history, such as would be likely to occur from a differential radial shrinkage of the earth.’ Evidence of thermal fluctuations is given by the varying composi- tion, specific gravity, and temperature of lavas from time to time emitted from vents. ‘‘The extrusion of lavas is largely due to increase of tempera- ture and consequent increase of volume. Relief by extrusion causes a reduction of temperature, and shrinkage takes place in the supplying reser- voir. This results in a subsequent period of quiescence, which lasts until the molten matter of the reservoir again becomes hot enough to compel extrusion.” The author holds tenaciously to the theory which he advanced in a former work,' that mountain ranges are caused by sedimentation and a subsequent heating of the sediments. Numerous cuts of models illustrating earth movements are introduced, with a full discussion of the experiments. NODE OF EXPEAN ATION: By an oversight, Maps I and II (Ten Mile Creek) of the paper of Mr. George C. Matson entitled “‘A Contribution to the Study of the Interglacial Gorge Problem,” which appeared in the February-March number of the JOURNAL, were omitted. They are inclosed with this number, and may be inserted between pages 140 and 141 of the February-March number. ; 1 The Origin of Mountain Ranges. MAP OF | VE MG cel 9 : FROM ITHACA AND DRYDEN SHEETS U.S GEQLOGICAL SURVEY. ° ‘ Miles Datum, Mean Sea Level Og Divide 0 (This map should be inserted in Vol. XII, No. 2, February-March, 1904, of the JOURNAL OF GEOLOGY, to face page 140.) yh PART OF TEN SMIEE CHEEK, Contour Interval, 20 ft. Datum, Mean Sea Leve/. weet Legend = Portage Sandstones and Shales Map I (This map should be inserted in Vol. XII, No. 2, February-March, 1904, of the JOURNAL OF GEOLOGY, to face page 142.) Ay oom ci fey : , i + te} : : . z z ‘ j § ; ' 5 oes a) ° ¥ : > F v f i 4 : b : Eee j z = ‘ We i . ; ; Hite H i U ; a * Z ied \ ? | * jen 5 . * ‘ a « ht bi * . . " s fs i ‘ I ‘4 . aoe 0 Sree i . rs ; : if ; 1 . s ; ; \ - HOCINNAE OF GEOLOGY NEA QUINTERO! A SERIES OF GENTLE FOLDS ON THE BORDER OF TE TAPPALACHTIAN SYSTEM THE conclusions of this paper have been reached in the study of the Watkins Glen quadrangle, New York. This quadrangle com- prises the four fifteen-minute quadrangles known as the Watkins, Ithaca, Elmira, and Waverly quadrangles. Extending from the Pennsylvania state line to the Seneca and Cayuga Lake valleys, it includes the southern portion of each. This area lies immediately north of the region of the Appalachian folds. The surface rocks of the quadrangle are the shales and sandstones of the Chemung, Portage, and Genesee formations. The rock strata over much of the area vary so little from the horizontal position that the dip can usually be recognized only by the very careful use of the clinometer or the level. Dips high enough to be conspicuous, and ranging from 8 to 55°, have been noted occasionally in various parts of the quad- rangle, but these high dips have in nearly all cases been found to be associated with small local aniiciines or faults, extending frequently only a few rods and possessing only iocal interest. With these higher dips the present paper is not concerned; but the interpretation of the more obscure dips ranging usually from r to 3 or 4° will be attempted. A careful study of the low dips characterizing the rocks over the major part of the quadrangle has shown them to have an important structural significance. They have been found to represent a series of low, approximately parallel, anticlinal folds, having the same general direction as the great mountain folds immediately south of them in Bradford county, Pennsylvania. t Published by permission of the director of the U. S. Geological Survey. Vol. XII, No. 4. 281 282 EDWARD M. KINDLE Very gentle dips ranging from 4 to 5°, but rising in a few instances to 10° or more, characterize these folds. Although very low, they belong to anticlinal folds, which are very persistent. Some of them have been traced entirely across the quadrangle. Five of these folds, separated by a corresponding number of synclines, have been rec- ognized. Beginning at the north, the folds will be designated by the SKETCH MAP OF THE WATKINS GLEN QUADRANGLE and adjacent portions of Pennsylvania showing location of anticlinal and synclinal axes. by E.M.Kindle Scale — 110 miles BHA DKORD * nie describes an eclogite in Silesia occurring with gabbro, amphibole, and serpentine, and forming bands separated by serpen- « “On an Ecolgite from Loch Duich,”’ Mineralogical Magazine, Vol. IX, p. 217- 2 “Notes on the Vicinity of the Upper Part of Loch Maree,” Quarterly Journal of the Geological Society, Vol. XXXVI (1880), p. 105. 3 ‘Die Serpentin-und Amphibole-Gesteine nérdlich von Marienbad in B6hmen,” Tscherm. min.-petr. Mitth., Vol. IX (1887), p. 89. ; 4 “Notes on the Eclogite of the Bavarian Fichtel-Gebirge,’ Transactions oj the New York Academy of Science, Vol. XVI, p. 24. 5 “Ueber ein Vorkommen von Eklogite in Schlesien,” Neues Jahrbuch, Vol. 1 (1889), p. 195. ECLOGITES IN CALIFORNIA 347 tine. The eclogite from the Tyrol Alps recently described by Hezner’ is composed primarily of clear red garnets, commonly in rounded grains, in a groundmass of emerald-green omphacite. The garnets vary from fine grains up to the size of peas, but the crystal form is seldom distinct. Hezner considers that this eclogite is chemically a gabbro or a variation of the same magma that furnished the gabbro. Eclogite, he thinks, is formed in the greatest depths and in the higher zones amphibolite—the garnet and the omphacite being amphib- olized. The foregoing brief extracts will give some idea of the eclogites of Europe and of their probable derivation. CALIFORNIA ECLOGITES. Probably the most typical eclogite in California is that found in the bed of Coyote Creek, about eighteen miles southeast of San José and some six miles east of north from San Martin. The outcrop of massive rock is exposed for about twenty feet in the edge of the stream. Apparently it breaks through the shale and jasper exposed at the foot of the hill only a few feet away, but the gravel of the creek entirely covers the contact. On the opposite side of the stream, and within a hundred yards, is a large mass of serpentine, which, how- ever, is not in contact with the eclogite. ~The most characteristic facies of the outcrop is that which shows a grass-green groundmass, thickly studded with dark red garnets several millimeters in diameter and showing distinctly the rhombic dodecahedron form. ‘The faces of the garnet are fresh and shining, with clear-cut edges. The reproduction of the photograph in Fig. 1 shows the structure of the rock, but the striking effect of the red garnets in the light green groundmass is unfortunately lost. Seams and segregations of glaucophane, sometimes bearing gar- nets, occur in the exposure. Prominent veins of a fine-grained red- dish mineral were taken in the field for inclusions of the nearby jasper, but its fusibility, and its isotropic character under the micro- scope, proved it to be a compact variety of garnet. Segregations of actinolite crystals are common, and chlorite frequently occurs. Some particles of chalcopyrite are seen in the rock, and a few parti- I Op. ct. 348 RULIFF S. HOLWAY cles of free gold were found in the granular garnet. The rock has attracted the attention of prospectors, and it is reported that an assay showed nearly two dollars of gold per ton. Ina few places there were inclusions, some 10-15™™ in diameter, of a reddish-brown mineral with cleavage faces giving an almost metallic luster. In the labora- tory the mineral was found to be infusible and to possess a hardness Fic. 1.—Eclogite from San Martin, Calif. Slightly reduced in size. of over six. The streak is yellowish-brown. . The crucibles were kept at the highest temperature of the furnace for about three hours, and then the drafts were closed and the whole allowed to cool slowly overnight. The eclogites fused easily to homo- geneous lava like obsidian. Under the microscope this showed a practically uniform isotropic character. With a high power points showing double refraction were found scattered through the slide, but no detail could be made out. A list of the constituent minerals found in California eclogites, and a brief mention of their properties as they are found in these rocks, will unify the preceding descriptions. First among the’ essential minerals is garnet. The color is usually a dark red in the hand specimen. In size the garnets vary from 2 to 5™™) for the well-developed dodecahedrons of the San Martin eclogite, t Petrology for Students, 1887, .p. 320. ECLOGITES IN CALIFORNIA 57 ios) to 10 or 12™™ for the rounded forms in some of the Calaveras expos- ures. Of the essential minerals garnet is evidently the oldest, for in the cracks are found glaucophane, omphacite, and hornblende. Yet these latter minerals are also found as inclusions together with rutile, apatite, feldspar, and quartz. Inclusions of glaucophane may be paramorphs of inclusions in the original matrix. Qualitative tests indicate that the garnets contain iron, calcium, and aluminum. Omphacite, the light green augite, usually occurs in aggregates of prismatic crystals, some a $™™ in length and without definite termi- nation. The bright polarization, absence of pleochroism, high extinction angle, and occasional augite cleavage serve to distinguish it. Liidecke deduced the formula (CaFe) 5i0,MgSiO, from his chemi- cal analyses. Smaragdite is an emerald-green actinolite that in the eclogites much resembles omphacite, being distinguished from it by the horn- blende extinction angle and usually by pleochroism. The blue soda hornblende, glaucophane, has the marked pleochroism already given and an extinction angle of C onc’ varying from 5 to 15°. It seems to be derived from hornblende in the eclogite in Calaveras Valley at least the hornblende crystals are bordered by it. Diller has already been quoted as to the change of pyroxene to glaucophane in the Port Orford rocks, some of which are here classed as eclogites. The analyses of some of the igneous rocks considered show that they contain enough soda for this change to occur without the addition of elements from the outside. The hornblende in the Calaveras eclogite show soda in the qualitative test, and also has a general resemblance to glaucophane in its pleochroism and in its orientation. Hintze has already been quoted in regard to a black pargasite that seems to agree closely with the Calaveras variety of hornblende in pleochroism and extinction angle. This pargasite contains, according to his analysis, 3.44 per cent. of soda—which is more than the per- centage of soda in the Port Orford glaucophane. Both garnet and hornblende are found replaced by chlorite— pale green in color, and showing low interference colors except in the pennine variety, which shows Prussian blue between crossed nicols. Mica seems to be absent from the Calaveras slides and is found in very small quantities in those from San Martin. In the Tiburon 358 RULIFF S. HOLWAY eclogite margasite is very plentiful and has been fully described by Ransome, as already cited. White mica, probably paragonite, occurs in the San José eclogite and in that from Sonoma county. Titanite, or sphene, is thickly distributed in some of the slides in very minute grains. It is a very light brown color and pleochroic in tints of that color. Rutile in large crystals 1o-15™™ in diameter is found only at San Martin, but it is found somewhat less freely than sphene in the shape of small irregular grains in nearly all the slides. It is yellowish to reddish-brown and somewhat pleochroic, and has very high relief. Epidote is rather irregular in its occurrence. When found, it exhibits lower interference colors than those usually described. Ransomet thinks that this is accounted for by a smaller proportion of iron, and that in chemical constitution the epidote may grade over into zoisite. Zoisite and cyanite were seldom found. Pyrite appears occasionally, but not so frequently as was expected from the descrip- tion of the European eclogites. All of the above minerals are possibly secondary in their occur- rence in eclogites. ‘The occasional feldspar and quartz may be the only minerals remaining unchanged from the original rock. The garnets and the epidote have probably taken up the line of the original feldspar, while the soda is to be found in the glaucophane or in the pargasite variety of hornblende. While the use of the term “eclogite” is now fairly definite, there is still a question of some limitations in its application. In constitution the rock must contain garnets in a matrix of omphacite, glaucophane, or hornblende, or of some mixture of these minerals. Hezner evi- dently would insist that there must be omphacite, for even with about equal proportions of hornblende he drops the term ‘‘eclogite” and uses “eclogite-amphibolite.”” The common accessory minerals are sphene, rutile, cpidote, apatite, zoisite, cyanite, feldspar, and quartz. While some metamorphosed sedimentary rocks approach eclogites in composition, the tendency seems to be to restrict the term to rocks that are clearly of igneous derivation. RuLIFF S. HOLway. BERKELEY, CALIF. WOpAcis ips SiO: TA DIERORLAL. Ir is always a pleasure to note a graceful expression of appre- ciation of long and faithful endeavor to promote our science—doubly so when this service has been given in a singularly quiet and modest way, without any apparent realization of its true merits. At a recent alumni dinner of the State University of Iowa, the former students of Professor Samuel Calvin, to the number of over two thousand, united in the commemoration of the completion of his thirtieth year as professor in that institution. The recognition took the form of a costly silver loving-cup, designed especially for the purpose of symbolizing the scientific achievements of the recipient. The cup is a classic Greek vase, sixteen inches in height, and stands on a base of serpentine five inches high. It is adorned with casts taken directly from fossils, with a drainage map of Jowa, with crossed geological hammers, a microscope, and the more conventional spray of laurel, owl of wisdom, and torch of learning—all in relief. One side bears an appropriate inscription in raised letters. Professor Calvin was elected to the chair of natural history in Jowa’s university thirty years ago. The chair has since been sub- divided into four distinct departments, Professor Calvin retaining the department of geology. As well known to the profession, he has been state geologist of Iowa during the last twelve years, and an admirable series of reports is appearing under his administration. ios) mn \o REVIEWS. Catalogue of the Ward-Coonley Collection of Meteorites. By HENRY A. WARD. Chicago, 1904. THE third catalogue of this collection has been issued within four years as a consequence of its rapid growth from 424 falls in 1g00 to 603 falls in t904. The weight of the present collection aggregates 2,495 “*. It contains 241 falls of siderites, 28 of siderolites, and 334 of aerolit s. The number of specimens is about 1600. The greater number of falls is from North America and Europe, but there are considerable numbers from Asia, Africa, Australia, Sandwich Islands, and South America. The catalogue is prefaced by a statement of the method by which the collection was gathered together and an account of the most noteworthy specimens. Under the entry of each fall is given the date and locality of fall, the name and description of the meteorite, the reference to the pub- lished description, together with the weight of the chief piece and the total weight of the material in the collection. There is in addition an alphabetical list of all known meteorites, and another giving their geographical distribution, arranged according to countries. This is followed by a statement of the latest revised classifica- tion of meteorites by Dr. Aristides Brezina, of Vienna, giving the system, composition, and name, with the list of meteorites belonging to each division. The Ward-Coonley collection also contains a library of over 800 titles, besides some minerals characteristic of meteorites, a small collection of their sections, specimens of terrestrial iron, and a series of casts of meteorites. The collection is at present exhibited in the American Museum of Natural History in New York city. ; 360 JOURNAL OF GEOLOGY JULY-AUGUST, 1904 DESCRIPTION AND CORRELATION OF THE ROMNEY FORMATION OF MARYLAND." CONDENSED DESCRIPTION OF THE ROMNEY FORMATION. THE lower member of the Romney formation in Allegany county is composed principally of fissile black shale, some of which weathers to a yellowish or buff color on long exposure. In comparatively fresh exposures, however, as in the railroad cuts at Twenty-first Bridge, the shales are either black or rusty-brown after some weather- ing. The black shales are shown to best advantage in these cuts, although on the Williams Road, three and one-half miles southeast of Cumberland, is perhaps the most nearly complete exposure of this division with an approximate thickness of 512 feet. In the lower part of some of these exposures are bands of very dark-colored thin limestone. The lithological characters of these shales agree closely with those of typical exposures of the Marcellus shales in New York state, and in addition they contain such characteristic species as Liorhynchus limitare (Vanuxem) and Agonitaites expansus (Van- uxem). The second member of the Romney formation, the Hamilton beds, has an approximate thickness of 1,100 feet, and is composed of shales and sandstones. In recent exposures the shales, generally bluish or bluish-gray in color, vary in composition from quite coarse arena- ceous to those that are fine and argillaceous. The sandstones, «Published by permission of Dr. William Bullock Clark, state geologist of Mary- - land. The data upon which this paper is based will appear in detail in the forth- coming Devonian volume of the Maryland Geological Survey. Vol. XII, No. 5. 361 362 CHARLES S. PROSSER — which on fresh surface are generally blue or gray in color, are not very coarse in texture, and the layers are often less than a foot thick. All of these rocks, however, on long exposure usually present along the highways a slightly greenish or yellowish-gray tint. ‘There are two prominent sandstone zones in this member of the formation, varying in thickness from about 30 to 75 feet. The lower one is from 500 to 550 feet above the base of this division, or from 1,000 to 1,050 feet above the base of the formation, while the upper zone is at or near the top of the formation. Both of these sandstone zones are clearly shown in the sections on the Williams Road and at Great Cacapon, and the upper one at Gilpin and above Corriganville. The shales in many localities are very fossiliferous, especially those between the two sandstone zones, and contain numerous specimens of such characteristic species of the New York Hamilton as Spirijer mucronatus (Conrad), S. granulosus (Conrad), Athyris s pirijeroides (Eaton), Tvropidoleptus carinatus (Conrad), Chonetes coronatus (Conrad), Phacops rana (Green), and other species. On account of the presence of numerous Hamilton species together with a litho- logic similarity and approximate stratigraphic position, this division of the Romney formation is regarded as equivalent to the Hamilton stage of New York. The estimates of the thickness of the Romney formation vary from about 1,600 to 1,650 feet. In Allegany county both the Mar- cellus shale and Hamilton stage are clearly shown; but farther east in Washington county the Marcellus shale or a part of it is wanting. This would indicate that the subsidence of the Onondaga land area began at an earlier date in Allegany than in Washington county. CORRELATION OF THE MARCELLUS SHALE. The lithological similarity of the thin, black shales forming the lower part of the Romney formation to the Marcellus shale of New York has been noted in the description of the Romney formation, and in other places in the Maryland volume. This is so marked in connection with its similar stratigraphic position that the north- ward continuation of these shales in Bedford county, Pennsylvania, were unhesitatingly called the Marcellus by Professor Stevenson in his geological report of that county. Following the northeasterly THE ROMNEY FORMATION OF MARYLAND 363 strike of the Devonian formations across Pennsylvania, a similar lithologic and stratigraphic shale has been noted by various geologists at different localities, until Monroe county in the northeastern part of the state is reached, where it was positively identified by Dr. I. C. White. The localities in the northeastern part of the state were later studied by the writer, who from the lithologic, stratigraphic, and paleontologic evidence fully accepted Dr. White’s correlation.* This practically carried the black shale into southeastern New York, where the identification of the black, fissile shale below the Hamilton beds as the Marcellus has not been questioned. Finally, in the Cum- berland basin of Maryland Mr. Schuchert positively identifies this shale as ‘‘the Marcellus stage of the Middle Devonic,”’ which he states “rests directly upon the eroded Oriskanian.”’? These shales in general are sparingly fossiliferous in Maryland and northern West Virginia; but there are occasional layers in some localities which contain a more abundant fauna. There may be a question whether the fossiliferous zones noted at certain localities are not stratigraphically above the very fissile bituminous shale so well exposed in the southern part of the Baltimore & Ohio Railroad cut at Twenty-first Bridge, which agrees so strikingly with the Mar- cellus shale of New York. The writer has studied to some extent the fauna found mainly in the shales which lithologically closely agree with the New York Marcellus. Dr. J. M. Clarke has probably a larger collection, obtained in part from the fossiliferous layers mentioned above, which he is elaborating, and consequently this is to be regarded as only a preliminary account of this fauna. Twenty-one species have been listed by the writer from these shales, three of which are restricted to Maryland.3 The other eighteen either occur in New York or are represented by closely affiliated species. ‘These species range in New York from the Scho- harie grit to the Chemung, inclusive, and the formations containing the largest number of them, which therefore become the most impor- tant in correlation, are as follows: Marcellus, ro identical, 4 affiliated; t Bulletin No. 120 (1894), U. S. Geological Survey, p. 4. 2 Proceedings of the U. S. National Museum, Vol. XXVI (1903), p. 422. 3 Tables giving the geological range and geographical distribution of the Romney species will be published in the Maryland Devonian volume. 264 CHARLES S: PROSSER Hamilton, 8 identical, 6 affiliated; Sherburne, 2 identical, 3 affliated; Ithaca, 8 identical, 3 affliated; and the Chemung, with 3 identical, and 2 affiliated. It will thus be seen that, so far as this fauna is con- cerned, it indicates the correlation of this shale with the Marcellus shale and Hamilton beds of New York, with the evidence somewhat in favor of the Marcellus, since it contains to identical species to 8 in the Hamilton. The weight of evidence in favor of correlation with the Marcellus shale is strengthened when the list is examined a little more closely. Liorhynchus limitare (Vanuxem), which, so far as ] am aware, is confined to the Marcellus shale in New York, and perhaps may be considered its most characteristic fossil, or at least its most distinctive Brachiopod, is found generally in the black, fissile shale constituting the lower part of the Romney formation in Maryland. Bacirites aciculatus (Hall) is known only in the Mar- cellus of New York, and the A gonzatites expansus (Vanuxem) is so characteristic of a thin layer of limestone in the lower Marcellus of New York that it has been named the Agoniatite limestone. In Maryland 1760 feet or more above the base of the black shales are thin limestones which also contain A goniatites expansus (Vanuxem). Finally, it may be said that, so far as the paleontological evidence is concerned, it shows a close relationship between the Maryland black shales and the Hamilton beds of New York, and Dr. John M. Clarke has already shown that such a relation exists in New York, since a large percentage of the species found in the Marcellus shale of that state occurs in its Hamilton beds.t’ The paleontological evidence, however, shows a still closer relationship with the Mar- cellus shale fauna of New York, which is supported by the visible continuity, lithologic similarity,? and stratigraphic position of the containing shales, so that the correlation of this Maryland black shale with the Marcellus shale of New York appears to be fairly well sus- tained. CORRELATION OF THE HAMILTON BEDS. The rocks overlying the Marcellus shale of the Romney formation, and extending northeasterly from northern West Virginia across t Kighth Annual Report of State Geologist of New York, 1889, pp. 60, 61. 2 For a summary of the various methods of correlation see Mr. GILBERT in Compte rendu, Fifth Session, International Geological Congress, 1893, pp. 151-54. THE ROMNEY FORMATION OF MARYLAND 305 Maryland and Pennsylvania to New York, have been much more frequently correlated with the Hamilton beds of New York. Pro- fessor James Hall and other paleontologists have identified collections of fossils from these rocks in northern West Virginia, and from inter- mediate localities between that state and New York, as composed of Hamilton species. If the various geological maps, reports, and papers describing the Devonian formations from West Virginia to New York are put together and considered, it will be found that this correlation is strongly supported by visible continuity. Furthermore, the stratigraphic position of these beds strongly supports this correla- tion. The paleontological data are as yet much more extensive regarding the Hamilton beds than for the Marcellus shale. The total number of species recorded by the writer from the Hamilton beds of Mary- land is 147, of which 21 are limited to Maryland, leaving 126 identical or closely related species which also occur in New York. An enumer- ation of the totals for the New York Devonian formations shows that 3 identical species occur in the Helderbergian series; 1 identical, in the Oriskany; 6 identical, in the Schoharie; 17 identical, doubtfully 4 more, and 2 affiliated, in the Onondaga; 47 identical, 1 more doubt- fully, and 7 affiliated, in the Marcellus; 92 identical and 32 affiliated, in the Hamilton; 2 identical, in the Tully; 4 identical and 1 affiliated, in the Genesee; 2 identical, in the Portage; 4 identical and 2 affiliated, in the Naples; ro identical and 1 affiliated, in the Sherburne; 55 identical, 2 more doubtiully, and 9 affiliated, in the Ithaca; and 18 identical, 4 more doubtfully, and 3 affiliated, in the Chemung. Adding these numbers, the total number of entries for each New York forma- tion is as follows: Helderbergian series, 3; Oriskany sandstone, 1; Schoharie grit, 6; Onondaga limestone, 23; Marcellus shale, 55; Hamilton beds, 124; Tully limestone, 2; Genesee shale, 5; Portage beds, 2; Naples beds, 6; Sherburne sandstone, 11; Ithaca beds, 66; and the Chemung beds, 25. Judging from the number of entries, it is then seen that the Maryland beds show the closest relationship with the Onondaga, Marcellus, Hamilton, Ithaca, and Chemung formations of New York; and especially with the Marcellus, Hamil- ton, and Ithaca. On examining the total number of entries for these three formations, it is found that the Marcellus has 44.3 per cent. as 366 CHARLES S. PROSSER many as the Hamilton, and the Ithaca 52.8 per cent. This is not remarkable, however, when it is recalled, in the first place, that a large percentage of the species in the Marcellus shale of New York continue into the Hamilton beds of that state, as has been shown by Dr. John M. Clarke; and, in the second place, that the Ithaca fauna is sequential to the Hamilton, and in the Ithaca region contains a large percentage of Hamilton species. When followed to the east- ward, and after the disappearance of the Tully limestone and Genesee shale in the Chenango valley, the writer has shown that a still larger number of the Hamilton species lived into Ithaca time, although part of them were represented by simply a few individuals which were the last feeble representatives of their species. ‘These rare individ- uals have been recorded in the range of the species, making the faunas of the Hamilton and Ithaca beds of New York seem more closely related than they actually are; and the same is true regarding the faunas of the Maryland beds and the Ithaca beds of New York. This explanation is sufficient to show that the above tabulation gives full expression to the closeness of the relationship which exists between the fauna of the Maryland beds and the faunas of the Marcellus shale and Ithaca beds of New York, as compared with that which exists between the fauna of the Maryland beds and the New York Hamilton fauna. Restating the tabulation, then, it is shown that there are more than twice as many entries common to the Maryland and New York Hamilton beds than to the Maryland and New York Marcellus; and nearly twice as many for the Maryland and New York Hamilton beds as for the Maryland and New York Ithaca. There- fore the paleontological evidence strongly supports the correlation of the Maryland beds, which represent in general the midddle and upper portions of the Romney formation, with the Hamilton beds of New York. Recently Professor H. S. Williams has published an extended account of what he calls the Tropidolepius carinatus fauna of the Hamilton formation.' Faunally he considers the Hamilton forma- tion as including the deposits between the top of the Onondaga lime- stone and the base of the Tully limestone of central New York, which t American Journal of Science, Fourth Series, Vol. XIII (1902), pp. 421-32; Bulletin: No. 210, U.S. Geological Survey, 1903, pp. 42-68. THE ROMNEY FORMATION OF MARYLAND 367 have generally been divided into the Marcellus shale and the Hamil- ton beds. Professor Williams compiled a list of twelve species for the Tvopidoleptus fauna which he called the “‘standard list of domi- nant species for the New York-Ontario province.’ Another list was also compiled, which he called a “revised list of dominant species of the Hamilton formation of eastern New York and Penn- sylvania, as expressed in 183 faunules,” which contained the twelve species given in the standard list and four additional ones. All of these sixteen species occur frequently in the Hamilton beds of Mary- land. Professor Williams, after an examination of the preliminary lists from the Hamilton beds of Maryland, published the following state- ment: In the list furnished me by Professor Prosser there appear 132 entries, 91 of which are positive identifications. Among the latter are found all of the dominant species of the Tvopidoleptus carinatus fauna, as estimated from the New York statistics. This is sufficient to establish the extension of the Tropidoleptus fauna, in its integrity, as far south in the Appalachian trough as Maryland. Other facts brought out in the Maryland Devonian report by Dr. John M. Clarke and the writer apparently show that the Hamilton beds of Maryland are succeeded by deposits and faunas similar to those succeeding the Hamilton of New York, and therefore it may be concluded that the deposits of the Hamilton beds from New York to West Virginia were brought to a close at about the same geological time. EUROPEAN EQUIVALENTS. The early attempts at correlating the Devonian rocks of the United States with those of Europe dealt only with the formations found in New York, which, in fact, has generally been the custom down to the present time. In 1842 Conrad published the statement that “the Ithaca group, Chemung group, and the Old Red Sandstone near Blossburg, in Pennsylvania, constitute the equivalents of the Devonian system as developed in Europe,’’ and contain a number of fossils characteristic of European Devonian strata.2 The same year Vanuxem stated that the last three groups of the ‘“‘ Erie Division”’ t Bulletin No. 210, U.S. Geological Survey, p. 67. 2 Journal of the Academy of Natural Science, Philadelphia, Vol. VIII, p. 232. 368 CHARLES “S~ PROSSER —viz., the Portage, Ithaca, and Chemung—“‘appear to correspond with the Devonian system of Mr. Phillips.”* The following year Professor Hall gave the base as somewhat lower when he stated that the Devonian system appears “to correspond to the Chemung and Portage groups, and also to include a portion of the Hamilton.’’? In 1847 Professor Hall stated that With the Schoharie grit, commences a series of strata containing fossils as distinct from those of the preceding formations, as these are from the lower division. We here, for the first time, recognize several species that are regarded as Devonian forms; and if zodlogical characters are to be paramount, we are compelled to unite all the succeeding strata as of Devonian age.3 Finally in 1859, he raised the question whether even the Oriskany sandstone might not be considered as of Devonian age. For he wrote as follows concerning the line’ of demarkation for the Silurian and Devonian systems. Shall the advent of the Oriskany sandstone, with its Spirifer of dichotomizing coste,be the division ? Or shall we look for some more marked and more readily defined and recognized feature for the distinction between what are regarded as two great geological systems ?+ So far as the writer is aware, de Verneuil in 1847 was the first geologist definitely to correlate the younger formations of the New York system with subdivisions of the Devonian system of Europe. He made the base of the Oriskany sandstone the dividing line between the Devonian and Silurian systems;5 correlated the Hamilton, Tully, Genesee, Portage, and Chemung with the formations of the Eifél and Devonshire, and the Marcellus with the shales of Wissenbach in Nassau, as is proved by their Goniatites, so analogous in form.°® In recent years several geologists have considered the correlation of the American Mesodevonian with European rocks of equivalent age, of which the following are the most important: t Geology of New York, Part III, p. 171. 2 Ibid., Part IV, p. 20. 3 Paleontology of New York, Vol. 1, p. xvii. 4[bid., Vol. III, Part I, p. 42. 5 Bulletin de la Société Géologique de France, Second Series, Vol. IV, p. 677; also American Journal of Science, Second Series, Vol. V (1848), p. 367, on the parallelism of the Paleozoic deposits of North America with those of Europe, translated by JAMES HALL. 6 Loc. cit., p. 678; and American Journal of Science, loc. cit., pp. 367, 368. THE ROMNEY FORMATION OF MARYLAND 269 In 1889 Professor H. S. Williams apparently correlated in a general way the American Middle Devonian with “the Ilfracombe [England] beds of Phillips, the Givétien limestone of Belgium, [and] the Strin- gocephalien shales or limestones of the Eifél and Hartz regions.’’! In 1888 Professor Williams examined in the field typical sections of the Devonian rocks of Devonshire, England, and later stated that “it appears probable that the limestones of South Devonshire repre- sent the general interval between the close of our Corniferous [Onon- daga] and the early part of our Chemung formation.”’? Professor Renevier in 1896 classed the Hamilton flags and Marcellus .shales together and regarded them as having been deposited during the same general period of time as the Tentaculite slates (lower part) of Thuringia, Hesse, Nassau, and Bohemia; the Wissenbach or Orthoceras slates of Nassau; the Lenne slates (in part) of southern Westphalia; and the schists with Phacops potieri of Brittany; all of which were correlated with the Couvinien age or stage, which he gave as the lower one of the Middle Devonian or Eifélien epoch or series.% Dr. Frech draws the line between the Paleodevonic and the Meso- devonic of New York at the top of the upper Oriskany sandstone, and considers the Mesodevonic as composed of the Ulsterian and Erian seris, in the latter of which are the Marcellus shales, Hamilton beds, and Stringocephalus beds of Canada. At an earlier date Dr. Frech, in his summary of the important occurrences of the Devon- ian, gave the Marcellus shale and Hamilton group as forming the upper part of the Middle Devonian, and correlated them as beginning in the time of the upper part of the Calceola sandalina stage and continuing through that of the Stringocephalus burtini of Rheinland.$ In this same table the Marcellus and Hamilton considered together are correlated with the upper part of the Eifélien (Calceola shales t Compte rendu, Fourth Session, International Geological Congress (London, 1888), 1891, Appendix A, p. 142; also issued as Report of the Sub-Committee on the Upper Paleozoic (Devonic), by H. S. WILLIAMS, C, 1889, p. 22. 2 American Journal of Science, Third Series, Vol. XXXIX (1890), p. 36. 3 Chronographe géologique, 2d ed. of Tableaux des terrains sedimentaires; Compte rendu, Sixth Session, International Geological Congress (Zurich, August, 1894); Lau- sanne, March, 1897). 4 Lethaea geognostica, 1, Lethaea palaeozoica, Vol. II, Part IV (1902), p. 690. 5 [bid., Vol. II, Part I (1897), Table XIX, opposite p. 256. 37° CHARLES S. PROSSER of Couvin) together with the entire Givétien (which is composed in ascending order of the red sandstone and conglomerate of Vicht and Stringocephalus limestone of Givét) of Belgium; while they are given as equivalent in England to the Ilfracombe beds, with probably additional ones below and above, of North Devon; and to the upper part of the Calceola shales of Hopes Nose and Ogwell House, suc- ceeded by the diabase and scale stone of the Ashfrington series and the Stringocephalus limestone of South Devon. In another part of the work Dr. Frech, in comparing the North American and Rhenish Devonian, said: In the Corniferous [Onondaga] limestones the faunal diversity is less sharply defined than in the lower formations; but in this case, as in the higher Hamilton group, still distinctly perceptible. The latter is often developed in the form of sandy marl and calcareous sand, and the peculiar faunal similarity with the Rhenish Lower Devonian partly rests upon this harmony in facies. But, on the other hand, the marl (Moscow shale), for example, where it forms on Cayuga Lake the greater part of the Hamilton, has a perfect agreement in facies with the * Calceola marl, and likewise the Encrinal limestone reminds one of a similar inter- stratified limestone. .. . . The fauna of the American Middle Devonian, whose chief representatives the Hamilton group contains, is, notwithstanding some corresponding features, yet, on the whole, so different that one must assume the existence of a special sea province also in Middle Devonian time differing from the Rhenish.? Finally, at the close of this section is the statement that the Mar- cellus shale corresponds to the lower part of the stage of the Maene- ceras terebratum? of Rheinland, which Dr. Frech puts in the stage of the Stringocephalus burtint. De Lapparent considers the Middle Devonian of North America as composed of the Corniferous (Onondaga) limestone, Marcellus shale, and Hamilton beds. The Marcellus shale he correlates with the upper part of the Eifélien stage and the lower part of the Givétien, while the Hamilton beds represent the remaining and greater part of the latter stage. He also gave the lower Marcellus shale as rep- resenting the upper part of the shales of Ogwell House, and then the remaining portion together with the Hamilton beds as synchronous with the Ilfracombe or Plymouth beds of Devonshire, England.4 Professor Kayser in the table of the Devonian formations of New 1 [bid., pp. 214, 215. 2 [bid., p. 216. 3 Traité géologique, 4th ed. (1900), p. 857. 4 [bid., p. 869. THE ROMNEY FORMATION OF MARYLAND 371 York, gives the Mesodevonic as composed of the Marcellus shale and Hamilton beds,’ but in the text he says: The American geologists generally still classify the Onondaga limestone as Lower Devonian; according to European experience, one would be rather inclined to classify it entirely or mostly as Midddle Devonian. The great simi- larity of the characteristic Spirifer acuminatus Con. with our S. cultrijugatus argues for this classification.? Regarding the classification of the Hamilton the professor says: Although the Hamilton shale locally might represent the entire Middle Devon- ian, yet, on the whole, it corresponds to the upper division. This is surely shown by the frequent overlying beds of the Tully limestone and Genesee shale, the first of which contians the Brachiopod fauna of our Iberg limestone (Rhynchonella venustula =cuboides, etc.).3 Finally, this writer has given the correlation of the Middle Devon- ian of Europe and North America in the following table: RHEINLAND AND BELGIUM. BOHEMIA. NORTH AMERICA. G3, Stringocephalus | Wissenbach G2; Hamilton beds, limestone, and G* Marcellus beds, Caleceola shales. Lenne slates. Mnenian limestone. Onondaga limestone+ Dr. Hermann Credner gives the Middle Devonian of New York as composed in ascending order of the Upper Helderberg (Onondaga), Marcellus shale, and Hamilton sandstone, shale, and limestone. The Upper Helderberg he correlates with the Eifélien and stage of the Calceola sandalina, and the Marcellus and Hamilton with the Givétien and stage of the Sivingocephalus burtini.s Sir Archibald Geikie considers the Middle Devonian of New York as composed of the Marcellus and Hamilton groups,° while the same division in Europe he gives as composed of the Eifélien and Givétien, with which he correlates the Marcellus and Hamilton.7 GENERAL DISTRIBUTION OF THE MESODEVONIAN. Rocks of the Mesodevonian age have a considerable distribution, aside from that of the eastern United States and Canada, for they t Lehrbuch der geologischen Formationskunde, 2d ed. (1902), p. 150. 2 [bid., p. 151. 3 [bid., p. 151. 4Ibid., p. 155. 5 Elemente der Geologie, oth ed. (1902), p. 447. 6 Text-Book of Geology, 4th ed., Vol. II (1903), p. 997. 7 Ibid., “The Geological Record,’”’ opposite p. 86r. 372 CHARLES S. PROSSER have been identified and described in Nevada; the dolomite of Mani- toba contains the European species Stringocephalus burtint; S pirijer mucronatus has been found upon the banks of the Albany River south of Hudson Bay; the fauna of the Hamilton shales occurs in the Mackenzie Valley from the Clear Water River to the Arctic Ocean; while it is also reported from the Porcupine River, a western tributary of the Yukon in Alaska, and perhaps also on Kouiou Island, in the southern part of that territory. In South America, in Brazil, in the province of Para,in the Ereré district, are beds which Katzer refers to the base of the Middle Devonian, and Dr. J. M. Clarke has stated regarding the fauna of the Ereré sandstone that it is remarkably free from species or representatives of subgeneric groups pre- vailing elsewhere in early Devonian faunas, and equally devoid of types which elsewhere pass upward into the later faunas; in other words, it is, with all its resemblance to the Hamilton, a more typical and better-defined Middle Devonian fauna than that;* ! while Dr. Frech reports Middle Devonian in Bolivia and Cleland from the Jachel River in Central Argentina.? On the eastern conti- nent Middle Devonian rocks occur in England in northern and south- ern Devonshire, in northern France and southern Belgium, in the region of the Vosges, in the Central Plateau and the Montagne-Noire of France, and in the Pyrenees and Spain. In central and eastern Europe they occur in the Eifel, Rheinland (Nassau), Hartz, Thur- ingia, Bohemia, Galicia, Russian Poland, the Carnic Alps, and on the Bosporus. These rocks also cover a large area of eastern Russia and the western slope of the Urals, extending to the border of Fin- land on the north. In Asia Middle Devonian rocks occur in Siberia, China, and on the south side of the Tian-Shan Mountains in Central Asia. In Australasia they are found in New South Wales, Victoria, and Tasmania; and they also probably occur in Africa.3 CHARLES S. PROSSER. OHIO STATE UNIVERSITY, Columbus, Ohio, July, 1904. t Archivos do Museu. Nacional do Rio de Janeiro. Vol. X (1899); author’s English edition (1900), p. 90. 2 Bulletin No. 206, U. S. Geological Survey, (1903) p. 19. 3 For this account of the distribution of the Mesodevonian the writer is largely indebted to DE LAPPARENT’S Traité de géologie, FRECH’S Lethaea palaeozoica, and Kayser’s Lehrbuch der geologischen Formationskunde. GRANITES OF NORTH CAROLINA. CONTENTS. INTRODUCTION. DISTRIBUTION OF THE GRANITES. TYPES OF THE GRANITES. Normal granites. Porphyritic granites. Relations between the porphyritic and the even-granular granites in the Mooresville -area. Granite-gneiss. RELATIONS BETWEEN THE THREE GRANITE TYPES. GRANITES OF THE COASTAL PLAIN REGION. General characters. Lithological characters. The Wilson pink granite area. The Elm City area. The Rocky Mount area. The Wadesboro-Rockingham porphyritic granite area. GRANITES OF THE PIEDMONT PLATEAU REGION. General characters. The northeastern Carolina Granite Belt. Lithological characters. The Raleigh area. The Greystone area. The Main Granite Belt. General characters. Lithological characters. The Dunn’s Mountain area. The Mooresville area. The Mecklenburg county areas. The Western Piedmont gneiss and Granite Belt. General characters. Lithological characters. The Mount Airy granite area. GRANITES OF THE APPALACHIAN MOUNTAIN REGION. General characters. Lithological characters. The Unakite area. t Published by permission of the state geologist of North Carolina. 373 374 THOMAS L. WATSON STRUCTURAL FEATURES OF THE GRANITES. Megascopic structures. Joints. Slickensides. Schistosity. Basic inclusions. Microscopic structures. Granophyric structure. Micropoikilitic structure. INTERSECTING DIKES AND VEINS. Basic igneous dikes. Granite dikes. Pegmatite and aplite. Quartz veins. RELATIONS BETWEEN THE JOINTS OF THE GRANITE AND DIKES OF Basic IGNEOUS ROcKs. AGE RELATIONS OF THE DIKES oF Basic IGNEOUS ROCKS. INTRODUCTION. THE following paper is based on a study of the granites and gneisses of North Carolina during the past field season, while I was engaged on the State Geological Survey in a field study of these rocks. The economic report treating of the granites has been prepared and will form a part of a general report on the building-stones of North Carolina to be published by the State Survey. A general summary setting forth the more important points in the petrography and struct- ure of these rocks is offered in the present paper as a separate contri- bution. DISTRIBUTION OF THE GRANITES. North Carolina is divisible naturally into three principal physio- graphic provinces, which, named in order from east to west, are: the Coastal Plain, the Piedmont Plateau, and the Appalachian Moun- tains. ‘These form a part of the continuation of the same provinces to the north and south of North Carolina; and they observe approxi- mate parallelism with each other and with the general trend of the coast line of the southeastern Atlantic states. The granites of North Carolina are distributed over parts of each of these three provinces; the areas being fewer and smaller over the Coastal Plain region, and more numerous and larger over the Piedmont Plateau region. Gran- 375 GRANITES OF NORTH CAROLINA = RISER SIGEAY 6 MOIDBY NVaivid iNONMGZIS ite N afew aa NOIDay NiVLNNON 376 THOMAS L. WATSON ite is extensively developed in many places, within the limits of the Piedmont Plateau, and it forms over many parts of the plateau one of the dominant rock types. The distribution of the principal gran- ite areas of the state is shown on the accompanying map. Over all parts of the state the rocks are profoundly decayed from weathering and, as a rule, are buried under a deep mantle of loose residual decay. Notwithstanding the extensive decay of the rocks, exposures of the moderately fresh granite are comparatively numerous and assume oftentimes very large dimensions. The granite forms unreduced residuals or ridges, as shown in Fig. 7, and it is further exposed in bowlders and ledges and flat-surface masses, both along the stream courses, and more or less remote from the streams on the interstream areas. (Figs. 2 and 3.) ) Numerous quarries have been worked in the outcrops of granite throughout the state, and the stone has been used for all purposes for which granite is ordinarily adapted. TYPES OF THE GRANITES. Based on texture and structure three types of the granitic rocks are distinguished: (1) the massive even-granular (normal) granites; (2) the porphyritic granites; and (3) the banded or schistose granites— granite-gneiss. Field and laboratory study develops close similarity in the three types in mineral composition, and, with one exception, the even-granular and porphyritic textures represent different phases of the same rock-mass. Moreover, the granite-gneisses differ essen- tially from the massive granites, from which they have been derived, only in the pronounced schistose structure subsequently induced by pressure-metamorphism. NORMAL GRANITES. The even-granular granites have wide distribution, and they compose largely the bulk of the granites occurring in North Carolina. As a rule, they vary from fine- to medium-textured rocks, rarely coarse, and are of pink to gray color. With only a few exceptions, light to medium and dark gray is the prevailing color of the granites. Variation in structure is from massive to schistose rocks. ‘They are mixtures of orthoclase and plagioclase feldspars and quartz, with a GRANITES OF NORTH CAROLINA Bil variable quantity of biotite and, in places, some additional horn- blende as the characterizing accessories. ‘The prevailing large amount of plagioclase feldspar in the rocks is a very noteworthy feature; rarely does this constituent sink to very subordinate proportions, but usually it is almost equal to, or even exceeds in amount, the potash Fic. 2.