LAG ae oe a re As ~ e) a ewe oe < gv ¥; @r ; ed hed bd RD | RAE LT ey a ye < Neth ee Woye E — a) Waa ee yyirut Sea Ph aalG ite” ae Pi ‘e ra = ce oe Ma ee Ue peeee Y vy oy ci See a Sd ots S isl gs Se | _ steed Wines ca EL ; Symi ews “bab Muar icy Saal ti eel Pc Lawe Serial ieee ~ PUY OS igo mye a E y) 6% ee rp} fil | See : ARS we a4 ws vir Rex a 4" 1 Paha ‘ ; + a “ 3 ; oy = As 4 ae es pl WO, ote eng tad : : ™N ) Ws eee | eons) jogs | oerss a es SDN IN ee | aay | PAL || pees d po 10a oe, Sere wee gy a gr" eaeribeon