—Granite bowlders on Dunn’s Mountain, near Salisbury, North Carolina. Showing weathering along the joint-planes. feldspars; and only in a few of the sections studied does it entirely fail. In many of the areas the poverty of the rocks in the ferromag- nesian silicates is a striking feature—a fact well illustrated in the large and important granite area in Rowan county, known as Dunn’s Mountain, and its southwestern extension. PORPHYRITIC GRANITES. In all respects save that of texture the porphyritic granites are identical with the normal granites. ‘They are usually coarser in tex- ture than the even-granular rocks, but mineralogically the two are 78 THOMAS L. WATSON Oo identical. With one exception, the two textures grade one into the other, representing different textural phases of the same rock-mass. Porphyritic granites have wide distribution over the state in asso- ciation with the even-granular phases of the rock. ‘The most impor- tant areas are: (1) The Wadesboro-Rockingham area to the east of Charlotte and near the South Carolina line, extending over contiguous parts of Anson and Richmond counties. Over parts of this area the feldspars are colored a rich olive-green, representing apparently a superficial phenomenon caused by some form of surface alteration. A second difference noted is in the distribution and character of the biotite, which occurs in hexagonal plates of single or grouped individ- uals occupying distinct areas. (2) The Gastonia area in Gaston county, west of Charlotte. (3) The Mooresville area in Iredell county. (4) The Cabarrus county area, three miles southwest of Concord; and a second area in the same county in the vicinity of Landers station on the Southern Railroad, to the north of Concord and near the Rowan county line. The porphyritic rock occurring three miles southwest of Concord is quite different in appearance from that of any other area of porphyritic granite known in the state. Here the rock is uniformly coarse-textured and largely feldspathic, composed of large bluish-gray feldspars without pronounced crystal outline. The rock contains hardly more than a trace of groundmass, but is made up chiefly of the large feldspars wrapped about each other and closely interlocked. Biotite in small irregular shreds is one of the chief minerals in the scant groundmass. (5) The Salisbury area, which forms one of the most extensive belts in the state and is traced over parts of Rowan, Davie, and Forsyth counties. Smaller and less important areas of similar porphyritic granite have been noted in several other counties in the state. Excepting the area three miles southwest of Concord in Cabarrus county, the porphyritic granite of the different areas mentioned above is closely similar in color, texture, and mineral composition. ‘The groundmass is usually a medium-coarse gray granite composed of feldspar, quartz, and biotite. Some variation in the amount of the last constituent, biotite, is noted from place to place, resulting in a corresponding variation of color from light to dark gray. The porphyritically developed mineral is potash feldspar. As a rule, GRANITES OF NORTH CAROLINA 379 phenocrysts are marked in the different areas by idiomorphic out- lines; and they are prevailingly of large size, measuring in extreme cases more than two inches long by one inch across. ‘They are usually white, with pinkish ones not uncommon, and, in most cases, they contain more or less included biotite, frequently as large in size as Fic. 3.—Granite pinnacle on the west slope of Dunn’s Mountain, near Salisbury, North Carolina. Height of pinnacle more than twenty-five feet. that of the groundmass constituent. Under the microscope other inclusions of the groundmass constituents are contained in some of the phenocrysts. For this and other reasons elsewhere’ stated by me, the phenocrysts of the Carolina porphyritic granites are regarded as having been formed largely, if not entirely, 7m place and not as often urged of intratelluric origin. Generally the phenocrysts do not grade into the similar ground- mass constituent, but they are, in most cases, conspicuously developed « “On the Origin of the Phenocrysts in the Porphyritic Granites of Georgia,” JouRNAL oF GEoLocy, Vol. IX (1901), pp. 97-122. 380 THOMAS L. WATSON and are sharply defined from the groundmass feldspar. No marked or definite orientation has been observed among the phenocrysts in any of the areas, but in some of the exposures a slight tendency toward such was indicated. Flow structure in the groundmass is entirely absent from all of the areas studied. The porphyritic feldspars almost entirely fail in places, reappearing again within a short distance in their usual abundance and of conspic- uous development. Over most of the areas the probable average ratio of phenocrysts to groundmass is approximately one to one, though extreme variations from this ratio in either direction occur. In the transitional phases of the rock where gradation from the porphyritic to the even-granular texture is observed, the phenocrysts are so few and are so widely scattered that the rock could hardly be termed porphyritic. On the other hand, in the porphyritic granite area three miles southwest of Concord, in Cabarrus county, the rock is composed almost entirely of the large feldspar individuals with very scant groundmass. In all cases the phenocrysts display good cleavage development and twinning on the Carlsbad law. In many of the areas where outcrops of the fresh or moderately fresh granite are few, the rock can be traced almost as readily by its residual decay. The loose phenocrysts are scattered over the sur- face in places in a partially altered condition, often split into smaller fragments along the cleavage directions. Relations between the porphyritic and the even-granular granite in the Mooresville area.—Like the other areas studied, the porphyritic and non-porphyritic granite of the Mooresville area differ from each other only in texture. Without going into detail descriptions of the two texturally unlike rocks, evidence is afforded at several places to the south and the southwest of Mooresville for regarding the one as intrusive in the other, and they do not represent separate facies of the same granite mass. Were the latter true, the line between the two rocks should mark a transitional zone from the porphyritic to the non- porphyritic granite, and not indicate in the few exposures examined a sharp contact. Furthermore, certain phenomena along this line would be difficult of explanation on this supposition. On the other hand, the evidence strongly suggests that the porphyritic rock is the oldest, and that the even-granular granite is intrusive in it. The field GRANITES OF NORTH CAROLINA 381 evidence supporting this belief may be summarized as follows: the sharpness of contact between the porphyritic and the non-porphyritic granite; the prevailing coarser texture of the porphyritic granite along the line of contact than that of the non-porphyritic rock; the banding in places along the contact; inclusions of the porphyritic granite in the even-grained granite; and the occurrence of probable apophyses of the fine-grained granite penetrating the porphyritic granite. GRANITE-GNEISS. True granite-gneisses are found in many places over the state, but perhaps the most typical areas are: (1) the unreduced residual locally known as Rockyface Mountain in Alexander county; (2) the area near Balfour station in Henderson county; and (3) the area on the Josey-Bogers places, three miles southwest of Faith in Rowan county. Wherever studied, the granite-gneisses are biotite-bearing, fine to medium texture, and closely resemble the massive granites in all essentials except that of schistose structure. Some effects of pressure- metamorphism are indicated either megascopically or microscopically in the granites of most of the areas studied. Granites of complete schistose structure were not definitely traced in any single area into entirely massive ones, although this may be the real condition in many cases. Lack of sufficient exposures of the rocks renders this uncertain in the North Caroling areas. * RELATIONS BETWEEN THE THREE TYPES OF GRANITE. Enough description has been detailed above under the even- granular, porphyritic, and schistose granite to make fairly clear, so far as the field study is possible, the relations between the three tex- turally and structurally unlike granites here distinguished. It seems reasonably certain, so far as I have been able to interpret the field relations, that the porphyritic and non-porphyritic granites represent, with one or two possible exceptions different facies of the same granite mass. The principal exception noted above is that of the Mooresville area in Iredell county, where the normal granite is intrusive in the porphyritic granite. The exact relations of the schistose (granite-gneisses) to the even- granular massive granites are less clear, and for lack of adequate 382 THOMAS L. WATSON exposures no definite statement can be made regarding whether the granite-gneisses are the more schistose transitional portions of a more massive granite stock. The completely schistose structure has nowhere been positively traced into the more massive granites, though this may be the true relation. That the areas of schistose granites mentioned above are the metamorphosed equivalents of the massive granites apparently admits of no reasonable doubt. As pointed out in my study of the Géorgia granites,’ the exact relationship is an important one as bearing on the question of the relative age of these rocks. If the schistose rocks (granite-gneisses) cannot be traced into granites of more massive structure, but the two are separate and distinct, then clearly two periods of intrusion must be admitted, separated by an interval of intense pressure-metamor- phism, which resulted in inducing the secondary schistose structure in the earlier massive granites, represented, if the postulate is correct, by the present granite-gneisses. If, on the other hand, the two rep- resent phases of the same rock-mass in which the action of dynamic forces has been greater in some parts of the massif than in others, then the same age must be assigned them. As elsewhere shown in my study of the Georgia granites, the first condition apparently obtained, and the granites are not all of the same age, but at least two separate periods of intrusion of nearly identical material are indicated. c GRANITES OF THE COASTAL PLAIN REGION. GENERAL CHARACTERS. The line marking the contact between the Coastal Plain and the Piedmont Plateau formations in North Carolina is a very irregular one. It enters North Carolina from Virginia a short distance east of the Warren county line and crosses the state in a general southwest direction, passing into South Carolina a short distance southwest of Wadesboro. To the east of this line are the loose unconsolidated formations of the Coastal Plain. In many places near the contact of the Coastal Plain formations with the rocks of the Piedmont Plateau, but falling well within the t “Granites and Gneisses of Georgia,” Bulletin No. 9-A, Geological Survey of Georgia, 1902. GRANITES OF NORTH CAROLINA 383 limits of the former, the thin veneer of loose unconsolidated sands and gravel has been stripped from the surface, mainly along the streams, exposing comparatively small, irregular, somewhat elongate areas of the crystalline rocks, composed either in part or in whole of granite. In such areas the general nature of the granite exposures is in the form of ledge and bowlder outcrops, and as flat-surface masses a short distance back from the streams. The inlirs of crystalline rocks mark the eastward extension of the Piedmont crystallines beneath the Coastal Plain sediments. Some of the schists and gneisses composing parts of these areas are the metamorphosed equivalents of original igneous masses. They include both acid and basic types. The principal granite areas of the Coastal Plain are exposed in Wilson, Edgecombe, and Nash counties to the east of Raleigh, and in Anson and Richmond counties in the vicinity of Wadesboro near the South Carolina line. They are composed of massive biotite gran- ites, In one instance hornblende-bearing, which vary from fine even- granular to coarse porphyritic rocks of gray to pink color. The area to the north of Elm City, in the extreme northern part of Wilson county, contains some hornblende in addition to the biotite. LITHOLOGICAL CHARACTERS OF THE GRANITES. The Coastal Plain granites are described separately by areas in order to bring out more clearly their close similarity to the granites of the Piedmont region. The Wilson pink granite area—The rock is uniformly coarse- textured, of moderate pinkish-red color, and displays a pronounced porphyritic tendency. Feldspar is greatly in excess of the other minerals. The largest feldspar individuals measure one to two inches long. ‘They are pinkish-red in color, exhibiting good cleavage development and twinned on the Carlsbad law. Microscopically the feldspar and quartz make up not less than 95 per cent. of the rock. A rough estimate indicates about 80 per cent. of the former and 15 per cent. of the latter. Of the feldspathic content about 50 per cent. is potash feldspar, most of which is orthoclase, with but little microcline; and 30 per cent. is plagioclase in very large, finely striated laths, whose extinction angles measured against the twinning bands indicate 384 THOMAS L. WATSON albite. Biotite is but sparingly present in the rock and is largely altered to chloride. Occasional small grains of magnetite and scattered inclusions of prismatic apatite, with some secondary musco- vite, kaolin, and calcite, complete the list of accessories. The weathered surface of the granite is ight in color. The feld- spars are dull and opaque, and have lost much of their decided pinkish- red color, so characteristic of the same constituent in the fresh rock. The Elm City area.—The Elm City granite is strongly contrasted with the Wilson pink granite described above. It is much finer in texture, of gray color, and contains more biotite and quartz than the Wilson pink granite. Feldspar predominates and consists of. the potash varieties and much striated acid plagioclase. Extinction angles measured on the plagioclase correspond to an albite of the composition Ab,,An,. Areas showing alteration to kaolin and mus- covite cloud some of the feldspars. Micropoikilitic structure is strongly developed in some of the larger individuals of potash feld- spar, the inclusions of which consist of large microscopic grains of quartz and striated feldspar. Quartz is of the usual kind, forming distinct areas of an interlocking finer mosaic. Biotite is distributed through the sections as small plates of brown color and strong pleochro- ism, bleached to green on alteration, and is further altered to chlorite and some epidote. In one of the thin sections a few additional small crystals of compact hornblende, showing the usual development of the prismatic cleavages, occur. The principal accessories include zircon, apatite, a little pyrite, titanite, and ilmenite. The granite shows evidence both megascopically and microscopically of dynamic metamorphism, manifested, in places, by a rather pronounced schistose structure. The quarry opening indicated penetration of the granite by a dike of very finely schistose amphibolite, closely jointed and containing many disseminated small grains of pyrite. Additional quartz veins penetrate the granite, wrapped, in cases, by films of thin layers of hornblende. The surfaces of the joints are slickensided, coated with a thin veneer of yellowish-green mineral substance, probably epidote in part. Considerable epidotization characterizes the rock in places. The Rocky Mount area.—Westward from Rocky Mount the granite is associated with an irregular but large body of crystalline schists GRANITES OF NORTH CAROLINA 385 which were derived in part from original igneous rocks. Contacts between the granite and the schists were nowhere observed, but the field evidence clearly suggests that the granite is the younger rock. Pegmatite veins, ranging from a fraction of an inch to more than six inches across, composed of pinkish feldspar and quartz, with a very subordinate amount of biotite, penetrate the granite. Nearly all gradations from those veins containing mostly feldspar to those containing mostly quartz occur. In some of the veins the quartz forms a narrow central band in the feldspar. Others still are banded with a fine-textured granite of the same mineral composition as the inclosing rock. ‘Some of these are true veins of segregation; others are less certain of interpretation, mainly because of the lack of expo- sures for examining them. Megascopically the rock is somewhat unlike that described above, the Elm City area. It is entirely massive, of medium texture, and gray color. In mineralogy it differs from the Elm City rock in the entire absence of hornblende, and in the presence of only a slight amount of plagioclase. Orthoclase and microcline have nearly equal distribution. Quartz forms a fine mosaic, occupying distinct areas. Intergrowths of micrographic structure are rather abundantly distributed through the sections in irregular small areas. The accessory minerals are biotite, chlorite, epidote, apatite, zircon, and magnetite. The Wadesboro-Rockingham porphyritic granite area— This is a large area of porphyritic granite lying partly in Anson and partly in Richmond county, near the South Carolina line and to the west of Charlotte. About one mile east of the westernmost exposure of the granite the Triassic sandstones first appear, overlying unconformably the crystalline schists. Between the granites and the sandstones is an area of variable crystalline schists, principally micaceous and quartz schists. Between Rockingham and the easternmost outcrops of the granite the rocks are concealed by Coastal Plain sands and afford no evidence of the nature of the underlying crystallines. In and about the town of Rockingham the principal outcrops indicate a greenstone schist much crushed and fractured, and strongly suggests derivation from an original basic igneous rock. ‘The schists to the east and west of the granite mass must be regarded as older in age 386 THOMAS L. WATSON than the granites, and they form the country rock into which the granite was intruded. Exposures of the granite are in the nature of huge bowlders, ledges, and flat-surface masses. Sections of the fresh and weathered granite are seen to advantage in the cuts along the Seaboard Air- Line Railroad to the east and west of Lilesville. It is a coarse-grained porphyritic biotite granite of gray color, with pinkish and yellow tones characteristic in places. The groundmass is medium coarse-textured, dark gray, granite, containing, as a rule, much biotite. This is subject, however, to some variation. In the railroad cut one and a half miles west of Lilesville the biotite has quite a different occurrence and distribution in the granite from that of any other exposure examined in the area. Here the biotite is usually in sharp idiomorphic hexagonal plates, distributed through the rock as single individuals and aggregates, which occupy distinct areas. Elsewhere it occurs as irregular shreds crowded close together and freely distributed through the groundmass. The phenocrysts are composed of potash feldspar, showing marked cleavage development and twinned on the Carlsbad law. They are usually of a pinkish hue, and may be either idiomorphic or allotrio- morphic in outline; and, they contain usually some included biotite. The ratio of phenocryst to groundmass is variable, but will probably average about one to three. The principal minerals are quartz, orthoclase, microcline, plagio- clase (near oligoclase), biotite, chlorite, apatite, zircon, magnetite, and a few other less common accessories. Microcline may fail in some sections and be present in large proportion in others. Plagio- clase is a constant constituent and is usually present in large amount. GRANITES OF THE PIEDMONT PLATEAU. GENERAL CHARACTERS. The Piedmont region in North Carolina is composed of a number of geologic belts, approximately parallel to each other and crossing the state in a general northeast-southwest direction. These belts are composed, as a rule, of unlike rocks, probably of different ages. A variety of granites are found within the limits of the Plateau region; always biotite-bearing, with additional hornblende in some areas, GRANITES OF NORTH CAROLINA 387 and muscovite when present in very subordinate amount. In tex- ture, variation is from even-granular to porphyritic; and in structure from massive to schistose. As a rule, some shade of gray prevails, though pink is largely characteristic of certain areas. The Plateau granites are described under the following belts: The Northeastern Carolina Granite Belt, the Main Granite Belt, and the Western Carolina Gneiss and Granite Belt. THE NORTHEASTERN CAROLINA GRANITE BELT. This area comprises a part or the whole of five counties located in the extreme northeast part of the Piedmont region, extending north- ward from Raleigh. With but few exceptions, the granites of this belt are partially schistose in structure. The characterizing accessory, biotite, is variable in quantity; imparting accordingly either a light or a dark gray color to the rock. In places the feldspars are of a pronounced pinkish hue and with the subordinate amount of biotite present, the rock assumes more or less of a mixed pinkish-gray color. The granites are usually even-granular in texture, though the porphy- ritic tendency is somewhat emphasized in places. ‘They do not differ - essentially in mineralogy, although hand specimens from different areas in the belt may bear no resemblance to each other. LITHOLOGICAL CHARACTERS. Two areas, differing somewhat widely in the hand specimens, but which may be regarded as representative of the granites of the belt as a whole, are selected for description, in order to make clear the general characters of the rocks as a whole. ‘These are the Raleigh area in Wake county and the Greystone area in Vance county. The Raleigh area.—Studied in the quarries opened within the eastern limits of the city of Raleigh, the rock is a fine even-textured, medium-gray, biotite granite-gneiss, completely interlaced by inter- secting pegmatites. The principal minerals are quartz, orthoclase, microcline, acid plagioclase near oligoclase, brown biotite, muscovite, zircon, apatite, chlorite, and epidote. Quartz is of the usual kind, and is often intergrown with feldspar in micrographic structure, clearly indicating that its period of formation began before that of the feld- spar closed. In addition, it occurs with other minerals, especially 388 THOMAS L. WATSON plagioclase, in the form of inclusions in the larger potash feldspar individuals developing micropoikilitic structure. Microcline nearly equals orthoclase in quantity, and plagioclase is hardly less than both. Carlsbad twinning is sometimes observed in the feldspar. Biotite is brown in color, strongly pleochroic, and is largely altered to chlorite and some epidote. An occasional shred of muscovite is intergrown with the biotite. The other accessory minerals present no noteworthy features. The larger feldspar and quartz individuals are partially or entirely surrounded by finer mosaics of the same minerals, denoting peripheral shattering from dynamic forces. Strained shadows in the minerals are further characteristic of the same pressure effect. In the northern part of Wake county, in the vicinity of Rolesville, a light pinkish biotite granite occurs, which differs from the Raleigh granite in the hand specimen, but does not differ essentially in mineral- ogy. Plagioclase is very variable, equaling in some sections the potash feldspar and nearly failing in others. Orthoclase and microcline are in nearly equal proportion. Small irregular areas of micrographic intergrowths of quartz and feldspar are quite abundantly distributed through the thin sections. The Greystone area.—The Greystone quarries are among the largest in the state. The typical Greystone granite is medium gray in color, of pronounced schistose structure, and in texture varies from fine- to medium-grained. More or less tendency toward a poorly defined porphyritic texture is manifested in some parts of the area. The feldspars are partly of a pronounced pinkish hue, which impart a slightly mixed pink and gray color to the rock. In some respects hand specimens from the Raleigh and Greystone areas bear some resemblance to each other; in others they are strikingly different. In mineral composition they are essentially identical. The principal minerals are quartz, orthoclase, microcline, micro- perthite, a little acid plagioclase near oligoclase, biotite, apatite, zircon, chlorite, muscovite, and kaolin. Microscopically the most noteworthy feature of the granite in this area is the very small amount of plagioclase present, which entirely fails in some of the thin sections. When present, the plagioclase individuals show broad twin- ning bands having extinction angles near that of oligoclase. Micro- GRANITES OF NORTH CAROLINA 389 cline is a constant constituent, and, though somewhat variable, it may equal in amount the orthoclase. Intergrowths of orthoclase with a second feldspar as microperthite are fairly characteristic. Simultaneous crystallization of the quartz and a part of the feldspar is clearly indicated in micrographic intergrowths of the two minerals, and in the development of the micropoikilitic structure in the potash feldspars. The drop-like inclusions in the feldspar consist of quartz and other feldspar species, principally plagioclase. The feldspars are partially clouded from irregular patchy areas of alteration to minute scales of muscovite and kaolin. Carlsbad twinning is not uncommon among the feldspars. Biotite is in irregular shreds of brown color and strong pleochroism, and in many of the sections it is extensively altered to chlorite. Occasional shreds of muscovite are intergrown with the biotite. Zircon inclusions are somewhat common to the three principal constituents, feldspar, quartz, and biotite. The effects of dynamic forces are manifested in the thin sections by peripheral shattering and recrystallization, and in strained shadows in the principal minerals, quartz and feldspar. Mega- scopically the arrangement of the minerals along approximately parallel lines is manifested in the development of a thinly schistose structure. The granite areas farther north in Warren-county bear no resem- blance in the hand specimens of the rocks to the Raleigh and Grey- stone granites, though they differ only slightly in mineralogy. In one exposure near Warren Plains biotite entirely fails, and muscovite is substituted. To the north of this locality biotite again assumes the réle of principal accessory with some muscovite present, and the granite has abundant garnets scattered through it. The granite of the Louisburg area in Franklin county resembles somewhat that of the Greystone quarries in the hand specimens, except that the former is entirely massive. Plagioclase occurs only sparingly in the granite of the Louisburg area, and microcline and microperthite entirely fail. In other essentials the granite from . the two areas is similar. Over. parts of the Louisburg area as well as at Greystone a slight porphyritic tendency is indicated in the rock. 390 THOMAS L. WATSON THE MAIN GRANITE BELT. GENERAL CHARACTERS. This belt is an irregular one in width, occupying the central part of the Carolina Piedmont region, and extending from near the Vir- ginia line in Person county, North Carolina, in a southwest direction across the state into Mecklenburg and Gaston counties, along the South Carolina line. The whole or a part of a dozen counties are included within the limits of this belt. Granite, either massive or schistose in structure, forms the principal rock type over the belt. Texturally two distinct phases of the granite are developed, an even-granular and a porphyritic granite, both of which have wide distribution. With the single exception discussed on p. 381, the two texturally unlike rocks represent phases of the same granite mass. In all other essentials the normal and the porphyritic granites are closely similar. Over many parts of the belt the granite manifests some evidence of the effects of intense dynamic forces, in many instances resulting in the partial or complete development of a secondary schistose struc- ture. In thin sections of the rock from those areas in which the granite megascopically appears entirely massive, evidence more or less pronounced of pressure-metamorphism is shown. The rocks are usually colored some shade of gray, light or dark, in accordance with the proportion of the ferromagnesian mineral present. In some a pronounced pink prevails, and the biotite, which is in very small amount, is not noticeable in the rock. Hardly with- out exception, the rocks are biotite granites, containing in several places additional hornblende, which is the principal accessory in the granite of several areas in Mecklenburg county, although biotite does not entirely fail in these. Muscovite as a primary constituent is sparingly developed, in a few localities, in association with biotite. Microscopically the granites are essentially the same in mineral composition as those described above and to the east of this belt. Plagioclase feldspar is a nearly constant constituent, though subject to some variation. At times it exceeds the potash feldspar in amount, and it does not entirely fail in but one or two of the thin sections. Optically it corresponds to a very acid plagioclase, albite or oligo- clase, or both. Orthoclase and microcline are both usually present GRANITES OF NORTH CAROLINA 391 in the sections, oftentimes in nearly equal proportions, though the microcline is subject to considerable variation. Simultaneous crys- tallization of the quartz and a part of the feldspar is usually indicated in the distributed areas of micrographic intergrowths of the two minerals and in the development of micropoikilitic structure in some of the feldspars, the inclusions consisting largely of quartz and plagio- clase. The usual accessory minerals common to granite are noted in the thin sections of the rocks. The porphyritic granites have been previously described on p. 378, and it is only necessary here to again emphasize the fact that they are in every case biotite-bearing without hornblende, and are the equivalents in mineral composition of the even-granular granites. Three areas, representing different types of the normal granites, are separately described below. These are: the Dunn’s Mountain area in Rowan county; the Mooresville area in Iredell county; and the occurrence of hornblende granite in Mecklenburg county to the north and the south of Charlotte. LITHOLOGICAL CHARACTERS. The Dunn’s Mountain area.—This is a granite ridge twelve to fourteen miles long, having a general northeast-southwest trend and located a few miles southeast of Salisbury. Numerous quarries have been worked over many parts of the ridge which afford excellent opportunity for obtaining fresh specimens of the rock. ‘Two distinct types of the rock are recognized; (a) a very light gray, nearly white, and (6) a pronounced pink. The two are intimately associated. They have the same mineral composition and texture, and are appar- ently phases or differently colored portions of the same rock-mass. The texture is medium even-granular and is fairly uniform. Dynamic metamorphism has manifested itself over all parts of the ridge in a faintly marked schistose structure, and on the Josey-Bogers places three miles southwest of the village of Faith the rock is completely thinly schistose. Over the north slope of Dunn’s Mountain proper pronounced shear zones of the crushed and laminated rock, narrow in width, are developed in the granite mass, striking N. 55°-70° E. The rock surfaces are slickensided, accompanied by considerable epidotization. 392 THOMAS L. WATSON The component minerals are quartz, orthoclase, microcline, at times microperthite, abundant plagioclase, a little biotite, magnetite, chlorite, epidote, titanite, rutile, and occasional garnet. The granite is largely a mixture of feldspar and quartz, with scant biotite as the third principal component. One of the most striking features in the mineral composition of the rock is its poverty in the ferromagnesian mineral and the excessive plagioclase, which latter constituent exceeds the potash feldspars in all the thin sections examined. Plagioclase is in large, stout, rudely prismatic forms, polysynthetically twinned, and corresponds in optical properties to albite and very acid oligo- clase. Both orthoclase and microcline occur, the latter very variable in quantity, being reduced to one or two grains in some sections and in fairly good proportion in others. Microperthite is distributed in small amount through some of the sections. Carlsbad twinning is observed, but it is not so frequent as in some of the granites from other localities in the state. In the pink phase of the rock the feldspars are usually filled with closely crowded dust-like particles of a reddish- brown color, which probably represent some form of iron oxide. Biotite, when present in the thin sections, is of the usual kind, and is distributed through the rock in small, very irregular shreds, altered largely to chlorite. Quartz occupies well-defined areas between the larger feldspars, forming an aggregate of interlocking fine grains, in which feldspar may or may not appear. Crushing and recrystallization from the action of intense dynamic forces are strongly emphasized in all of the thin sections. Finer mosaics of quartz and feldspar border the larger individuals of the two minerals and fill the interspaces. ‘The larger laths of plagioclase are fractured and broken, and in many instances curved and bent, with irregular fractures and strained shadows common to the other essential components. | The Mooresville area —The porphyritic granite of the Mooresville area has been previously described on pp. 380 and 381. Attention is here directed to the even-granular granite, which is a very fine textured rock of dark gray color. It is strongly contrasted with that of the Dunn’s Mountain area described above. Orthoclase and microcline are in nearly equal amount. Unlike the other areas, plagioclase is very sparingly present, not more than a few grains GRANITES OF NORTH CAROLINA 393 being observed in any one of the sections. Some microperthite occurs. Biotite of deep brown color and strong pleochroism is uniformly distributed in large amount through the sections, altered principally to chlorite, a colorless mica, and occasional epidote. Areas of micro- graphic structure are abundantly distributed through some of the sections, clearly indicating the overlapping of the periods of forma- tion of the quartz and feldspar. Pleochroic titanite in crystals and grains is quite freely developed in one of the sections. Inclusions of apatite and zircon are fairly constant. The Mecklenburg county areas.—Biotite-bearing hornblende gran- ite occurs one mile east of Davidson in the extreme northern part of the county, and again five to six miles south of Charlotte in the extreme southern part of the county. Hand specimens of these granites are entirely different from, and bear no resemblance whatever to, those of the types previously described. They are of medium texture and gray color, and in the locality south of Charlotte a few bowlders have been worked off for monuments. Much plagioclase is present in association with the potash feldspar. Compact hornblende with strong cleavage development is the principal ferromagnesian mineral present. It is accompanied by more or less biotite of the usual kind, which varies greatly in amount. The granite to the east of Davidson contains large lath-shaped crystals of the hornblende distributed through it, which measure in extreme cases as much as one and a half inches in length. All grada- tions in the size of the hornblende individuals down to the smallest grains occur. Chlorite is one of the chief alteration products of the ferromagnesian constituent. Much titaniferous magnetite occurs along with some of the usual minor accessories. THE WESTERN PIEDMONT GNEISS AND GRANITE BELT. GENERAL CHARACTERS. The principal rocks of this belt are gneisses of variable mineral composition. A residual of biotite granite-gneiss, closely resembling in the hand specimens and in mineral composition the well-known Lithonia area of granite-gneiss in Georgia,’ occurs in Alexander t “Granites and Gneisses of Georgia,’ Bulletin No. gQ—A, Geological Survey of Georgia, 1902, pp. 125 ff. 304 THOMAS L. WATSON county near Hiddenite. Large areas of biotite granite of medium to coarse texture are located in the northern part of the belt. The Mount Airy area in Surry county, near the Virginia line, may be taken as the type. LITHOLOGICAL CHARACTERS. The Mount Airy granite area.—Quarries are worked one and.a half miles north of the town of Mount Airy, on a ridge slope of con- tinuously exposed granite. In mineral composition the Mount Airy granite does not differ essentially from some of the types already described, though hand specimens of the rock bear little or no resem- blance to each other. The usual granitic minerals are noted, such as quartz, orthoclase, microcline, plagioclase, biotite, muscovite, zir- con, apatite, epidote, chlorite, and magnetite. In addition to these, several grains of allanite have been noted in one of the thin sections. Orthoclase is in excess of the microline in all of the sections examined, accompanied by a larger proportion of finely striated acid plagioclase. Zonal growth and Carlsbad twins are beautifully developed in some of the feldspars. Biotite of brown color and strong pleochroism is the principal accessory, and is altered into chlorite, a colorless mica, and some epidote. Much secondary muscovite derived from the alteration of the feldspar and biotite is present. The large individuals of quartz and feldspar indicate peripheral shattering in finer-grained mosaics of the two minerals filling the interstices. GRANITES OF THE APPALACHIAN MOUNTAIN REGION. GENERAL CHARACTERS. Areas of granite and granite-gneiss resembling essentially in min- eral composition some of the granite types described above are known over parts of the mountain region of the state. Those of the extreme northwest part of the state have recently been mapped.and described by Keith.t Since, with perhaps a single exception, those of the other parts of the mountain region are not unlike certain types already described in this paper, only one type is of special interest here, namely, unakite, which occurs in the vicinity of Hot Springs, t Geologic Atlas of the United States, ‘Cranberry Folio, North Carolina— Tennessee.” U.S. Geological Survey, 1903. GRANITES OF NORTH CAROLINA 395 Madison county, North Carolina, and the contiguous part of Cocke county, ‘Tennessee. LITHOLOGICAL CHARACTERS. The Unakite area.—The unakite area in Madison county, North Carolina, and Cocke county, Tennessee, is the type locality first described by Bradley in 1874. I quote in full Bradley’s description» This name [Unakite] is proposed for a member of the granitic series, from the Great Smoky Mountains, a portion of the Unaka range of the Blue Ridge, which range forms the boundary between Tennessee and North Carolina. The speci- mens thus far seen are from the slopes of the peaks known as ‘‘The Bluff,”’ “‘Wal- nut Mountain,” and ‘‘Max’s Patch,” Cocke county, Tenn., and Madison county, N.C. The rock is said to occur also in Yancey county, N. C., but in a compara- tively inaccessible region. The character relied upon for the separation of the species is the constant replacement of the mica of common granite, or the hornblende of syenite, by epidote. The amount of this ingredient present is quite variable, in some cases even exceeding one-half of the whole mass. The feldspar present is orthoclase, of various shades of pink, forming from one-fourth to perhaps one-third of the whole. The quartz is mainly white, but occasionally smoky; its isolated portions form but a small part, say one-fourth, of the mass; it is veined in structure, but this is probably not a constant character. Small grains of magnetite are scat- tered through the rock, but not so thickly as in many granites. No other ingredi- ents have as yet been detected. Mr. G. W. Hawes has determined the specific gravity at 2.79. The rock is very compact and takes a high polish, and will doubtless prove to be a valuable material for ornamental architecture. The deep weathering of all the rocks of the Southern Appalachians has caused the covering of most of these mountain slopes with deep beds of débris, which conceal most of the solid outcrops; and the dimensions of the bodies of uanakyte are therefore as yet unknown. Apparently forming part of the same series, there are heavy beds of specular iron ore; and the whole series is referred with little doubt to Archaean age. The outcrops of the granite over the surface, as traversed by me, indicate an area of about twenty-four square miles, lying mostly in Madison county, North Carolina, beginning about five miles south- west of Hot Springs. Further detailed work will probably extend the boundaries of the area considerably beyond those mentioned here. Two types of the granite occur in the area, both of which contain epidote. The bulk of the granite or main body of the rock is a dark t American Journal of Science, Vol. CVII (1874), pp. 519, 520. 396 THOMAS L. WATSON pinkish-green epidote granite of medium-coarse texture and fairly schistose or foliated in structure. This is not the unakite proper, but may be properly designated the epidote-bearing rock. It varies from a typical granite in which quartz is present in the usual amount to a nearly quartzless rock of the same color and texture. The unakite proper is a coarse-textured rock composed of yellow-green epidote, dull pink or red feldspar, and quartz. It is not entirely uniform in color and composition, but it grades into a highly feld- spathic rock of pink color on the one hand, and an epidotic rock of yellowish-green color on the other, with still a third gradation observed into pure quartz. Thin sections of the epidote-bearing granite show the principal minerals, quartz, orthoclase, and microcline, in about equal propor- tion, a little plagioclase, biotite, epidote, chlorite, rutile, zircon, apatite, magnetite, pyrite, and kaolin. The epidote distributed through the sections is wholly a secondary product derived from the interaction of the biotite and feldspar. It occurs in the form of minute microscopic granules thickly crowded together, in many cases lying next to the biotite. ‘The mass of granules, when forming an area large enough to be visible megascopically, appears as a single large epidote individual rather than as separate microscopic granules. Besides epidote, the feldspars are altered to a colorless mica, and in some instances the original mineral is completely obscured by the alteration products. At times the feldspars are much fractured and broken, the fissures of which are now filled with another mineral substance. Sometimes alteration has progressed along the lines of fracture in the feldspar, and patches and stringers of deep-green mica line them. ‘Thread-like filaments of rutile, broken into minute segments, are crowded together in the quartz anhedra. Some periph- eral shattering from pressure-metamorphism is indicated in narrow zones of fine-grained mosaics of quartz and feldspar partially sur- rounding these two minerals. Strained shadows and fractures are common to both the larger quartz and feldspar grains. The principal difference microscopically between the unakite and the epidote-bearing rock is that of extreme epidotization of the former, further marked by the absence, in those sections studied, of both plagioclase and ferromagnesian minerals, identified as such. GRANITES OF NORTH CAROLINA 307 Since both of these minerals are present in the epidote-bearing rock, their absence in the unakite is very likely due to extreme epidotization of the unakite, completely altering both the mica and the plagioclase in case these were originally present. That mica or some ferro- magnesian constituent was originally present in the unakite it seems necessary to assume in order to meet the requisite conditions of for- mation of so large a quantity of epidote. The principal component minerals in the unakite are quartz, ortho- clase, epidote, rutile, titaniferous iron oxide, and secondary musco- vite, kaolin, and leucoxene. Rutile is contained in the quartz anhedra as inclusions of hair-like filaments. ‘The orthoclase has a pronounced pink color in the hand specimens which disappears in the thin sec- tions. Here, as in the epidote-bearing rock, conclusive evidence is furnished of the wholly secondary nature of the epidote. Its largest occurrence is in the replacement of the feldspar individuals in the form of a complete mass of microscopic granules, entirely obscuring in many cases the feldspar substance, but preserving the outline of the latter in a more or less perfect manner. The granular masses of epidote appear in the hand specimens as single large epidote anhe- dra. Other feldspar individuals show patchy areas and scattered granules of epidote over their surfaces. Still a third occurrence of the epidote is in irregular broken bands or stringers, following the network of fractures in both the feldspar and the quartz. In some cases these stringers ramify outward from the granular masses of epidote, replacing the feldspar. Again the granular masses often show irregular margins traced outward from the more compact mass into scattered granules of epidote. All gradations between these occurrences of the epidote are traced. Very little of the feldspar is entirely free from epidotization. Some peripheral shattering accompanied by much fracturing of, and strained shadows in, the quartz and feldspar, indicating the effects of dynamic forces, characterizes to some degree all of the thin sections. It seemed quite conclusive, from the different exposures studied of the unakite in its relations to the inclosing epidote-bearing granite, that the unakite is of distinct vein character which can be referred 398 THOMAS L. WATSON very likely to the segregation type.* It does not exist in quantity large enough to be worked over any part of the area traversed by me; hence it has only scientific interest. STRUCTURAL FEATURES OF THE GRANITES. MEGASCOPIC STRUCTURES. Joints.—With possibly one or two exceptions, the Carolina gran- ites are characterized by a strong development of the jointed structure, the planes of which break the rock into polygonal blocks of different sizes, as indicated in Fig. 4. The most noteworthy exception is that of the extensive granite area near Mount Airy, in which no visible jointing is apparent. Careful measurements of the joint-planes were made in all the granite openings visited in the state, and the results can be summarized as follows: Those joints whose planes lie in the northeast and the northwest quadrants respectively, and composing the major jointing; and two minor sets whose planes strike east-west and north-south. In the northeast and the northwest quadrants the limits of variation in the strike of the joint-planes are N. 10° E. or W., to N. 80° E. or W. Out of the total number of joint-planes meas- ured fifty-six lie in the northeast quadrant and forty-five in the north- west quadrant; while nineteen strike in a north-south direction, as against sixteen having an east-west strike. Slickensides.—As a rule, the joint-planes show smooth, more or less polished and striated surfaces, indicating considerable movement in the rocks since the formation of the joints. Striz are developed in a thin coating of yellow to yellowish-green mineral substance, derived from certain minerals in the granite and produced by the rubbing together of the two sides along the plane. Schistosity.—Between the perfectly schistose granites (granite- gneisses) and the perfectly massive granites, nearly all gradations in t Since this paper was written, PHALEN (‘‘A New Occurrence of Unakite,”’ Smith- sonian Miscellaneous Collections, Vol. XLV (1904), pp. 306-16) has published a pre- liminary paper on the occurrence and petrography of unakite at Milams Gap in Virginia. The rock in which the unakite occurs in the Virginia locality is reported by the author to be hypersthene akerite. ‘The epidote of the unakite is secondary, replacing pyrox- ene and feldspar, both plagioclase and orthoclase. The author believes the unakite has originated_from the akerite by hydrometamorphism, aided perhans by dynamic disturbances. GRANITES OF NORTH CAROLINA 3909 schistose structure are shown. Not in the same mass, however; for not in a single instance has such been observed. Whether the entirely schistose structure is more or less sharply and completely separated from the massive, or whether they grade from one into the other, it is impossible to say, owing largely to the lack of sufficient exposures of the rocks. Even in most of those granites which megascopically Fic. 4.—Vertical jointing in granite. City of Charlotte quarry, North Carolina. appear massive, more or less evidence is shown in the thin sections of the effects of dynamic forces, in peripheral shattering or granula- tion, and in strained shadows and fractures in the quartz and the feldspar. In such cases no semblance of rearrangement of the min- eral components along certain definite lines appears. Basic inclusions.—The granite in many of the quarries contains inclusions of basic segregations which were formed from the cooling magma. ‘These are invariably of darker color and finer grain than the inclosing granite, and are entirely massive. Sections from them resemble closely the inclosing granites in mineral composition, except 400 THOMAS L. WATSON that biotite is greatly in excess, while quartz may be somewhat dimin- ished in quantity. It frequently requires the closest examination, megascopically, to detect more of the minerals than the dark ferro- magnesian components, except in a few cases where an occasional porphyritically developed feldspar has formed, so fine-grained and very dark in color are the schlieren. Their distribution in the granites is subject to much variation. From some quarries they are entirely absent, in others they are developed only occasionally, and in others still they are so abundant as to exclude the rock from use in certain higher grades of work. Variation in size is from a fraction of an inch to more than a dozen inches across. Irregularity of outline is usually characteristic of them, variation being from roughly oval-shaped and round to greatly elongated areas, one of whose diameters is several times that of the second. MICROSCOPIC STRUCTURES. Peripheral shattering or granulation and recrystallization has already been mentioned. It is only necessary here to call attention briefly to two fairly constant microstructures in the granites, observed in similar rocks elsewhere, namely, granophyric and micropoikilitic. Grano phyric structure—Granophyric intergrowths of feldspar and quartz are either sparingly or abundantly developed in most of the thin sections examined of the Carolina ‘granites. ‘They display the usual form and development observed in granites in general. So far as it was possible to determine, the intergrown feldspar may be either of the potash or plagioclase species. ‘The areas may be inclosed by the larger feldspar individuals, or they may be formed at or near the contact between the larger feldspar and quartz grains. Their character clearly indicates simultaneous crystallization of the two minerals—an overlapping in the period of formation of the feldspar and quartz. ‘This conclusion finds added confirmation in the micro- poikilitic structure described next below. Micropotkilitic structure.t—In thin sections of the granites from both North Carolina and Georgia, certain of the large feldspar 1G. H. Witttams, “On the Use of the Terms Poikilitic and Micropoikilitic in Petrography,”’ JouRNAL OF GEOLOGY, Vol. I (1893), pp. 179 ff. This paper con- tains numerous references to published accounts of this structure. GRANITES OF NORTH CAROLINA 4o1 individuals are frequently filled with crystals, or grains of other min- erals, arranged without reference to one another or to their host. Hardly without exception, the inclosing mineral is a potash feldspar, and the included minerals are quartz and plagioclase, with an occa- sional shred of biotite. In the phenocrysts of the porphyritic granites this order is frequently reversed, and biotite is often inclosed in larg- est amount. Variation is from several scattered included grains in the feldspar to hosts fairly filled with the inclusions. The inclosed quartz grains are rounded in outline, drop-like in form, while the striated plagioclase may vary from similarly rounded grains to lath- shaped forms. Because of its occurrence in devitrified glasses, Williams* ascribed in some cases a secondary origin to the very abundant micropoikilitic structure in the ancient acid lavas of South Mountain, Pennsylvania, and Maryland. There is no evidence for regarding the structure in the Carolina and Georgia granites of other than primary origin—a fact which aids, as I interpret it, in formulating the order of separa- tion of the feldspars and quartz from the magma. When viewed in connection with the granophyric structure described above, the two microstructures furnish conclusive evidence of the overlapping in the period of crystallization in the potash and plagioclase feldspars and the quartz in the southern granites. A tendency was noted in many of these sections toward maximum development of the two structures in the same thin section. This seeming association was by no means a constant feature, for many sections in which granophyric inter- growths were present indicated the entire absence of the micropoi- kilitic structure. INTERSECTING DIKES AND VEINS. Genetically the intersecting materials are of two kinds, true dikes and true veins. These contrast quite strongly in some instances, and in all cases they show differences to some degree in both texture and composition. BASIC IGNEOUS DIKES. Dikes of basic igneous rocks, principally diabase or its altered form, were observed penetrating the granites in most of the impor- tant areas in the state. A majority of these were noted from surface t G. H. Wrtrams, American Journal of Science, Vol. XLIV (1892, 34 S.), p. 482. 402 THOMAS L. WATSON exposures of the rocks, and hence more or less deeply decayed, while many of them were exposed in the quarry openings, which afford excellent opportunity for observing their relations to certain struc- tural features of the granite discussed below (Fig. 5). In several quarries in widely separated parts of the state a series of some half- dozen or more dikes were observed. Where exposed in the quarries, Fic. 5.—Dike of diabase penetrating granite in the city of Charlotte granite quarry, North Carolina. Part of a second dike is visible in the extreme left of the view. none of the dikes exceed fifty feet across, and usually they are less than four feet. Lithologically two types are indicated. The first is usually an unaltered massive diabase, the thin sections of which show under the microscope the essential minerals and structure of diabase. The pyroxene in this type may be considerably altered in some of the dikes, and in hand specimens the rock presents a rather pronounced green- ish hue, resulting from the alteration. ‘The second type is more or less completely thinly schistose, and is largely composed of horn- GRANITES OF NORTH CAROLINA 403 blende, some feldspar and quartz, and may contain much additional pyrite. The rock is a typical amphibolite derived from an original diabase or diorite or both. The structural relations of the dikes to each other and to the inclosing granites clearly indicate at least two different periods of penetration of the granites by the basic rocks, and therefore different ages for the dikes. More or less complete evidence of schistose structure in the inclosing granites is sometimes noted in those cases where cut by dikes of similar structure. I take it that the period of intrusion of the dike antedates that of the dynamic disturbance induc- ing schistosity into both dike and granite alike. In those cases where the dike rock remains massive and the inclosing granite shows evidence of schistose structure, the reasonable conclusion is that the basic dike material penetrated the granite after the period of dynamic disturbance closed. GRANITE DIKES True granite dikes of normal composition and usually fine texture are numerous in certain areas, but only in one or two instances have they been observed penetrating the granite masses. In the main granite belt of the Piedmont region dikes of this character are quite frequent, penetrating the surrounding rocks, and they must be regarded as apophyses from the granite massifs (Fig. 6). As a rule, they vary only a few degrees from the vertical; are irregular in out- line; fine-grained in texture; composed of light to pink feldspar and quartz, with subordinate amount of mica, which may entirely fail at times; and they range from a few inches to several feet in width. PEGMATITE AND APLITE. Pegmatites are present in large numbers in some quarries and entirely fail in others. As a rule, they do not attain very large size, but are narrow, apparently deep-seated, and of aqueo-igneous origin. Others are limited in extent, surrounded entirely by the granite, marking in such cases true veins of segregation. ‘Texture and com- position of the two are identical. They are chracterized by the usual coarse crystallizations of feldspar and quartz, with subordinate stout, platy, black biotite. The feldspar may be pink or white, showing 404 THOMAS L. WATSON good cleavage development, and twinning on the Carlsbad law. Feldspar is the most abundant constituent, composed of microcline and orthoclase, with little plagioclase; quartz frequently occurring in very small amount. Muscovite has been observed occasionally; and tourmaline, garnet, and the rarer minerals sometimes associated with pegmatites, are strikingly absent. Fic. 6.—Granite dike west of Salisbury, North Carolina. In the Raleigh city quarries where the only true aplite has been observed, the aplite and pegmatite are associated as banded aplite- pegmatite, the aplite being in contact with the pegmatite on one side and the inclosing granite on the other. They are very light-colored rocks, containing but little biotite; are fine-grained; and are only a few inches across. Orthoclase, microcline, very little plagioclase, microperthite, quartz, biotite, muscovite, chlorite, rutile, magnetite, and kaolin appear in thin section. Microscopically they are potash aplites. GRANITES OF NORTH CAROLINA 405 QUARTZ VEINS. Quartz veins of small dimensions cut the granite in a number of quarries; usually, in the ones where pegmatitic intrusions are strongly developed. They are not numerous in any of the openings, and, as a rule, they will not measure more than a few inches across. Veins of quartz of considerable dimensions are numerous over the crystalline Fic. 7.—Stone Mountain, Wilkes county, North Carolina. A granite residual. Several others are shown in the distance on the extreme left of the view. area of the state, and they can be readily traced over the surface for considerable distances by partially disintegrated outcrops and abund- ant angular fragments which litter the surface. RELATIONS BETWEEN THE JOINTS IN THE GRANITE AND THE DIKES OF BASIC ROCK. Referring to the direction in the strike of joints discussed on p. 23, it will be observed that the planes of most of the joints lie in the north- east and the northwest quadrants, respectively. Likewise careful measurement and tabulation of the dikes of basic rocks show that the 406 THOMAS L. WATSON direction of strike of most of them is in the same quadrants as the strike of the joints. Furthermore, in nearly every quarry exposing dikes of basic igneous rocks the strike of the dikes and that of one set of joints were coincident, probably indicating that for the areas men- tioned the jointing has exercised some influence in the cutting direc- tion of the dikes. Not only is this true of the dikes penetrating the granites, but it is equally true of the Triassic sandstone belt, where coincidence in strike of the diabase dikes and that of jointing in the sandstone is very strikingly shown. Careful measurements in the strike of the joints and the dikes were made in the numerous openings over the sandstone belt, which can be summarized as follows: Variation in major jointing is from N.15°-60° W. to N. 20°-60° E., with minor sets noted in some open- ings striking N.-S. and E.-W. Likewise, variation in strike of the dikes is from N. 20°-60° W. to N. 20° E., with a few striking N.-S. In every opening the strike of the dikes and that of the joints for a given direction was found to be coincident. Whether this will apply in general to those dikes beyond the limits of the fresh rock exposures it is not possible to say at present, as the jointing is entirely obscured by the deep residual decay covering the fresh rocks. AGE-RELATIONS OF THE DIKES OF BASIC IGNEOUS ROCKS. As noted above in the basic dikes penetrating the crystalline rocks, strongly contrasted structural differences in the’ dike rocks obtain. Many of them are entirely massive and unaltered, bearing little or no evidence of pressure-metamorphism, while a large proportion of them are completely schistose, and are otherwise mashed and closely jointed. The ferromagnesian constituent in the latter is usually partially or completely altered. Both classes of the dikes often pene- trate original massive igneous rocks that are now more or less schistose in structure. These facts afford a strong and sufficient basis for regarding the basic rocks of the dikes of different periods of intru- sion, and therefore, of different age. The massive dikes penetrating the more or less schistose rocks must postdate in age the period of t Data kindly furnished by Mr. F. B. Laney. GRANITES OF NORTH CAROLINA 407 disturbance inducing the schistose structure in the inclosing rocks. Likewise the schistose dikes were intruded at an earlier period and prior to the metamorphism of the inclosing rocks, for the field evidence indicates that the schistose structure in the dike and the inclosing rock is the result of the same forces. Until the age of the inclosing rocks is definitely determined, that of the more schistose dikes must largely remain conjectural. As stated, the dikes must antedate the period of pressure-metamorphism affecting the inclosing rocks, for both dike and inclosing rock are similarly affected. The sandstones of Triassic age occupying the marginal position along the eastern border of the Piedmont region, are cut by a system of typical massive, unaltered, diabase dikes. The dikes conform, as a rule, to northeast and northwest directions, and are coincident in strike with that of the joints in the sandstones for these directions. Nowhere have the dikes been observed to cut rocks younger than the sandstone, and their age is accordingly definitely fixed as middle ‘Mesozoic. They are correlated with flows of the same composition and age in New Jersey, New York, and the Connecticut Valley region, and with similar dikes in Virginia and Georgia to the north and south of the Carolina area. The dikes of the Carolina sandstone belt are traced into the neighboring crystalline rocks of the Plateau region, where they have wide distribution. Beyond the limits of the sandstone belt, in the crystalline areas penetrated by the dikes, close similarity in texture, structure, and composition of the massive dikes obtains, and their relations to the inclosing crystalline rocks make it reasonably certain that they are of the same age as those penetrating the Triassic sand- stones. THomas L. WATSON. GEOLOGICAL LABORATORY, Denison University. GEOLOGICAL NOTES ON THE VICINITY OF BANFF, ALBERTA. Banrr, the easternmost of the resorts established by the Canadian Pacific Railroad in the mountains of the Northwest, lies a little east of the axis of the Rocky Mountain range, on the Bow River, at an altitude of 4,521 feet. The surrounding mountains rise to heights of 8,000 to 10,000 feet and upward. Structure of the mountains—The Rocky Mountains in Alberta contrast with the same range in the United States in that folding and overthrust faulting are their predominant features. Their structure closely resembles that of the southern Appalachians, con- trasting strongly with the Basin type where normal faulting is the rule. In Alberta parallel ridges of folded Carbonic limestones are the prevailing features. ‘These are underlain by Cambrian sandstones, and overlain by coal-bearing Cretaceous sandstones. About Banff the general trend of the ridges is northwest-southeast. Drainage.—The normal drainage of the region is similar to that often noted in the Appalachian region; the channels being established either along the strike of some soft layer, or cutting across the ridges at right angles to the. strike. In such a region stream robbery is common, and one river will, at an advanced stage of adjustment, present a series of right-angled turns, wind-gaps often indicating former channels. Glaciation.—The Canadian Rockies have been heavily glaciated at a comparatively recent date. The glacial action appears to have been that of very large valley glaciers, rather than of the continental ice-sheet. Local moraines are common, and the larger valleys are bordered by glacial terraces. The drainage in the vicinity of Banff presents several interesting features. The Bow River, after flowing southwest along the strike of the Cambrian sandstones, turns abruptly northeast, cutting a gap in the Sawback, Vermilion, and Cascade ranges (see map, Fig. r). It then turns southeast again, flowing along the strike of a Cretaceous 408 NOTES ON THE VICINITY OF BANFF, ALBERTA Reduced from Banff. ins near f the Rocky Mounta ic map o Fic. 1.—Topograph Canadian Geological Survey. 410 I. H. OGILVIE infold. At Banff the Bow River is joined by the Spray River from the southeast and by the Cascade River from the northwest. Lake Minnewauka outlets westward into the Cascade River, its eastern end leading through a wind-gap to the Ghost River. The Bow Valley about Banff is drift-filled. Westward from the station its course is through gravel, largely stratified (Fig. 2). At the a Fic. 2.—Still waters and alluvial deposits of the Upper Bow River above Banff; Mount Edith in the distance. station the river turns abruptly southeast, cutting a little canyon along the strike of Carbonic shale and forming the Bow Falls. At its junction with the Spray it again enters drift and again turns north- east. ‘There is here presented the abnormal feature of a water-fall along the strike, followed by quiet water when the course is at right angles to strike and up dip. Forty-mile Creek flows southeast along the strike between Ver- milion and Sawback ranges. It cuts across a gap in Vermilion NOTES ON THE VICINITY OF BANFF, ALBERTA AIl range and cuts off the end of Cascade Mountain. As it enters the Bow Valley it turns abruptly west, flowing over drift for a mile and a half, and emptying into Vermilion Lake. These various abnormalities are the result of two distinct causes: first, adjustment to the soft Cretaceous infold of the lower Bow Valley; second, glaciation. Fic. 3.—Devil’s Canyon; a postglacial gorge of the Cascade River. Lake Minnewauka (Devil’s Lake) is a body of water twelve miles long and about half a mile wide. Its sides are precipitous, except close to the water’s edge where glacial gravels are found. Alluvial cones occur, projecting into the lake. As noted by Dawson in 1885," this lake basin presents every appearance of having formerly been a river valley. Its eastern end is now filled with drift, talus, and allu- vial cones, but its level floor and steep sides extend to Ghost River on the east. Dawson, in the report cited, suggests that this was the t Report B, Geological Survey of Canada, 1885, p. 141. AI2 I. H. OGILVIE preglacial course of the Bow River. If this were the case, a drift- dam, or indications of an ice-dam, should be found, and the valley beneath the drift should present older features in the preglacial than in the present Bow Valley. _ An investigation of the western end of the Lake Minnewauka disclosed hills of morainic aspect, evidently deposited at the junction of an ice-lobe descending the Cascade Valley with one coming from the Minnewauka Valley. Westward from these morainic hills is an overwash plain which descends westward and southeastward to form the upper terrace of the Cascade River. Lake Minnewauka is held up by this moraine, and its existence as a lake undoubtedly dates from the time of retreat of the ice. But that this dam could have diverted the Bow River is not possible. The outlet of Lake Minne- wauka flows through drift for about a mile; it then joins the Cascade River, and together they flow through a postglacial gorge cut in Car- bonic limestone. ‘This gorge is known as the Devil’s Canyon. ‘The surface of this limestone is drift-covered, and the top of the gorge undoubtedly represents the height of the valley floor before the ice- invasion, also the amount eroded by the ice. ‘This level is 5,000 feet. The limestone cut by the Devil’s Canyon extends in a low ridge for more than a mile and blocks any considerable outlet or inlet of the Minnewauka Valley. Any river flowing in or out of this valley in preglacial times must have had its floor above 5,000 feet. Five miles southwest of the lake the Bow River flows mainly over gravel. The Cretaceous beds are occasionally exposed, but the valley bottom is mainly in drift. The valley floor at this point is 4,350 feet, and the preglacial river in this valley could not have been higher. It therefore follows that no preglacial river could have flowed from the Bow to Lake Minnewauka Valley. Moreover the lower Bow Valley (below the junction with Cascade River) is older physiographically than the Lake Minnewauka Valley. The sides are less precipitous, the cross-section, though steep sided, more nearly U-shaped. In the opinion of the writer, the present drainage at this point is due, not to glacial agencies alone, but to preglacial adjustment to the soft Cretaceous beds. The Minnewauka Valley and the Bow Valley above the junction with Cascade River were probably once NOTES ON THE VICINITY OF BANFF, ALBERTA 413 v one valley. Stream robbery then took place, owing to the advantage possessed by the present lower course of the Bow, a stream on the Cretaceous beds, over the upper part, the upper Bow-Minnewauka River. This adjustment was preglacial. The Bow was thus drawn away into its southeastward course in preglacial time, and a divide was established at the western end of Minnewauka Valley. A river Fic. 4.—Bow Valley, from the Banff Springs Hotel; the preglacial course of the Spray River. At the left the Bow is emerging from its postglacial canyon. continued to occupy this valley, emptying into Ghost River, until obliterated by the ice-invasion. ‘The basin may have been deepened by ice-gouging; at all events, as the ice retreated northward, a lake was formed at its western margin, and an outlet was formed to the west, the eastern end of the gorge being still ice-filled. By the time the ice had retreated altogether, the westward outlet was too well established for a return to the eastward drainage. ‘The present level of the lake is lower than that of Ghost River. AI4 Ihs Tele) OG IME Ie, The abnormalities in the vicinity of Banff village appear to be due to glacial agencies only. Forty-mile Creek takes its abrupt westward bend at the point where its old valley is filled by drift. The Bow takes its southward turn around Tunnel Mountain for the same reason, namely, damming of its old valley by drift, its present rapids and fall being postglacial. South of Tunnel Moun- tain the Bow leaves its gorge and turns into the preglacial valley of the Spray (Fig. 4). The Spray itself has been pushed out of its valley at this point by talus from Mount Rundle. I. H. OcILviz. GEOLOGICAL DEPARTMENT, Columbia University. GLACIAL AND POST-GLACIAL HISTORY OF THE HUDSON AND CHAMPLAIN VALLEYS. OUTLINE. INTRODUCTORY STATEMENT. TOPOGRAPHY OF EASTERN NEW YORK AND SOUTHERN NEW ENGLAND. SUBLACUSTRINE OR SUBMARINE GLACIAL DEPOSITS IN THE HUDSON AND CHAM- PLAIN VALLEYS. Hupson VALLEY SouTtH OF HIGHLANDS AND LOWLAND WEST OF PALISADE RIDGE. Brooklyn-Perth Amboy Moraine. Drift of Long Island and Staten Island inside above moraine. Drift of lowland west of Palisade Ridge and inside above moraine. Drift in the Hudson Valley south of Highlands and north of Long and Staten Islands. Haverstraw. North Haverstraw. Croton. Oscawanna-Crugers-Peekskill. Jones Point and State Camp. Roye Hook. HIGHLANDS OF THE HUDSON. Drift of the Highlands. Hupson VALLEY NORTH OF THE HIGHLANDS. Drift at Newburg—New Windsor and Fishkill—Dutchess Junction. High-level terrace north of Fishkill. Drift at Carthage Landing, Low Point, and Roseton. New Hamburg gravel plateau and Wappinger Creek stratified drift. Camelot kames. Drift north of Camelot to north of Catskill. Drift north of Catskill to north of Glens Falls. Old lake-floor or old sea-floor. Gravel plateaus and deltas. Elevations above old lake-floor or old sea-floor. Depressions below old lake-floor or old sea-floor. Lake basins. Valleys. Submerged channel of the Hudson. Extra-morainic channel. Channel inside the moraine. 415 416 CHARLES. EMERSON PEET Wave-wrought features in the Hudson Valley. Fossils in the Hudson Valley and in the lowland west of the Palisade Ridge. Buried soils. WESTERN PASSAGE FROM HUDSON TO CHAMPLAIN VALLEY. EASTERN PASSAGE FROM HUDSON TO CHAMPLAIN VALLEY. CHAMPLAIN VALLEY. East side, Lake Champlain. West side, Lake Champlain. Clay plain. Gravel plateaus and deltas—upper and lower series. Wave-wrought terraces. Upper series. Lower series. Fossils. Moraines. Eskers. The streams and their valleys. Fort Edward Valley. Whitehall-Putnam Station Valley. Submerged Poultney-Mettawee Valley. Erosion of tributaries to Poultney-Mettawee stream in southern Cham- plain region. [Outline to be concluded.) INTRODUCTORY STATEMENT. THE plans for the investigation the results of which are here pre- sented were first made under the direction of Professor R. D. Salis- bury. At the beginning of the actual work, in the absence of Professor Salisbury from the country, the work was pursued under the direction of Professor T. C. Chamberlin, and has been continued under his direction up to the present time. ‘The writer’s interest in the subject was first aroused while engaged in detailed mapping of the Pleistocene deposits of the Palisade Ridge of eastern New Jersey in 1893 and 1894, under Professor Salisbury’s direction.? Subsequently, in the preparation of the Pleistocene maps for the New York City Folio 1 A brief summary of this paper was presented before the Geographic Society of Chicago in March, 1904. An alternative hypothesis bearing on crustal movement entertained at that time has replaced one favored then. 2See RoLLIn D. SALISBURY AND CHARLES E. PEEt, ‘Drift Phenomena of the Palisade Ridge,” Annual Report of the State Geologist of New Jersey, 1893. 3See Pleistocene maps of the New York City Folio, U. S. Geological Survey, by ROtiin D. SALISBURY, assisted by HENRY B. KUMMEL AND CHARLES E. PEET. Fic. 1.—Reference map. The numbers refer to places in the following list. On all the maps each place has the same number. 133, Addison; 104, Addison-Rutland County Line; 67, Amsterdam; 27, Annsville; 2, Arthur Kill and Perth Amboy; so, Athens; 78, Ballston Lake; 82, Ballston Spa; 129, Beekmantown; 111, Bouquet River and Wadham’s Mills; 35, Breakneck Mountain; 139, Burlington and the Winooski River; 125, Cadyville; 58, Cairo; 48, Camelot; 44, Carthage Landing and Low Point; 33, Cold Spring and Foundry Brook; 1009, Cole’s Bay; 36, Cornwall; 60, Coxsackie; 22, Croton Landing; 108, Crown Pt. Center; 126, Dannemorra, 95, Dunham’s Basin; 41, Dutchess Junction; 131, East Creek; 10, East River; 4, Elizabeth River; 8, Engle- wood; 119, Ferrona; 97, Fort Ann; 92, Fort Miller, 33, Foundry Brook and Cold Spring; 55, Glasco; 87, Glen Brook and Glen Lake; 7, Hackensack and Hacken- by Na ake sack River; 120, Harkness Station; 30, Highland Falls; Pode 90, Hoosic River delta; 89, Hopkin’s Pond; 32, Indian , REY Brook; 88, Jenkins Mills, Queensbury, and Round pA dhty a Pond; 190, Jones’ exebs FARA ) [SS aN Point; 118, Keese- | GC Leste ville; o, Kill van Iga Kull; 141, La Moille po Ge eae Next Se River and West Mil- ee oe ton; 83, Lonely Lake; : pas >a 9 59 44, Low Point and ; a ONS ere WR iy aN as pe a Carthage Landing; pees Re 140, Mallett’s Bay; i a 80, Maltaville; 40 Marlboro; 77, Mechanicsville; 142, OG | pie Missisquoi River; 85, Moreau Pond; 127, Morrison- Celtel LP ae ville; 93, Moses Kill; r05, Mount Defiance; 122, Mount seo Cina as Etna; 5, Newark Bay; 61, New Baltimore; 46, New S COO AS. Hamburg and Wappinger Creek; 63, New Scotland; 65, NS 6, Newtonville; 37, New Windsor; 114 North Branch of “™S = Bouquet River and Tower’s Forge; 62, Oniskethau, : ‘SQ Oe Spraytkill, and South Bethlehem; 20, Ossining (Sing Sing); 6, Passaic River; 2, Perth Amboy and Arthur a CATSKILL -. Kill; 121, Peru; 117, Port Douglas; 88, Queensbury, ee ies Round Pond, and Jenkins Mills; 1, Raritan River and sf Bay; 115, Reber; 54, Rondout; 39, Roseton; 79, Round Lake; 88, Round Pond, Queensbury, and Jenkins Mills; to4, Rutland-Addison County Line; 123, Salmon River and Schuyler’s Falls; 81, Saratoga Lake; 68, Schoharie Creek; 123, Schuyler’s Falls and Salmon River; 20, Sing Sing (Ossining); 132, Snake Mountain; ror, South Bay; 62, South Bethlehem, Spraytkill and Oniskethau; 124, South Plattsburg; 14, Sparkill Valley; 62, Sprayt- kill, South Bethlehem, and Oniskethau; 28, State Camp; 71, St. Johnsville; 18, Stony Point; 75, Teller Hill; 106, Ticonderoga; 114, Tower’s Forge and North Branch of Bouquet River; 128, Treadwell Bay; 52, Ulster Park; 25, Verplanck’s Point; 64, Voorheesville; 111, Wadham’s Mills and Bouquet River; 46, Wap- pinger Creek; 47, Wappinger Falls; 130, West Beek- mantown; 50, West Park; 112, Whallonsburg; 116, Willsboro; 139, Winooski River; 3, Woodbridge Creek; 96, Wood Creek; 141, West Milton and La Moille River. 418 CHARLES EMERSON PEET and for the report on the Glacial Geology of New Jersey,! under Professor Salisbury’s guidance, work bearing on the problems here involved was carried out in 1897 and 1901. The main results here presented were in hand before the latter date, and the advance since then has been mainly in determining the crustal movement and in the analysis of facts bearing on the origin of the Hudson water body. To Professor Salisbury the writer is under obligation for the opportunity of detailed study of the Pleistocene formations of New Jersey and adjacent portions of New York, for early training in methods of investigation and mapping of those formations, and for suggestions in the original plans for the work, the results of which are here pre- sented. ‘To Professor Chamberlin the writer is under obligation for assistance with difficulties encountered in this investigation, and for continued inspiration to perseverance in searching out the truth. Neither Professor Chamberlin nor Professor Salisbury is responsible for opinions here expressed or for any failure to arrive at the truth. GENERAL STATEMENT OF TOPOGRAPHY OF .EASTERN NEW YORK AND SOUTHERN NEW ENGLAND. Southern New England has been described as an upland rising gradually inland from the sea and reaching elevations of 1,500 to 2,000 feet in southern New Hamsphire and Vermont.? Above this upland there rise higher elevations such as Mt. Monadnock, and groups of elevations such as the Green Mountains and the White Mountains. Below the upland, valleys have been sunk, a small amount near the sea, but deeper farther inland. These valleys are broad on the soft rocks and narrow on the harder rocks. Without assuming an identical history, this picture may be trans- ferred to eastern New York, where, as a first approximation to the truth, the country may be pictured as a rolling surface rising inland from the narrows at Long Island and Staten Island. Above this sur- face there are elevations, such as the Adirondacks and the Green Mountains. Below it there are depressions, such as the Hudson and Champlain Valleys. t See Glacial Geology of New Jersey, by ROLLIN D. SALISBURY, assisted by HENRY B. KuMMEL, CHARLES E. PEET, AND GEORGE N. Knapp (Vol. V of the Final Report of the State Geologist, 1902). 2Davis, Physical Geography of Southern New England. GLACIAL AND POST-GLACIAL HISTORY 419 The Hudson Valley has three natural divisions: (1) the part south of the Highlands from Peekskill to the narrows at Brooklyn; (2) the Highlands from Peekskill north to near Fishkill; and (3) the broader Hudson Valley from near Fishkill to north of Glens Falls. North of the Hudson Valley is the Champlain Valley, to which two passages lead, one by way of Lake George and the other east of the Lake George pass by way of southern Lake Champlain. The broader Hudson Valley and the Champlain Valley have been con- sidered the northeastward continuation of the Greater Appalachian Valley. SUB-LACUSTRINE OR SUB-MARINE GLACIAL DEPOSITS IN THE HUDSON AND CHAMPLAIN VALLEYS. In the bottom of the Hudson and Champlain Valleys there has been built in recent geological times a plain mainly of clay, with margins frequently of gravel and sand.t This plain has the form of an old lake-floor or old sea-floor. The clay plain is best seen in the north- ern part of the Hudson from Catskill north. In the southern part of the valley—from Poughkeepsie south—it is either absent entirely, or is to be seen only in limited areas, and generally covered with gravel and sand. ‘The clay in both the Hudson and Champlain Valleys is laminated, with alternate “fatty” and sandy lamine having a thick- ness of one-fifteenth of an inch or more. (See Fig. 2.) The lamine are sometimes grouped into beds a few inches in thickness, sepa- rated by rather prominent lines of parting. In one place ripple marks were seen. (See Fig. 3.) The clay often shows faulting and frequently shows joint structure conspicuously. (See Fig. 2.) It is sometimes contorted. The upper part of the clay is generally yellow in the Hudson Valley and brownish-red in the Champlain Valley. The tThe clays and gravels and sands of the Hudson Valley have been described in considerable detail by F. J. H. MERRILL AND HErnricH RIEs in the Tenth Annual Report of the New York State Geologist, and by Mr. Ries in the Bulletin of the New York State Museum, Vol. III, No. 12. The former report was in hand in the field, and while the interpretation placed on these deposits is quite different, the writer wishes here to make general acknowledgment of its aid in his studies. Specific acknowledg- ment is made in the proper place where this report has been drawn upon for facts beyond the writer’s personal observation. The writer also had the aid in the field of the article by S. PRENTISS BALDWIN on “‘ The Pleistocene History of the Champlain Valley,” American Geologist, Vol. XIII (1894), pp. 170-184. 420 CHARLES EMERSON PEET lower part is blue. Bands and blotches of yellow often occur in the midst of the blue layers in the Hudson Valley. The clay often con- tains concretions. In thickness the clay varies from a small amount Fic. 2.—Showing joint structure in the laminated clay of the Hudson Valley. to 215 feet in the valley of the Hackensack and zero to 243 feet™ in the Hudson Valley, where it is commonly 80 to 100 feet or more thick. In the Champlain Valley it is known to have a thickness as great as 60 to 75 feet, but is generally thinner than in the Hudson Valley. tH. Ries, Bulletin N.Y. State Museum, Vol. III, No. 12, p. 184. GEACIAE AND POST-GLACIAL HISTORY A421 The clay overlies till (Fig. 4), gravel and sand, or rock surfaces, which are frequently striated. It fits into the irregularities of the till, or the stratified gravel and sand which sometimes appears to have the form of kames. In the upper Hudson the underlying stratified sand and gravel often has a high angle of dip, generally southward, but sometimes in other directions. These layers are interpreted as repre- senting deposits made by the ice waters in the standing body of water as the ice was retreating. ‘This structure can frequently be seen from the car windows of the New York Central Railroad. The marginal Fic. 3.—Ripple marks in the clay at New Windsor. deposits of gravel and sand have the form of plateaus of two distinct classes: Class t.—Gravel terraces and plateaus with undulatory topography on the edge toward the Hudson, which sometimes assumes a more or less kame-like or morainic form, or with the edge next to the Hud- son higher than the edge next to the valley wall, and with the dip of the layers of gravel and sand toward the valley wall and down- stream. This phase of the drift is the characteristic phase in the Highlands, and is not accompanied by a clay plain. Class 2.—The second phase of the stratified drift consists of gravel plateaus and terraces with the undulatory edge toward the 422 CHARLES EMERSON PEET valley wall, and the smoother and lower edge toward the Hudson.* The inner and higher edge is sometimes marked by distinct kames or by moraine. ‘The structure is delta-like. The layers usually dip at high angles toward the Hudson, and the coarse gravels and sands grade rapidly down the dip of the layers into fine laminated clay. (See Fig. 5, A and B, and Fig. 6.) In the clay and over the clay there are sometimes masses of till. (Fig. 4.) In the till there are sometimes masses of clay. Over the clay there often is coarse gravel with a subdued undulatory topography, and the contact of the Fic. 4.—Showing till both above and below the clay at Haverstraw. gravel and clay is of such a nature as to indicate that the gravel has been forcibly pressed against the clay surface. This phase of the gravel plateaus is the characteristic phase of the Appalachian Valley division of the Hudson Valley, and is usually associated with a wide clay plain. .These two classes may be referred to as high-level ter- races. On the whole, they increase in altitude from south to north, but not at a uniform rate or continuously. They indicate a water tSome of the smoother plateaus may properly be called subaqueous overwash plains. See R. D. SALISBURY AND HENRY B. KUMMEL, Annual Report of State Geolo- gist of New Jersey, 1893, pp. 266-68; and R. D. Sartspury, Glacial Geology of New Jersey, pp. 130-33. GLACIAL AND POST-GLACIAL HISTORY 423 body in the Hudson and Champlain Valleys as the ice was retreat- ing after making the Brooklyn-Perth Amboy moraine. Below the level of the high-level gravel plateaus there are two classes of deposits: (1) secondary deltas; (2) river terraces. The former have been recognized with certainty only in the northern Hudson. The latter occur in the northern Hudson Valley and in tributary valleys both in the northern and southern parts of the Hudson. In these lower terraces pebbles of clay occur rarely, evi- dently derived from the erosion of the higher clay deposits. Fic. 5.—Diagrammatic sections: [A, of Haverstraw gravel plateau from west to east; B, of Newburg delta and moraine at the left and Dutchess Junction gravel plateau with morainic east edge on the right; C, a section similar to and about one mile south of D; D, Roseton on the left and the northern part of the Low Point deposits on the right. Par- allel horizontal lines represent clay. In the Hudson Valley this old sea- or lake-floor plain is naturally divided into three portions roughly corresponding with (1) the por- tion south of the Highlands, (2) the Highlands, and (3) the Hudson Valley north of the Highlands. Deposits in a lowland west of the Palisade Ridge will be described in connection with Division 1. HUDSON VALLEY SOUTH OF THE HIGHLANDS AND LOWLAND WEST OF PALISADE RIDGE. From the narrows at Brooklyn northward to the Highlands the land rises gradually, as a slightly rolling upland, best represented by the even crest of the northern part of the Palisade Ridge, and by the 424 CHARLES EMERSON PEET level to which the hilltops reach east of the Hudson.t Below the level of this upland to the west of the Palisade Ridge there is a low- land. Below the surface of this lowland 100-150 feet there are Fic. 6.—Photograph showing gradation of gravel and sand down the dip of the layers into clay, in the part of the Newburg delta north of the Quassaic. The work- man’s shovel marks the point where one of the layers of gravel and sand at the left changes into clay. valleys? in which there are deposits of gravel, sand, and clay, pres- ently to be described. In the valleys in the southern part of this tNew York City Folio, U.S. Geological Survey; Geography by R. E. DoDGE AND BAILEY WILLIS, p. 1. 2Physical Geography of New Jersey, R. D. SALISBURY, p. I41. NE WO YT ORT : AND WLeCINiTxX. Fig. 7—Model of New York City and vicinity. [Copyright by Edwin E. Howell; printed by permission.] 426 CHARLES EMERSON PEET lowland there are salt waters, which in their widest expanse consti- tute Newark Bay. (See Fig. 1, No. 5, and Fig. 7.) The upland surface represented by the even crest of the Palisade Ridge is a remnant of the Cretaceous peneplain. ‘The lowland sur- face represents a later peneplain developed on the softer rocks of the Triassic area.t The valleys in this lowland represent erosion in pre-last glacial and post-Pensauken time. ‘The stratified drift in these valleys was deposited largely by ice waters on the retreat of the ice from the Brooklyn-Perth Amboy moraine. Below the upland surface and at the east base of the Palisade Ridge is the Hudson Valley, now occupied by the waters of the Hudson estuary. Along the sides of the valley and below the waters of the Hudson estuary there are deposits of stratified gravels, sands, and clays similar in origin to those in the lowland west of the Pali- sade Ridge. Below the waters of the Hudson estuary and of Newark Bay there are certain submerged channels which are shown in Fig. 8 and will be referred to later. BROOKLYN-PERTH AMBOY MORAINE. Across the southern end of both the Hudson Valley and the low- land west of the Palisade Ridge there is the massive and complex ridge which forms the Brooklyn-Perth Amboy terminal moraine.’ It is popularly referred to as the backbone of Long Island, and _ it also makes the more conspicuous elevations of the southwestern part of Staten Island. Through this moraine there are two gaps—one at the south of the Hudson and the east end of Staten Island called the Narrows, and the other at the west end of Staten Island occupied by Arthur Kill. (See Fig. 8.) t This is called the Somerville peneplain by Professor W. M. Davis, and the pre- Pensauken peneplain by Professor R. D. Salisbury (Joc. cit., pp. 114-15). 2See R. D. SatisBury, Glacial Geology of New Jersey, Chap. 9, and New York City Folio, U. S. Geological Survey. T. C. CHAMBERLIN, Third Annual Report, U.S. Geological Survey, 1881-82, pp. 377-79; WARREN UpHAm, American Journal of Science, 1879, pp. 81-92 and 179-209; G. H. Cook AND J. C. Smock, Geological Survey of New Jersey, Annual Reports for 1877, 1878, and 188o. “AoTTeA PEpteds ‘vr troany yseq ‘or tyMy wea YY ‘6 {poomasuy ‘g SIeALI pue A}o—yovsuexoeyH ‘4 saa oressed ‘9 ‘Avg yvmon ‘S Sraary Yleqeziyq ‘by Syseig epliqpoom ‘f “TOY Anyjzy pue Aoqury yyeg ‘2 thkeg uvjrey ‘T “sOR]q Ul SPUULYD oUTIeUICNs ‘o}TYM Ul UMOYS V1v SvoIv LT, “Ip poynens pue Ty pextu ‘> ‘pues ounp jo svore ][PUIs oUIOS SUIpNpoUr pur ‘py ur papnpout JOU Sauey WIA WYP peyyes Jo svore ‘g ‘sor oy} Jo seovjd Supypey Yeu OF poaoyfod sv YTYA svore oTUPeIOUL PUL SIO OUIVA SUIpNpUr lap JayxpTyy Jo sieq ‘py :aNaoa'T [spapy.D Kanang jsv0Q “SQ 94} WoIF poutr}qo vIVp Worf poov.} 19M spoUULYS pesiourqns ey, ‘svore TW sv Apjaed pur sowey sv Apjavd peddevur 919M YSIYA payIeUL iv s}jaq awios puLs] SUCT UO “gz eT _q UT papNpoUr jou nq *OULIOF VY} FO JX9} OY} UL POUOTUUT 91M YOTTM UMOYS IIL SVIIR IIUTLIOUL PUP S}[9q SUIvY SuIpNypout FLIP IayOTG} Jo syaq ewog “Avamng [eoVsofoox) *S “AQ PY} Jo oNoT MD Y4oA Ma NT WY} Wor, pur Kasaar man {o k80j0ax Jp19v]4 S AANGST[eS “Cl “AY JO Sz VV q wosy usye],] lp oy} yo dew—'s “OI fe) % A 20 LM \ Sy tS ry aoo ; > ° Ry co BS S&S See. oy 0) 00 09 >) —— Sy. ap. 08 WR Go BE 'S1O\S0.50, obo P~-0.9 ; Lora, Raz O50 90) SBL50°0 606% 9 GAS do S0y Or Om. 350 2o OO lor ice ze) oOo BASE 5 > O26 428 CHARLES EMERSON PEET DRIFT OF LONG ISLAND AND STATEN ISLAND INSIDE THE BROOKLYN- PERTH AMBOY MORAINE. Inside or north of the Brooklyn-Perth Amboy moraine a number of positions taken by the ice on its retreat are marked by moraines or kame belts, or other similar phenomena. Near the west end of Long Island at least two, and probably three, such belts are rep- resented more or less discontinuously. (See Fig. 8.) Possibly one such position is represented on Staten Island. DRIFT OF THE LOWLAND WEST OF THE PALISADE RIDGE AND NORTH OF THE BROOKLYN-PERTH AMBOY MORAINE. On the higher part of the lowland west of the Palisade Ridge and inside the Brooklyn-Perth Amboy moraine, there is an extensive series of belts of thicker drift, with more or less distinct morainic topography, or elongate belts of kames with the aspect of moraines, which are frequently bordered by plains of gravel and sand with the form of overwash or outwash plains. In some places such aggra- dation plains have no definite kame or morainic areas at their source. On the lower part of the lowlands there is a complex series of gravel and sand plains or plateaus, some of which head in kames, but others have ice-molded, but kameless, sources. Some of the plains have delta-like forms and delta-like struc- tures. The elevations of these plains at the south are 20-40 feet, while farther north plains whose structure is unknown have eleva- tions of 80-100 feet.t These deposits are found north from the latitude of Hackensack and Englewood well toward the northern border of the state. Underneath the gravel and sand of these plains, or spread out to the southward with little overlying sand or gravel, : there is laminated clay which frequently has a thickness of 100 feet and sometimes as great as 215 feet. This clay extends south of the lat tude of Hackensack and Englewood, and is also found to the north.’ t See R. D. SALISBURY AND C. E. PEET, “‘ Drift Phenomena of the Palisade Ridge,” Annual Report of State Geologist of New Jersey, 1893, pp. 195-210; and idem, “Drift of the Triassic Plain of New Jersey,” Glacial.Geology of New Jersey (Final Report of State Geologist, Vol. V), Chap. 12, and especially pp. 506-13, 595-628, 632-42. aThe areal distribution of a large part of these deposits is shown in the maps of the New York City Folio, U. S. Geological Survey. See also Fig. 8 of this article. 100006 GO CG0) YJe} oo) osoo8ocess 20025000 90099009) 8090960 Fic. 9.—Map showing some of the areas of stratified drift along the Hudson from north of Sing Sing to north of Camelot. LEGEND: A, moraines and kames, generally. In some places ice-shaped drift-forms that are neither moraine or kames are thus indicated. B, ice-moulded stratified drift. C, gravel plateaus made by ice waters. D and E, secondary deltas. F, clay chiefly, but other forms of stratified drift are included; knowledge is more certain where lines are not broken. G, approximate limits reached by the standing water. H, boundary lines showing approximate limits of the features inclosed. The white space next to the streams marks swamps, flood-plains, and low-level terraces. 14, Sparkill Valley; 15, Haverstraw and Minis- -ceongo Creek; 17, North Haverstraw and Cedar Pond Brook; 18, Stony Point; 19, Jones’ Point; 20, Sing Sing (Ossining); 21, Croton Point and Croton River; 22, Croton Landing; 23, Oscawanna; 24, Crugers; 25, Verplanck’s Point; 26, Peekskill—village and creek; 27, Annsville; 28, State Camp; 29, Roye Hook; 30, Highland Falls; 31, West Point; 32, Indian Brook; 33, Cold Spring and Foundry Brook; 35, Breakneck Mountain; 36, Cornwall; 37, New Windsor; 38, Newburg and Quassaic Creek; 39, Roseton; 40, Marlboro; 41, Dutchess Junction; 42, Fishkill and Fish Kill; 43, Walcotville; 44, Carthage Landing and Low Point; 46, New Hamburg and Wappinger Creek; 47, Wappinger Falls; 48, Camelot; 49, Poughkeepsie. 430 CHARLES EMERSON PEET DRIFT IN THE HUDSON VALLEY SOUTH OF THE HIGHLANDS AND NORTH OF LONG ISLAND AND STATEN ISLAND. The deposits of drift of special significance in this area occur north of Ossining (Sing Sing), where the valley begins to broaden out into Haverstraw Bay. Most of the deposits south of here are covered by the “Surficial Geological Maps” of the New York City Folio, and are mentioned in its text. ‘The deposits of drift of special significance in this part of the valley belong to the two classes of high-level terraces mentioned above (p. 228). The features of Class 1, similar to those found in the Highlands, are found (1) in a terrace at 120-100 feet A. T. from north of Sing Sing to south of Croton River mouth; (2) at Jones Point; and (3) features of similar import occur at Roye Hook near State Camp, Peekskill. These features are shown in Fig. 9, the first between Nos. 20 and 21, the second at No. 19, and the third at No. 29. To which class some of the features of a high-level terrace from Peekskill toward Osca- wanna belong is a question. (See Fig. 9, Nos. 23, 24, 25). The features of Class 2, similar to those prevailing in the Appalachian Valley part of the Hudson, occur at Croton Point and Croton Land- ing on the east side of the river, and from Haverstraw to north Haverstraw on the west side. (Fig. 9, Nos. 21, 22, and 15 and 17.) Haverstraw.—The deposits at Haverstraw of interest in connec- tion with this paper are of four types: (1) the narrow moraine at the foot of the Palisade Ridge; (2) a gravel plateau with undulatory topography and delta-like structure at an elevation of less than 120 feet A. T., which is underlain by laminated brick clays, yellow in color above, and blue below; (3) low-level gravel and clay. 1. The moraine: The position of the ice-edge as it rested near the northern base of the Trap Ridge, which farther south makes _ the Palisades of the Hudson, is marked in the southern part of Hav- erstraw by a narrow moraine with a west-by-north and east-by-south trend, parallel approximately to the trend of the Trap Ridge. Where well-defined this moraine has a width of one-fourth to one-half of a mile. It has been traced for a distance of about two miles, from a point about one mile southeast of Thiell’s Station to within some- thing less than three-fourths of a mile northwest of Short Clove. It extends farther east in the form of a ridge, but with less definition. GLACIAL AND POST-GLACIAL HISTORY 431 as Short Clove is approached. At its best the moraine shows a relief between hillocks and hollows of between 20 and 30 feet. (See Fig. 9, No. 15). Some of the hillocks are composed largely of strati- fied drift, but, so far as exposures show, a considerable part of the moraine is made up of till which is prevailingly of gneissic materials. There are places, however, where it shows a conspicuous red color from the abundance of Triassic constituents. 2. The Haverstraw gravel plateau has been sometimes called the Haverstraw delta. It extends from about the lower edge of the moraine above mentioned north- ward to Cedar Pond Brook, and descends from an elevation of a little less than 120 feet on the west to 40-60 feet on the east, where Fig. 10.—Cross-section of the Appa- iPMallsmottaapruntlyatomalower 21a) ee aes) aay AB, uplands; CD, general valley lowland. Hori- plains bordenne they Mainisceongo” Jonraiinine shows clay. and Cedar Pond Brooks. The topography of the plateau for the most part appears at first sight to be quite plain, with a general slope eastward and southeastward. In the northern part, however, it becomes quite undulatory, and there are some depressions as great as 20 feet in depth, one of which is situated in a long conspicuous ridge extending north and south parallel to the road west of it, leading from West Haverstraw to North Haverstraw. ‘This ridge is separated from the higher land to the west by a considerable trough-like hollow. On closer examination much of the surface, which at first sight appeared to be quite plain, is found to be affected by ridge-like undulations and hollows with a relief of 6, 8, and to feet These undulations extend eastward nearly to the abrupt east front. (See Fig. 5, A.) Structure and materials: Exposures are abundant in this plateau. In the higher western portion numerous gravel and sand pits show stratified sands and gravels more or less horizontal in the upper few feet, and dipping at high angles below in an easterly and southerly direction. A little eastward these stratified gravels are said to over- lie clay, and still farther eastward exposures are of such a nature that it is quite certain that the gravels and sands grade into finer materials, and into clay with alternate layers of fine sand, and finally into the laminated brick clays. 8 u dive s H & owed F banwcnen see ess cneree 7 = ‘ 2 u 432 CHARLES EMERSON PEET While stratified materials appear to prevail in the capping of the clay, a number of places are to be found where the material is not stratified, and where it has the character of a compact till showing indications, at the contact with the clay, of having been subjected to a pressure which forced the clay and the till together. The contact surface is irregular. (Fig. 4.) The clay surface rises and falls as much as 2 feet in a distance of 10 feet, while the layers of clay show contor- tion. in the upper part. In earlier observations compact lumps of clay were noted in till-lke material. Before it was firmly established that till overlies the clay, the writer did not feel sure that in the extraction of the clay for brick-making the lumps of clay had not become artificially mixed with till. It seems reasonably certain now, however, that the observed clay lumps in the till were not introduced artificially. Underlying the clay stratified gravel and sand was observed in some places with a topography which suggested buried kames. In other places till was observed with some constituents that are not present in thg overlying till. Flowing springs and flowing wells arising from this underlying gravel and sand indicate that the water has access to these porous layers at higher levels. North Haverstraw.—The gravel plateau at North Haverstraw very likely was once continuous with that described south of Cedar Pond Brook. It has a general elevation of something less than 120 feet A. T. on its west side, and a large part of its total area is 100 feet A. T. It descends abruptly 100 feet to the meadow along Cedar Pond Brook on the south, and on the northwest it descends abruptly to a plain at about 20 feet A. T. On the northeast the descent is to a kame-like gravel knoll at about 50 feet A. T. On the east there is a similar knoll. Each of these knolls is indicated by a black circle at No. 17 in Fig. 9. On the southeast the descent is to a terrace-like form at 4o feet, and farther south to a ridge of gravel of uncertain ongim at Oo) feet. The west and northwest sides of the plateau have a gently undula- tory topography, and the surface of the plateau farther east has some ridge-like irregularities that suggest the presence of the ice during its deposition. Materials and structure: There are no good exposures in this conspicuous plateau, although exposures occur in the lower gravel GLACIAL AND POST-GLACIAL HISTORY 433 forms to which the steep edges of the plateau descend. The indica- tions from the surface and from well data are that it is composed of gravel and sand. A well on the top of the plateau near the west side at the house of Mr. Neely is reported to have penetrated 120 feet of gravel and sand from an elevation of 100 feet A. T. without reaching rock. The kame-like gravel knoll at the northeast edge of the plateau is situated south of the North Haverstraw Station of the West Shore Railroad. An exposure shows about 20 feet of coarse gravel. Far- ther south, east of the railroad, where the gravel likewise has a kame-like form, the layers dip north and south and east. North of these kames toward Stony Point clay occurs up to 40-60 feet, and has a form which may be due to erosion. Low-level gravel, and clay. What appear to be low-level terraces derived from the erosion of the higher gravels have been observed in the form of discontinuous shoulders on the north side of the Haverstraw gravel plateau on the slope to Cedar Pond Brook at elevations of 60and 4o feet. A 4o-foot level on the southeast side of the North Haver- straw plateau, and also a 60-foot level southwest from the 40-foot level, may represent a product later in origin than the plateau itself. It may be said, however, that where the level of stratified drift varies so greatly as it does in this region it is not easy to determine posi- tively that the shoulders of limited area and uncertain relations are of later origin than the higher gravels, and do not represent remnants of undulations descending to the lower levels made while the ice was present. A wide plain extending along the water front from near Short Clove to Grassy Point, with an elevation by the topographic map of less than 20 feet A. T., has been so worked over in the making of bricks that it is difficult to say what was its original height and extent. Some exposures have been observed in it in which the materials included rounded pebbles of clay, evidently derived from the erosion of the higher level clays. In exposures near at hand and at about the same level the layers have a southerly dip at a high angle, thus sug- gesting lower-level delta deposits made from the erosion of the higher gravels and clay. At Grassy Point (south of No. 17, Fig. 9) there are deposits in this lowland which were made in the presence of the ice. No exposures which reveal the structure occur in the 60-foot and 434 CHARLES EMERSON PEET 4o-foot levels south of Cedar Pond Brook. The 4o-foot level at North Haverstraw is well exposed. In no part of the exposure was the delta structure seen which is so common in the high-level gravels, Fic. 11.—The Catskills and the lowland in the Appalachian part of the Hudson Valley. {From A. P. Brigham, Geographic Influences on American History; courtesy of Ginn & Co.] and there is said to. be no clay underlying the stratified gravel here for a considerable depth at least. Interpretation: (1) Ice present with its edge resting on the mor- aine at the foot of the Trap Ridge and with a general west-by-north and east-by-south direction on the west side of the Hudson. GLACIAL AND POST-GLACIAL HISTORY 435 (2) The ice retreated so that, either at one time or at successive positions, its edge occupied the Haverstraw and North Haverstraw gravel plateaus. (3) The ice waters discharged into a standing body of water and built up the deposits of gravel and sand, with the steep dipping layers of gravel rapidly grading into sand and then into clay. The clay was deposited on the irregular surface of the drift previously deposited. Fic. 12.—Part of the Newburg delta, on the south side of the Quassaic. Looking west. (4) The ice, either by re-advances, or because of more favorable conditions in some places than in others while continually present, worked over the clays, producing some of the contortions observed, and involving masses of clay in the till which it deposited over the clay in favorable places. The water-worn gravel was in places brought under pressure, and the contact of the clay with the gravel was thereby made more intimate. (5) On the retreat of the ice and the fall of the water-level the higher-level gravels were eroded by the Minisceongo and Cedar Pond Brook, thus producing deposits at lower levels. Whether there are remnants of lower-level deltas cannot be confidently stated. ‘They would naturally be expected. (6) After deeper erosion by the streams than the present Hudson level, submergence took place, thus drowning the mouths of the tribu- tary streams. 436 CHARLES EMERSON PEET Croton.—On the east side of the river at Croton Point, and Croton Landing, deposits of similar import occur, but not identical in detail, with those on the west side of the river. Lack of space forbids detailed description. Oscawanna—Crugers—Peekskill—Clays and gravels: The clays and gravels in the city of Peekskill and south to Oscawanna show phenomena which in some features are similar to, and in other features are unlike, those at Haverstraw and Croton. They are evidently deposits made later than the last-mentioned deposits. Their relations, however, to any marked and definite position of the ice-edge are not so well shown. The approximate area of these deposits is shown in Fig. 9 between Nos. 23 and 25. The deposits in this region are notable for their irregularity in level. The clay surface varies from an elevation approaching 100 feet A. T. to tide-level. It is in most cases covered with sand, or gravel, and in a broad statement it may be said that the finer materials predominate at the south, while at the north the materials overlying the clay include both fine materials and coarse gravels with bowlders up to five feet in diameter. In one or two instances till has been found overlying the clay in this region. The structure of the stratified materials overlying the clay even at high levels does not generally show the high angle of dip so common in the high-level gravels at Haverstraw and Croton. The high-level terrace has an elevation of 100-120 feet A. T. Parts of it, however, reach lower levels—6o-80 feet A. T., and per- haps less. It is somewhat difficult to discriminate between what may properly be called high-level terrace and what may properly be called low-level terrace. Terraces at low levels exist at 60, 40, 30, and 10-20 feet A. T. In general, these terraces are covered by gravel and sand. The gravel is sometimes very coarse, containing bowlders as large as five feet in diameter. The relations to the clay are not always distinguishable. It is not an uncommon relation, however, to find these low-level gravels much lower in level than clay in the immediate vicinity. It is questionable, however, if in all cases this relation has been brought about by the erosion of the higher-level deposits. It is conceivable that some of the low-level deposits are original deposits by the ice water, which simply failed to build up at ‘TIDST sasoyy ‘£6 Sr9q TAT “MO 26 YPY waWeg ‘16 ‘toaNYy d1S0OH pe sen sop ee £ ‘zg $ ‘1g foyTAe At 10g faye unoy ‘64 sayry uojsyeg ‘g4 ‘ayrasorueyoay 44 SAory, ‘94 ‘ITH 49 : pea a eee rea chig ee enti Reo Wet cneone MON ‘£9 {YaaIQ NeyoYsUO pue [ry Wradg ‘woysyjog ee 0) STL el ea ea i a le} ‘sdurems SyIeUt sureas ay} 0} xou douds azTY[M OYJ, *SO.ANJVaF pasopoUur 9y} JO SIT] ayeutrxoidde Surmoys seul Arepunoq H pace Unt eee ge “uayoIq JOU o1B SOUT] a1IYM UTE}I90 JIOU SI aSpayMOoUY “pephpour o1v Ip poyryesjs Jo surtoy Joyo Inq Ayeryp Avy ‘uy ee YL weg ayy re ‘seyjap Arepuoges ‘YF pur q *S.1a]@M aot Aq apvur sneered paw ‘J "yp WAIOJ-]JUIp padeys-aor ssovjd owios uy “AT[vsdues ‘souey puv soureiowl ‘pf :aNaoa’] B oidde ‘+ .19]VM SUTPUR\s 9} AQ payoverd SHUT] 9}VUUTX “Q0LAaP SIT} AQ pataA0d svaiv oy} UL papnpour oq Avur syisodap Jo}yeM-adt 9} JO 9WOS [I payuens pepjow-sor ‘g *payeoIpul sNy} atv savvy JO 9UTeIOUT Joy}oU o1v yey} s “sSutidg eBoye1eg Jo Y}IOU 0} YOVPOYIS YING Jo epnjye] Wolf UOSpNY IY} SuoTe Up peylyesjs Jo svoie edriourd SutmoyS— EI ‘org Sry S00) ao afore 8099 Boeco Pee jeonc9 e000 AO e0co 2060 fQio%,? lo9 809 9000 soeee Alses 00 By woge ogess le00G0 YO 9 ° 2000) fot o$ Soose ooo 438 CHARLES EMERSON PEET these particular places to the level reached in adjacent places. It may be said that this relation is more clearly shown by phenomena seen elsewhere. The facts in this locality are sufficiently doubtful, at any rate, to justify caution in asserting the erosion origin of the irregularities in the clay. However, there are some facts which seem to indicate a considerable amount of erosion of the clay in this region and subsequent deposition of gravel in the clay channels. Jones Point and State Camp near Peekskill.—At Jones Point a narrow terrace less than a mile in length occurs on the right side of the Hud- son. It is made up mainly of stratified gravel and sand and a little clay, and formerly contained more clay.!' The terrace has an elevation at its north end of about too feet A. T., and at its south end of about 60-80 feet, and at present, in places at any rate, is higher toward the Hudson than toward the valley wall. See Fig. 9, No. 19. At State Camp (28) near the mouth of the Peeks Kill there is a gravel plateau with an elevation of too feet A. T. whose flat top is used by the New York state militia as a camping and parade ground. This plateau continues up the valley of Sprout Brook with some breaks, to a point about one mile northeast of Annsville (27), where it has an elevation of 140-160 feet A. T. It was not studied beyond this point. It also extends northward up the valley of the small tribu- tary north of State Camp and southeast of Wallace Pond. Its greater development occurs, however, at State Camp on the right bank of Peeks Kill. A small remnant occurs farther upstream on the left bank. While the plain surface as a whole rises upstream, the northern portion of it at-State Camp slopes eastward—a fact of some significance perhaps in determining the history of the plateau. The exposures in this plateau show gravel overlying clay. The clay reaches higher above sea-level near the extremity of the plateau than farther upstream, thus indicating a gradation of coarse materials from upstream into the fine clay. Low-level terraces occur farther downstream on both right and left sides of the stream at 30-40 feet Aa: Roye Hook.—West of the plateau near the State Camp Station of the New York Central Railway there is an isolated hill of small dimensions, reaching an elevation of 100 feet A. T. This is called ™See H. Ries, Tenth Annual Report State Geologist of New York, p. 114. GLACIAL AND POST-GLACIAL HISTORY 439 Roye Hook (29, Fig. 9). Atits base, both on the north and on the south, there are low-level terraces at about 30 feet. In the Roye Hook hill a large gravel and sand pit shows about ro feet of fine sand and gravel overlying 5 feet of silt, both sand and silt being hori- zontally stratified, or nearly so. But under the silt there is about 85 feet of coarse gravel and sand, with layers dipping at a high angle eastward and southeastward. ‘The gravel and sand in this lower portion of the exposure were not observed to grade into clay, as they were observed to do at Croton, and as they apparently do at Haver- straw. The phenomena at Jones Point and Roye Hook are more nearly allied to deposits which occur in the Highlands than they are to those that occur in the broader part of the valley to the south. These characteristics are discussed in the description of the Highlands phe- nomena. THE HIGHLANDS OF THE HUDSON. As indicated above in the description of the general features of the Hudson, south of Newburg and Fishkill the Hudson leaves the Appalachian district and crosses through the plateau to the south in that part of its course designated the Highlands of the Hudson. The features of the rock valley here differ radically from those at the north, and in place of a broad dissected lowland between distant uplands, like that in the Appalachian Valley, or of the broad amphi- theater between distant uplands like that at the north edge of the Palisade Ridge at Haverstraw, we have the upland descending abruptly from elevations of 1,100 and 1,400 feet to the waters of the estuary, which is here generally from four-fifths to seven-tenths miles wide, and reaches.a maximum width of one and three-fifths miles. In this narrow valley the gravel plateaus are present, but if there is a Clay plain, it is covered by the waters of the estuary, or is pepresenice by a few limited remnants only. The gravel plateaus in the Highlands of the Hudson have charac- teristics which are typical of Class 1 above described, and indicate the presence of the ice in the valley while these deposits were accumu- lating. ‘There are situations however, where streams of water came from the ice outside of the immediate Hudson River valley and deposited their loads on slopes toward the center of the valley. Such 440 CHARLES EMERSON PEET phenomena occur where streams head northward from their debou- chure, and are represented, it is believed, by a part of the State Camp deposits on the southern edge of the Highlands. ‘The gravel plateaus of the Highlands that were examined are (1) at West Point and south toward Highland Falls (30); (2) at Cold Spring (33) and south toward Garrison. The stratified drift at Roye Hook near State Camp, per- haps part of the State Camp plateau where the surface slopes east- Fic. 14.—Looking across the clay plain of the upper Hudson. [Photograph by W. S. McGee.] ward, and the Jones Point plateau, on the southern edge of the Highlands have characteristics similar to those in the Highlands and indicate a similar disposition of the ice in respect to the valley when they were building. North of West Point in the Highlands there is little left of any former valley filling, if the valley ever was filled. One or two miles south of Storm King, on the right bank of the Hudson, a remnant of clay in a terrace form was observed at an elevation estimated at 60 feet, and nearer Storm King another remnant was observed. On the left bank there are several remnants of clay or gravel, the most conspicuous of GLACIAL AND POST-GLACIAL HISTORY 44I which is a terrace of coarse gravel on the south side of Breakneck Mountain (35, Fig. 9) with an estimated elevation of 80 feet. The weight of evidence in this part of the Hudson indicates that the ice protruded as a tongue down the valley, and that it was influential in shaping the edge of the plateaus toward the Hudson. An alternative hypothesis is that the ice retired in a northeasterly direc- Fic. 15.—Looking eastward from the bluffs north of Albany across the trench of the Hudson, cut into the clay plain. tion; that the plateaus on the west side of the valley were made first and had their east edge shaped by the ice-front; and that later those on the east side of the valley were constructed and had their east edge shaped by the ice. This alternative hypothesis, however, does not explain the eastward-dipping layers of the plateaus on the east side of the valley, nor does it explain the apparent lower elevation of some parts of the terraces on the east side next to the valley wall. It would explain successive undulations from lower levels near the Hudson to higher levels away from the Hudson, such as are found at Cold Spring (33, Fig. 9). 442 CHARLES EMERSON PEET HUDSON VALLEY NORTH OF THE HIGHLANDS. The valley north of the Highlands has been considered the north- eastward continuation of the Greater Appalachian Valley. It lies between the upland surface on the east, which in New England is called the New England Plateau, and on the west it is limited by the Alleghany Front which is the steep slope from the Alleghany Plateau to the Greater Appalachian Valley, and of which the dissected edge is called the Catskill Mountains. Below the level of these eastern and western plateau surfaces there is a broad lowland. Below the level of this lowland surface there are deep valleys. In the bottom of these valleys there are the deposits of stratified drift which have the form of an old lake- or sea-floor. Below this old lake- or old sea-floor there are valleys and other depressions. In the bottom of the Hudson Valley is the Hudson estuary, and beneath its waters is the sub- merged Hudson channel described below. In this region and farther to the southwest, upland and lowland surfaces in the Appalachian district have been interpreted as representing peneplains at two or more levels.t| The interpretation of the preglacial history of the valley is beyond the scope of this article. ‘The writer wishes to get before the reader a general picture of the region only, without refer- ence to the preglacial history. (See Figs. ro and 11). In the Appalachian part of the Hudson the drift phenomena may be described in seven sections: (1) the Fishkill-Dutchess Junction Fig. 9, Nos. 42, 41), and Newburg-New Windsor deposits (38, 37); (2) high-level terrace north of Fishkill south of 44 and north of 42; (3) deposits from Carthage Landing to Lowpoint on the east side of the Hudson and at Roseton on the west side (Nos. 44 and 39); (4) New Hamburg gravel plateau and stratified drift on Wappinger Creek (Nos. 46 and 47); (5) Camelot kames (No. 48); (6) deposits . north of Camelot from Poughkeepsie to Catskill; (7) deposits from north of Catskill to north of Glens Falls (Fig. 13). The deposits in Division 1 are similar in import to those in Division 7, and are very much like the phenomena at Haverstraw and Croton. tSee R. D. Satispury, Physical Geography of New Jersey, pp. 8-14, 83-85, 94- 98; BatLey WILLIS, ‘‘Northern Appalachians,” National Geographical Society Mono- graphs, Vol. 1; C. W. Hayes, ‘Southern Appalachian,” zbid., and W. M. Davis, Pro- ceedings of the Boston Society of Natural History, Vol. XXV (1891), pp. 318 et seq. GLACIAL AND POST-GLACIAL HISTORY 443 The phenomena in Division 2 are somewhat like those from south of Peekskill toward Oscawanna, and indicate the presence of the ice in the valley near the north edge of the terrace when the deposits found here were being made. The Carthage Landing-Lowpoint and Roseton deposits indicate the presence of the ice-in the valley when adjacent stratified drift at higher levels was accumulating. The New Hamburg gravel plateau and Wappinger Creek stratified drift indicate the presence of the ice in the valley at the edge of the New Hamburg plateau while the ice in the higher lands was retreating through the Wappinger Creek Valley. I. NEWBURG-NEW WINDSOR AND FISHKILL-DUTCHESS JUNCTION. These places are situated on opposite sides of the river. On each side the Haverstraw phenomena are 1epeated. Gravel plateaus under- lain by clay are situated next to the river, and morainic phenomena occur on the higher land away from the river. The gravel plateau at Newburg is more delta-like in form than either at Haverstraw or Fishkill-Dutchess Junction. (See Fig. 12.) In the latter places the surface is marked by undulations similar in kind to, but more sub- dued than, those in the moraines on the adjacent higher land. Masses of till are found at Dutchess Junction in the clay, and ripple marks are found in the clay at New Windsor. (See Fig. 3.) Interpretation: The interpretation of these deposits is similar to that of the Haverstraw deposits. It is difficult to see how the ice of a single ice-lobe can retire from a valley both to the eastward and to the westward, either simultaneously or successively, unless it be by the making of an embayment in its front. The evidence here, like that at Croton Point and Haverstraw, points to an embayment of the ice-front with at least wings of ice extending farther down the Hudson oneachside. If the interpretation of the Jones Point, Roye Hook, and West Point-Cold Spring phenomena be correct, it would seem that the front changed from .a protruding tongue in the Highlands and immediately south, to an embayment at Newburg—New Wind- sor and Dutchess Junction—Fishkill. This is similar to the inter- pretation of a protruding ice tongue south of Croton Point and Haverstraw and an embayment at those places. The level to which the waters of the Hudson water body reached 444 CHARLES EMERSON PEET seems to be 140 feet. It may have reached somewhat higher. The only means of determining this limit seems to be the maximum limit of the delta structure. The presence of kames or moraine does not fix an upper limit to the water, as indicated by the North Haverstraw kame-like bodies, at low levels and by kames at Camelot and other places to be mentioned hereafter. Low-level terraces on the Fishkill and Quassaic have the same import as those at Croton. It is a question how much erosion has taken place in the Hudson Valley itself. If the ice was on the valley sides and the waters discharging into the valley which was free from ice, it would be expected that the valley would be filled entirely across up to the level required by the slope of the deposits from each side. Since the clay would naturally take a very gentle slope if the valley was free, it would be expected to fill to a height somewhat less than the height of the clay on each side, but not very much less. Alto- gether it seems probable that the erosion here has been more than 100 feet, while in regions immediately north it is very much less than this. Il. HIGH-LEVEL TERRACE NORTH OF FISHKILL. One and one-quarter miles north of Fishkill a sand-and-gravel- capped clay terrace at 100 feet at its outer edge and 120 feet at its inner edge extends northward for about one mile, with a width of less than one-quarter of a mile up to three-quarters of a mile. The overlying gravel and sand have a depth of 6-10 feet, and the layers dip south. Near its north end gravel and sand occur at a higher level, with a slightly undulatory topography, which may mark the ice-edge on the land when the terrace was building in the Hudson water body. (See Fig. 9 south of No. 44 and north of 42.) Ill. CARTHAGE LANDING-LOWPOINT AND ROSETON. At the northwest margin of the above-mentioned 100-120-foot terrace a lower terrace of gravel-capped clay occurs at 20-40 feet A. T. The clay in this terrace varies in elevation from sea-level or below to 40 feet above sea-level. The gravel above the clay is coarse and contains some sink-like depressions. (Fig. 5,C.) It extends from a point one and one-quarter miles south of Carthage Land- GLACIAL AND POST-GLACIAL HISTORY . 445 ing to a point three-quarters of a mile north of that place. Fora half mile at its southern end it borders the undulatory gravel area mentioned above in connection with the roo-120-foot terrace. Nearer its north end it is separated from a higher deposit of sand-capped clay by a till area which has a slightly undulatory topography. (Fig. 5, D.) Opposite this low terrace, at Roseton and north toward Danskammer Light, a gravel-capped clay deposit occurs at a slightly higher level, with a terrace form. There are two hypotheses to explain these relations: (1) The deposits at Roseton once extended entirely across the valley and have since been eroded, and the 20—40- foot terrace is the product of erosion of the higher deposits. (2) The second hypothesis is that the Roseton gravel-capped clay terrace and the high-level clay deposit on the east side of the Hudson were never continuous, but that the ice occupied the valley when they were made, and stood on the 20—40-foot gravel terrace, but failed to build it as high as the 100-120-foot terrace to the south, or the Roseton terrace to the west and the clay to the east. The sinks in the 20-40-foot terrace, and the faint undulations in the till on he slope between the 20-40-foot terrace and the higher clay, bear this out. If this is the correct interpretation, it is a case similar in kind to the kame-like knolls near the North Haverstraw gravel plateau, and also like the phenomena just south of Peekskill near Verplancks. Under this interpretation the 20-40-foot terrace may or may not once have extended entirely across the valley. Under the first hypothesis both the high-level and the low-level terraces formerly extended entirely across the valley. It may be objected to the second hypothe- sis that when the ice had retired to New Hamburg, and the clays and sands and gravels were accumulating, some of the finer mate- rials at least should have been carried south into this unoccupied part of the valley. This argument would be especially strong if the Hudson ‘water body was subject to tides whose ebb would tend to carry the fine detritus down into this space. It is, however, not at all necessary to believe that the valley between these two terraces was unoccupied at this time. If the valley was occupied, it was by a mass of stagnant ice. There is evidence elsewhere that such masses of stagnant ice were left in the valley. 446 CHARLES EMERSON PEET IV. NEW HAMBURG GRAVEL PLATEAU AND WAPPINGER CREEK STRATIFIED DRIFT. A gravel plateau occurs at New Hamburg (46), which has a width in a north-south direction of about one-half mile, and connects with stratified drift which extends upstream to the northeast four miles or more, and spreads out to broader dimensions. It has an eleva- tion at its west edge of too feet A. T., and a plain surface which rises upstream; and at Wappinger Falls (47), at an elevation of 140 feet A. T., it has a topography which indicates the presence of ice. Farther northeast kames occur in small areas. The edge of this plateau is steep at the northwest side, but farther south, and east of New Hamburg village, it falls off in undulations, which indicate the presence of the ice during its formation. Exposures in this plateau in this lower portion show layers of gravel and sand wih dips west in the western part, and east in the lower portion of the eastern part. Apparently this gravel and sand grades into clay farther east. It will be noted that this would be upstream as the Wappinger Creek now flows. The easterly dipping layers are interpreted as the fore-set beds (of Davis) and the clay as the bottom-set beds. ‘The westerly dipping layers are interpreted as the back-set.beds. Farther upstream, just east of the Print mills, near the point where the surface topography indicates the co-operation of the ice in forming the plateau, gravel and sand in considerable thickness overlie silt and yellow clay which ~ reach 60 feet A. T. or more. Blue clay was observed farther south on the left bank of the creek. The layers of gravel and sand dip west and southwest. Low-level terraces: In the village of Wappinger Falls a lower terrace at 40-60 feet is to be seen, and also one at 20 feet. Both are made of gravel 2-3 inches in diameter. Farther downstream a ter- race occurs at 30 feet A. T. V. CAMELOT KAMES. Near Camelot kames unassociated with the gravel plateau occur in a valley tributary to the Hudson within 20-40 feet of sea-level, and in the immediate Hudson Valley within 60 feet of sea-level. (Fig. 9, No. 48.) GLACIAL AND POST-GLACIAL HISTORY 4A7 VI. NORTH OF CAMELOT TO NORTH OF CATSKILL.’ From north of Camelot to Catskill the study of the deposits was made largely in passage, so that the relations of the deposits to the successive positions of the ice-edge cannot be stated. These fugitive observations indicate a general higher altitude of the deposits next to the valley side and a lower altitude next to the Hudson. They also indicate an alternate increase and decrease in the elevation both in he present Hudson bluffs and farther back from the river, and this is interpreted as indicating more than one stand of ice in this area. South of Poughkeepsie there is a pitted plain at 140 feet A. T. with a steep descent toward the Hudson, the origin of which is unknown. Gravel was observed at various points along the West Shore Railroad from West Park to Ulster Park, at elevations of 100 and 140 feet A. T. Clay was observed nearer South Rondout at 140 feet, and in South Rondout clay underlies sand which has an elevation of 150 feet. At Kingston along the Hudson a sand-capped clay terrace occurs 220 feet A. T., while farther west along the West Shore Rail- road a plain at 180-200 feet A. T. is coated with sand and gravel that are said to be underlain by clay. North of Kingston the clay has an elevation of 100-140 feet in the Hudson bluffs, and a higher elevation west of the bluffs, where sand covers the surface. At Glasco the clay has an elevation of 140 feet in the river bluffs and 180 feet along the West Shore Railroad two and one-half miles to the westward. At Saugerties and north toward West Camp it has an elevation of 140-160 feet. At Catskill it has an elevation of 100-120 feet, and a plain surface which has been trenched by the Catskill and its tributaries. It has been described here by Professor W. M. Davis, who has also described the delta of the Cats- kall.? VII. NORTH OF CATSKILL TO NORTH OF GLENS FALLS. Old lake-floor or old sea-floor—tIn the Appalachian Valley part of the Hudson north of Catskill, and perhaps from north of Pough- keepsie the first approximation to a correct picture of the topography t For places mentioned in this division see Fig. rt. 2Proceedings of the Boston Society of Natural History, Vol. XXV (1891), pp. 318-35. 448 CHARLES EMERSON PEET is that of a plain descending gradually from the valley sides to the bluffs (of clay generally) bordering the present Hudson. ‘The eleva- tion, onthe whole, increases from the south, where it is 100 or 120 feet to 180 feet, toward the north, where it is 220-240 feet along the bluffs of the present Hudson. From these elevations marking the bottom of the trough or the meeting of the slopes of the sea- or lake-floor, the plain rises eastward and westward to the higher land marking its limits. (See Fig. 14.) Gravel plateaus and deltas—Above the plain on the east and on the west rise gravel plateaus, some of them delta-like in form, which represent approximately the level of the watets when the floor above described was being built up. These gravel plateaus and deltas are found at the following places: On the left bank—(z) South Schodack and northwest at 340-360 feet; (2) Troy, 300 and 360 feet (?); (3) Hoosick River, 360 and 280-300*" (?) feet; (4) Batten Kill, 380-400 (?) and 300 feet; (5) At Glens Falls and vicinity, 460 (2), 389, 340*, and 280-300* feet. On the right bank—(1) South Bethlehem, at 300, 240*, and 200-220* feet; (2) at Maltaville, 340-360 feet; (3) at Saratoga and vicinity, 400 (?), 320-340, 300, and 260 feet; (4) southwest of Glens Falls, 340* and 380 feet—a continu- ation of the Glens Falls levels of the left bank. The plateaus descend abruptly toward the plain. The layers of coarse gravel and sand of which they are generally made may some- times be seen to dip at high angles in the same direction and to grade down the dip into the fine materials, and into the clay which makes up a large part of the plain. They are like those of Class 2 of the lower Hudson. The height of the gravel plateaus above the level of the floor varies from 160 to 180 feet on the north to 100-120 or 140 feet in the southern part of this area. While these gravel plateaus descend abruptly toward the old lake- or old sea-floor, and the stratification bears the significant relation to the clay plain mentioned above, the relation of the gravel plateaus up the slope of their surface is just as significant. When traced back they are found to connect, t The stars indicate secondary deltas. Perhaps the 300-foot Troy delta should be considered secondary. Perhaps a part of the 300-foot delta on the Batten Kill is not secondary. It has been seen only in its northern part. Its southern part as rep- resented on Fig. 13, No. 91 is hypothetical. GLACIAL AND POST-GLACIAL HISTORY 449 (1) with valley trains leading down from positions marking the ice- edge during its retreat: or (2) they head directly in kames or other morainic developments which are just as significant in showing the relation of the gravel plateaus to the source from which the material was derived, and from which emanated the waters that transported the delta materials to their present resting place. ‘These kames and moraines have been traced back from the immediate Hudson Valley in a few cases, and have been seen to develop into fairly well-defined morainic topography, while in the lower lands the morainic phenom- ena are more subdued. Such kames and moraines are found at the following places: (1) At the eastern.and northeastern margin of the South Schodack gravel plateau. The East Greenbush kame area mMakesay part. ot this belt." (See Pigs 135 Nos. 73, 74.) (2) The Teller hill kames (No. 75), which are fronted by clay without an intervening gravel plateau. (3) The line of kames and moraine extending from North Albany to Newtonville (No. 65.) (4) Kames at Troy (No. 76). (5) Glen Lake-Hopkins Pond kame belt north of Glens Falls (Fig. 18, Nos. 87 and 89. (6) Kames between New Scot- land and Voorheesville (Fig. 13, Nos. 63 and 64.) (7) Kames at Saratoga Springs (Fig. 13, No. 84.) (8) The Moreau Pond belt of kames (No. 85.) That the relation of this kame belt to the gravel pla eau east of it is similar to the relation of the kames and gravel plateaus mentioned above, is not certain. The surface of the gravel plateaus is sometimes marked by ridges or by deep sinks. The clay plain at the outer edge of some of these plateaus has a higher level than at the same edge of the plain farther north, or the reverse may be true. The clay plain sometimes fronts kame areas without an intervening gravel plateau, and the top of the kame area may be lower than the level of the gravel plateau immedi- ately adjacent. On some of the streams, notably the Hoosick River, deltas occur without the undulatory surface. From north of the lati- tude of Mechanicsville to the northern part of Saratoga Springs, the western part of the lowland is occupied by a succession of gravel plateaus (area 50-75 square miles), with discordant levels which are separated by depressions having a general northeast-southwest direc- tion, and in the bottom of which are lakes such as Round Lake, Saratoga Lake (with a length of 5-8 miles and a width of 2 miles), 450 CHARLES EMERSON PEET and Lonely Lake. Probably Ballston Lake is thus situated. (See Fig. 13.) Under several of these gravel plateaus, over wide areas, clay is - reported. In the bottom of the Round Lake depression there is till, and a limited coarse gravel area with deep sinks in it, although clay and stratified drift rise in steep faces to the north and to the west, and perhaps in other directions. It is believed that these northeast-southwest depressions were occupied by masses of stagnant ice while subsequent plateaus to the northwest were being built, and that the northeast-southwest trend of the depressions indicate some- thing of the direction of the ice-edge as it retreated. At Glens Falls? and north a succession of gravel plateaus is believed to mark the successive positions of the ice-edge in a similar way, but they are not separated by wide depressions, or depressions of any kind so gen- erally. Below the level of these gravel plateaus are secondary deltas derived from higher gravels by erosion. It is believed that they have been recognized at South Bethlehem (two levels—one on Sprayt Kill and one on the Oniskethau) (Fig. 13, No. 62), on the Hoosick River (Fig. 13, No. 90), on the Batten Kill near Schuyler- ville (Fig, 13, No. 91), and on the Hudson in the vicinity of Glens Falls (two levels) (Fig. 18, Nos. 94 and 86.) Elevations above old lake- or old sea-floor—Above the plain in situations not confined to its borders there rises another class of eleva- tions, some of which were islands in the sea or lake, when the gravel plateaus and deltas were being made. These hills sometimes are drift hills. More often they are drift-covered rock hills. Such islands may be seen in the following places:* (1) hill northeast of Saratoga; (2) highland east of Saratoga Lake; (3) hills north and south of Mohawk River; (4) hills southeast of Albany; (5) hills south of New Baltimore and at numerous places southward, where the elevations are ridges elongate in a north-south direction. Some of these hills may have formed shallows only, at the highest stage of the Hudson water body, but most of them were distinct islands and served to break the water body up into several more or less separate portions. There were doubtless other shallows or islands, and the elevations «See also WARREN UpHAM, American Geologist, Vol. XXXII (1903), pp. 223-30, 2 For these elevations see Fig. 13. GLACIAL AND POST-GLACIAL HISTORY A451 that made shallows at the highest stages of the water body must have produced islands and peninsulas at lower stages. Depressions below old lake- or old sea-floor.—Below the level of the sea-floor or lake-floor plain there are two classes of depressions: (1) lake basins and similar depressions not occupied by lakes; (2) valleys produced by erosion. ‘The depressions occupied by lakes are those like the Saratoga Lake basin and Round Lake basin, the origin of which has been referred to above and will be discussed below. The basins may be small—a few yards across and shallow; or they may be like the basin of Saratoga Lake, 5-8 miles in length and 1-2 miles in width, and with a depth below the plain surface of 60-100 feet, plus the depth of the water in the lake. The second class of depressions below the old lake- or old sea-floor are the valleys of the present streams, of which the Hudson is the chief. (Fig. 15.) Below Troy the Hudson is now an estuary, but above that place the tide does not reach. ‘The course of the Hudson is inter- preted as marking the trough of the depression down to which the old sea- or lake-floor sloped from each side, and which was followed by the main stream when the floor emerged from the waters in which it had been built. Some of the details in the course of the Hud- son may be explained as due to the greater building out along one side of the lake or sea, as for example the westward bend of the Hudson opposite the mouth of the Hoosick River, where the building out of the delta from the eastern side of the valley crowded the depres- sion farther out into the midst of the plain than to the north or to the south beyond the influence of the delta deposits. This westward bend amounts to about two miles. There is no doubt that similar explanations will account for the fact that the Hudson is nearer one side of the valley than the other in different portions of its course, and for other details; but it is probable that other slopes of the floor were the resultant of the interaction of several factors—the original topog- raphy of the valley, the rate of deposition by the agencies building up the floor, and the length of time the building continued. At any rate, the course of the Hudson may be considered a consequent course determined by the slopes of the old lake- or old sea-floor across which it flowed when the floor emerged from the lake or sea that had occu- pied it. Since that time it has trenched its course below the level of 452 CHARLES EMERSON PEET x the plain 150-170 feet north of Athens, 190-200 feet at Albany, and 80 feet near Fort Edward. This channel is now covered by tide- water to a depth varying from 10-25 feet at Albany to 15-50 feet north of Athens. This depth of water is included in the above esti- mates of erosion, from Albany south. The tributaries of the Hudson have courses that are as significant as that of the main stream itself. The larger streams and the longer small streams that head in the country beyond the limits of the plain and above the level of the Hudson water body descend in general by rather steep slopes to the plain, cross the plain on gentle gradients and descend abruptly near their mouths to the Hudson (Fig. 16, A). On See Oe eee Fic. 16. A, profile characteristic of streams tributary to the upper Hudson, or flowing either into the Poultney- Mettawee or Fort Edward Valley; B, profile characteristic of streams crossing the clay plain and flowing into northern Lake Champlain. the smaller streams in general this steep slope is not far from the Hud- son, and on the larger streams in general it is less abrupt and is farther back from the Hudson. The courses of these streams like that of the Hudson have been determined with few exceptions by the topography of the old lake- or old sea-floor, and one may picture the streams as extending their courses across this floor following the slope of the land as the Hudson water body receded. To these consequent streams subsequent tributaries have no doubt been added, and to the original consequent character of the stream gradients has been added the steep gradient found at the mouth of so many of the streams. The full explanation of this will be taken up later, but it may be said here that this steep gradient is due to so rapid a cutting down of the valley of the Hudson that the smaller and weaker tributaries were unable to keep pace with it. The reason for this will appear later. In some cases the failure of the tributaries to keep pace with the Hud- son in its downcutting is due partly to the disadvantage of a hard rock-bed, with which the Hudson did not have to contend. GLACIAL AND POST-GLACIAL HISTORY 453 SUBMERGED CHANNEL OF THE HUDSON. The submerged channel extends from Troy, the limit reached by the tides, through the Narrows of Brooklyn and beyond out to the outer edge of the overwash plain fronting the Brooklyn moraine. Beyond this there are a number of channels, but none can be con- sidered a continuation of the channel through the Narrows until opposite Sandy Hook, where a channel begins that can be traced southeastward to the forty-one-fathom line. (See Fig. 17.) Whether this should be considered a continuation of the Narrows channel will be discussed later. Extra-morainic channel.—This channel can be traced from a point opposite Sandy Hook south- eastward and ends at the forty- one-fathom line,’ eighty nautical miles from Sandy Hook. Indepth below the plain into which it is cut it varies from zero to 90 feet. Fic. 17.—Submerged extra-morainic Ten miles from Sandy Hook it has channel of the Hudson. a depth of 48 feet, and increases, Stilt gon fasme tton south-eastward to go feet, and — Lindenkohl in the U.S. Coast and Geodetic Survey Report for 1884.1; again decreases to zero feet, at the forty-one-fathom line. Beyond this channel, after an interval, there is a much deeper ravine, with which this paper is not concerned. It is 25 miles long, 3 miles wide, and has a maximum depth of 2844 teers Channel inside the moraine.—The channel inside the moraine is covered by the waters of the Hudson estuary which vary from ro to 216 feet in depth. On the whole, the minimum depth increases from Troy to just north of Newburg. South of here the water is shallow to the Highlands. In the Highlands it is deep, and from North Haverstraw south it is shallow again, but on the whole grows deeper to the Narrows. Throughout its entire length there is a very great variation in the depth, however. There are certain “‘deeps” which tA, LINDENKOHL, American Journal of Science, Vol. CXXIX (1885), pp. 475-80: 2 LINDENKOHL, /oc. cit. 454 CHARLES EMERSON PEET vary from a maximum of 216 feet at West Point to 120 feet and less. (See Fig. 8 for the submerged channels near New York city.) WAVE-WROUGHT FEATURES IN THE HUDSON VALLEY. There is an entire absence of wave-wrought features in the Hudson Valley, so far as known, with the possible exception of some gravel ridges on the low-level delta at Fort Edward east of Glens Falls, and some indistinct terraces on the west side of the valley south of Ballston. BURIED SOILS. An old soil with an elevation close to sea level has been observed on the clay surface south of Hackensack. It is overlain by ten feet of sand, the lower part of which contains clay a few inches in thick- ness, associated with fragments of leaves and woody stems. The soil is leached to a depth of one foot. On the east side of Newark Bay, south of the Lehigh Valley Rail- road, a bed of peat or peaty soil is buried by 10-30 feet of sand, much and perhaps all of which is a wind deposit.? FOSSILS IN THE HUDSON VALLEY, AND IN THE LOWLAND WEST OF THE PALISADE RIDGE. The only fossils that have been found in deposits of the Hudson water body in the Hudson Valley are: (1) Sponge spicules, fresh- water diatoms, and worm-tracks at Croton;3 and (2) leaves of Vaccinia oxycoccus at Albany.4 In the lowland west of the Palisade Ridge, near Hackensack, leaves and woody stems have been found in a bed of stratified sand and clay which underlies 8 feet of sand, con- taining a few gneiss bowlders, and overlies an old soil having an ele- vation close to sea level.’ 1“ Drift Phenomena of the Palisade Ridge,” Annual Report of the State Geo.ogist oj New Jersey (1893), p. 207. : 2 Loc. cit., p. 205, and GEORGE H. Cook, Geology of New Jersey (1868), p. 227. 3H. Ries, Bulletin of N. Y. State Museum, Vol. III, No. 12 (1895), pp. 119, 120. 4 Described by Dr. James Eights in 1852 as probably Mitchella repens. Pro- fessor B. K. Emerson thinks they are probably Vaccinia oxycoccus, which are the most abundant leaves in the clays of the Connecticut. See Mcnograph of U. S. Geo- logical Survey, Vol. XXIX, p. 718. s‘‘See Drift Phenomena of the Palisade Ridge,’’ Annual Report of State Geolo- gist of New Jersey (1893), p. 207. “yaoIg Isvq ‘IET t4aj}UID JUIOY UMOL ‘BOL :peoyY 3911S ‘Lor SeSorapuoory, ‘gor ‘umpyeg ‘ySor ‘vouvyaq Junoy[ ‘Sor faut AJUNOD puvpNy-uosippy ‘VOI suoNeIg Weujndg ‘Cor ‘Sulpuey] uosueg ‘zor ‘Avg yMog ‘IOI +1OATY AouyNog ‘oor ‘Teyoyty A £66 tzaary samvqoy “ZO Suu 10,7 ‘L6 tyxaaID poo, ‘96 ‘uiseg s,treyung ‘$6 tpieapy Woy ‘FO spuog surydoyy ‘6g ‘sy suryuef pur ‘puog punoy ‘Kmqsusand ‘gg {yooig wep pure eyeyT uals ‘4g ‘s|[ey sua[x4) ‘9g :sa0e[d BSULMOT[OF 9Y} O} Jojor stoquinu sq J, ‘seyjap Arepuoses MOYS BOI Jo Svayj}IOU pue LOT Jo JsBaYINOS Sa[II1 JOIAVOY OY], “S9IV19} [OAI|-MO] PUL ‘sure[d-pooy ‘sdureas syievur suIvaij}s 94} 0} Jxou adeds OIIYM OY], “PasojoUl SeInyvafy OY} JO S}TUT] ajeunrxoidde Surmoys sour Arepunoq ‘Fy “JoyeA\ Sutpuvjs oy} Aq poyovat SzTUT] ayeurxoidde ‘5 ‘wayxo1q JOU d1k SOUT] JIOI[M Ule}19 JIOU ST OSpa[MOUY “pepnpuy sv Lp paylyea}s Jo suIoF Joy} Nq ‘ G S Is (2) Ae ey 42 8 ga Lop, >5 a ae PpNaaUDYA a @ ime 9 ae, SS g aA Q S S) Ay ¢ Ra ae ‘a wer p UE i Sy ° b g Con, °COCHEAGUE PHVSIOGRAPHIC STUDIES IN PENNSYLVANIA 475 peneplains exist in this area, and similar features in the adjacent portions of Maryland have been described by Abbe.’ The moun- tains rise abruptly out of the valley 1,000 to 1,500 feet. The western range consists of a series of parallel ridges composed of quartzite and sandstone of Medina and Clinton age. The crests of these ridges form a very even and level sky line, which suggests at a glance that they are remnants of an old peneplain. As one approaches from the east North Mountain comes first into view, but extends into the area only about 6 miles. It is a double or canoe-shaped mountain, ending at the south in Parnell Knob. It is a closely folded syncline of quartzite, inclosing overlying shale in the center. The rocks are nearly vertical, and the quartzite ridges are conse- quently narrow and sharp. Although at a distance their tops fall in line with the crests of the more distant ridges and present a level sky line (see Fig. 2), on closer examination irregularities appear. The eastern limb of the syncline is cut through by water-gaps at several points, and its original height has been reduced so that the crest is irregular. Its highest points are slightly over 1,800 feet. The western limb is more even, but is cut by a wind-gap at the southern end. Its top lies between 1,850 and 1,950 feet elevation, except near the northern border of the area, where it is locally 2,040 feet. At the southern end of the syncline two knobs, whose summits are broad and rounded, rise to 2,000 feet, and they may well represent portions of the old peneplain but slightly eroded. The next ridge to the west is another canoe-shaped mountain, less compressed than the North Mountain syncline. It ends in Jordan Knob, a companion peak to Parnell Knob, and extends about the same distance into the quadrangle. The eastern limb of this fold is very steep and the ridge narrow and sharp-topped. In the western limb the dips are flatter and the mountain is broader and more massive. The eastern ridge has an altitude of about 1,950 feet. The general altitude of the western ridge is about 2,000 feet, but the highest peak rises to an altitude of 2,250 feet near the northern boundary of the quadrangle. The higher summits of this ridge are rounded, and in this respect possess the characteristics of an old 3 “Physiography of Maryland,” Maryland Weather Service, Vol. I, Part 2, pp. 157-61. 476 GEORGE W. STOSE peneplain, whereas the intervening portions of the ridge are reduced to a sharp crest. To the west is a group of ridges striking entirely across the area and forming the main mass of this western range. At the north the structure is complex and the ridges are irregular. To the south the mountain is composed of two distinct monoclinal ridges, with a deep, broad, synclinal valley between. ‘The eastern limb, forming Cove Mountain, is vertical, and has been weakened by faulting, so that the summit is a knife-edged crest of jagged quartzite beds. Although Fic. 2.—North Mountain from the plateau west of Chambersburg. some of its highest points reach 1,800 feet, its average height is between 1,600 and 1,700 feet, and its crest has a decided wavy or comby appear- ance. Just beyond the southern border of the quadrangle this ridge swings to the east, across the end of a flat anticline, in what is known as Cross Mountain, and then extends northward again a short dis- tance as Two Top Mountain. From the valley one observes an abrupt change from the low, comby crest of Cove Mountain to the uniform, level crest of Cross Mountain (see Fig. 3) which has an altitude of about 2,000 feet. It is apparent that Cove Mountain once stood at approximately this altitude, but has been lowered by the active erosion of the relatively narrow exposure of upturned rocks. The broad mass of gently dipping strata in the anticline of PHYSIOGRAPHIC STUDIES IN PENNSYLVANIA 477 Cross Mountain has resisted erosion, and its flat, broad top is prob- ably a remnant of the old peneplain surface. Northward beyond Cove Gap the structure of Cove Mountain is synclinal, but is complicated by faulting. Although some of its summits rise to 2,050 feet, its general elevation is not over 1,800. The western monoclinal ridge, Tuscarora Mountain, is more massive than Cove Mountain because the rocks are not so steeply inclined and have withstood erosion better. Consequently it has a greater altitude, ranging from 1,950 to 2,050 feet, and a smoother The middle-ground is the gently rolling upland north of St. Thomas. crest line. Northward this monoclinal mountain forks by the develop- ment of a synclinal valley and another monoclinal ridge on the western side, the eastern ridge becoming anticlinal for a short distance. The anticlinal portion of the ridge is 2,450 feet in elevation, and is very broad and flat-topped. From its summit one looks over the tops of the other ridges, whose general level is 300 and more feet lower. It is evident that this peak was a monadnock in the old pene- plain. Its broad, rounded character is due to its partial reduction to the graded plain. Northward the eastern limb of the anticline is faulted off and the ridge continues as a monocline at about 2,250 feet elevation. The monoclinal ridge, which branches off to the west, has an 478 GEORGE W. STOSE average elevation of 2,100 feet, but rises to the height of the anticlinal ridge near their junction. The rocks become steeper toward the north, and the ridge contracts correspondingly in size and altitude. South Mountain, on the east side of the valley, is more massive than the western range. The strata composing it are several thousand feet of interbedded Cambrian quartzites and shales, overlying older volcanics. The quartzites form the higher ridges, and the shales and volcanics are covered by a thick mask of the quartzite bowlders which prevent the erosion of deep valleys. The ridges are not straight, parallel, and even-topped like those of the western range, but are offset by cross folds and faults, and are cut through by transverse drainage, so that the crest is composed of numerous round-topped, elongate knobs and short ridges. Consequently they do not present so level a crest line as the western range. In the heart of the mountains extending beyond the eastern border of the Chambersburg quadrangle, two high and very level tracts occur. Sandy Ridge and Snowy Mountain, south- east of Montalto, are broad and level-topped, and have a general altitude of about 2,000 feet, with a small knob rising to 2,100 feet. They are composed of nearly horizontal quartzite, forming the flat top of the anticlinal uplift of South Mountain. The other tract is Big Flat, occupying a large area on the top of the mountains north- east of Fayetteville. This also is the flat crest of an anticline developed to the west of the main axis of the mountain, and producing a promi- nent offset in the mountain front opposite Fayetteville. This plateau extends for 7 or 8 miles beyond the limits of the quadrangle at a general altitude of 2,000 feet, attaining an elevation of 2,100 feet at two points. These two level tracts are undoubtedly remnants of the old peneplain, preserved at a height of 2,000 feet. The mono- clinal ridges along the front of the mountain, which once stood at this same altitude, have been reduced by erosion to 1,700 and 1,900 feet. This peneplain, observed in both South Mountain and the Tusca- rora Mountain group, has long been recognized in this part of the Appalachians, and has been described by several geologists. Davist named it the Schooley peneplain from its characteristic development on t “Geographic Development of Northern New Jersey,” Proceedings of the Boston Society of Natural History, Vol. XXIV, p. 377. PHYSIOGRAPHIC STUDIES IN PENNSYLVANIA 479 Schooley’s Mountain, N. J., and Willis‘ later called it the Kittatinny peneplain after the mountain of that name. It has been traced east- ward under the deposits of Cretaceous sediments on the eastern border of the continent, and is therefore shown to have been formed by aerial erosion while the land stood approximately 2,000 feet lower than at present. The plain extended over the present Cumberland Valley, where its surface was composed largely of limestone and shale, Fic. 3.—The Schooley peneplain preserved on Cross Mountain in the center, with the lower comby ridge of Cove Mountain on the right. but in part, probably, of overlying quartzites. These rocks have been removed during subsequent uplift and erosion. The resistant quartzites forming the mountains have withstood erosion, and rise approximately to the level of the former peneplain. The present altitude of the plain in South Mountain is about 2,000 feet, and in Tuscarora and associated mountains from 2,000 to 2,100 feet, which may indicate that in the uplift there was tilting toward the southeast. At lower altitudes in the mountainous areas there are broad, flat, terrace-like features, which probably represent prolonged stages of erosion during local halts in the elevation of the land. A broad flat t“The Northern Appalachians,” National Geographic Monographs, Vol. I, p- 189. 480 GEORGE W. STOSE 3 miles in extent occurs at 1,600 feet elevation in the heart of South Mountain east of Montalto, and another at 1,350 feet elevation at Monterey, near the state line. Both of these are cut in the softer, ancient volcanic rocks underlying the quartzites. There are also numerous broad, flat divides at elevations from 1,450 to 1,750 feet. A broad gap near the southern end of North Mountain, at an elevation of 1,550 feet, represents the abandoned outlet of the stream which formerly flowed longitudinally along the North Mountain syncline at this altitude, but was captured by Wilson Run at Franklin Gap. In Path Valley the shale spurs extending out from the mountain on the east are roughly terraced, two of the most prominent benches lying at 1,100 and 1,200 feet. Cowan Gap, in Tuscarora Mountain, through which Little Aughwick Creek formerly flowed into Path Valley, is a broad gap at 1,200 feet. Correlation of these features has not been attempted. Mr. Campbell, in his article on the Geographic Development of Northern Pennsylvania and Southern New York," states that through- out the great valley (which includes the Cumberland Valley) the limestones which occupy the southeastern side are eroded deeper than the shales which occupy the northwestern portion. ‘This is not the case in the Mercersburg-Chambersburg quadrangles and adjacent areas. ‘The rocks in the eastern portion of the valley are the lower members of the Cambro-Ordovician limestone, and contain many hard, siliceous beds and several resistant sandstones. Some of these produce prominent hills from 750 to 850 feet in altitude, a few rising to 900 and 1,000 feet. ‘The upper beds of the limestone series, however, are purer and dissolve more readily, so that in the central portion of the valley adjacent to the areas of overlying shale the surface is uniformly lower. Out of this lowland the shale hills rise abruptly, with steep escarpments, forming what may be called low plateaus. There are two main belts of shale in the Mercersburg-Chambers- burg quadrangles: one west of Chambersburg and Greencastle, which crosses the quadrangles from north to south, and is cut longitudinally by East Conococheague and Back creeks; the other a smaller north and south belt in the vicinity of Mercersburg. ‘These hills, or plateaus, are in general very level-topped (see Fig. 2), although the t Loc cit., p. 283. PHY SIOGRAPHIC STUDIES IN PENNSYLVANIA 481 rocks composing them are broadly folded and sharply plicated, and vary from fissile shale to soft but tough sandstone, which are alike planed off. The northern portion of the larger shale belt, lying between East Conococheague and Back creeks, is the highest of the shale tracts. It is a narrow plateau extending from the junction of the two creeks to beyond the northern limits of the quadrangle. Along its axis it has a nearly uniform elevation of 7 50 feet, but rises at one point in the north to 780 feet, and decreases to 700 feet near its southern apex and on its margins. It has been considerably dissected along its borders, especially on the western side, where the tributaries of Back Creek have trenched it deeply, but the intervening spurs still retain their level character and to some extent their original height. The descent from the plateau level to the stream bottoms on either side is very steep and abrupt. This is especially pronounced on the east side, where the valley is a limestone lowland. On the west the shale tract extends beyond Back Creek, and the plateau character con- tinues at a somewhat reduced altitude, but attains 720 feet elevation at the northern border of the quadrangle. There are no other extensive level tracts at this altitude in the area, but there are several scattered terraces and hilltops which approxi- mate this height. West of Mercersburg, near Cove Mountain, the shale hills rise to 700 feet, and a little farther north, where an inter- montane stream debouches at Cove Gap, an apron of mountain rock has been spread over these hills at an altitude of 740 feet. ‘The same is true of a small 720-foot terrace just east of Fort Loudon, which was covered by mountain wash from Bear Valley when its outlet was at this altitude. Another hill covered with water-worn gravel occurs at Franklin Gap in North Mountain, the outlet of Wilson Run. This, however, has an altitude of 800 to 820 feet. The east branch of Little Antietam Creek, which occupies a re-entrant valley in South Mountain in the southeast corner of the area, has bowlder-covered terraces on both sides of the stream at 760 feet elevation. A similar but more extensive level tract covered by quartzite bowlders, occurs at Black Gap, in South Mountain, where the East Conococheague leaves the mountains and enters the valley. Here a very level plain extends for four miles along the creek at an elevation of 840 feet. 482 GEORGE W. STOSE Nowhere on its surface is bed rock exposed, and only one or two out- crops were found in stream cuttings on its borders, indicating a deep deposit of stream wash. Inthe limestone area on the eastern side of the valley there are numerous hilltops which approximate 740 to 760 feet in altitude, but many rise higher, and there is no apparent uniformity It is clear that the valley in this region was once a nearly level plain, which has been deeply eroded, leaving remnants at approxi- mately 700 to 750 feet elevation. Campbell! has described a pene- plain preserved throughout this region on the shale at approximately this altitude. -He has named it the Harrisburg peneplain from its typical development at Harrisburg, Pa., and assigns to it an early Tertiary age. It seems strange that the peneplain should be pre- served in the Chambersburg plateau, which is exposed to such active erosion from tributaries of two large creeks, whereas in more favorable areas only small remnants remain, but this is probably due to a soft but tough sandstone which is locally interbedded and infolded in the shale of this hill and has aided in its preservation. Nearly all the other areas occurring at this altitude were protected by a covering of stream gravels. At Black Gap the gravel-covered plain has a uniform altitude of 840 feet. The capping of gravel and cobble, judging from the nearest bed rock observed, is about 60 or 80 feet thick. East Conococheague Creek, which issues from the mountains at this point, is a large stream and drains a considerable portion of the moun- tains to the east. On leaving its confined channel in the gap and entering the open plain its velocity would be slackened, its transport- ing power lessened, and a portion of its load of mountain rock would be dropped. In this way a delta has been built up 60 or 80 feet above the general level of the peneplain. The great extent and level character of this delta bears evidence of a prolonged halt in the uplift of the land and of active erosion on the headwaters of the stream. The delta at Franklin Gap, which also stands at 800 to 840 feet ele- vation, was similarly built up above the plain. Other local delta gravels at about this height occur at smaller gaps along the mountain front. In Path Valley two well-marked terraces on the shale at 820 feet may represent this stage but at a higher level in the narrow tribu- tary valley. t Loc. cit., pp. 283-91. PHYSIOGRAPHIC STUDIES IN PENNSYLVANIA 483 The larger streams have very meandering, tortuous courses, not due to present aggraded conditions, since they occupy deep gorges in the broad shale areas and their grade is about too feet in 6 to to miles air line. ‘These crooked streams originated on a graded plain, an additional evidence of the existence of a peneplain at about the Harrisburg level, and this mature drainage was rejuvenated by uplift and cut deep, sinuous valleys. ‘The spurs between the bends of the streams in the shale areas are terraced at various levels, ranging in altitude from 580 to 680 feet. Many of the terraces and slopes along East and West Conococheague creeks are covered with quartzite bowlders transported from the mountains by these streams during the cutting of the gorges, but no such deposits have been found on the surface of the Chambersburg plateau. As to the presence in this area of a lower peneplain of later Tertiary age equivalent to the Somerville peneplain of Davis, as suggested by Campbell,* the evidence is not so clear. Along both Back Creek and East Conococheague Creek terraces at 680 feet are very con- spicuous, and the shale plateau near St. Thomas also attains this alti- tude. ‘To the south the level tops of the central shale belt are all at 600 to 620 feet, and in many cases this altitude is maintained to the ends of the spurs between the creek bends. Other spurs have been lowered to 580 and 560 feet. At Upton the 600-foot plain is very pro- nounced and extends several miles on to the limestone area to the northwest. In the western shale belt the upland is at 600 to 620 feet elevation, with a few higher tables previously mentioned nearer the mountains. ‘The limestone tracts adjoining the shale have in general been reduced to a rolling lowland about 550 to 560 feet in altitude, which probably indicates a more recent epoch of erosion affecting these soluble rocks. The limestone area along East Conococheague Creek from Chambersburg south, however, stands at 600 to 620 feet elevation, forming a very level tract covered largely by stream gravel. Of these later erosion features the most marked is the 600-foot plain which forms plateaus of relatively wide extent and level charac- ter. This, if any, represents the lower peneplain of late Tertiary age which Campbell correlates with the Somerville plain of Davis. The t Loc. cit., p. 287. 484 GEORGE W. STOSE 680-foot terrace probably represents an earlier halt of short duration and the 560-foot lowland a very recent broadening of the limestone valleys. There are recognized in this area, therefore, the Schooley pene- plain at 2,000 to 2,100 feet elevation forming the mountain summits, the Harrisburg peneplain at 750 feet on the higher shale hills, a pene- plain at 620 feet on the lower shale hills possibly equivalent to the Somerville plain of New Jersey, and intermediate uncorrelated ter- races at various levels. GEORGE W. STOSE. WASHINGTON, D. C. UBER DIE GEGENSEITIGEN BEZIEHUNGEN ZWISCHEN DER PETROGRAPHIE UND ANGRENZENDEN WIS- SENSCHAFTEN.* In wenig anderen Naturwissenschaften haben seit dem letzten Drittel des vorigen Jahrhunderts so tief eingreifende Verinderungen Platz gegriffen, wie in der Petrographie; erst in den jiingsten dreissig oder vierzig Jahren sind jene feineren Untersuchungsmethoden ersonnen, ausgebaut und fruchtbar gemacht worden, denen sie einen Theil ihrer heutigen Gestaltung verdankt, vor allem die Herstellung der Diinnschliffe, die Benutzung des Mikroskops sowie die Ver- werthung anderer optischer Instrumente. Und mit dem Maass der dadurch gewonnenen neuen thatsachlichen Erkenntniss wuchs auch das Bestreben, unter Beriicksichtigung geologischer Beobachtungen die Einsicht in den causalen Zusammenhang petrographischer Erscheinungen und in genetische Verhaltnisse zu vertiefen, mit dem Descriptiven das Speculative zu verbinden. Im Verlauf jener Zeit ist nebenbei die Anzahl der Forscher auf diesem Gebiete ganz ausserordentlich gestiegen, zum Theil in Folge der Anregung und Unterstiitzung, welche die inzwischen neu errichteten Institute dar- boten, waihrend die Aufsammlungen und Arbeiten der geologischen Landesanstalten das Untersuchungsmaterial ins Ungemessene ver- mehrten. Die petrographische Literatur, vordem ausser in Deutsch- land fast nur in England, Frankreich und Skandinavien gepflegt, hat einen sozusagen internationalen Charakter angenommen und die Vereinigten Staaten sind, nachdem einmal eine Schaar ausgezeich- neter junger Gelehrten in Europa Ausbildung und Interresse gewonnen hatte, durch selbstandige und unabhangige Weiterarbeit mit in die allererste Reihe getreten. Es gibt gar keine Wissenschaft, welche fiir sich ganz allein, ohne passive oder active Beeinflussung bestehen kénnte; wie sie alle zu ihrem Ausbau die Aufnahme von Ergebnissen verwandter Disci- plinen bediirfen, so spendet auch jede wiederum von ihren eigenen Resultaten etwas zur Foérderung anderer. « Address presented at the International Congress of Arts and Science, Universal Exposition, St. Louis, September 22, 1904. 485 486 FERDINAND ZIRKEL Wenn die Petrographie sich mit dem Material, weleher die aussere feste Erdkruste zusammensetzt mit den Gesteinen, beschaftigt, so ist es nicht zweifelhaft, dass Mineralogie und Geologie, Physik und Chemie die zunachst verwandten und solche Wissenschaften sind, welche im Dienste der Petrographie durch friedliche Assimilation selbst zur Petrographie werden, wie jeder Stein, den man zum Gebiaude benutzt, dadurch ein Baustein wird, mag man ihn sonst noch nennen, wie man will. Handelt es sich um die beiden Fragen: erstlich, was alles tragen die benachbarten Wissenschaften zum Ausbau der Petrographie bei, und zweitens, was vermag umgekehrt die Petrographie aus dem Umfange ihrer eigenen Erfahrungen abzugeben, um auf angrenzen- den anderen Gebieten Verstandniss von gesetzlichen Erscheinungen zu schaffen oder dic Losung von Problemen anzuregen, so scheint es, dass unsere Wissenschaft wohl im Ganzen mehr als Empfangerin, denn als Geberin dasteht, wenn auch nicht in demselben Maasse genahrt und unterstiitzt, wie es bel jenem grossen Complex von heterogenen Disciplinen der Fall, den man die moderne Geographie nennt. In einer Beziehung liegt die Sache freilich ganz anders, bei dem Verhaltniss der Petrographie zur Mineralogie. Jeder, der in den letzten Jahrzehnten auf beiden Gebieten thatig war, oder gar, wie es bei mir zutrifft, auch den modernen Aufschwung der Petrographie noch mit erlebt hat, wird zugeben, dass fiir speciell petrographische Zwecke unternommene Studien unendlich mehr mineralogische Frucht getragen haben, als es umgekehrt der Fall. Zwar waren in den fiinfzigen Jahren schon vereinzelte zusammenhanglose Ver- suche gemacht worden, isolirte Mineralien mit dem Mikroskop zu betrachten, Versuche aber, die bei der damals herrschenden Gleich- giiltigkeit, Verstandnisslosigkeit oder Skepsis sozusagen ohne jede weitere Bedeutung blieben. Die methodische und verallgemeinerte mikroskopische Untersuchung der Mineralien hat jedoch erst bei den Diinnschliffen derjenigen eingesetzt, die eine Rolle als Gemeng- theile von Fe/sarten spielen und in deren Erkenntniss der Hauptauf- gaben der Gesteinskunde beruht, so dass alle diese Forschungen in viel grdsserem Maasse um specifisch petrographischer als um specifisch’® mineralogischer Zwecke willen unternommen worden sind. Und alles, PETROGRAPHIE u. ANGRENZENDE WISSENSCHAFTEN 487 was nun fiir die Gesteinsmineralien mit wachsendem Eifer fest- gestellt wurde; die Lage der optischen und Elasticitatsaxen in ihnen, die Brechungsquotienten und die Absorptionscontraste, die Cohisi- onsverhiltnisse, die Gesetze ihrer Zwillingsbildungen und die Beschaf- fenheit ihrer feineren Structur, die Natur der in ihnen enthaltenen festen und fliissigen mikroskopischen Einschliisse, die Erscheinungen der Zersetzung und Verwitterung, die Umbildung in neue epigene- tische Substanzen—alles dieses ist nun auch der eigentlichen Mine- ralogie zu Gute gekommen. Auf die Entwickelungsgeschichte zahlreicher Mineralien ist erst Licht gefallen, als man veranlasst war, die /pelrographischen Vorkommnisse derselben zu_studiren. Wie sparlich waren, bevor die Gesteinskunde sie in ihr Bereich zog, unsere Kenntnisse von ‘Titaneisen, Sillimanit, Cordierit, Zoisit, Tridymit, tiber Nephelin, Leucit, Melilith und viele Feldspatharten, iiber die Glieder der Pyroxen-Amphibolgruppe; wie diirftig wiirden die Lehrbiicher der Mineralogie erscheinen, wenn alles aus ihnen hinweggenommen ware, was auf Grund von _petrographischen Arbeiten jetzt ihren Inhalt bereichert und anziehend macht. Jene petrographisch-geologische Theorie, durch welche Bunsen die ver- schiedenartige chemische Zusammensetzung der eruptiven Felsarten erklaren wollte, spiegelt sich wieder in der geistvollen und frucht- bringenden Auffassung Tschermak’s von dem Aufbau der triklinen Feldspathe aus zwei chemisch differenten aber isomorphen End- gliedern. Dass bei allen diesen mineralogisch-petrographischen Studien phystkalische Methoden unausgesetzt zur Geltung kommen, ver- steht sich von selbst. Wenn aber auch so die optisch-physikalischen Instrumente zum Gemeingut der Petrographen geworden sind, so sollte nicht ttbersehen werden, dass die Letzteren fiir ihre speciellen Zwecke gewisse derselben eigens ersonnen, an anderen werthvolle Verbesserungen angebracht haben, was Alles wieder der eigentlichen Physik zu Gute kommt. Und sodann, dass ein betrachtlicher Theil der optischen und thermischen Gesetze tiberhaupt erst ergriindet oder bestitigt werden konnte an Objecten, die dem Steinreich ent- stammen. Das physikalische Verfahren bei der Fractionirung heterogener Gemenge mittels schwerer Fliissigkeiten ist durch seine Anwendung auf petrographischem Gebiet jetzt hoher Vollendung 488 FERDINAND ZIRKEL entgegengefiithrt worden. Nur zum Theil petrographischer, vor- wiegend geologischer Natur sind die Untersuchungen, welche sich bestreben, die Gesetze der Mechanik anzuwenden auf das Gesteins- material, welcher einer Deformation, Torsion, Zerreissung anheimfiel. Schon sehr lange besitzen wir chemische Bauschanalysen von Gebirgsarten, ferner sog. Partialanalysen der in Sauren léslichen oder zersetzbaren und der davon unangegriffenen Antheile, Analysen der enizelnen isolirten Gesteinsmineralien, wenn alles dies auch anfangs vielleicht nur mehr als ornamentale Verbramung der Gesteins- beschreibung betrachtet und vielfach von wenig erfahrenen Novizen unternommen wurde, dann auch eine Periode der Vernachlassigung eintrat, wo das rapid wachsende Studium der Kohlenstoffverbin- dungen als ein verlockenderes und méglicherweise finanziellen Gewinn bringendes Gebiet erschien. Augenblicklich ist die Anwendung der analytisch-chemischen Untersuchungsmethoden auf das petrogra- phische Material, in ihrer unabweisbaren Bedeutung nicht hoch genug anzuschlagen, mehr denn je zur Geltung gekommen, und wie von jeher stehen mit Recht die massigen Eruptivgesteine und die Krystallinen Schiefer im Vordergrunde des Interesses. Ja, in den letzteren Jahren scheint man gar in der Beriicksichtigung der chemischen Specialitaten dann zu weit zu gehen, wenn man auf Grund von geringfiigigen Differenzen unter den einwerthigen: oder unter den zweiwerthigen Metallen oder zwischen beiden gleich Veranlassung nimmt, neue belastende Namen fiir diese tiberhaupt nicht st6chiometrisch zusam- mengesetzten Gesteinsmassen aufzustellen. Hochst werthvolle und zahlreiche Einzelbeitrage sind jetzt im Laufe der letzten Zeit auch von Seiten der U. S. Geological Survey geliefert worden, viele Hunderte von Analysen nach immer mehr vervollkommneten und den strengsten Anforderungen gentigenden Methoden, nach Methoden, die auch gezeigt haben, dass als ausserst selten geltende Stoffe, wie Vanadin, Baryum, Strontium, sich in den meisten oder fast allen Eruptivgesteinen finden, Molybdan zwar sehr sparlich, aber unerwartet haufig. Hier ist vor allem der Name des verdienstvollen Hillebrand zu erwaihnen, dessen ‘“‘ Praktische Anleitung zur Analyse der Silicatgesteine” einen formlichen Schatz von Erfahrungen und Fingerzugen enthalt. Sehr richtig hob er hervor, vie wiinschenswerth eine Wechselwirkung zwischen chemischer PETROGRAPHIE u. ANGRENZENDE WISSENSCHAFTEN 489 und mikroskopischer Untersuchung sei, die beide haufig noch getrennt ausgefiihrt werden und dass, wenn das Studium der Diinnschliffe allemal den Analysen vorausginge, die letzteren viel leichter und exacter erledigt wiirden. Die chemisch-petrographische Literatur ist jiingst um ein wahr- haft grossartiges Werk bereichert worden, was ebenfalls diesseits des atlantischen Oceans mit bewundernswerthem Fleiss ausgearbeitet wurde. Henry Washington hat es, ein Nachfolger von Justus Roth, aber von moderneren Gesichtspunkten ausgehend, zu Wege gebracht, alle in den 16 Jahren von 1884-1900 ver6ffentlichten Analysen von Eruptivgesteinen und Tuffen zusammenzustellen und kritisch zu verarbeiten; neben den einleitenden Bemerkungen iiber Auswahl der Objecte, Materialmengen, Maass der Genauigkeit und der Ausfiihr- lichkeit, Irrthumsquellen u. s. w. ist vor allem wichtig der erste Versuch, den Werth der Analysen gerecht und unbefangen zu taxiren: von ahnlichen Erwagungen aus, nach denen der Credit eines kauf- mannischen Geschafts beurtheilt wird, unternimmt er es, im Hinblick auf den Grad der Exactheit und der Vollstandigkeit, die Analysen in 5 Gruppen zu bringen, welche in absteigender Folge die Praedicate: excellent, good, fair, poor, bad erhalten, ein sehr dankenswerthes Beginnen, hoffentlich zugleich ein Mahnruf an die Analytiker. Bei der Veranstaltung chemisch-petrographischer Analysen han- delt es sich einmal darum, tiberhaupt die Zusammensetzung einer Felsart festzustellen, sowohl um die procentarische Betheiligung der verschiedenen Stoffe daran zu erkennen, als auch eine Einsicht in die Stellung zu gewinnen, welche das Vorkommniss innerhalb gewisser chemischer Reihen einnimmt. Wahrend in einem normalen Ver- bande sich ein constantes Steigen und Fallen der Stoffe geltend macht, sind in dieser Hinsicht namentlich bemerkenswerth die eigen- thiimlichen Ultraglieder, z. B. die ganz abseits stehende Gruppe der trotz grosser Basicitat fast thonerdefreien und alkalifreien, aber enorm magnesiareichen Eruptivmassen, und ein anderes, vielleicht noch auffalliger aus dem allgemeinen normalen Rahmen heraus- fallendes Glied mit kaum 20 pro cent. Kieselsdure, alles andere fast nur Thonerde, dennoch aber ein achtes Durchbruchsgestein. Die neuere Zeit hat, auch zu dem Zweck, die Verwandtschaften hervortreten zu lassen, viele Bestrebungen hervorgebracht, formel- 490 FERDINAND ZIRKEL ahnliche einfache Ausdriicke fiir die chemische Gesteinszusam- mensetzung zu gewinnen, sowie graphische Methoden zu ersinnen, wodurch das Verhiltniss der einzelnen Stoffe, welche meist als aus den Gewichtsprocenten berechnete Molecularproportionen erscheinen, zur Darstellung gelangt, und der Ort angegeben wird, den eine Analyse inmitten einer Schaar von anderen einnimmt. Loewinson-Lessing, Pirsson, Michel-Lévy, Migge, Brogger, Becke, Iddings, Osann haben auf diesem weiten Gebiete der chemisch-classificatorischen Formu- lirung, der Graphik und Topik verschiedene specielle Vorschlage gemacht. Der zweite Hauptzweck der chemischen Gesteinsanalysen besteht darin, die an einem Material erfolgten substanziellen Verdnderungen nachzuweisen, indem es mit demjenigen verglichen wird, an welchem dieselben nicht eingetreten sind. So haben die chemisch-analy- tischen Methoden jene grosse Summe von Kenntnissen aufgehauft, die sich beziehen auf den gesetzmassigen Verlauf der einfachen Ver- witterungen und der complicirteren Zersetzungen, welche unter dem Einfluss der allerwegen wirksamen Agentien und der dadurch zuniachst beschafiten carbonatischen und silicatischen Lésungen von Statten gehen. In das Verstindniss von diesem stillen Spiel der chemischen Verwandtschaften und von dem gegenseitigen Austausch der Stoffe in den Felsen und Erdschichten zuerst Ordnung gebracht zu haben, ist das unvergingliche Verdienst des grossen Meisters Gustav Bischof. Aber auch fiir die Einsicht in andere mehr locale Umwandlungen innerhalb der Gesteinswelt muss die Chemie helfend zur Seite stehen. Einmal da, wo in Folge des Durchbruchs von emporgedrangten Erup- tivmassen die angrenzenden Gebirgsschichten oft auf weite Er- streckung hin in denjenigen verinderten Zustand versetzt worden sind, den man den contactmetamorphischen nennt. Soweit die Einwir- kung des activen Eruptivgesteins auf die passive Umgebung in diesen hofahnlichen Umwandlungsgebieten erkannt werden kann, von der Grenze beider an, wo die Energie des Metamorphismus am inten- sivsten ist, bis dahin, wo die letzten f4ussersten Spuren in das unver- andert gebliebene Nebengestein ausklingen, findet sich des betroffene Material, je nachdem es sich solchen Einfliissen gegentiber mehr oder weniger empfanglich verhialt, in dieser oder jener Weise alterirt, indem an Hunderten tiber die ganze Erde verstreuten Orten auf PETROGRAPHIE u. ANGRENZENDE WISSENSCHAFTEN 401 eine in den grossen Ziigen iibereinstimmende Art sein Mineralbestand und auch sein Gefiige zu einem anderen geworden ist. Auf che- mischem Gebiete muss hier entschieden werden, ob es sich dabei um eine blosse moleculare Umlagerung der in dem Nebengestein vorhanden gewesenen Stoffe handelt, oder ob dasselbe auch eine wesentliche Veranderung seiner chemischen Zusammensetzung dadurch erlitten hat, dass die Eruptivmasse bei der Erstarrung etwa Stoffe aus sich ausschied und in dasselbe hinein abgab. Grosse Reihen von ver- gleichenden Analysen schienen, wenigstens fiir die Tiefengesteine, das Erstere zu bekraftigen, dass in der Regel diese contactmeta- morphischen Ereignisse erfolgen ohne Zufuhr und Abfuhr von Sub- stanzen, dass das active Gestein blos durch seine Eruption, durch die von ihm ausgetibten physikalischen Bedingungen des Drucks und der Temperatur wirkte, nicht auch durch die jeweilige Beschaff- enheit seiner eigenen Masse. Franzésische Forscher sind freilich im Gegensatz dazu der Ansicht, dass auch bei den tiblichen contact- metamorphischen Umwandlungen, z. B. von Thonschiefer in Hornfels, Fruchtschiefer, Garbenschiefer neu zugefiihrte Stoffe in dem Substrat eine Rolle spielen; dass Letzteres im Contact mit intrusiven Diabasen thatsaichlich der Fall, wurde schon frith durch chemische Analysen erwiesen. Und wenn bei dem Durchbruch gewisser Granite durch ein Nebengestein sich das letztere, abgesehen von den sonst gewohn- lichen Alterationen, mit neugebildetem Turmalin, Topas, Zinnstein, Axinit, fluorhaltigem Glimmer in immer wiederkehrender geschlos- sener Gesellschaft ausgestattet zeigt, so ist es nicht zweifelhaft, dass die Entstehung dieser, vielfach an Spalten gebundener Mineralien in Verbindung gebracht werden muss mit einer die Granit-eruption begleitenden fumarolenahnlichen Aushauchung von fluor- und bor- haltigen Dampfen, also thatsaichlich eine Aussendung fremder, che- mischer Stoffe in die Umgebung hinein stattgefunden hat. Nun gibt es aber auch noch eine andere Art der Gesteinsmeta- morphose als die durch Contact bedingte: die gebirgsbildenden Druckkrijte waren es, die in weiten Regionen das Material, auf welches sie wirkten, zusammengepresst, gestaucht, zermalmt haben, wobei es dann in der Regel zur Erwerbung eines anderen, nament- lich schieferigen Gefiiges und daneben auch zur Herausbildung eines abweichenden Mineralbestandes gekommen ist. Dabei erhebt sich 492 FERDINAND ZIRKEL aber die wichtige Frage, wie es mit der chemischen Beschaffenheit solcher Druckproducte bestellt ist. Auf ein unzulangliches Material gestiitzt und im Banne von tendentiés erwiinschten Vorstellungen hat man den Satz ausgesprochen, dass auch mit den weitestgehenden Umgestaltungen in Structur und Mineralfiihrung eine chemische Verinderung von nennenswerthem Betrage micht verbunden sel. Es ist das Verdienst von Reinisch, durch eine umfangreiche Analysen- reihe dies als Irrthum aufgedeckt und fiir die druckmetamorphisch umgebildeten Orthoklasgesteine, fiir gepresste Diabase gezeigt zu haben, dass sie in gesetzmissiger Weise sogar einer recht erheblichen chemischen Veranderung unterlegen sind. Fir die einzelnen Stoffe konnen bei normalem und gepresstem Gestein, welches den Gewiissern Unmengen neuer Angriffspunkte bietet, die Differenzen so bedeutend werden, dass von einer unyersehrten Erhaltung des chemischen Bestandes keine Rede mehr ist und dass damit das friiher versuchte zuriickschliessen aus der Analyse des Druckproducts auf das urspriing- liche Gestein—bei dem ganz verwischten chemischen Bilde des letzteren—einfach unmoglich wird. Diese Beispiele zeigen, welch ein unentbehrliches Hiilfsmittel die chemische Analyse fiir petrographische Probleme darstellt. Aber die Fiille des Dankes liegt doch nicht ganz allein auf der einen Seite, es lassen sich vielmehr auch gewisse Beziehungen anfiihren, wo umgekehrt die Chemie zu einiger Erkenntlichkeit Veranlassung hatte, indem sie durch die Petrographie nicht ganz gleichgiiltige Anregung zur Verscharfung oder Erweiterung ihrer eigenen Methoden erhalten hat. Vor die Aufgabe gestellt, auch die nur ganz spurenhaft in den irdischen Gesteinen vorhandenen Elemente nachzuweisen, mussten die Chemiker darauf Bedacht nehmen, jene Reactionen ausfindig zu machen, wodurch diese Elemente am besten in ihrer Gegenwart erkannt, am scharfsten von einander getrennt und am sichersten quantitativ bestimmt werden kénnen. Die auf Begehr der Petrographie zu ihren Gunsten unternommenen Arbeiten z. B. von Hillebrand sind so der ganzen analytischen Chemie zu Gute gekommen. In Diensten der Petrographie fand Gooch die neuen Trennungsmethoden fiir Titan, Lithium, Bor und seinem erfinderischen Geschick verdanken die Chemiker den Gebrauch des PETROGRAPHIE u. ANGRENZENDE WISSENSCHAFTEN 493 perforirten Platinfiltrirtiegels und des gekrimpten Platintiegels zur Wasserbestimmung. Der Muineralreichthum der Stassfurter Salz- lager hat van’t Hoff Anregung geboten zu seinen Jahrelangen wichtigen Untersuchungen tiber Gleichgewichtsverhaltnisse, Loéslichkeitscurven und Bildungsbedingungen von Hydraten, Doppelsalzen und Pro- ducten des doppelten Umtausches. Neben der tiblichen Makrochemie ist letzthin auch eine Mikro- chemie erwachsen und ausgebaut worden. Hier unternimmt es das mikroskopisch bewaffnete Auge, die an. dem zu priifenden Object erfolgenden Verinderungen und die Natur der neu hervorgerufenen Producte zu erkennen. Haben die Reagentien auf ein winzig kleines Partikelchen oder ein Loésungstrépfchen gewirkt, so kommt es vor allem darauf an, beim Verdunsten zwar nur mikroskopische, aber so charakteristisch krystallisirte und optisch wohl gekennzeichnete Producte der Reaction zu erhalten, dass sie zur zweifellosen Erken- nung des dieselben bedingenden, in der Probe enthaltenen Elements verwerthet werden k6nnen, Wenn diese specifisch mikrochemischen Methoden, fiir zahlreiche Elemente ausserst befriedigend ersonnen und sehr haufig angewandt, jetzt der qualitativen Analyse zu Gebote stehen, so mége das Historische nicht vergessen werden, dass sie zuerst als etwas Neues lediglich um petrographischer Zwecke willen in die Wege geleitet wurden. Boricky war es, der 1877 bei seinen Gesteins- untersuchungen auf die Idee kam, die Mineralpartikel mit Kiesel- fluorwasserstoff zu behandeln, um Fluorsiliciumsalze der Alkalien, der alkalischen Erden u. s. w. zu erhalten, die durch ihre unter- scheidbaren Formen das in sie eingetretene neue Element verrathen. Immer weiter um sich greift die Uberzeugung, dass eine grosse Anzahl petrogenetischer Probleme ihr Verstandniss finden wird auf dem Boden derjenigen Wissenschaft welche, obschon sie dem Namen nach zwischen zwei anderen steht, doch in letzterer Zeit mit ihren hochbedeutenden Errungenschaften eine férmliche Selbstandigkeit beanspruchen darf, der physikalischen Chemie. Dass ihre Grund- zatze, Gesetze und Arbeitsmethoden fiir petrographische Gebiete fruchtbar gemacht werden kénnen, zum Theil schon verwerthet worden sind, mégen folgende skizzenhafte Andeutungen darthun. Es ist eigenthiimlich, dass ein Begriff, der als ein anscheinend neuer, durch seine Aufstellung grosses Interesse erweckte, der der 494 FERDINAND ZIRKEL festen Lésungen, in der Petrographie schon lange Zeit vorher als eine selbstverstaindliche Thatsache gegolten hat. Wir haben von jeher gewusst, dass, indem das Lavamagma eine schmelzfliissige Lésung mit wechselndem Verhaltniss der Mischungstheile ist, auch sein chemisch identisches, homogenes, festes amorphes Erstarrungspro- duct, welches entsteht, wenn die moleculare Beweglichkeit aufhort, bevor krystallinische Ausscheidung beginnt, dass das natiirliche Glas nichts anderes sein kann, als eine unterkiihlte festgewordene Losung. Bei den natiirlichen Silicatschmelzfliissen wird nicht mehr ein Gegensatz von gelésten Korpern und einem Lésungsmittel von bestimmter stéchiometrischer Zusammensetzung angenommen, es sind gegenseitige wahrscheinlich dissociirte Lésungen; die Specula- tionen tiber die Natur der auch ihrer Herkunft nach stets ganz pro- blematisch gewesenen Lésungsmittel sind dadurch belanglos gewor- den. Die Gesetze, welche die Krystallisation aus wasserigen Losungen beherrschen, miissen, wie schon Bunsen hervorhob, auch giiltig sein fiir schmelzfltissige. Zweifellos steht auch die Consolidation von natiirlichen Schmelzfliissen unter der Herrschaft der von Gibbs fiir Salzlosungen aufgestellten Phasenregel; aber in Folge der Compli- cationen, welche durch die Gegenwart so vieler im Magma gelost vorhandener Verbindungen bedingt werden, diirfte es schwerlich gelingen, auf diesem Gebiete ihre Wirkung auf die Ausscheidungs- folge zu specialisiren. Als ein altes Hauptproblem gilt die Rezhenfolge, in welcher die einzelnen Mineralgemengtheile eines gleichmassig kornigen Erup- tivgesteins festgeworden sind, oder, genauer ausgedriickt, in welcher sie zu krystallisiren angefangen haben. Dass es keineswegs, wie Rosenbusch glaubte, die zunehmende Aciditaét ist, sondern nach dem Hinweis von Lagorio weit mehr die Natur der Basen, wodurch diese Succession in normalen Fallen geregelt wird, diirfte nicht mehr auf Widerspruch stossen. Diejenigen Stoffe, welche am leichtesten sittigen, werden zuerst zum Aufbau der Ausscheidungen verbraucht, die am schwersten sattigenden zuletzt Und experimentell ist nach- gewiesen, dass die abnehmende Reihenfolge in der Sattigungsfahig- keit ftir schmelzende Silicatlésungen lautet: Eisenoxyde, Magnesia, Kalk, Natron, Kali und Thonerde, welche erst relativ spat in das PETROGRAPHIE u. ANGRENZENDE WISSENSCHAFTEN 495 Moleciil der verschiedenen Gemengtheile eintritt, dann Kieselsiure selbst. Doch legen hunderte von wohlstudirten Beispielen vor, wo die darnach construirte Reihenfolge: Eisenerze, Olivin und rhom- bische Pyroxene, monokline Pyroxene, Amphibol und Biotit, Anor- thit, Kalknatronfeldspathe, Nephelin, Albit und Aegirin, Orthoklas, Quarz nicht eingehalten wird, sei es, dass diese Reihe an gewissen Stellen eine Umkehrung erfahrt, oder dass Mineralien, welche erst nach einander hatten krystallisiren sollen, gleichzeitig ausgeschieden vorliegen. Blos zweierlei scheint ganz festzustehen: erstlich, dass in den kieselsiurereichen Gesteinen solcher Art mit Gehalt an Quarz dieser in der Regel mit zu den letzten Verfestigungen geh6rt, und sodann, dass die Trager der nur dusserst sparlich oder spurenhaft in dem Magma vorhandenen Stoffe, der Phosphorsaure, Zirkonsaure, Titan- siure u.s. w., also Apatit, Zirkon, Rutil, Titanit, IImenit, Perowskit zu allererst zu krystallisiren anhuben, wenn sie auch in einigen Fallen, gleichwie die Erze, eine nicht unbetrachtlich lange Ausscheidungs- dauer besitzen. Es ist fraglich, ob die friihe Festwerdung dieser Accessorien, wie oft geglaubt wird, in der That auf ihrem geringen Mengenverhaltniss beruht, denn weil die Lésung dann fiir dieselbe verdiinnt erscheint, hatten.sie wohl eigentlich gerade umgekerht erst ganz spat auskrystallisiren mtissen. Da man auch nicht, in etwas drastischer Weise, dem Magma das Bestreben zuschreiben kann sich dieser Fremdkérper gewissermassen zunachst zu entledigen, so ist zur Deutung der Thatsache vielleicht eher anzunehmen, dass jene Mineralien in der silicatischen Lésung bei niedrigeren Temperaturen besonders schwer léslich sind. Die Ursachen fiir jenes abwechslungsvolle Verhalten der Haupt- mineralien in ihrer Krystallisationsfolge sind zum guten Theil noch recht unbekannt. Die Behandlung der Frage gestaltet sich aber dadurch besonders schwierig, dass man bei Experimenten und theo- retischen Erwaigungen nur mit zwei Substanzen in Lésung zu operiren pflegt, wahrend ein Silicatgesteinsmagma in der Regel iiber vier Substanzen gleichzeitig gelost enthalt. Es wurde hingewiesen auf die Thatsache, dass in gewissen L6- sungen der Temperaturspielraum fiir das Herausfallen einer Verbin- dung, z. B. Leucit, ein engbegrenzter, fiir das Auskrystallisiren einer 496 FERDINAND ZIRKEL anderen Verbindung, z. B. Augit, unter sonst gleichen Umstanden-ein viel weiter begrenzter sein kann, so dass sich aus dem gleichen Magma je nach der Temperatur der Augit bald vor, bald nach dem Leucit auszuscheiden vermag. Auch Meyerhoffer hat gezeigt, dass je nach dem labilen Gleichgewicht aus derselben Schmelze bald a, bald 6 zuerst krystallisiren kann. Noch durch ein weiteres Moment kann die Reihenfolge der Aus- scheidungen verindert werden, durch den Druck. Da nach der iiblichen Auffassung die gesteinsbildenden Mineralien sich beim Erstarren aus ihrer Schmelze contrahiren so muss, wie Sorby und Bunsen nachwiesen, verstarkter Druck diese Contraction beférdern, d. h. die Krystallisation beschleunigen. Die damit zusammenhang- ende Verschiebung des Erstarrungspunktes erfolgt aber dann bei verschiedenen K6rpern ungleichmassig, so dass zwei Kéorper, die unter einfachem Atmospharendruck verschiedenen Erstarrungs- punkt haben, unter erhodhtem Druck, unter welchem die Schmelz- punkte naher zusammengeriickt sind, gleichzeitig erstarren kénnen, wahrend unter noch stairkerem Druck der vorher rascher erstarrende zum langsamer erstarrenden werden kann, auf Grund dessen sich die Reihenfolge der Ausscheidungen dndert, z. B. zwischen dem leichter schmelzbaren Augit und dem _ schwerer schmelzbaren Orthoklas. Nach Doelter kénnte auch die Krystallisationsgeschwindigkett in so fern von Belang sein, als der Vorsprung, welchen die schwerere Loslichkeit einer Substanz a fiir ihre friihere Ausscheidung hat, ein- geholt oder tiberholt wird durch die raschere Krystallisationstendenz einer leichter léslichen Substanz b. Tritt dies nicht, wie es der Fall sein sollte, allenthalben ein, so liesse sich als Ursache dafiir vielleicht die abweichende Viscositaét der Magmen anfihren, mit welcher sich die Krystallisationsgeschwindigkeit andert; sollte die Viscositat, die Zunahme der inneren Reibung, der Auskrystallisation von @ und 6 gleichmassig entgegenwirken, so wiirde jener Vorsprung nicht so leicht oder iiberhaupt nicht eingeholt werden. Andere physikalisch-chemische Fragen auf diesem Gebiete sind die warhscheinlich zu verneinende, ob und wie weit die Reihenfolge der Ausscheidungen beeinflusst wird durch die relaizven Mengenver- hiltnisse der Bestandtheile; ttber die noch wenig untersuchte Wir- PETROGRAPHIE u. ANGRENZENDE WISSENSCHAFTEN 497 kung der sog. Imp/-krystalle; sodann, ob in gewissen gleichmissig-fein und in bestimmtem Verhaltniss gemengten Agegregaten zweier Mi- neralien, die sich an den Felsarten betheiligen, etwa das Product einer eutektischen Mischung im Sinne Guthrie’s, analog den Kryohy- draten, vorliegt. Weiterhin tiber die Rolle, welche die sog. Mineral- isatoren, die “agents minéralisateurs” bei der Verfestigung des Magmas spielen, jene darin vorhandenen, zum Theil gasigen Stoffe, welche auf die Auskrystallisation rein katalytisch zu wirken scheinen, d. h. dieselbe beférdern, ohne selbst dabei verwandelt zu werden und ohne in die bei ihrer Gegenwart sich bildenden Substanzen einzutreten. Besser sind wir, namentlich durch Iddings, dariiber unterrichtet, durch welche Ursache die so haufigen magmatischen Corrosionen und Resorptionen, die Wiederauflésungen bereits ausgeschiedener Gemengtheile bedingt werden, wobei es sich um Verschiebung des Gleichgewichtszustandes zwischen der festen und fliissigen Phase handelt. Ganz besonders wird aber auch die Mithiilfe der physikalischen Chemie Noth thun bei der Erklarung der Differenzirung der Mag- men, der weithin verbreiteten Erscheinung, dass umfangreiche Erup- tivmassen, auch miachtige Giange, sich gespalten haben in ein saureres, vorwiegend alkalisches, auch thonerdereicheres und in ein basischeres, an Eisen- und Magnesiasilicaten reicheres, an Thonerde und Alka- lien armes Theilmagma, wobei das erstere fast immer im Centrum, das letztere als basische Randfacies an der Peripherie lagert. Die Entstehung dieser Theilmagmen muss wiahrend des Fliissigkeits- zustandes durch Diffusionen in entgegengesetzter Richtung vor sich gegangen sein und so handelt es sich insbesondere um zwei Fragen: 1., welche Krafte tiberhaupt die Separation in die abweichend beschaf- fenen polar entgegengesetzten Theilmagmen, das Zusammengehen der zweiwerthigen Metalle mit einander und mit wenig Silicium, das der einwerthigen mit mehr Aluminium und mehr Silicium veranlasst haben, und 2., weshalb das acidere Theilmagma nun gerade die centrale, das basischere die peripherische Stelle eingenommen hat. Mehrere Einwendungen lassen sich dagegen erheben, hier das von Soret aufgestellte, von van’t Hoff ausgebaute Princip zur unmit- telbaren Anwendung zu bringen, dass der oder die Bestandtheile, mit denen eine Solution nahezu gesattigt ist, sich an den kalteren Stellen 498 FERDINAND ZIRKEL anzuhiufen streben; der Satz ist z. B. nur fiir die Vertheilung eznes gelésten Stoffs in einem Lésungsmittel nachgewiesen. Guy und Chaperon’s Satz, dass die Schwere mitwirkt die Homogeneitat einer Solution aufzuheben, kann hier keine Giltigkeit beanspruchen, denn dann miisste das schwerere basischere Theilmagma in einem unteren, das leichtere acidere in einem oberen Niveau erscheinen und der Gegensatz zwischen Centrum und Peripherie wiirde dadurch nicht erklart. Wer aber in schwer verstandlicher Weise die Erscheinung der randlichen Basicitaét auf eine Einschmelzung angrenzenden Nebengesteins zuriickfiihren will, setzt sich mit einer Unmenge von Thatsachen in Widerspruch und verneint ausserdem tiberhaupt eine stattgefundene Differenzirung. Einen wesentlichen Schritt zur Forderung der Erkenntniss hat Brégger gethan, welcher in speciellen Fallen nachwies, dass es wohl nicht die einzelnen Stoffe, sondern bestimmte stéchiometrische Ver- bindungen gewesen sind, welche eine entgegengesetzte Diffusions- richtung verfolgt haben, indem sich die kieselsdturearmen Eisen- Magnesia-Kalksilicate in der einen, die kieselséurereichen Alkali- Thonerdesilicate in der anderen Richtung bewegten und dass ausser- dem diese Verbindungen als solche den Mineralien der Eruptivge- steine entsprechen, in denen, wie bekannt, ja auch Alkalien, Alu- minium nebst Calcium einerseits, Magnesium, Eisen nebst Calcium andererseits zusammenzugehen pflegen. Soseien es die am schwersten léslichen Verbindungen, welche nach der Abkiihlungsflache hin diffundiren und in so fern hange die Differenzirung zusammen mit den die Krystallisationstendenz beherrschenden Gesetzen. Ahnlichen Vorstellungen scheint sich Harker hingegeben zu haben. Hierin liegt gewiss eine wichtige Erlauterung, welche aber nur Thatsachen erkennt, keine eigentliche Erklarung gibt, und immerhin bleibt es eine Frage, worin denn nun jene treibende Kraft besteht, vermége deren der melanokrate Pol gerade eine peripherische, der leukokrate eine centrale Position einnimmt. Uber die Unterschiede in den Diffusionsconstanten der betreffenden Verbindungen ist nichts bekannt. Um ahnliche Gegensitze, wie sie zwischen Centrum und Rand in- einem und demselben Massiv bestehen, handelt es sich da, wo in einer Gegend viele sog. complementére Gdange aufsetzen, acidere neben PETROGRAPHIE u. ANGRENZENDE WISSENSCHAFTEN 499 basischeren, welche dann aufgefasst werden, als spaltenerfiillende Producte einer in tibereinstimmender Weise zur Geltung gekommenen Differenzirung eines plutonischen Magmas. Eine hoch wichtige Frage ist es, ob das plugaducsiae Eruptiv- magma beim Ubergang in den starren krystallinischer Zustand eine Verminderung oder eine Vermehrung seines Volumens erfahrt. Gustav Bischof, Mallet und David Forbes haben sich auf Grund von experimentellen Untersuchungen bei basischen Fliissen fiir das Ein- treten einer Zusammenziehung um ca.1/1o der Masse ausgesprochen, womit auch die Entstehung von Contractionsrissen in der verfestigten Lava tibereinstimmt. Die vielcitirten Versuche von Barus wurden in der Weise ausgefiihrt, dass das feste Gestein zuriickgeschmolzen und dabei im Einklang mit dem Vorhergehenden nun eine Volum- vermehrung der Schmelze constatirt wurde. Die Frage hat aber dadurch wieder die Aufmerksamkeit in hohem Grade auf sich gezogen, weil sie bei der neuen Vulkantheorie von Stiibel eine wesentliche Rolle spielt. Stitbel laugnet, dass der Druck der sich contrahirenden Erdkruste auf den eigentlichen gluthigen Erdkern die vulkanischen Erscheinungen bewirke; er halt dafiir dass es die in der langsam erstarrenden sog. Panzerdecke der Erde restlich erhalten gebliebenen und nestahnlich abgefangenen relativ kleinen Reservoirs von gluthfliissigem Magma sind, welche dadurch zur Eruption auf einem Ausbruchscanal gelangen, dass bei dem Erstarrungsprocess eine Volumvergrésserung eintritt. Da er nun aber, angesichts der bisherigen Ergebnisse, es doch selbst als wohl- begriindet anerkennen muss, dass wenigstens der Schlusseffect umgekehrt in einer Contraction besteht, so glaubt er es als héchst wahrscheinlich annehmen zu diirfen, dass innerhalb des Erkaltungs- processes eine voriibergehende Phase der Schwellung, der Volum- vermehrung, sich einstellt; experimentell ist aber dariiber gar nichts bekannt. Ein weiteres physikalisch-chemisches Princip, welches petro- graphische Vorgange innerhalb der Sedimentgesteine erklart, ist das Bestreben, die vorhandene Oberfldche fiir eine Summe neben einander gelagerter gleichartiger Individuen méglichst zu verklemnern. _ Nur wenn die Beriihrungsfliche zwischen einem Krystall und_ seiner gestittigten Lésung ein Minimum ist, scheint das Gleichgewicht 500 FERDINAND ZIRKEL zwischen beiden erreicht zu sein. Befeuchtet man das Pulver lés- licher Salze und lasst es langere Zeit stehen, so nimmt die Masse eine deutlich krystallinische Zusammensetzung aus grésseren Individuen an, ein Theil der kleinen Partikelchen wiachst in seinen Dimensionen auf Kosten der anderen, welche dabei als solche aufgezehrt werden. Auf ahnliche Weise wird auch durch Rekrystallisation nach solchem Gesetz und mit solcher Wirkung die Structurbeschaffenheit der- jenigen grosskérnigen Marmore gedeutet, fiir welche es wahrschein- lich ist, dass sie friiher ganz dichte Kalksteine dargestellt haben, indem unter der Gegenwart kohlensadurehaltigen Wassers die kleinen Kornchen das Bestreben haben, durch gegenseitige Assimilation und durch Umlagerung ihrer Moleciile zu gleichartiger Orientirung in einander aufzugehen und sich zu grosseren Individuen auszuwachsen. Ferner wird so das bisweilen ziemlich grobe Korn der Alteren Salz- bildungen verstandlich, wahrend die Absatze der Jetztzeit aus den Salzseen fast dicht ausfallen, ebenso das Wachsthum des Gletscher- korns vom Firn abwarts bis zum unteren Ende des Eistroms. Da die Petrographie einen Theil der Geologie bildet, so ist die enge Verbindung selbstverstaindlich; beide erginzen sich gegenseitig und eine Geologie ohne Petrographie gibt es nicht, wie auch keine Petrographie, welche die auf anderen Gebieten der Geologie gemach- ten Enfahrungen vernachlassigen kénnte. Davon hier specieller zu reden diirfte jedoch in ahnlicher Weise nicht erforderlich sein, wie wenn man das Verhiltniss von Palaeontologie und Geologie ausein- andersetzen wollte. So steht die moderne Petrographie heute da inmitten eines reichen Kranzes angrenzender Wissenschaften, von hiiben und driiben fliesst Anregung, Erkenntniss und Belehrung im gliichlichen Wechsel zusammen. “Wenn aber auch von unserer Wissenschaft kaum das stolze Anerbieten an die Nachbarn ausgehen darf: “do ut des, ich gebe, damit Du gibst,” so lasst sie doch nicht vergeblich und ohne ihrerseits zu viel zu versprechen, die bescheidenere Bitte erklingen: “da ut dem, gib Du mirgdann gebe ich auch etwas.” FERDINAND ZIRKEL. LEIpzic, GERMANY. AN OCCURRENCE, OF GREENSTONE SCHISTS IN THE SAN JUAN MOUNTAINS, COLORADO. :! In the course of a recent examination of the Needle Mountains quadrangle by the United States Geological Survey, a series of meta- morphic rocks was encountered that differs in many respects from those occurring near by in the Animas Canyon, which have been con- sidered to be of Archean age. The region is near the southwestern limits of the San Juan Moun- tains of southern Colorado, which are made up largely of Tertiary volcanic rocks. In that portion which is known locally as the Needle Mountains, and which lies in parts of San Juan and La Plata counties, the younger lava flows and breccias are absent, and ancient crystal- line or metamorphic rocks have been exposed by the dissection of a dome-like uplift, in which all of the sedimentary formations, as late at least as the last of the conformable Cretaceous beds, have been involved. These rocks, which are all of pre-Cambrian age, are granites, schists, and quartzites; the ones to be described, which may be referred to conveniently as greenstones, occur at the southern side of the uplift. During a hurried visit to the region in 1go1, one of the members of the party, who was familiar with the Marquette and Menominee greenstones, called attention to the similarity of these rocks to those of the Lake Superior region. In the next field season a more detailed study was made of the complex, and in the laboratory specimens of the Needle Mountains rocks were compared with those collected by the late G. H. Williams in the Marquette and Menominee locali- ties, as well as with specimens described by Cross from near Salida, Colorado, in the Arkansas Valley. Occurrence of the greenstones——The greenstones are found for a little over seven miles in a north-and-south direction on both sides of Vallecito Creek, midway between its head and the point where it joins Pine River. From east to west the area occupied by the green- stones is not more than two and a half miles wide at most. To the t Published by permission of the Director of the U. S. Geological Survey. . 501 502 ERNEST HOWE north and east they are sharply bounded by fault contacts with sedimentary rocks of pre-Cambrian age, or have been intricately infolded with them; another fault separates the greenstones on the west from a large mass of granite, while to the southward they disap- pear beneath the Paleozoic sedimentaries. Many of the exposures are of well-banded schistose rocks, and it was only natural that they were at first assumed to be a part of the great Archean complex of schists and gneisses known to occur near by in the Animas Canyon. On closer study, however, it soon became evident that there were marked differences between the two series of rocks. ‘The Irving formation, as the greenstone complex has been named, from Irving Peak, was found to consist, not only of the schists first observed, but also of massive basic rocks sometimes possessing a porphyritic structure, others partly mashed or brecciated, and a few distinctly granular, while no well-defined system of bedding or stratification could be made out. All of the rocks are of a dull green- ish color and appear to have undergone extensive alteration. At two places massive quartzite was found, and at a number of localities extremely siliceous schists occur, some of which have undoubtedly been derived from quartzites through dynamic metamorphism, while others must have been originally granites or closely allied rocks. A single band of siliceous magnetite some fifteen feet in thickness was also observed near the northern end of the series. Most of the exposures are a dull leaden-gray or green, in sharp contrast to the lighter-colored granites and quartzites to the north and west, and their somber tones add to the gloomy aspect of the valley sides which have,in many places, been swept by destructive forest fires. The dull monotony is occasionally relieved by dikes of bright red granite porphyry or pegmatite near the contact with the granite mass to the westward. The only marked variations in the Irving formation itself are the comparatively rare occurrences of a very light gray gneiss or mashed quartzite. Description of the rocks.—The majority of the rocks found in the Vallecito section display considerable textural variety, but appear to be o nearly the same mineralogical composition. Hornblende, chlorite, epidote, and rarely biotite can be recognized megascopically in nearly all, and usually these dark minerals appear to be in excess GREENSTONE SCHISTS IN SAN JUAN MOUNTAINS 503 of the lighter silicates; what feldspar there is has evidently been much altered. The most significant features of the rocks in different parts of the complex are the variations in texture and structure which are conspicuous in the field and still more so when the rocks are exam- ined microscopically. As will be shown later, these are due partly to original textural differences in the rocks themselves, and in part to their dynamic metamorphism that has produced, in some instances, finely laminated schists in which all traces of original structure have been obliterated. In the field and laboratory two distinct kinds of massive rocks have been recognized, and transitions between them and the schists may be followed in many cases. Rocks of the first kind are granular and of a medium texture. The second consist of porphyries with phenocrysts of eldspar and hornblende in a very fine-grained ground- mass, or are extremely dense, without phenocrysts and resemble the finer-grained diabases. A number of rocks of intermediate textures have been found, but the two groups may be considered as fairly well defined for purposes of description. Massive granular greenstone or metagabbro.—Occurrences of strictly granular rocks are not numerous. ‘The best examples have been found in a restricted area which includes Irving Peak and its southwestern flanks near the northern limits of the complex. The texture of these rocks is practically the same as that of many medium- grained gabbros, the only differences being that laths of plagioclase cannot be made out, and that the dark amphibole, although it appears to be in the form of blades or lath-shaped crystals, is, in reality, fibrous. ‘The average rock is moderately coarse, even-grained, but specimens from the summit of Irving Peak show a tendency toward the formation of fine-grained segregations richer in hornblende. On fresh surfaces the color is a dark bluish- to greenish-gray. A microscopical examination shows at once that the rocks are more or less metamorphosed. The chief constituents are horn- blende, plagioclase, and generally a very little biotite and magnetite. The plagioclase is seldom fresh enough to permit of an exact deter- mination of its character, but the large extinction angles indicate labradorite. The feldspar and hornblende are present in about equal quantities in large irregular patches rather than grains. The 504 ERNEST HOWE feldspars seem, in many cases, to have possessed crystal boundaries and were often typically lath-shaped.- At present, even in the fresh- est specimens, they are much altered to calcite, muscovite, zoisite, and epidote, and individual grains have been bent and broken, but are not crushed. Hornblende, which is, as a rule, quite fresh, is of the pleochroic fibrous variety, uralite, and there can be little doubt as to its secondary origin; it sometimes appears almost massive, but the borders are extremely ragged, and minute blades and needles are scattered throughout the rock and often penetrate the feldspar areas. Biotite is unimportant and occurs more in the nature of an accessory. — Although from their mineralogical composition these rocks might be classed as diorites, and although no trace of pyroxene has been found, still from their resemblance to rocks of other regions, especially in the Menominee district, where all the stages in the change from gabbro to hornblende rock may be observed, they are to be regarded as gabbros in a rather advanced state of alteration. Porphyritic and fine-grained greenstones.—The mineralogical com- position of the porphyries and greenstones of finer grain is essen- tially the same as that of the granular rocks, the hornblende being possibly a little more prominent. Alteration, however, is generally more advanced, but strangely enough the original structures are often well preserved, even where the feldspars have been completely saussuritized or changed to other secondary minerals, and the dark silicates have been altered to chlorite. The structure which seems to prevail in all of the finer-grained rocks is the ophitic, so typical of diabase, in which laths of plagioclase lie, as it were in a groundmass of pyroxene and magnetite. In the case of the greenstones the out- lines of the feldspar laths are distinct, and the crystals often appear to radiate from a central point. What should correspond to pyrox- ene is, in the greenstones, either a mesh of uralite needles or chlorite, together with smaller grains of undeterminable feldspar and occa- sionally quartz. Here, there can be no doubt that the rocks were derived from diabase or diabase porphyry by the well-recognized change of pyroxene to hornblende and the alteration of labradorite to saussurite, accompanied by the development of epidote and calcite. GREENSTONE SCHISTS IN SAN JUAN MOUNTAINS 505 Between these altered massive rocks and the finely laminated schists, a series of rocks may be found which illustrates the various stages of dynamic metamorphism, and satisfactorily proves the close relation of the schists and massive greenstones. The bending and fracturing of the feldspars is followed by more complete crushing, and the hornblende is broken up and redistributed in small parallel blades through the rock. In extreme cases recrystallization has probably taken place, and the minerals in such rocks are, as a rule, much fresher than in many of the less mashed varieties. Greenstone schist—The more or less completely schistose rocks which make up the major part of the Irving series differ but little from many of the schists of the Archean. They are fine-grained and well laminated and of a dark greenish-gray color. ‘The micro- scope shows them to be made up of pleochroic green hornblende in excess of finely granulated feldspar, and usually biotite, a little quartz, and magnetite. The parallel arrangement of the minerals, especially of the blades of hornblende and biotite, is very striking. Feldspar sometimes occurs merely as interstitial grains between the dark sili- cates, but usually it is present as a fine mosaic in long-drawn-out lenses or bands. In some cases chlorite has completely replaced hornblende, and muscovite has been developed at the expense of part of the feldspar. Siliceous schists— At a number of localities within the greenstone area schists and gneisses of a much less basic character have been found. ‘They are light gray or nearly white in color, the dark silicate is biotite, and the feldspar, when recognizable, orthoclase and micro- cline, with only small quantities of plagioclase. Quartz is abundant, and the microscope shows that magnetite, muscovite, and rarely augite or hornblende, may occur as accessories. In all cases the mashing has been sufficient to destroy original textures and develop an excellent schistosity or lamination. Except in one or two doubtful cases, to be noted later, these rocks appear to have been formerly intruded into the greenstones as granites, and subsequently to have shared with the older rocks in the mashing and deformation of the region. The exceptions just mentioned are some unusually siliceous rocks whose relations to the greenstones are not altogether clear. 506 ERNEST HOWE Like the others, they are very completely mashed and contain both feldspar and biotite, but the amount of quartz is largely in excess of that of all the other minerals. ‘The occurrence of massive quartzite with the Irving greenstones has already been referred to, and it seems more than likely that these siliceous schists are quartzites that lay within zones of great mashing and suffered with the rest of the rocks. Structure.—It has not been possible to find any evidence of original bedding in the Irving greenstones, although frequently, on account of their schistosity, the rocks appear to be stratified. ‘This banding, though predominant, is not a constant feature, and transitions may be observed from schists to unmashed rocks which have the compo- sition and textural characteristics of intrusives. In addition to the schistosity, which is generally vertical and with a northwest-southeast strike, the massive greenstones, more especially near their contact with the Algonkian conglomerates and quartzites, have been frac- tured and brecciated. These conditions were probably brought about at the time that the shearing and complicated infolding took place between the Algonkian sediments and the Irving greenstones. In a very few cases it has been possible to recognize dikes of compact greenstone cutting either coarser rocks of the same sort or siliceous schists. The petrographical examination of the rocks failed to show that they had occurred as surface flows, all of the characteristics being those of intrusive rocks. The only possible exceptions are certain rocks near the eastern border of the greenstone area. ‘They are con- siderably altered and have been mashed, and their field appearance suggested that they were flow-breccias or tuffs, with fragments or drawn-out lenses of greenish porphyry lying in a dense cementing material of a darker color. Unfortunately, decomposition has pro- ceeded too far for a microscopical examination to be of value; second- ary minerals are about the only ones that can be made out, but the texture suggests crushing or mashing. The occurrence of these rocks seems to be restricted to the region immediately adjoining the con- tact with the Algonkian conglomerates, where, as has been said, complicated infolding and fracturing have taken place, and it is more reasonable to suppose that these rocks are friction breccias, subse- quently mashed, than that they are of pyroclastic origin. GREENSTONE SCHISTS IN SAN JUAN MOUNTAINS 507 Age oj the greenstones.—The oldest rocks in contact with the Irving greenstones are the conglomerates and quartzites of the Algonkian to the north and east. The actual relations of the two formations have been obscured by faulting or infolding, but the greater age of the Irving is shown by the quantities of schistose and massive greenstone pebbles in the lower portions of the Algonkian conglomerate. The character of these pebbles also supplies infor- mation in regard to the age of the mashing of the Irving greenstones, the greater part of which evidently took place before the deposition of the conglomerate. Later movements occurred after Algonkian time and resulted in the further fracturing and mashing of the green- stones, as well as the lower portions of the conglomerate. As to the lower age limits, little can be said with certainty, except that the rocks have been less affected by dynamic metamorphism than any that are known to occur in the neighboring Archean areas. They have been regarded as of early Algonkian age, separated from the younger Algonkian sediments by an erosion interval of unknown extent. That the Irving, as now exposed, represents but a small part of a much greater series of rocks which formerly existed seems certain. Evidence of this is to be found in the great thickness of Algonkian conglomerates, for, although greenstone pebbles are con- spicuous in many places, quartzite débris is more abundant, and lenses or beds of magnetite or jasper bowlders are often seen high up in the section, all of which indicates the destruction of an earlier terrane, the only traces of which that may now be recognized being the comparatively rare beds of quartzite and magnetite associated with the Irving greenstones. These same remnants are believed to be inclusions in the diabase and gabbro which intruded sediments that have since been worn down to supply the materials for the younger series of quartzites. Comparison with rocks of other localities-—The collection of rocks from the Menominee and Marquette regions made by the late G. H. Williams, and described by him in a bulletin of the United States Geological Survey,t has been examined in connection with the pres- ent study, and a certain similarity between the two series of rocks t G. H. Wittiams, “‘The Greenstone Schist Areas of the Menominee and Mar- quette Regions of Michigan,” Bulletin No. 62 (1890), U.S. Geological Survey. 508 ERNEST HOWE has been found in a number of cases. Professor Williams states that ‘‘there is considerable evidence to show that the greenstones, both of the Menominee and Marquette regions, solidified at the sur- face, under subaérial or subaqueous conditions.”” ‘There are, how- ever, a number of instances where the rocks are clearly intrusive, and it is these that the Irving greenstones most closely resemble. The character of the metamorphism appears to have been the same in both regions, and the rocks from which the greenstones were derived were undoubtedly very similar. The most important differ- ence is that tuffs and surface flows have been recognized in the Lake Superior complex, while no rocks of this nature have been seen in the Irving. Flows or fragmental deposits of igneous origin may, of course, have been originally present, but all traces have been destroyed. The Irving greenstones and those occurring near Salida in the Arkansas Valley, some 125 miles to the northeast, possess a number of common characteristics; in fact, the description by Mr. Cross? of . the hornblendic members of the Salida rocks might be applied directly to those in the Needle Mountains. In both localities granular horn- blendic rocks occur and also denser rocks in which the microscope reveals an ophitic structure. The only rocks found in the Salida section which differ greatly from those of the Irving are ones which are in some instances sufficiently well preserved to indicate a struc- ture and composition like that of rhyolite, and which in extreme cases of metamorphism suggest finely mashed micaceous quartzites. ‘There are not sufficient data in regard to the field occurrence of these rocks to make it possible to say whether they were originally surface flows or intrusives, but in regard to the series as a whole, Cross is inclined to believe that they represent “‘a great series of surface lavas erupted in Algonkian time.” , In a recent monograph of the United States Geological Survey, Bayley? has redescribed the greenstones of the Menominee region under the name of ‘‘Quinnesec schists” and has referred them to t On a Series of Peculiar Schists Near Salida, Colorado. Proceedings af the Colorado Scientific Society, Vol. IV (1893), pp. 286-93. 2, W.S. BAyLey, “The Menominee Iron-Bearing District of Michigan,’ Mono- graph 46 (1904), U. S. Geological Survey. _ GREENSTONE SCHISTS IN SAN JUAN MOUNTAINS 509 the Archean. The greenstones of the San Juan Mountains are most clearly associated with Algonkian rocks, and, aside from their litho- logical similarity, there seem to be no good reasons for attempting to correlate such distant occurrences as the Quinnesec and Irving schists. ERNEST Howe. WASHINGTON, D. C. AN OCCURRENCE OF TRACHYTE ON THE ISLAND OF HAWAII! THE Hawai an Islands are commonly described as consisting almost wholly of basaltic lavas, though in truth there is a known wide range in the rocks thus broadly characterized. ‘These lavas have issued as flows from many centers, building up great volcanic mountains. The eruptions have continued for a long time with gradually shifting scene of action, and agencies of degradation have reduced some ancient basaltic mountains to mere reefs, while other volcanic piles are now like deeply dissected models, showing their constitution to the very core. With all this opportunity to examine the products of Hawaiian volcanoes of various epochs and at many centers, no one has, to the writer’s knowledge, found and described any other than basaltic or related lavas as existing in the entire group. However, during the summer of 1902, while the writer was engaged in a reconnaissance of the Hawaiian Islands for the Geological Survey, he was fortunate enough to discover an occurrence of trachyte which, from several standpoints, seems to merit description. The locality at which this unique rock was found is on the island of Hawaii, at the northern base of Mount Hualalai, one of the great basaltic volcanoes of the island, the last eruption of which took place in 1801, from a low-lying point near that at which the trachyte occurs. The accompanying map shows the geographic position of the trachyte locality. Mount Hualalai is a basaltic cone of the Mauna Kea type, rising 8,000 feet above the sea, consisting mainly of lava flows, but dotted with numerous small cinder cones and punctured by perhaps as many remarkable “pit craters.” The cinder cones are most numerous near the summit of the mountain, and, so far as the writer is aware, the products of all these recent and local outbursts are basaltic in character. ‘The later lavas on the northern slope are olivine-rich plagioclase basalts. « Published with the permission of the Director of the United States Geological Survey. 510 TRACHYTE ON THE ISLAND OF HAWAII Sil Near the northern base of Mount Hualalai is a tuff cone, notably larger than those on the upper slopes of the mountain, and forming a very striking feature of the landscape as seen from the north. This cone, called Puu Waawaa, seems at first, like many others, an excres- Makolea Ft Fic. 1.—Map of a portion of the island of Hawaii. (From map issued by Hawaiian Government Survey.) cence on the surface of the larger volcano, but the material of which it is made is the trachytic lava to be described, and it is a matter for future observation to determine whether the situation relative to Mount Hualalai is not, as the writer suspects, quite a matter of chance. The base of Puu Waawaa is at about 3,300 feet above the sea and its summit several hundred feet higher. Adjacent to the cone of Puu Waawaa on the north, and extending west of north for 512 WHITMAN CROSS three miles, is a very clearly defined bench or terrace with abrupt western border and northern end. A small point on its surface gives the name Puu Anahulu. This bench is indicated on the map by the hachures marking its steep western slope. On the south the terrace of Puu Anahulu merges with the low grade slopes of Mount Hualalai and on the east with the surface of still smaller angle down which lava flows have come from Mauna Loa, some thirty miles away. Some of the late flows from Mount Hualalai have spread over the terrace, and in places have fallen in a stony cascade down the steep western bank. Others from the same source have passed between the terrace and the cone of Puu Waawaa. Some from Mauna Loa have also poured over portions of the bench or passed around its northern end to the sea. Puu Waawaa has been encircled by lavas from Hualalai which have cut it off from the terrace. The terrace bench of Puu Anahulu is made up, so far as obser- vations go, of an agglomeratic aggregate of large and small fragments of the rock to be described. In general, this rock is much decom- posed by kaolinization, only the larger fragments still consisting in part of dark gray material. The softer, light-colored rock has been used for road metal on the new government road from Waimea to Kona, which passes over this bench. Both the bleached and the darker rock exhibit a rude schistosity due to a parallel arrangement of minute feldspar tablets, like that common in phonolite and in some trachytes. Puu Waawaa is scored by numerous radial ravines penetrating its slopes for fifty feet or more, and in all of them examined the typical structure of a tuff cone is revealed.. The cone seems to be made up of tuffs of varying texture, well stratified, and dipping nearly with the slope of the cone. Most of the tuff is made up of ash or fine gravel, with angular fragments irregularly distributed through it, which rarely exceed a few inches in diameter. The fragments consist of brown pumice, dark aphanitic, or black obsidian-like rocks, with some showing a mingling of the latter _ materials. The dark aphanitic fragments are not unlike some dense basalts of the island in appearance, yet resemble also the freshest rock from the bowlders of Puu Anahulu. TRACHYTE ON THE ISLAND OF HAWAII 513 Thin sections of the obsidian show it to be a colorless glass contain- ing streams of feldspar microlites in some parts and free from them in others. The dull aphanitic streaks and masses are largely crystal- line, with more or less of fine magnetite dust and ferritic globulites, and a colorless glassy base of variable amount. To ascertain the chemical character of this rock an analysis was made by Dr. Hillebrand of the black glassy portion, with the result given in column 1 of the subjoined table. It is clearly an almost anhydrous, trachytic obsidian of close relationship to many other known rocks, of which analyses are also given in the tables, a resem- blance which will be discussed in a later paragraph. There is every reason to suppose that the associated pumice is of the same character as the obsidian, since no other rocks were observed among the fragments of the tuff." Turning now to the rock occurring in fragments at Puu Anahulu, microscopical examination shows it to be a felsitic trachyte, holo- crystalline in the specimens examined, of typical trachytic texture. The feldspar microlites forming the greater part of the rock are mainly very small, averaging but 0.02-0.05™™ in thickness, less than 1™™ in length, and, while some are prismatic, many are more or less tabular in shape, causing by their nearly parallel fluidal arrangement the characteristic semi-schistose texture of the rock, which has already been commented upon. The feldspars are fresh, and among them may be recognized sanidine, anorthoclase, and probably albite. The sanidine forms somewhat larger microlites than the anorthoclase and albite, but a distinct porphyritic texture was not observed. The feldspars make up more than four-fifths of the trachyte. The norm of the obsidian analyzed indicates a possible content in nephe- lite for the perfectly fresh rock, but a careful search for this mineral failed to reveal it in the sections of the somewhat altered Anahulu trachyte. The remainder of the rock consists of magnetite and apatite, in small amounts, and two or more undetermined minerals which occur for the most part in the angular interstitial spaces between t The writer’s observations at Puu Waawaa were necessarily quite limited and leave much to be desired. He did not have time to explore the cone thoroughly, but from the facts noted on the north and east slopes, it seems probable that Puu Waawaa is composed solely of tracyhte tuff. 514 WHITMAN CROSS ANALYSES OF HAWAIIAN TRACHYTES AND ALLIED ROCKS. No. I 2 3 4 5 6 7 8 SIOR PAT AC aeers Rees rears 62.19] 62.11) 64.28) 62.70] 60.50) 57.52) 60.39] 63.09 AO sree cris conte tone 17.43|w 15.97} 16.40) 16.86] 18.46) 22.57] 18.44 Bes Ognarntonvecesa sec sie ee TO5\U22.0 7) ZO ts. O7] e222 |e Ok Ae |er2 Oo NAD Ee mba ae ooo No sas 2.64/w Soul Bog: Aaya Boma BA” se5<(6 Mig Op aiomias sonirretickers OQoAO|oavgde 0103) 10270], slip bs Co\P Onnes Onno CaO ie ieee cn esninsee OI9)| VOLES || sCIS|| (COsOG|!) OD OS Bowl xeloeall \ Ce) Nias Osi oatesi leader eathay cue eee 8.28) 16:80)" 97/28) 7.03) (0.46) 87458) 8uaal eos Nes Oe rrr acse ley ene et ston: Og ZEA = GO| SoA) - Gotal “Aces ala 7]| G28 Te Oe or eeiiien ta cmeritiseaale ote 0.39 ty 6o| 2:20] 9-79} 41-40] 1.80) 0.57! 0.62 des Oe acter a ses taes Pe I ce O.14 Tree oer ea Bl psittaci 0.21 COes Hee eee COMI) HRUCO "ls Samscihocgooec CN o.cios we EES lesa onc ADO PERM Ac obi oie a Gn eos OUR ide renee Doe) OLA HO. 7/5 OcOAle cas oo 0.45 LO ps cc av oo ae res OO eee he ely cle ie eet mrcigereecoal |e epee ree er Coste | a tea | eee a 1S O Sanita os acre on baa OeiMiles sioec OOS | nee Crile eens was aig coos c SO; bole aase Gugolow.ord oo 8010/0 LOTS sooty eala do C.crolhod Goal osoeoo colle Soka a ellos abo alle 666.6.0 Gli go Reais hear geataretoliattee oper eeeaaniaete OsSD2IE Sect ca einti cosets mea staal epee eel neem Cre Ogseeueds a cee eer ERAGE EN cars asl NSS s ccereiecd eeateateise oa ous cee viel | Ast meee I eet tert | eee INT @ ae eo es eta ticle are THOME Ale, gota tet ctr lNev sae Ae tae oP allie CHUTE Ae tora cede Dare IMEI eee cses ons A atotel Op Gallade clo tal Gea pena: OsAC)|a cao ole ONOS|Ereeavee Ba Orava fun screws OFO2 eres MOTE Deepa. Setel| Uses geeks lint fevicn cl ueee heres | eee a SEO! saitesiy are aged seanietne vey abear' NONEM|esee ET ACE ssi se ieres cl hepe ee onal hails pebealel ences eohreel lomencesteae Ts Ot ere ons eee tracey ater PACE! Riese co Ye a oa oe || ee arte [een OQOROB|s sae 100. 33/100.53/100.77| 99.64] 99.95|100.77 i No. 9 Io II 12 13 I4 15 16 SIO Remora e sha seve ay 61.08} 66.22) 63.24] 66.50] 63.20] 60.11} 63.76} 66.06 AE © Fierpasemces teas al ences acl 18.71] 16.22] 17.98] 16.25] 17.45] 19.01] 17.37] 16.46 Be OBS sive qiore minutes ACOH ABAC}S)) MR Oy | Oxovilly (Col. VASO ©, 10] DoWs BGO Waiae ama este coenc CEOS CSuwOl, “Catersl\> Cy UO)|a con o Oe gy|| aoseay| 365 ©) IMG Oates cuts tant (ates onelts ©108| = 08/771) 50.103) TOL TS On 75) On23| 0503s Os To CAO MG ee cease me ise ok TE. 50/2 1.22] 0.8) HOWSG||a ls4o] OOO|hst72-OK70) Nas Orsgscn 5 atin sess eres 8.68} 6.49] 6.27] 7.52] 6.90) 6.53] 6.69] 6.81 Kis Ohinc ae canedait aac he Mee = eG Gort Gobel Gack! oxo! Gaz SoS2 FS Oe Bache is deca tee 2) 2Tl SO) 24) PONS Os SO|k NON5O|L E37 OnA© 6 Tel OL y packers meal (hers OnOS| rons 7lleciom ah lene secret eaten cea (eens Sie (OG Fieraicue renee best cli tae | earl Vouk ee lea ed aber. EO Ae a cen cy (oye yl veel letal ance ee DO pastieite oct one lee sra es OstG|) 20/22 oO. 28l Ton 70) | On4 Oe OOO) TOn7 Ole mmr UA © EIR ee Tea ENB rar seat denen Sele demir cl htt pial beeerreld Lf seia.o 5 lso.0:6 0-8 PaO ernie ck Nolensnarens shar eelcne ee ee OLeTO MIRO} 22 RUN eas | bnler sical tel ret OpUOlons eo. SO sags eos croc tes cera cu | ewnteene OOD ake Jetise nj |[ervar treet | nree ave el | Oe sel ge Soyo ral ane eae (Ges oe has ter a Aero ae OH TA ORO E See liners natal lomarare ol loners amen lets coeriael | Seams Crs Oye eee ae Sekar ict seedvell ounerat jilleneiees MONE) | oisaee eel |seneriatell ae eee cee dated enter INTC Se Pe CIE SP ease cea (nee Mi erent hn r ars | erie Mewes y allarm at ere ii ais aipral [Seni ald IME © ieee cisco eusaetst ste: axe He Sian ore hae aveione Ol OSA soesodilocoace 0.37| 0.55 Ba ONG rey odie wath aray sult SSO eM OnPe A Ommialtye Gis eeailenied oellaecoe aalleia Sie-olalle'o6-3.0'9 STO asea vary eet Ieee OHO] Oboe sia'sSaollaoe so dllacaoosilos oso cllescio'oo ENO i pe mate e allan a Ry etna eae Ses ERAGE A Aisi st arie caller ga fev | ese cekone teal teaeeeer One| Coenen 99.86) 99.97/100.14]100. 46/100. 14|100.07] 99.28)100.35 TRACHYTE ON THE ISLAND OF HAWAII ys LIST OF ROCKS IN TABLE OF ANALYSES. Rock Locality Analyst 5 - IDAKE NIKO 6 coe a eba66 soc . Glaucophane-sélvsbergite Hornblende-sélvsbergite. . le diguinmite meyer ear: b. INOW ENaloioteic. oo unig oo biome ee leitchtielditenemrt yarn wm Pulaskitetee give seein os seehonoliteyaecn ice . Quartz-syenite-porphyry . OS OI ANBWNH meBiotite=tnrachyiterr stoi: eS tiwianlteney sis seine: meNordimarkite- meee PE BOStonitesaecie Geka eie . Pyroxene-syenite........ . Riebeckite trachyte...... Puu Waawaa, Hawail........ Puu Anahulu, Hawaii........ Cape Ann, Massachusetts... . Laugendal, Norway......... OB, NOMEN Vooncoscco Ga Guce Gran Nonwayeneecae see: Litchfield, Maine...... ebiacteen Salem Neck, Massachusetts.. . Devil’s Tower, South Dakota. Bearpaw Mountains, Mon- Yellowstone Park, Wyoming Laugendal, Norway.......... fonsenas, Norway cise. «ei Laugendal, Norway......... Ksunsamoy Finland.) 5550+... Berkum, Germanys. W. F. Hillebrand W. F. Hillebrand H. S. Washington L. Schmelck V. Schmelck L. Schmelck L. G. Eakins H. S. Washington L. V. Pirsson H. N. Stokes .| W. F. Hillebrand V. Schmelck G. Forsberg V. Schmelck N. Sahlbom H. Laspeyres the feldspars and much less abundantly in minute stout prisms. The two most distinct minerals are respectively colorless and clear pale yellow, of strong refraction and double refraction, approximately like diopside or acmite. Extinction is parallel for the colorless prisms and never more than 10° for the yellow ones. There is no visible pleo- chroism in either, hence lovenite seems to be excluded. The norm of the glassy rock shows the acmite molecule, but no recognizable acmite, aegirite, or riebeckite has been noticed. As the analysis shows zirconia and titanic acid in determinable quantities, it is possible that rare or unknown zirco- or titano-silicates are present. Clear glass and dark globulitic areas are present in small amount. A glance at this trachytic rock suggests its close relationship with the obsidian of Puu Waawaa. It is not sufficiently fresh, as to its dark constitutents, to warrant full analysis, but the feldspars are only slightly attacked, and a partial analysis by Dr. Hillebrand, given in column 2 of the table, demonstrates sufficiently the practical identity of the two rocks. The position of the obsidian in the Quantitative Classification oj Igneous Rocks? is shown by the norm calculated from the analysis, which is as follows: 1 WHITMAN Cross, JoSsEPH P. Ipprncs, Louis V. Prrsson, AND HENRY S. WASH- INGTON, Quantitative Classification of Igneous Rocks, Chicago, 1903. 516 WHITMAN CROSS Orthoclase’ - . - - 29.47 | Salic molecules, 86.74 Albite tee - - - a) Sy is Sal. 6.09 Nephelite. - - - = - 5.40 Fem. Acmitey) = : : > 4.62 ) Na,siO, - - - - =) 0,05 Diopside - - : sii aes 2.88 elite | Olivine : : g a - 2.63 + Femic molecules, 12.45 Fayalite - - - - - 0.61 | Ilmenite - - - - - 0.76 Apatite - - : : = 0.34 | 99-19 Etc. - - - - - 0.62 99.81 As the ratio between salic and femic molecules of this norm is 6.92:1, or slightly less than 7:1, the rock falls within the Dosalane Class. Since the feldspars predominate over nephelite to an extreme degree and there is no anorthite in the norm, the rock falls in the per- felic order germanare, and the peralkalic rang umptekase. Soda strongly dominates potash, and thus the glass belongs in the subrang um ptekose, but is so close to the corresponding subrang of the Persa- lanes that the position is best shown by the name nordmarkose- um ptekose. The partial analysis of the trachytoid rock shows that to belong also in the subrang um plekose. Local interest of the discovery.—The discovery of this lava rich in alkali feldspar, so far removed petrographically from the basalts and allied rocks hitherto supposed to be the only products of volcanoes on the island of Hawaii, raises primarily a question of much local interest. The modes of occurrence, described above, suggest that there may have been quite extensive eruptions of these lavas. Unfor- tunately, these rocks were found just at the close of my visit to the island of Hawaii, and I was unable to search for further occurrences. The terrace-like bench of Puu Anahulu is certainly a remnant of a topography carved out of the trachytic rock, and it may well be that beneath the basaltic flows from Mauna Kea, Mauna Loa, and Mount Hualalai there is an ancient trachytic island. So far as | am aware, however, no fragments of trachyte have been found in the TRACHYTE ON THE ISLAND OF HAWAII 517 more recent basaltic lavas. In those of Mount Hualalai I observed only inclusions of dunite, olivine-gabbro, pyroxenite, and_ basic augite-andesites. If further exposures of the trachytoid rocks are found, it seems to me probable that they will be in the area of the Waimea plain which extends practically from Puu Anahulu for twenty miles north- easterly to the north base of Mauna Kea, or in the northern and oldest basaltic section of the island, the Kohala Mountains. _ Trachytic cones like Puu Waawaa may exist near the bases of Mauna Kea or Mount Hualalai. Their materials, if of the black glass or aphanitic rock described, might naturally be considered as basaltic, if not subjected to chemical examination. Puu Waawaa has the appearance of being a cone built up by explosive eruption of one short period, like the scores of basaltic cones which dot the slopes of Mauna Kea and Mount Hualalai, and which seem to belong to a period of local volcanism following that of lava outpourings. The materials of Puu Anahulu is agglomerate- like and is certainly not a part of a simple flow, but its crystalline texture shows also that it is not like the tuffs of Puu Waawaa. In all probability, therefore, there was in the comparatively remote past of the island of Hawaii a period of trachytic eruptions of a magnitude not now to be ascertained, but possibly of much importance. A broader significance of this discovery is connected with the history of the Hawaiian group. While no one of the larger Hawaiian islands has been thoroughly investigated as to the range in com- position of the lavas which have built it up, it is hardly probable that trachytoid rocks occur upon them. It is, therefore, highly inter- esting that the most recent island of the group, Hawaii, should exhibit this unique rock. In discussion of this question, it must be understood that the island of Hawaii is not only the largest, but is also, in the current view, the youngest, and lies at the extreme southeastern end ofa zone of oceanic islands which geologically belong together, extending for fifteen hundred miles or more to the west-northwest, toward the coast of Japan. The larger islands, commonly referred to in speak- ing of the Hawaiian group, are all located in the eastern end of this zone. ‘The smaller ones, stretching out for more than a thousand 518 WHITMAN CROSS miles, farther, are, in part at least, known to be remnants of former volcanic piles now reduced by erosion in some cases to mere reefs. So far as I have found statements concerning this long train of islets, they are basaltic, excepting the coral-reef rock. The inference drawn by many geologists who have considered the matter is that all these islands are the products of a long cycle of volcanic eruptions producing similar lavas. It appears to me plausible to assume that the earliest eruptions occurred at or near the western limit of this zone, and that, in a general way at least, the centers of activity have developed suc- cessively farther and farther to the east or southeast, until now the only active loci of eruption are those of Mauna Loa and Kilauea on the island of Hawaii. A somewhat different view was taken by J. D. Dana, who has said: There is reason for believing that the fires along the Hawaiian line broke out all together at some time in the long past, but only Hawaii has kept on piling up lava streams from that remote time of outbreak until now, and hence has come the altitude of these loftiest volcanic mountains of the Pacific.t Dana recognized, however, that Kauai and Oahu, the most north- westerly of the larger islands, exhibit great erosion. It is also a fact that both these islands are dotted by cinder cones, some of them with craters of model-like form, situated in such independent relation to the topographic forms produced by the great erosion that their later age 1s unquestionable. Whatever the point of first eruption may have been, this volcanic cycle began in a remote geologic epoch, not now determinable, but there are grounds for believing that it may well have been in the early part of the Tertiary period. The chief evidence bearing on this point consists in the destruction of the older centers by enormous and long-continued erosion, the existence of the later tuff cones men- tioned, and the fact that the raised reef of beach deposits abutting against the Diamond Head crater, on the island of Oahu, contain fossils considered by Dr. W. H. Dall to be of late Pliocene or early Pleistocene age. ‘The crater of Diamond Head, however, is but one of many small centers of tuff eruption, belonging to the comparatively recent and feeble epoch of volcanic activity. Oahu may be, more- over, one of the younger islands of the long Hawaiian chain. ¥ Characteristics of Volcanoes, pp. 357, 358. Bs TRACHYTE ON THE ISLAND OF HAWAII 519 The magnitude of the erosion by which the mountains and valleys of Kauai and Oahu have been carved out of former basaltic piles can- not now be measured. All writers upon these islands have recognized that erosion has been great. The imposing cliffs or pali which face the sea upon both islands, the canyons with walls 2,000-3,000 feet in height, the general ruggedness of mountain contours, are comparable in scale with the same features of the Rocky Mountain country. The erosion periods of the two districts must also be comparable. ‘The former volcanic centers are rudely indicated by the attitude of lava flows, but enormous sections of the old volcanoes have been engulfed by faulting or wholly destroyed by erosion. The discovery by Dall of marine fossils in a raised reef or beach rock about Diamond Head, the well-known tuff crater near Honolulu, fixes a datum point in the history of Oahu which is manifestly of great importance. Close determination of the age of the fossiliferous deposit is at present impossible because, to quote Dr. Dall, we have no standard of comparison in the whole Polynesian region by which the species could be compared with those of Tertiary beds of known age; but the fossils have every characteristic of those generally assigned to the Pliocene or upper Miocene in their general aspect, and state of fossilization. ... . To sum up, it is concluded that the reef rock of Pearl Harbor and Diamond Head lime- stones are of late Tertiary age, which may correspond to the Pliocene of west American shores, or even be somewhat earlier.* The view of Dr. Dall that “the whole mass of Diamond Head had been slowly deposited in comparatively shallow water and gradually elevated without being subjected to notable flexure” seems to the writer incorrect for various reasons, some of which have been pointed out by Dr. J. C. Branner? and Dr. S. E. Bishop. This point is not here at issue, for whatever the relations of the marine deposits to the Diamond Head tuffs, they exist and are younger than the great erosion of Oahu. Even if of early Pleistocene date, they indicate that the enormous denudation of the Oahu volcanoes must be referred to the late Tertiary and the lava eruptions themselves to a still earlier period. From the considerations above presented it appears that the ‘I Loc. cit., pp. 58, 60. 2 American ournal of Science, Vol. XVI (1903), pp- 306, 307- 3 American Geologist, January, 1901. 520 WHITMAN CROSS Hawaiian volcanic province was characterized by basaltic or allied. lavas for a very long geologic period, and that in comparatively recent times the first highly siliceous and feldspathic magma appeared. Even now the basaltic emissions continue with no known recurrence of the trachytic magma. General significance of the occurrence.—The long sequence of basaltic or allied lavas, which has probably occupied a long geologic period, is in striking contrast with the succession of widely differing lavas, ranging from highly siliceous and alkalic rocks like rhyolite or trachyte, to basic basalts, produced during the same time in various continental provinces adjacent to the Pacific Ocean. ‘To be sure, there is much greater variety in the dark Hawaiian lavas than is implied in the common term given them, but this trachytic magma is certainly highly exceptional, if indeed there has ever been another eruption of equally salic material in the history of the group. To the modern petrographer this occurrence must at once suggest many interesting problems of petrogenesis. One of these will be the query as to whether magmatic differentiation has produced this unusual lava or not. The extreme advocates of this process, con- cerning which so much has been assumed and so little proved, will no doubt treat it as a matter of course that this magma, so exceptional for the Hawaiian province, is one of the complementary products of differentiation. This view has, however, not much to support it in the history of the older centers, and until further examination of the chemical characters of the ‘“‘basalts”’ of this province have been made elaborate discussion of this point has no good foundation. It remains a matter for speculation as to why the process of differentiation has not long ago produced a greater variety of lavas, or, assuming that the differentiation has taken place, why the products have not been emitted at the more recent centers of volcanic activity. Directly connected with these questions arises also that of possible funda- mental differences between the subjacent magmas of the Hawaiian and the continental provinces. In all of these problems the relations of the Hawaiian trachyte to similar rocks of other localities will be of use. The table already presented gives a series of analyses of allied rocks, and in that follow- ing are the norms calculated from those analyses. The latter table TRACHYTE ON THE ISLAND OF HAWAII 521 brings out the fundamental chemical relations of these varied rocks and their positions in the Quantitative System of Classification. The current names assigned to these rocks are those of the authorities quoted. The nordmarkose-umptekose rock of Hawaii is shown by the table to be nearly akin in its chemical characters to other rocks of varying occurrence and texture in widely separated parts of the world. The textural differences and those due to the distinguishing roles played by quantitatively subordinate constituents have led to a variety of names, obscuring the fundamental similarity of the magmas. Considering the trachyte of Puu Waawaa as a possible differen- tiate of an underlying parent magma, it is interesting to note that the associations of the chemically similar rocks of the table are widely different. Several of the rocks compared belong in the grorudite- tinguaite series described by Broégger from the Christiania region of Norway.*' Others are from the similar petrographic province of NORMS OF ROCKS COMPARED WITH THE HAWAIIAN TRACHYTE. UMPTEKOSE 3 res| #2 | ede an SE fos ea MOLECULES oe SHS 8 5 Bi a eats) |) pecs ae 7a Catena OF NORM “es Ea *v ‘S £08 BS oe Wide tel Seba a = ae eg | ass £:0 Ons HT D=ms! ata =) ov Hes aw a) yy bo ois} EO) Stew) Fo. | s 0H - Ay iS gos mB oes Ks 2 4 Hw | ras ee OA @orunidumtarrar ce Aeien: cers eeh a tes rcie wha es Leal" eeeepel| Aare =) |) QUENI. ooo Sdee coud oo cmoasabobede sa ici Saal eee h Seno ccleags .O gp (O)arlneliSe 565-6 0 ob op doom n oan ade aanon 23 70M sie la )32) 25s OnOml 2 Onis CAST lle Jen Sy tic sha aie ete cee i ne On 3) 5 5On | 4 Os tales zeae cma CulNepielitesy ac arse am eyes ais Ssnt, aes. Ghar Wal Nena ANCA aera 5-4 AN aKOVAIONUCy e ad org Acploakoke. Cort oe ORC ROI 4.4 Ose} || ioe ae | Memite (& Naz SiO;)..........-.:...... sate QeOle ce ar. On Ome se2 ge OtlrerssilicatesMumaruistrcaenst ieee, < «ors. tas Wee SON LOR 6.5 BAG Bila) Osatles@i noah; CiCionp ae ono oa dono obOnG 4.9 4.3 Bou 1.6 Teg SRlUp A A tltenegn wena veree Neon tenecesrsaet tasers sptereceeakt Seba oe Wibitanitiewetcutancimcwisnpy een cerca oe Miotalesalicamoleculesieers a. nsis 6 scle s os 86.0 | 86.1 | 84.6 | 85.4 | 86.8 Motalstemicamoleculesmeytima- eis ecicreee ss Wit] see) Il aizkeah | EGS) |) Waly 1 Die Eruptivgesteine des Kristianiagebietes, “Die Gesteine der Grorudit-Tin- guait-Serie”’ (Kristiania, 1894). 522 WHITMAN CROSS NORMS OF ROCKS COMPARED WITH THE HAWATIAN TRACHYTE—(Con? d) NORDMARKOSE = —- B. | ae se (a a 3 ea a => = 3 vg Weyl Sp ele (hole te le lee le") ee ls Sm|Be| BS|S | a>(S/B | 8S | mele, SS(ESl|Se|s |es|ea|~g| 22) 28) ee MOLECULES B56 /S5\as Ss saa Ene a & as a os oF NorM 22%) G)38| 26 (68 | BeIT alee ls SEI SA Sl Sel Esl. 2/88 (Bel | ss Ss (RZe)e¢| oo) Se]we| s B21. a | og ao l/HS|sS|e |SS|bo)/ 8b) Cola les Se)\ss/EPls | oy | So] o8| Sol 88) ss BE geo l|eSla (ee lan | ee | 2 a2 | 2A oe ise ee) g- |ge ee ae ae | ee |e a |S |$Ale [so |e | Sei Se | sa14 : Gils Se ess) i ees : ee he Ae o Conimehin. go 455ccoc05 5008 Ti eeane loess Sale 2.9 Bol Oulartz yeewer, se Ro recon scaler ae Poe l| Claymore || Che S| Tact ss ele'ac alles cllacor0 QB Orthoclase................ 31. 7/35-0]32-8135 - 6/32 .8/32.8/33.9/27.2/31.1/28.4 4a Albite.......-.0.... seen. 55-0148. 2152.4/52.4/53.5]52-9/51-9]54.0/59.2/52.4 CalmiNe phelite Ser: ferecmtre shee ci ee Godill5 6 5'6\| Bo@|lo.c valloo6 $54 6|| Oo Gi) eo alll O) AMONtMIte ee wre sistas eta aLt lo sealeooaladadisacs|) ciadllccool| Moll BO) is 7 % Acmite (& Na, SiO3)....... Soooll iocll! OSC) OsGl) Bolecasl| Bslococllodocllace.o oF @thersilicatessa 4m Go sa) vas Oo BoB) WoO) BsO} Bs@\ wok) Bois 29 | Oxides of iron, etc.......... Gaclll Co)aC}| sill thee sean! Ah@)| itns}] orl Alay]! Oo) Fie PUA matter 2.ce. ey ssc ouaicneysie essai rds Foal |e vatgeweal age eal cea seed eu centca beame baie Sintra | aeegen oorerper ease albtanitenetcu met ree Tse Sse one S| eeeeves eke es eal cea eee | eee Motalisalicsmoleculesiry-rtei 1. cee 1.1/88.6/89.9|90.0190.8/92.9/91.2/92.8|93. 6 : 9 9 -9}9O.9190. 5/92 .9/9T. 1192.9/93-4)95 Total femic molecules...... Beads Oxeliteis@| CsA) Soudl slaail GO| Gea FoBlOoGl) Boe Essex county, Mass., where there is a great variety of rocks, dis- cussed in their genetic relations by Washington.t No one of the rocks of the table occurs in an association comparable with that of the Hawaiian trachyte. The occurrence here described seems to be a notable one, requiring consideration from the advocates of the view that a natural or genetic classification of igneous rocks may be worked out from the study of their associations. The recent tabulations of chemical analyses of igneous rocks by Washington? and Iddings,$ t “The Petrographic Province of Essex County, Mass.,” Journal of Geology, Vols. VI and VII (1898, 1899). 2 “Chemical Analyses of Igneous Rocks, 1884-1900,” Professional Paper No. 14 (1903), U. S. Geological Survey; ‘The Superior Anlayses of Igneous Rocks from Roth’s Tabellen, 1869-1884,” Professional Paper No. 28 (1904), U. S. Geological Survey. 3 “The Chemical Composition of Igneous Rocks, Expressed by Means of Dia- grams, etc.,”’ Professional Paper No. 18 (1903), U. S. Geological Survey. TRACHYTE ON THE ISLAND OF HAWAII 523 according to the Quantitative System, bring out many similar signifi- cant facts, to be taken into account in this discussion. They show the essential identity of many magmas having widely different asso- ciations, and indicate that the designation of a given rock or magma as belonging in the train (Gejfolgschajt) of certain parent magmas can at most be true only of certain peculiar or aberrent modifications, as a complete statement available for systematic classification. WHITMAN Cross. > PHYSIOGRAPHIC;: PROBEEMS OR ODay: In looking ahead and endeavoring to see in what ways our knowl- edge of the earth’s surface can be increased, the fact should be borne in mind that physiography is one of the younger of the sciences. In truth, the new geography, or physiography as it has been christened, is of such recent birth that its limits and its relationship to other sciences are as yet in part, indefinite. Accepting the conservative view, that physiographers should confine their studies to the earth’s surface, but have freedom to investigate the causes producing changes of that surface, whether coming from without or arising from forces at work within the earth, my task is to suggest ways in which man’s knowledge of his dwelling-place may be enlarged. INHERITANCES. Although the scientific study of the earth’s surface can with sufh- cient accuracy be said to be less than a century old, and to have attained the greater part of its growth during the past half-century, the fact must be freely admitted that, preceding the recognition of physiography as one of the sisterhood of sciences, there was a long period of preparation during which man’s physical environment, and the many changes to which it is subject, attracted attention and awakened interest. The more or less general and diffuse descriptions of the earth’s surface embraced under the term “physical geography,” when vivified by the idea of evolution, became the more definite and concrete physiography of today. Physiography from this point of view may perhaps be justly designated as scientific physical geography. New thoughts grafted on the previously vigorous stem have borne rich fruits, but in many instances inherit much of their flavor from the original trunk. One of the important duties of the physiographer is to select all that is of value from the inheritance that has come to him, whether of fact, or theory, or suggestion and give it a place in his systematically classified records. t Read before the section of Physiography, International Congress of Arts and Science, Universal Exposition, St. Louis, September 21, 1904. 524 PHYSIOGRAPHIC PROBLEMS OF TODAY 525 In the physical geographies on our library shelves, in books of travel, in transactions of learned societies, etc., pertaining to the era preceding the time when physical geography became a science, there are numerous records of facts, concealed perhaps in part in dreary cosmogonies and exuberant theories, which in many instances are of exceptional value because, in part, of the date at which they were observed. One of the leading ideas in scientific geographical study is the recognition of the wide-reaching principle that changes are everywhere in progress. Many, if not all, of the changes referred to have an orderly sequence, and constitute what may be suggestively termed life-histories. In writing the biographies of various features of the earth’s surface the observations made a century, or many cen- turies, ago have a peculiar, and in some instances an almost price- less, value, because of the light they furnish as to the sequence of events. In this and yet other ways the records left by past generations of geographical explorers contain valuable legacies. In attempting to winnow the wheat from the chaff of physical geography, the physi- ographer should avoid the conceit of youth, and fully recognize the work of the bold and hardy pioneers who blazed the way for the more critical and better-equipped investigators who came later. NOMENCLATURE. One of the reasons for the slow growth of knowledge concerning the earth’s surface during the centuries that have passed was the fact that the objects which claimed attention were, to a great extent, desig- _ nated by terms derived from popular usage. The language of geography, in large part of remote antiquity, was adopted from the parlance of sailors, hunters, and others in the humbler walks of life, and retained its original looseness of meaning. The change from geographical description to scientific analysis, which marked the birth of physiography, necessitated greater precision in the use of words. This change is not yet complete, and physiography is still hampered in its growth and usefulness by a lack of concrete terms in which tersely and accurately to state its results. In the nomenclature of physiography today the words inherited from physical geography by far outnumber the technical ferms since introduced, and to a large extent still retain the indefiniteness and lack of precision that charac- 526 ISRAEL C. RUSSELL erize the multiple sources from which they were adopted. One of the pressing duties of the scientific student of the earth’s surface, and one which on account of its many difficulties may well be reckoned among the physiographic problems of today, is the giving of fixed and precise meanings to the words employed in describing and classi- fying the features of the earth’s surface A scientific physiographical nomenclature is of importance, not only to the special students of the earth’s surface, but through them to communities and patrons. ‘The diverse interpretations that have been given to such seemingly simple terms as <‘shore, “lake: “river?” > hille? "mountains 2 dividers etc., as is well known, have led to misunderstandings, litigations, international disputes, and even threatened to bring on war between highly civilized nations. A duty which physiographers owe, not only to their science in order that its continued advancement may be assured, but to communities in payment for the terms borrowed from them, as well as for the general good, is a systematic effort accurately to define the words and terms now used to designate the features of the earth’s surface. Careful attention needs to be given also to the coinage of new terms when their need is definitely assured. An appropriate task for a group of physiographers would be the prepara- tion of a descriptive geographical dictionary, suited to the wants of both the specialist and the layman. While considering the advantages of a language of science, its disadvantages should also be recognized. The histories of all sciences show that, as they became more and more precise, and as their nomenclature grew so as to meet their — internal requirements more and more completely, they at the same time, on account of the very precision and accuracy of their language, became more and more circumscribed and farther and farther removed from the great mass of humanity for whose use and benefit they exist. Not only this, but a science dealing with facts of vast public importance and filled with instructive and entertaining matter—nay, in itself even poetic and as fascinating as the pages of a story-book—has, in not a few instances, been rendered difficult to understand, and even repellant to the general reader, by a bristling array of esoteric terms built about it like an abatis. Between the two extremes—on the one hand, a science without PHVSIOGRAPHIC PROBLEMS OF TODAY 527) words in which to speak concisely and accurately, the condition in which the physiographer finds himself at the present time; and, on the other hand, a science with a language so technical and abstruse that it seems a foreign tongue to the uninitiated—is there not a happy mean? Such a much-to-be-desired end seems to be within the grasp of the physiographer. By giving precision to, and defining the bounds of words inherited from physical geography, and adding to the list such terms as are strictly essential in the interest of economy of time and space, or for accuracy—such contributions, so far as practicable, to be chosen from the language of everyday life—it would seem as if a nomenclature could be formulated which would at the same time meet the requirements of the scientific student and enable the general reader of average intelligence to receive instruction and inspiration from the talks and writings of the especially qualified interpreters of nature. EXPLORATION. Physiography, to a great extent, is still in the descriptive stage of its development, but the descriptions demanded are such as discrimi- nate and select the essential, or suggestive, from the confusing wealth of secondary details frequently present. The records should also include comparisons between the objects described and analagous topographic or other physiographic features, and, within safe and reasonable limits, be accompanied by explanations of their origin and life-histories. One of the important functions of physiography, as a more mature growth of physical geography, is to continue.and render more com- plete the exploration of the earth’s surface and to conduct resurveys where necessary. Geographical exploration has, as is well known, been carried on viogrously, although spasmodically, in the past, and: the areas marked “unknown” on our globes have become smaller and smaller, and more and more isolated. The more critical physiographic studies, however, which have for their object not only the descrip- tion of coast lines, mountain ranges, plains, etc., but a search for the records of their birth, the discovery of their mode of development and their assignment to a definite place in the complex whole, termed man’s environment, has progressed but slowly. In this stricter sense, the unknown areas on the earth’s surface embrace regions of continental 528 TSRAEENG. RUSSHLE extent. It is this latter method of geographical exploration and survey which now demands chief attention. The terms “exploration” and “‘survey” are here used advisedly, as two divisions of physiographic field-work may justly be recognized. These are: first, travel in which physiographic observations are inci- dental to other aims, or perhaps the leading purpose in view, as during a physiographic reconnaissance; and, second, detailed surveys and critical study of definite areas or of concrete problems. Each of these subdivisions of the great task of making known the beauties and harmonies of man’s dwelling-place has its special functions. From the observant traveler we expect comprehensive and graphic descriptions of the regions visited, rendered terse by the use of well- chosen terms, in which the more conspicuous elements of relief, and other physiographic features, and their relation to life, shall be clearly and forcibly presented. In order to render this service, the traveler should not only be familiar with the broader conclusions and fundamental principles of physiography, but skilled in the use of its nomenclature. The chief contribution to the science of the earth’s surface demanded of the explorer of new lands is a careful record of facts. When a journey becomes a reconnaissance with physiography as its leading feature, it is not only an advance into a more or less unknown region, but an excursion into the realm of ideas as well. It is during such explorations, when one’s mind is stimulated by new impressions, that hypotheses spring into existence with greatest exuberance. While most of these springtime growths are doomed to wither in the more intense heat of subsequent discussion, their spontaneity, and the fact that the mind when not oppressed by a multitude of details grasps significant facts almost by intuition, make the suggestions of the explorer of peculiar value. The detailed work of physiographic surveys falls into two groups: namely, the study of definite areas, and the investigation of specific problems. In each of these related methods the desirability of record- ing facts by graphic methods is apparent. ‘The demand for accurate maps as an aid to both areal physiography and the study of groups of specific forms, or the functions of concrete processes, needs no more than a word at this time. With the growth of physiography the time has come when the work of the individual explorer, who from force of PHYVSIOGRAPHIC PROBLEMS OF TODAY 529 circumstances endeavors to follow many of the paths he finds leading into the unknown, is replaced to a large extent by well-organized and well-equipped scientific expeditions. It is to such systematically planned campaigns, in which the physiographer and representatives of other sciences mutually aid each other, that the greatest additions to man’s knowledge of the earth’s surface are to be expected. ‘The most extensive of the unexplored or but little-known portions of the surface of the lithosphere, in which a rich harvest awaits the properly equipped expedition, are the sea-floor and the north and south polar regions. As is well known, splendid advances have been made in each of these fields, but, as seems evident, much more remains to be accomplished. In the branch of physiography appropriately termed “oceanog- raphy” the problems in view are the contour of the sea-floor, or its mountains and deeps, plains and plateaus, the manner in which each inequality of surface came into existence, and the various ways it is being modified. In both of these directions the interests of the physiographer merge with those of the biologist and the geologist. One phase of the study of the ocean’s floor which demands recognition is that the topographic forms there present are such as have been produced almost entirely by constructional and diastrophic agencies, free from complications due to erosion which so frequently obscure the result of like agencies on the land. For an answer to the question: What would have been the topography of land areas, had there been no subaérial decay and denudation? the topography of submarine regions furnishes at least a partial answer. The sounding line in the Caribbean region has furnished examples of topography due, as it seems, mainly to differential movements of blocks of the earth’s crust bounded by faults, which have not been modified by subaérial denuda- tion. Ina similar way, as is to be expected, a survey of other portions of the at present water-covered surface of the lithosphere will supple- ment our knowledge of the emerged portions of the same rock-envelope, and assist in an important way in the deciphering of their histories. In the Arctic and Antarctic regions, where all is unknown, systematic research may be expected not only to extend many branches of physio- graphic study, but to bring to the front as yet unsuspected problems. The larger of the unexplored regions of the earth, however, are not 30 ISRAEL C. RUSSELL On the only portions of our field of study that demand attention. New ideas, new principles, and fresh hypotheses make an unknown country of the most familiar landscape. ‘The definite formulation of the base- level idea, the suggestive and far-reaching principle embraced in the term “geographic cycle,” the planetesimal hypothesis as to the origin of the earth, etc., furnish new and commanding points of view, or, as they may be termed, primary stations in the physiographic survey of the earth’s surface by means of which previous local surveys may be correlated and corrected. In the search for problems, the unraveling of which may be expected to advance the scientific study of the earth’s surface, the writings of travelers, the pregnant suggestions of those who make reconnaissances into the realm of unknown facts and of unrecognized ideas, as well as the precise and accurate pictures of portions of the earth’s surface presented on the maps of the topographer and the charts of the ocean- ographer, point the way to still greater advancements, and furnish inspiration to those who follow. FUNDAMENTAL CONCEPTS. While physiography deals with the surface features of the earth, the fact that in those features is expressed to a great extent the effects of movements originating deep within the earth, leads the student of continents and oceans to ask of the geologist and the physicist puzzling questions as to the changes that are taking place in the great central mass of our planet, and even in reference to the origin of the earth itself. So intimately are the various threads of nature-study inter- woven that the full significance of many of the surface features of the earth cannot be grasped and their genesis explained until the nature and mode of action of the forces within the earth which produce surface changes are understood. The growth of physiography up to the present time has been largely influenced by the far-reaching ideas of Laplace and others in reference to the nebular origin of the solar system. In all of the questions respecting secular changes of land areas in reference to the surface level of the ocean, the origin of corrugated and of block mountains, the fundamental nature of volcanoes, etc., the controlling idea has been that the earth has cooled from a state of fusion, and is PHYSIOGRAPHIC PROBLEMS OF TODAY 531 still shrinking on account of the dissipation into space of its internal heat. With the recent presentation of the planetesimal hypothesis by Professor Chamberlin, a radically different point of view is furnished from which to study the internal condition of the earth. The new hypothesis—which has for its main thesis the building of a planet by the gathering together of cold, rigid, meteoric bodies, and the compres- sion and consequent heating of the growing globe by reason of gravita- tional contraction—is suggestive, and seems so well grounded on facts and demonstrated physical and chemical laws that it bids fair not only to revolutionize geology, but to necessitate profound changes in methods of study respecting the larger features of the earth’s surface. One of the several considerations which make the planetesimal hypothesis appeal forcibly to the inquiring mind is that it employs an agency now in operation, namely, the process of earth-growth through the incoming of meteoric bodies from space; and for this reason is welcomed by uniformitarians, since it is in harmony with their under- standing of a fundamental law of nature. In many, if not all, questions respecting the origin of the atmosphere, the ocean, continents, mountains and volcanoes, and the secular, and to a marked extent in certain instances, the daily changes they experience, it is evident that the planetesimal hypothesis necessitates a revision, or at least a review, of some of the fundamental conceptions held by physiographers. The objection will perhaps be advanced that to make such a radical change of front on the basis of a young and as yet untried hypothesis is not wise. ‘The reply is that the older hypothesis has been tried and to a marked extent found wanting, and that the new conception of the mode of origin of the earth demands considera- tion, not only as affecting a large group of basement principles of interest to the physiographer, but with the view of testing the planetesi- mal hypothesis itself by physiographic standards. The problems interlocked with the mode of origin of the earth, in which the physiographer shares an interest with the geologist, are the rate at which the earth’s mass is now being increased owing to the ingathering of planetesimal, and the chemical and physical and per- haps life conditions of the incoming bodies; the temperature of the earth’s interior, and the surface changes to be expected from its 532 ISRAEL C. RUSSELL increase or diminution; the results of gravitational contraction in reference to movement in the earth’s crust; the extrusion of gases and vapors from the earth’s interior, and the resultant changes in progress in the volume and composition of the atmosphere and hydrosphere. In these and still other fundamental conceptions of the primary causes of many of the changes in progress on the earth’s surface the plan- etesimal hypothesis seemingly furnishes the cornerstone of a broader physiography than has as yet been framed. IDEAL PHYSIOGRAPHIC TYPES. During the descriptive stage of the study of biology the relation- ships among plants and animals were the chief end in view, and as a result of the conditions confronted, a systematic classification of animate forms under species, genera, families, etc., was formulated, which has been of infinite assistance during the more philosophical investigations that followed. Biological classification was facilitated, as learned later, by the fact that with the evolution of species there was concurrent extinction of species. As the tree of life grew, its branches became more and more widely separated. Throughout the many changes the surface features of the earth have experienced there has also been development, not of the same grade, but analagous to that recognized in the realm of life; but the process of extinction has been far less complete than in the organic kingdom, and the connecting links between the various groups of topographic and other physiographic forms produced have persisted, and to a conspicuous extent still exist. The task of arranging the infinitely varied features of the earth’s surface in orderly sequence, or systematic physiography, is thus far more difficult than the similar task which the flora and fauna of the earth present. The utility of classification is fully recognized by physiogra- phers, and various attempts have been made from time to tine to meet the demand, but thus far without producing a generally accept- able result. Remembering that a scheme of classification of topo- graphic and related forms is to be considered as a means for attaining a higher end, namely, the history of the evolution of the surface features of the earth, and should be of the nature of a table of contents to a systematically written treatise, the task of preparing such an index PHYSIOGRAPHIC PROBLEMS OF TODAY 533 becomes of fundamental importance to the physiographer. Since extinction of species among physiographic features has probably not occurred, and connected series of forms which grade one into another confront us, the practical lesson taught by the success of schemes of biological classification seems to be that ideal physiographic types should be chosen correlative with species among plants and animals. By “ideal physiographic types” is meant complete synthetic examples of topographic and other physiographic forms, which will serve the réle of well-defined species in the study of the surface features of the earth. Ideal types may be likened to composite photographs. They should combine critical studies of many actual forms, within a chosen range, and in addition be ideally perfect representatives of the results reached by specific agencies operating under the most favorable conditions. Like the idealized personalities of history and religions, the types of physiographic forms might well be more perfect than any actual example. When such idealized types shall have been chosen after careful study, described with care, and illustrated by means of diagrams, maps, pictures, models, etc., a comparison with them of actual examples on any portion of the earth for the purpose of identification and classification would be practicable. A well- arranged catalogue of ideal types would be an analytical table of contents to the history of the evolution of the features of the earth’s surface, and constitute a scheme of physiographic classification. In illustration of what is meant by an idealized physiographic type: We find in nature a great variety of alluvial deposits, now designated as alluvial cones or alluvial fans. ‘They present a wide range and infinite gradations in size, shape, composition, structure, angle of slope, degree of completeness, stage of growth or decadence, etc. Complications also arise because of the association and inter- growth of such alluvial deposits with other topographic forms. In constructing the ideal type the characteristics of many of the most perfect actual alluvial cones, aided by a study of the essential features of similar artificial structures, should be combined in an ideally perfect and representative example which would serve as the type of its specie. All actual examples might be compared with such a type, their specific and generic relations determined, and their individual variations noted. In like manner, other topographic forms, ranging 534 ISRAEL C. RUSSELL from the more concrete species—such as constructional plains, cinder cones, sea cliffs, river terraces, etc.—to the more complex forms—as, for example, mountain ranges, mountain systems, and yet larger earth- features—could be represented by ideally perfect examples free from accidental and secondary complexities and accessories. While individual examples of idealized topographic and other features of the earth’s surface would serve as species, their arrange- ment under genera, families, etc., offers another problem, in which relationship or genesis should be the controlling idea. The selection of idealized physiographic types, as just suggested, has for its chief purpose the reduction of endless complexities and intergradations to practicable limits. It is a method of artificial selection so governed that, while no link in the chain of evolution need be lost to view, certain links are chosen to represent their nearest of kin and serve as types. A danger to be marked by a conspicuous signal, in case this plan for aiding physiographic study is put in prac- tice, is that it may tend toward empty ritualism. To give the idealized types chosen for convenience of classification an appropriate atmos- phere, the fact that changes are constantly in progress—that moun- tains come and go even as the clouds of the air form and reform— should be ever present in the mind. The process of evolution without concurrent extinction which characterizes the development of physiographic features finds expres- sion also in related departments of nature, as, for example, in petrog- raphy, where, as is well known, it has greatly delayed the framing of a serviceable and logical system of classification. Indeed, the prin- ciple referred to may be said to be one of the chief distinctions between the organic and the inorganic kingdoms of nature. THE PRIMARY FEATURES OF THE EARTH’S SURFACE. The primary features of the earth’s surface may consistently be defined as those resulting from the growth and internal changes of the lithosphere, while the modifications of relief resulting from the action of agencies which derive their energy from without the earth may be termed secondary features. ‘The primary or major character- istics of the earth’s surface, so far as now known, may be ranked in three groups, in accord with the agency by which they were principally PHYSIOGRAPHIC PROBLEMS OF TODAY 535 produced; namely, diastrophic, plutonic, and volcanic physiographic features. Each of the groups presents many as yet unsolved problems. Diastrophic jeatures—Under this perhaps unwelcome term are included a large class of elevations, depressions, corrugations, faults, etc., in the surface portion of the lithosphere due to movements within its mass. ‘The causes of the changes which produced these results are as yet obscure, and, although a fruitful source of more or less romantic hypotheses, may in general terms be referred to the effects of the cooling and consequent shrinking of a heated globe, or, under the terms of the planetesimal hypothesis, reckoned in part among the results of gravitational condensation. However obscure the fundamental cause, the results in view are real, and among the larger of the earth’s features with which the physiographer deals. They are the greater of the quarry blocks, so to speak, which have been wrought by denuding agencies into an infinite variety of sculptured forms. Included in the list, as the evidence in hand seems to indicate, are continental platforms, oceanic basins, corrugated and block - Mountains, and many less mighty elements in the marvelously varied surface of the lithosphere. Not only the study of the shapes of these features of the earth’s surface, but the movements they are still experi- encing, and their transformations through the action of denuding agencies, claim the attention of the physiographer. While it may be said that the investigation of the method by which the primary relief of the lithosphere have been produced, falls to the lot of the geologist or the geophysicist, the physiographer is also interested in the many profound problems involved. The geologist and physiographer here find a common field for exploration, and can mutually assist each other. The task of the physiographer is to describe and classify the elements in the relief of the lithosphere due to diastrophic agencies, discriminate them from deformations due to other causes, and restore so far as practicable the forms that have been defaced by erosion. He can in this way assist the geologist by presenting him with the results of diastrophism free from accessories. With pure examples of the forms produced, the geologist will be better able to discover the causes and their mode of action, which have produced the observed results. Although much has been accomplished in the way of determining which elements in the relief of the lithosphere are due to diastrophic 536 ISRAEL C. RUSSELL agencies, only a small part of the difficulties to be overcome have been met. ‘The aim in view is the attaining of a knowledge of what would have been the shape and surface features of the solid earth, had there been no modifications by internal causes except diastrophism, and no changes in relief by erosion or other surface agencies. Included in this branch of physiography is the shape of the earth itself, in the study of which the physiographer became .a geodesist. The earth’s shape, and its primary surface features due to diastrophism, form the logical, basis for physiographic study, in which ideal types of topo- graphic forms declare their usefulness. In the geographical museums of the future, at the head of the long series of models of physiographic types illustrating the species, genera, families, etc., of the earth’s surface features, should be placed ideal examples of the most typical elements of relief due to diastrophism. Physiographers cannot rest content with the study of the shape of the lithosphere and of its surface relief, in which so much of the history of the earth is recorded, and refrain from searching for the deeper meanings these facts suggest, but must have freedom to invade the province of the geologist, the astronomer, the physicist, the chemist, and other subdivisions of the science of the cosmos, in search of truths bearing on his special line of work. This is particularly true in connection with the special department of physiography in hand, in which many of the branches of the river of knowledge have their sources. Plutonic jeatures.—Intimately associated with the irregularities of the earth’s surface due to a decrease in its volume, and, as our reasoning tells us, dependent primarily on the same cause and at present only partially differentiated from them, are surface elevations and depressions, produced by the migration of portions of the earth’s central magma from the deep interior toward or to the surface. A convenient but arbitrary subdivision of the matter forced outward from the earth’s interior is in vogue among geologists, and rocks of plutonic and of volcanic origin are recognized. ‘To the physiographer the distinction referred to is more suggestive than it appears from the point of view of the geologist, since the recognition of differences between topographic forms produced by the injection of fluid or plastic magmas into the cooled, rigid outer portion of the earth, and PHYVSIOGRAPHIC PROBLEMS OF TODAY Bey: topographic forms resulting from the extrusion of similar matter at the surface, is of genetic significance. The simile was used above between the quarry blocks taken to the studio of the sculptor and the portions of the earth’s surface brought by diastrophic movements within the sphere of influence of denuding agencies. There are two other primary classes of physiographic quarry blocks; one produced by intrusions of highly heated plastic or fluid magmas into the earth’s crust, which cause upheavals of the surface above them, and the other due to extrusions of similar material at the surface, as during volcanic eruptions. The first of these two series of earth-features has received much less attention from physi- ographers than the second series. Surface elevations due to local intrusions are well illustrated by the reconstructed forms of the Henry Mountains and the similar information in hand concerning several other regions. The topo- graphic forms referred to have a conspicuous vertical measure in comparison with their breadth of base, and their prominence gained for them earlier recognition than in the case of related, and in part far more important, plutonic changes. It is to be remembered, however, that every intrusion of a magma into the earth’s crust is, theoretically at least, accompanied by a change in the relief of the surface above. What surface changes accompany the lateral movements in the rocks invaded by a dike has eluded search and seemingly escaped conjecture. The surface changes produced by an extensive horizontal injection of a magma, as when intruded sheets are found in stratified terranes, is a matter of inference rather than of observation. Intrusive sheets are numerous, and the surface changes in topography, and consequently of drainage, that accompanied their production must have been important, but definite examples are wanting. Critical studies are needed in this connection, both by physiographers and by geologists, in order that the widely extended movements which have been observed in the surface of the lithosphere may be referred to their proper cause. How do we know, for example, that the many recorded changes in the relation of the land to sea-level may not in part be due to the injection of magmas into the earth’s crust, instead of diastrophic movements as commonly supposed. The activity of volcanoes at the present day is warrant for the hypothesis that the concurrent process of sub- 538 ISRAEL C. RUSSELL surface injection is still in progress, and is today producing changes _ in the geography of the earth’s surface. Of still more importance to the physiographer than the surface changes known, or legitimately inferred, to have resulted from the formation of dikes, laccoliths, and intruded sheets are the elevations and possibly concurrent depressions of the surface of the lithosphere caused by still greater migrations of portions of the earth’s central magma outward and into or beneath the rigid surface rind. Con- cerning these regional intrusions, as they may be termed, the geologist has furnished suggestive information. We are told, for example, that the granitic rocks from which the visible portion of the Bitter Root Mountains in Idaho have been sculptured are intrusive. The now deeply dissected granitic core of this mountain range measures not less than three hundred miles in length and from fifty to over one hundred miles in width. ‘The area occupied by intrusive granitic rocks in the Sierra Nevada is seemingly still greater than in the case just cited, and other regional intrusions of even mightier dimensions are known in the vast region of crystalline rocks in Canada and else- where. ‘The covers of sedimentary or other material which formerly roofed these vast intrusions in the instances now open for study have for the most part been removed by denudation, although instructive remnants of metamorphosed terranes occurring as inliers in the granitic areas sometimes persist and reveal something of the nature of original domes of which they formed a part. The surface changes in relief produced by the migration of magmas measuring thousands, and in many instances, as we seem justified in concluding, tens of thousands, of cubic miles, from deep within the earth outward, but failing to reach the surface, must be reckoned as of major physiographic importance. The very magnitude of the features of the earth’s surface due to such intrusions has served to conceal their significance. We look in vain in our treatises on physiog- raphy for so much as a mention of them. Perhaps the excuse will be offered that the modifications in relief referred to are commonly grouped with the results produced by diastrophic agencies; but, if so considered, a differentiation seems necessary, and the significance of the topographic forms resulting from intrusions of various kinds clearly recognized. PHYSIOGRAPHIC PROBLEMS OF TODAY 539 In our dreamed-of museum of ideal physiographic types, mighty domes raised by regional intrusions, broad uplifts with perhaps sharply defined boundaries, elevated by relatively thin intruded sheets, as well as steep-sided domes with relatively small bases, con- cealing laccoliths, and the still smaller covers arching over plutonic plugs, will demand a place in the group of type examples of primary unsculptured elements in the relief of the lithosphere. Volcanic features Elevations on the surface of the lithosphere due to the presence of material extruded from volcanic vents have long been recognized, but the specific, or, as perhaps may be con- sistently claimed, generic, differences among them has only recently claimed attention. Of primary importance in the classification of topographic forms of volcanic origin is the fact that volcanoes are both constructive and destructive in their action. Among the results of constructive action are included the changes produced by effusive, fragmental-solid, and massive-solid eruptions, each of which has furnished a wide range of primary topographic forms. The catalogue of recognized types includes lava plains and plateaus, cinder and lapilli cones, lava cones and domes, lapilli and dust plains, together with many minor structures, such as “spatter cones,” “lava deltas,” “lava gutters,” “lava levees,” and the various surface details of lava streams due to the flow of still mobile magmas beneath a stiffened crust which ranged in physical consistency from a highly plastic to a rigid and brittle condition. With these more familiar forms are to be included also the results of massive-solid extrusions, of which the ‘‘obelisks” of Mont Pelé are the most striking examples. Our present list of destructional topographic forms due to volcanic eruptions includes decapitated cinder, lapilli and lava cones, and subsided and broken lava domes, calderas, crater rings, etc., together with cones of various kinds breached by outflowing lavas; and, as minor features, the floated blocks sometimes carried on lava streams, or the moraines oj lava flows, as they may suggestively be termed, the subsided and broken roofs of lava tunnels, etc. The interesting contributions made during the past decade to the list of topographic forms resulting from the action of volcanic agencies are highly suggestive, and warrant the belief that still more numerous and equally important results in the same direction will reward more 540 ISRAHE (Gy RUSSEEL extended and more careful search. The progress of physiography would evidently be accelerated by a systematic review and a more definite classification of the topographic forms, both constructional and destructional, known to have resulted from volcanic agencies, and a more critical selection of types to serve as species than has as yet been attempted. From such a catalogue something of the under- - lying principles governing the many ways in which the relief of the earth’s surface has been modified, and is still being changed through .the agency of volcanoes, -would make themselves manifest, and pre- dictions rendered possible which would facilitate further study. The analogy between lava streams and rivers, on the one hand, and glaciers, on the other, suggests interesting and instructive methods for con- sidering the entire question of the movements of liquids and solids on the earth’s surface. While the topographic changes produced by volcanic agencies are of chief interest to the physiographer, they lead him to profound speculations in reference to the nature of the forces to which they are due, the source and previous condition of the matter extruded during eruptions, and the study of the existing relations between the earth’s interior and its surface. ‘The great, and as yet but partially answered, questions: Whence the heat manifest during volcanic eruptions ? What is the source of the energy which forces lava to rise from deep within the earth through volcanic conduits to where it is added to the surface, perhaps ten to twenty thousand feet above sea-level? and, what is the source or sources of the steam discharged in such vast quantities during eruptions of even minor intensity ? are as of great interest to the physiographer as they are to the geologist, and furnish another illustration of the unity of nature-study. From the new point of view furnished by the author of the planetesimal hypothesis, the many questions the physiographer is asking concerning volcanoes and fissure eruptions are rendered still more numerous by the sugges- tion that these fiery fountains are the sources from which the ocean and all the surface waters of the earth have been supplied. This startling revelation, as it seems, makes a still more urgent demand than had previously been felt for quantitative measures of the vapor discharged from volcanic vents. Nor is this all; with the steam of volcanoes is mingled various gases, and the mode of origin of the PHYVSIOGRAPHIC PROBLEMS OF TODAY 541 earth’s atmosphere as well as the changes it is now undergoing, is a theme in which the physiographer is profoundly interested. Volcanic mountains are numbered among the most awe-inspiring - of topographic forms; the solid additions which volcanoes make to the surface of the lithosphere are in view, and the contributions to the atmosphere of vapors and gases from the same sources are tangi- ble facts; but another phase of the great problem is also of interest to the physiographer, namely, what changes take place in the rigid outer shell of the earth by reason of the transfer of such vast volumes of material as are known to have occurred from deep within the earth to its surface. The magmas which have been caused to migrate and come to rest for a time, either as intrusions within the earth’s outer shell, or as extrusions on its surface, are measurable in millions of cubic miles. In connection with the profound questions concerning the formation of folds and fractures in the earth’s crust, an agency is thus suggested comparable in importance with loss of heat, as under the nebular hypothesis, or with gravitational compression, as explained by the planetesimal hypothesis. In the many discussions that have appeared as to the adequacy of earth contraction to account for the origin of mountains of the Appalachian type, I have been able to find but one mention of the réle played by the transfer of matter from deep within the earth outward, and in part its extrusion at the surface, in causing folds in the crust from beneath which it was derived. Problems of fundamental importance are outlined by the considerations under review. To the immediate question, What is the best plan for enlarging our knowledge of the physiography of volcanoes? the reply seems pertinent: Press on with the study of both active and dormant or extinct examples. In this connection it should be remembered that, while the individual volcanoes and volcanic mountains which have been critically studied can be enumerated on the fingers of one’s hands, those which are practically unknown number many thousands. The fact that Mont Pelé and La Soufriére of St. Vincent during their recent periods of activity furnished examples of at least two important phases of volcanic eruptions not previously recognized is an assur- ance of rich returns when other eruptions are critically investigated. While it is difficult to formulate the precise questions so numerous 542 ISRAEL GG. RUSSEEL are they, to be asked of volcanoes, whether active, dormant, or dead, and in various stages of decay and dissolution, it is plain that all .the facts that can be learned concerning them should be classified and put on record, and their more obvious bearings on the fundamental questions concerning the condition of the earth’s interior, and the changes there taking place, pointed out. In this connection—and as is true in all branches of research—the fact may be recalled that energy expended in discovering, classifying, and recording facts decreases the time and force necessary for the framing of multiple hypotheses. With an abundance of well-classified and pertinent observations in hand, the nature of the thread on which the gems of truth should be strung usually declares itself. Résumé.—On a previous page of this essay the desirability was suggested of recognizing ideal types with the aid of which the multi- tudinous surface features of the earth could be classified and studied. Thus far we have considered the elements in the relief of the earth’s surface which have resulted from changes within its mass. We term them primary physiographic features, because their birth precedes the modifications of the lithosphere due to agencies acting externally. They are (1) the topographic forms resulting from contraction on account of cooling, or of condensation owing to growth in mass; (2) the surface changes produced by intrusions of magmas into the earth’s outer shell; and (3) the results of volcanic eruption. Among ‘the more important idealized models in our future physiographic museum there should be displayed continental platforms, oceanic basins, corrugated mountains, block mountains, domes of various and some of vast dimensions upraised by intrusions, volcanic cones, lava plateaus, etc. These are the major physiographic types, or the larger monoliths from which the rock-hewn temples of the earth have been sculptured by forces acting on the surface of the lithosphere and deriving their energy mainly from the sun. Resulting from surface changes come a vast array of both constructive and destructive physiographic features, which may consistently be termed secondary. Under secondary features may be included also relational topographic forms, such as islands in water, in glaciers, and in lava fields. In the study of the primary features of the earth’s surface the work of the physiographer is most intimately linked with that of the geolo- PHYSIOGRAPHIC PROBLEMS OF TODAY 543 gist, but, on passing to the secondary feature, the influence of life becomes apparent, and the relation of man to nature is in the end the leading theme. SECONDARY PHYSIOGRAPHIC FEATURES. The most familiar features of land areas, as is well known, are those which owe their existence to the work of moving agencies resident on the earth’s surface, such as the wind, streams, glaciers, waves, currents, etc. The forces at work are set in motion by energy derived from without the earth, and the material worked upon is brought within the range of their activities by forces resident within the earth which cause deformations of, or additions to, its surface. The earth-born primary physiographic features are thus modified by sun-derived forces, and a vast array of secondary modifications of relief are produced which give variety and beauty, particularly to those portions of the lithosphere which are exposed for a time to the air. The study of secondary physiographic features has produced a rich and abundant harvest, especially during the last quarter of a century, and the returns are still coming in with seemingly an accel- erated rate. ; The themes for study are here mainly the various processes of erosion and deposition of the material forming the outer film of the lithosphere, and the characteristics of the destructive and constructive topographic forms produced. With the knowledge gained concerning the changes now in progress on the ocean’s shore, in the forest, by the river side, on the snow-clad and glacier-covered mountains, etc., the physiographer seeks to decipher the records made in similar situa- tions during the past. ‘Two groups of problems are in sight in this connection; one is concerned with observing, classifying, and recording the changes now in progress; and the other has for its chief aim the translation, in terms of the agencies now at work, of the records left by past changes. We find that today the same area is being inscribed perhaps in several different ways. The surface of the earth, like an ancient manuscript, is frequently written upon in different directions, and with different characters. It is the duty of the physiographer to translate this ancient palimpsest, and deduce from it the history of the development of the features of the earth’s surface. It has 544 TSRAEE Cy RUSSELL been said that ‘geology is the geography of the past,” but to the physiographer this formula has a yet deeper meaning. There is a physiography of the past, of venerable antiquity, which has begun to receive attention. Ancient land surfaces, buried during geological eras beneath terranes which were deposited upon them, have here and there been exposed once more to the light of the sun, owing to the removal by erosion of their protecting coverings. In northern Michi- gan, for example, one may gaze on the veritable hills and valleys which were fashioned by the wind, rain, and streams of pre-Potsdam days of sunshine and shower. These fossil landscapes invite special study, not only on account of their poetic suggestiveness, but as fur- nishing evidence, supplementary to that afforded by organic records, ripple marks, shrinkage cracks, etc., as to the oneness of nature’s processes throughout eons of time. The consideration of past physi- ographic conditions, the tracing of former geographic cycles, the study of the concurrent development of primary and secondary physi- ographic features, the causes and effects of past climatic changes, and the influences of these and still other events of former ages on the present expression of the face of nature, offer not only a fascinat- ing, but a far extended field for research. One especially important development of the study of past physio- graphic conditions, and the manner in which they merge with the present phase of the same history, is the connection between the life of the earth and its control by physical environment. ‘The present and past distribution of floras and faunas affords important data supplementary to those recorded by abandoned stream channels, glacier scorings, elevated and depressed shore lines, desiccated lake basins, and other physical evidences of former geographic changes. In the excursions into the domain of the unknown, here suggested, the physiographer seeks the companionship and counsel of both the geologist and the biologist. PHYSIOGRAPHY AND LIFE. In the study of the relation between physiography and the present state of development of living organisms on the earth, it is convenient and logical to recognize two great subdivisions: the one, the control exerted by physiographic features on the distribution of plants and PHVSTOGRAPHIC PROBLEMS. OF TODAY 545 animals; and the other, the reaction of life on its physical environ- ment, and the modification in the relief of the lithosphere and the geography of its surface thus produced. Although man is embraced in each of these categories, there are sufficient reasons for consider- ing his relations to his environment separately from those of the lower forms of life. The dependence of life on its physical environment has received much attention from botanists and zodélogists, and is perhaps the leading thesis now claiming their attention. So important is this branch of study that a name “ecology,” has been coined by which to designate it. The phase of nature-study thus made prominent pertains to the marvelously delicate adjustment that has been found to exist between the distribution of life and the nature of the region it inhabits. Among the interesting themes involved are topographic relief, degree of comminution and disintegration of the surface blanket of rock waste, depth and freedom of penetration of water and air into the life-sustaining film of the earth’s surface, and the concurrent changes in life with variations in these and other physical conditions. In this most fascinating branch of study the ecologist borrows freely of the physiographer, and makes payment in peat bogs, living vegetable dams in streams, organic acids service- able for rock disintegration and decay, deposits of calcium carbonate and silica in lakes and about springs, vast incipient coal beds in the tundras of the far north, and numerous other ways. From the physiographic point of view, however, the many and intricate ways in which life leads to modifications in the features of the lithosphere are of more direct interest than studies in ecology. Much has been accomplished in this direction, but it is evident that as yet but partially explored paths leading through the borderland between biology and physiography remain to be critically examined. In connection with the changes in progress on the earth’s surface, due to the influence of organic agencies, and the application of that knowledge in interpreting past changes, the study of the influences exerted by the lowest forms of life in both the botanical and the zoological scale seems most promising to the physiographer. The secretion of calcium carbonate and silica by one-celled organisms, as is well known, has led to the accumulation of vast 540 ISRAEL C. RUSSELL deposits like the oozes on the sea-floor, beds of diatomaceous earth, deposits about hot springs, the so-called marl of fresh-water lakes, etc. A review of the several ways in which such accumulations are formed, and an extension of the search in various directions, give promise that other and equally wonderful results flowing from the activities of the lowest form of life would be discovered. ‘The mode of deposition of iron, and perhaps of manganese, the generation of hydrocarbons, the origin of extensive sheets of seemingly non-fossil- iferous limestone and dolomite, the method by which the beautiful onyx marbles are laid down, film on film, the nature of the chert so abundant in many terranes and so conspicuous in the surface waste of extensive regions, and other equally important deposits which exert a profound and frequently controlling influence on topographic forms, seemingly demand study with the hypothesis in mind that they owe their origin to the vital action of low forms of plants or animals. Not only the concentration of mineral matter by one-celled organisms, but the part played by similar organisms in the comprehensive pro- cesses of denudation, also invites renewed attention. Many of the organisms in question do not secrete hard parts, and hence are incapable of directly aiding in the concentration of inorganic solids on the surface of the lithosphere. If not assisting in the building of physiographic structures, the suspicion is warrantable that they are engaged in sapping their foundations. The wide distribution of one- celled organisms—and indeed, as one may say, their omnipresence on the earth’s surface—and their seeming independence, as a class, to differences in temperature, light, and humidity, enable them to exert an unseen and silent influence, not suspected until some cumu- lative and conspicuous result is reached. ‘The importance of bac- teria in promoting decay, and in consequence the formation of acids which take a leading part in the solution and redeposition of mineral substances, the réle played by certain legions of the invisible hosts in secreting nitrogen from the air and thus aiding vegetable growth and perhaps to be held accountable also for the concentration of nitrates in cavern earths, the part others play in fermentation, and the diseases produced in plants and animals by both bacteria and protozoa, render it evident that an energy of primary importance to the physiographer is furnished by these the lowest of living forms. PHYVSIOGRAPHIC PROBLEMS OF TODAY 547 Physiographers were given a new point of view when Darwin explained the part played by the humble earthworms in modifying the earth’s surface. As it seems, still other advances in our knowledge of the changes in progress in the vast laboratory in which we live may be gained by studying the ways in which organisms far lower in the scale than the earthworm are supplying material for the building of mountains or assisting in the leveling of plains. In brief, a review of the inter-relations of physiography and life, shows that from the lofty snow-fields reddened by Protoccocus, to the bottom of the ocean, the surface of the lithosphere is nearly everywhere enveloped in a film teeming with life. In part the vital forces at work are reconcentrating material and adding to the solid framework of the globe, and in part, but less obviously, aiding in rock decay and disintegration. Throughout this vast complex cycle of changes new physiographic features are appearing, others disappearing, and one and all, to a greater or less degree, are undergoing modifications. The wide extent of the changes in progress, and their known impor- tance in certain instances, are justification for the belief that the physiographer as well as the ecologist will find many problems of fundamental importance to his science in the inter-relations of life and physiographic conditions. PHYSIOGRAPHY AND MAN. Go forth, subdue and replenish the earth, is the language of Scripture. The observed results show that, while man strives to bend nature to his will, he himself is a plastic organism that is molded by the many and complex external forces with which it comes in con- tact. Here again two groups of themes present themselves to the physiographer: one, embracing the influences of environment on man; - and the other, the changes in the features of the earth’s surface, brought about by human agencies. In the first the physiographer can aid the anthropologist, the historian, the socialist, etc.; and in the second, which is more definitely a part of his own specialty, he searches for suggestive facts throughout the entire domain of human activities. It is in these two directions that the student of the earth’s surface finds the most difficult and the most instructive of the prob- lems in which he takes delight, and the richest rewards for his efforts. 548 TSRABEN GA ROS SHEL, The control exerted by physiographical environment on human development is so subtle, so concealed beneath seemingly accidental circumstances, and its importance so obscured by psychological con- ditions, that its recognition has been of slow growth. ‘The countless adjustments of both the individual man and of groups of men in com- munities, nations, and races to physical conditions, is so familiar that the sequence of causes leading to observed results passes as a matter of course and to a great extent fails to excite comment. The due recognition of the influence of physiographic environment on history is now coming to the front, and, as is evident, the rewriting of history, and especially the history of industry, from the point of view of the physiographer, is one of the great tasks of the future. The problems in this broad field are countless, and the end in view is similar to those embraced in dynamical physiography, namely, the study of the various ways in which man is now influenced by his physical environment, with the view of interpreting the records of similar changes in the past and of predicting future results. Or more definitely formulated: peoples have reached a high degree of culture under certain multiple conditions of environment, while other peoples, exposed to other combinations of conditions, have remained stationary, or retrograded and become degenerate. What are the essential conditions in con- trol in the one case or the other? Can predictions be made as to what the results of a given combination of physical conditions on a given community will be, in spite of that other and still more mobile, and as yet but little understood, group of conditions embraced under the term psychology? Many profound questions, in the solution of which the physiographer unites his efforts with those of the student of the humanities, present themselves for study during the century that is yet young. Within the broader questions just suggested are many others that are more concrete and definite, and of vital importance to mankind, which can be conveniently grouped under the term economic physi- ography. The problems which here present themselves share their chief interests with the engineer. They relate to plans for transpor- tation in all of its various forms, drainage, irrigation, water supply, sanitation, choice of municipal locations, control of river floods, selection of cities for homes, farms, vineyards, factories, etc. In PHYSIOGRAPHIC PROBLEMS OF TODAY 549 every branch of industry a critical knowledge of the physical condi- tions, both favorable and adverse to the economic ends in view, and of the limitations of the daily, seasonal, and secular changes they experience, is of primary commercial importance. Although the money value of truth should be a secondary consideration to the truth- seekers, a critical study of the influence of environment on industry is as truly a matter of scientific research as any of the less complex and less directly utilitarian branches of physiography. The reaction of human activities on physiographic features pre- sents two great groups of problems. These embrace, on the one hand, the far-reaching and frequently cumulative effects of man’s interference with the delicate adjustment reached in natural condi- tions before his influence became manifest; and, on the other hand, the effects of such changes on man’s welfare. A change amounting to but little less than a revolution in the long-established processes by which the features of the earth’s sur- face are modified and developed, accompanied the advancement of man from a state of barbarism to one of civilization, and is most strikingly illustrated when men skilled in the arts migrate to a pre- viously unoccupied region. This new factor in the earth’s history demands conspicuous changes in the methods of study usually employed by physiographers, and makes prominent a series of investigations the full significance of which is as yet obscure. The wholesale destruction of forests, drainage of marshes, diversion of streams, building of restraining levees along river banks, tillage of land, abandonment of regions once under cultivation, the introduction of domestic animals in large numbers into arid regions and the conse- quent modification, and frequently the destruction, of the natural vegetal covering of the soil, and many other sweeping changes incident to man’s industrial development, are fraught with consequences most significant to the student of nature, and of profound import to the future of the human race. From the point of view of the physiog- rapher, the ultimate result of these great changes in the surface con- ditions of the earth can to a great extent be expressed in one word, and that word is desolation. In view of the suicidal lack of fore- thought manifest in the activities of peoples and, as experience shows, increasing in many directions in destructiveness with industrial prog- 550° ISRAEL C. RUSSELL ress, the problems that confront the physiographer are not only what far-reaching changes in the surface condition of the land result there- from, but how the ruin wrought can be repaired, and how human advancement can be continued and its deleterious consequences on the fundamental conditions to which it owes its birth and develop- ment be avoided or lessened. Considerations which lessen the horrors of the regions crossed by industrial armies are that nature, no matter how severely torn, has great recuperative power and tends to heal her wounds; and also that man, through the science of agriculture par- ticularly, although greatly modifying natural conditions, is able to reconstruct his environment and, so long as intelligent care is exer- cised, adjust it to his peculiar needs. In the relations of physiography to man, as the above hasty sketch is intended to show, the themes for research are many and important. As a suggestive summary, they include the review of history with the aid that modernized physical geography furnishes; the recognition of a strong undercurrent due to inorganic conditions in the political, social, and industrial develop- ment of peoples; the incorporation of physiographic laws into the formulas used by the engineer in all of his far-reaching plans; the calling of a halt in the wanton destruction of the beauties of nature, and the providing of a check on the greed of man which casts a baneful shadow on future generations. Great as are the results to be expected from a better knowledge of the mode of origin of the earth, its defor- mation by internal changes, and the removal and redeposition of material by forces resident on its surface, the combined results of all these studies culminate in the relation of man to his environment. IsRAEL C. RUSSELL. UNIVERSITY OF MICHIGAN. WIDESPREAD OCCURRENCE OF PAYALITE IN CER- TAIN IGNEOUS ROCKS OF CENTRAL WISCONSIN. In central Wisconsin, in the vicinity of Wausau, forming the south- ern portion of the pre-Cambrian district of the northern half of the state, is a wide variety of intrusive igneous rocks. Chief among these intrusives, which constitute perhaps 75 per cent. of the rock forma- tions for many miles around Wausau, is a complex series of holo- crystalline rocks ranging from granite, high in silica, to quartz syenite, nepheline, and sodalite syenites, and related basic, alkali-rich rocks. Associated with this series are an older series of basic rocks and a still older flow of rhyolite, to both of which the various members of the granite-syenite series bear the same structural relations. The habitat of the fayalite is in certain phases of the granite- syenite series, and it is not known to be in the other varieties of igneous rock of the region. The mineral fayalite, it will be recalled, is the pure iron olivine, iron chrysolite, having the formula 2FeO, SiO,, and the theoretical composition, ferrous oxide 70.6, and silica 29.4 per cent. Considered in its entirety, the series in which the fayalite occurs resembles in many respects, though not in all, other well-known series of nepheline-bearing and related alkali-rich rocks, such as those in Arkansas, in the Christiania region of southern Norway, and in Essex County, Mass. The composition of some of the phases of the central Wisconsin alkali-rich rocks is shown in the subjoined table. The analyses were made by Professor W. W. Daniells, professor of chemistry in the University of Wisconsin. In the analyses attention is called to the fairly regular increase in amount of alumina and the alkalies as the silica decreases. The iron is abundant and is fairly uniform in amount, with the exception of that in the granite rich in quartz. With respect to the lime and magnesia, these constituents, like the iron, do not differ appreciably in the various phases. Attention is especially called to the uniformly low content of magnesia, which is without doubt one of the causes of the development of the ferrous silicate, fayalite, in the magma. 55! 552 SAMUEL WEIDMAN I II II IV Vv VI SIO asker es 76.54 67.99 61.18 57-48 54-79 54-76 UO Moats clon eget 13.82 15.85 19.72 20.04 22.87 24.72 ISAO veviaia iaeeeiaae: 1.62 \ 6 Beit 5.64 1.74 A G3 Fe Que edge taco ies off 5 +3 Tea2 B70 3.24. 25 Mn@)isc sae etine apie Es Bean ok trace see MUEXO). ci5on 6 ac ‘ 0.01 O.41 trace °.40 1.92 0.10 Ca@eans 0.85 1.78 2.64 1.70 trace its (O7/ Nas Ore cri ae nena A: 4.32 220 5.28 iOS 10.75 10.38 TF OMA ema ene As ai 4.81 5.66 2.65 4.06 7 Te Ole ete enh oes 0.20 0.30 ©2322 ORIZIG Hae NEU hale 0.55 Clete. ah 0.54 sue EP ieoties orn eee 0.08 T Aa O An ots aie 0.07 Motaliy sass tees aOOROr7 90.71 99-83 100.17 99-75 99 - 63 I. Granite. Quarry granite known as the Wausau red granite. Consists mainly of albite, orthoclase, and quartz. II. Amphibole granite (5298).! Consists mainly of microperthite, quartz, and an amphibole rich in ferric oxide, and alumina. III. Hedenbergite-quartz syenite (5917). Consists mainly of microperthite, quartz, hedenbergite, arfvedsonite, lepidomelane, and fayalite. ITV. Amphibole syenite (6011). Consists mainly of orthoclase and bluish- green amphibole, and some magnetite, mica, fluorite, and zircon. V. Sodalite-nepheline syenite (6426). Consists mainly of anorthoclase, nepheline, sodalite, and hedenbergite, with some fluorite. — VI. Nepheline syenite (5829). Consists mainly of nepheline, orthoclase, hedenbergite, and some fayalite. It should be understood that not all the phases of rock represented in this alkali-rich series is indicated in the above table. Nor is the range in content of silica supposed to be outlined. A%girite is a common constituent of many of the rocks, but analyses of the egirite- bearing phases have not been made. ‘Those who have studied the alkali-rich magmas can best appreciate the amount of work necessary, and the complexity of the problem met with, in describing these interesting rocks from the mineralogical standpoint. Some of the prevailing amphiboles, pyroxenes, and micas, as well as other minerals, have been separated from the rocks and analyzed, and hence where varietal names are used such use is based upon knowledge of the chemical composition. The object of the present paper is merely to describe the general character, occurrence, and signification of 1 The numbers refer to specimens in the Wisconsin Survey collection. FAVALITE IN CERTAIN IGNEOUS ROCKS 553 the fayalite found in various phases of the series, rather than a com- plete description of these rocks as a whole. The nature of the chemical and mineral composition of the series will be briefly referred to again. In the meantime it may be of interest to show how the true character of the fayalite was deter- mined. Under the microscope it appears in ordinary light as a -honey-yellow mineral, and under crossed nicols it has bright polar- izing colors, resembling olivine. ‘The minerals associated with it, however, such as an abundance of quartz, orthoclase, and albite, were not the ones usually found with olivine. The composition of the rock, with its very small amount of magnesia and the alteration product of the mineral wholly to iron oxide, free from serpentine, led to the idea that it might be fayalite. A rough separation of the mineral was made by means of crushing the quartz syenite (like III) bearing considerable of the unknown mineral, and passing the finely crushed material through a Retgers solution of nitrate of thallium and nitrate of silver. A considerable amount of the heaviest material of the rock, which contained, besides the fayalite, some feldspar and pyroxene, was analyzed to see if any appreciable amount of magnesia was present, and to see if the theory that the yellow mineral was fayalite was probably correct. The material was analyzed by Dr. Victor Lehner, and is seen to have been largely iron and silica, and almost entirely free from magnesia, as follows: SiO, = - = - - - - 32 10% Al,O; See He Nats ahaa a oh VE 3-54 Fe,O, - - - - - - 19.86 Fe® - = = - - - 37.56 MgO - - = = - - = ORO 7, CaO - - - - - - 0.57 IPO)n = - - - - - - none MnO - = = = - - trace AKA - - = - - - none Undetermined - - - - - 6.30 100.00 After this rough test, a careful separation of the yellow mineral was made. With the Retgers solution the heavy minerals of the crushed rock were separated. It was impossible to separate with 554 SAMUEL WEIDMAN the Retger’s solution the fayalite from the iron oxide, as the two were so intimately mixed by alteration and intergrowth. ‘The iron oxide, which was strongly magnetic, was very largely removed by means of a magnet, and finally sorting out the honey-yellow fayalite from the magnetic iron oxide by hand had to be resorted to. After considerable work, 0.3610 grams of the nearly pure fayalite was separated, which was analyzed by Dr. Victor Lehner, and found to be as follows: SOR te - - - - - = BO Ge He,O; - - = - - = 0.23 HeOr> = ; > = = - = 62.09 Undetermined - - - - - 3.91 TOO .00 The materials associated with the fayalite, and necessarily included in the analysis, were small particles of feldspar, quartz, pyroxene, and amphibole. The undetermined portion in this second analysis, _ 3.91 per cent., as indicated by comparison with the first rough analysis, is very probably mainly alumina and the alkalies which may be assumed to be combined with perhaps 4 or 5 per cent. of silica to form the associated silicate minerals. An additional source of silica, probably 1 or 2 per cent., was due to small included fragments of quartz. After deducting, therefore, an amount of silica—6 or 7 per cent.—present in mineral other than fayalite, it will be seen that the ferrous oxide and silica occur in approximately the proportions found in fayalite. Under the microscope the fayalite has the appearance of this mineral as usually described. In ordinary light it has a light stone color, with yellowish-green or honey-yellow tints. It is distinctly pleochroic, the green tinge changing from a light to a darker shade. The double refraction is strong, probably stronger than for olivine. It apparently has two fairly well-defined planes of cleavage, as illustrated in Fig. 1, which are probably parallel to the pinnacoids, ool, and o1o. The fayalite occurs in the quartz-syenite, having a silica content of 61.18 per cent., and to a very small extent in the amphibole granite, bearing 67.99 per cent. silica. In phases of the quartz syenite it probably constitutes from 1 to 5 per cent. of the rock, and was noted : FAVALITE IN CERTAIN IGNEOUS ROCKS 555 in many thin sections of this variety from widely different parts of the district. feldspars, orthoclase, albite, and microperthite; with the calcium- iron-amphibole, arfvedsonite; with the calcium-iron-pyroxene, heden- In the rocks bearing quartz it is associated with the bergite; and with iron mica containing large amounts of potassium. In phases of the non-quartzose and nepheline-bearing rocks of the series the fayalite also contributes a fraction of 1 per cent. to as much as 5 per cent. of the rock. In some of the nepheline-rich rocks the faya- lite is the only dark-colored mineral noted in the thin sec- In these rocks the faya- lite occurs with orthoclase and tions. microperthite, nepheline, soda- lite, the soda amphiboles of the riebeckite, and crocidolite type, and the pyroxene, hedenbergite, and the potash-iron mica lepido- melane. The various rocks in which the fayalite occurs is thus seen to have a considerable range Fic. 1.—Photomicrograph of fayalite in The fayalite, light- colored, is surrounded on nearly all sides by Por- tions of it are in contact with feldspar and calcium-iron nepheline syenite; x 60. a black border of magnetic iron oxide. hedenbergite. The greatest length. of the fayalite in this section is along the ¢ axis, parallel to which is indistinct parting or cleavage. The areas within the fayalite are colored dark by photography, and in reality are yellowish-brown and _ bluish mineral The yellowish-brown inclusions in certain chemical constitu- inclusions. ents, such as silica, alumina, and the alkalies, and also in composition. The area over which these rocks mineral are in irregular areas, and are probably vel- lowish hydroxide of iron, géthite, formed by alteration of the fayalite. The bluish inclu- sions are finely striated crystals of a bluish ee : : amphibole. are distributed is quite exten- sive. The amphibole granite and quartz syenite cover several hun- dred square miles, and the nepheline-bearing and related basic rocks, from fifteen to twenty square miles. The fayalite, in the various phases of rock in which it occurs, has the associations and relations of a normal, original constituent of 556 SAMUEL WEIDMAN the rock. It does not occur in veins, segregations, or cavities, but as a common constituent distributed through the rock, like the quartz, feldspar, nepheline, and other common rock-forming minerals with which it is associated. Like the other abundant minerals associated with it, it does not occur in idiomorphic crystals, but assumes shapes in its development due to the mutual interference of surrounding minerals. Where present with an abundance of dark-colored silicates, it is usually in direct contact with them, but in quartz syenite it occurs in direct contact with quartz, and in certain phases of the nepheline syenite it is entirely surrounded by nepheline, or nepheline and feldspar. The fayalite has not been observed to occcur with the alkali- iron pyroxene, egirite, which is a common constituent of several phases of the nepheline-bearing rocks. While it may be stated with certainty that these two iron-rich minerals do not occur in the same thin sections of rock examined—and this includes a considerable number—with this special relation in mind, yet a final conclusion upon this point, it is believed, should be held in abeyance. The pyroxene, egirite, it will be recalled, contains a high per- centage of ferric oxide and but a small proportion of ferrous oxide. On the other hand, the fayalite contains ferrous oxide only, and the pyroxenes, amphiboles, and micas abundantly associated with the fayalite, as shown by the analyses, contain a marked excess of ferrous oxide over ferric oxide. It may be, therefore, that in those magmas, or portions of magmas, very poor in magnesia, which contain a large excess of ferrous oxide over ferric oxide, the chemical conditions for the development of fayalite exists, and in those containing an abundance of ferric oxide the conditions for the development of egirite are present and that necessarily these two minerals will not be expressed in the same phase of rock. A deficiency of oxygen as well as of magnesia may therefore be necessary for the development of fayalite. Fluorite is a persistent, though not an abundant, constituent of the various phases of this series. ‘The fluorite is colorless, has low birefringence and perfect octahedral cleavage, and on account of its extremely low index of refraction, produces the apparent anoma- lous appearance of having a very high index of refraction. ‘The FAVALITE IN CERTAIN IGNEOUS ROCKS 557 occurrence of fluorite in the various members of this series is but another expression, in addition to that of the prevalence of the fayalite, of the probable common magmatic source of the various phases of rocks here referred to as belonging to the same series. The idea has prevailed, to some extent at least, that fayalite, on account of its extremely basic character and its association with quartz in certain rocks, and with spherulites and lithophysae, prob- ably had only an aqueo-igneous origin in rocks rather than ordinary igneous origin. In this connection it may be well to call attention to the various occurrences of fayalite previously noted. It has often been observed in crystalline slags. The first natural occurrence of fayalite observed’ was in 1839 in nodules in volcanic rocks on the island of Fayal (whence the name is derived), one of © the islands of the Azores. In the Mourne Mountains of Ireland it occurs,” filling drusy cavities in granite. Professor J. P. Iddings,3 has described its occurrence with tridymite in spherulites and litho- physae in the obsidian of Obsidian Cliff, Yellowstone National Park; and Iddings and Penfield? have described its occurrence in the recent obsidian flows of the Lipari Islands. It also occurs in ‘“vugs,” in a soda granite at Cheyenne Mountain, Colorado,’ and in a pegmatite dike in hornblende granite at Rockport, Mass.° In the latter place, as described by Penfield and Forbes, it occurs as a lenticular shell of varying thickness, twelve to sixteen inches in diameter, filled on the inside with loose earthy material and enveloped by a layer of magnetite one inch thick. It is a fact worthy of note that this Rockport granite is a phase of the alkali-rich rock of Essex County.? It is also found in cavities in the lava of Vesuvius erupted in 1631, large crystals, apparently in 7 t Poggendor}’s Annalen, Vol. LI, p. 160. 2 DELESSE, Bulletins de la Société géologique de France, Vol. X (1853), p- 571- 3 J. P. Ippincs, Seventh Annual Report, U. S. Geological Survey (1885), pp. 270, 270. 4J. P. Ipprncs anp S. L.. PENFIELD, American Journal of Science, Vol. XL (1890), P- 75: 5 W. E. HIDDEN AND J. B. MactnrosH, ibid., Vol. XLI (1891), p. 439, and Vol. XXIX (1885), p. 250. 6S. L. PENFIELD AND E. H. ForBes, ibid., Vo!. LI (1890), p. 129. 7H. S. WASHINGTON, JOURNAL OF GEOLOGY, Vol. VII (1899), p. 466. 558 SAMUEL WEIDMAN occurring with sodalite and orthoclase. Lacroix has also noted the occurrence of fayalite in the trachyte lava of Mont Doré. In all these instances, with the possible exception of the last mentioned, it will be noted that the fayalite occurs either in cavities, veins, or segregations, and on this account it has been thought that it must necessarily be a product of aqueo-igneous origin. It may be well, therefore, again to call attention to, and emphasize, the fact that not only does the fayalite occur as a persistent, and in places a fairly abundant, constituent of these various rocks described from central Wisconsin, but that it occurs here as a rock constituent under perfectly normal conditions. Fic. 2.— Photomicrograph of quartz- The presence of fayalite as syenite; X60. The crystals shown are faval- ite, much fractured, on the right, hedenbergite : : on the left, both surrounded in part by bluish- number of phases of this series green arfvedsonite. The hlack mineral is would seem to indicate that, magnetite, and the colorless crystal is apatite. an essential constituent of a mineralogically and chemi- cally, these phases are unique. And not only are the fayalite-bearing phases unusual, but also the other phases of the series, since they con- tain no magnesia-bearing minerals, and contain, besides the quartz, feldspar, and feldspathoids, only calcic-iron alumino silicates or alkali- iron alumino silicates, or combinations of them. It is a well-known fact, of course, that all the alkali-rich rocks, such as those associated with nepheline-bearing rocks, contain much less magnesia than lime, and that not only is the proportion of magnesia to lime in these alkali rocks small as compared with the proportion in the usual basic rocks, but that it is also much smaller than occurs in average igneous rocks. In comparison with other somewhat similar and well-known alkali-rich rocks, such as the nepheline-bearing and associated rocks . of the Christiania region of southern Norway, described by Brégger,° °W. C. BROGGER, Zeitschrift fiir Krystallographie und Mineralogie, Vol. XVI (1890). FAYVALITE IN CERTAIN IGNEOUS ROCKS 559 it may be stated that the latter differ radically from the alkali rocks of Wisconsin in their much higher content of magnesia. In Brégger’s series there is, on the average, five or six times as much magnesia as in these, otherwise similar, Wisconsin rocks. The nepheline syenites and quartz syenites of Arkansas, described by J. Francis Williams,’ are more like these Wisconsin rocks, but the Arkansas rocks, otherwise similar, have much more magnesia than those here described. The nepheline-bearing. and associated rocks of Essex County,? Mass., are generally comparatively high in magnesia and are closely related to gabbro and diorite. When the mineralogical composition of these various alkali rocks, from different parts. of the world, is compared, a much more striking dissimilarity is to be noted. The difference, of course, does not lie in the nature of the quartz, feldspar, nepheline, and sodalite, but with the dark-colored silicates; and since the former constitute much the larger portion of the rock, the difference is necessarily accentuated in the less abundant iron silicates. In these Wisconsin rocks, therefore, the low content of magnesia finds expres- sion in the development of fayalite, and of pyroxenes, amphiboles, and micas comparatively low or free from magnesia. Whether or not fayalite or other magnesia-free silicate develops in the magma “probably depends also, as already stated, upon the oxide condition of the iron. A deficiency of oxygen in the magma, as well as of magnesia, therefore, may be the controlling factors in the development of fayalite; and in those portions of the magma where oxygen was more abundant the more complex ferric-oxide-bearing silicates, such as egirite, were formed. In comparison with other somewhat similar, alkali-rich rock series, such as the syenite of southern Norway, Arkansas, and Essex County, Mass., it may be stated that fayalite has not been observed in any of them. Furthermore, the pyroxenes present in these rocks are gener- ally the magnesia-rich varieties, such as augite or diopside, which tend to grade to egirite through augite-egirite and diopside-egirite. The alteration of the fayalite in some instances is quite extensive, the process of alteration consisting merely in the breaking up of this tJ. F. Wi1iams, Arkansas Geological Survey, Vol. IT (1891). 2 H. S. WASHINGTON, JOURNAL OF GEOLOGY, Vol. VII (1899), p. 463. 560 SAMUEL WEIDMAN Fic. 3.—a. Hedenbergite-quartz-syenite (6639); X20. The minerals shown are: alkali-feldspar (je), quartz (q), fayalite (fa), hedenbergite (/), arfvedsonite (a), and magnetite, the black mineral. There is also present a small amount of fluorite and apatite. b. Hedenbergite-mica-syenite (6597); X20. The minerals shown are: alkali- feldspar (fe), brown mica (m), hedenbergite (/), fayalite (fa), and magnetite. A small amount of fluorite and apatite is also present. c. Hedenbergite-nepheline-syenite (6600); X20. The minerals shown are: alkali-feldspar (fe), nepheline (7), hedenbergite (2), fayalite (fa), and magnetite. d. Fayalite-nepheline-syenite (5275); X20. The minerals shown are: alkali- feldspar (fe), nepheline (7), hedenbergite (7), fayalite (ja), and magnetite. e. Fayalite-nepheline-syenite (5275); X20. The minerals shown are: alkali- feldspar (fe), nepheline (7), fayalite (fa), and magnetite. FAYVALITE IN CERTAIN IGNEOUS ROCKS 561 ferrous silicate mineral into the oxide of iron and oxide of silica, the silica being carried away in solution, probably as hydroxide of silica, and the oxide of iron, which is apparently always magnetic, left as residual material. Quartz as an undoubted residual product of the alteration was not observed in any case. ‘The alteration of the fayal- ite to magnetite, as indicated in the photomicrograph, takes place about the borders of the crystals, and along the fractures and cleavage. Not all the iron oxide associated with the fayalite, however, is of secondary origin. ‘There can be little doubt that some of the iron oxide occurring as rims surrounding the fayalite, as well as that occur- ring as inclusions within the fayalite, is of primary origin. The dis- tribution of the iron oxide, probably mainly magnetite, is such as to indicate that most of it occurring in the rocks is an original separa- tion from the magma, for it is by no means entirely confined to the vicinity of the fayalite, or of the other dark-colored constituents, but often appears to be entirely independent of them. (See Figs. 1-3.) Rims of hedenbergite also surround the fayalite (Fig. 3, d), and some- times the reverse is true and fayalite and iron oxide nearly completely surround hedenbergite (Fig. 3, a). Furthermore, these associations are just as abundant in perfectly fresh rocks, as illustrated in phases from a well taken fifty or sixty feet from the surface, as in specimens collected from the rapids of the Wisconsin River at Wausau, and elsewhere, immediately beneath the soil. While much of the fayalite appeared to show more or less altera- tion, still in most cases the alteration was never complete, at least in crystals of the ordinary, or average, size, even when the fayalite- bearing rock was immediately associated with the soil, or running water. Attention has already been called to the alteration of the large nodules of fayalite in pegmatite, sixty feet below the surface, at Rockport, Mass., where the magnetite was observed as forming a shell about the fayalite. In all other cases noted in literature, the alteration of fayalite has been quite generally observed as being to magnetic material, and in this regard it is quite similar to the altera- tion of olivine, in which case the secondary products are usually * serpentine and magnetite. SAMUEL WEIDMAN. Mapison, WIS. REVIEWS. The Clays and Clay Industry oj New Jersey. By HEINRICH RIES AND HENRY B. KUMMEL, ASSISTED BY GEORGE N. KNAPP. (Geological Survey of New Jersey, Vol. VI.) Pp. xxi+548; LVI plates, and 41 figures in text. In 1868 the Geological Survey of New Jersey, Professor George H. Cook, state geologist, published a full report on the clays of the state—a report which was of great service in its day. ‘Ten years later a further report on some of the more important clay districts of the state was issued. These earlier reports were long since exhausted. Furthermore, the develop- ment of the clay industry and the progress of knowledge of the geology of the state, have created a demand for the new report on the clays of the state. The present volume is in response to this demand, and brings the subject up to date. The importance of the clay industry in New Jersev is shown by the fact that in 1902, the last year for which statistics are available, the total value of the clay products of the state was more than $12,600,000. The economic phases of the subject are primarily the work of Dr. Ries, while the geological questions involved are primarily the work of Dr. Kiimmel, who, however, has availed himself of the data in possession of Mr. Knapp, who became familiar with the Cretaceous and later strata of the state in connection with his work on the Pleistocene of southern New Jersey. The volume is divided into four parts as follows: (1) Clay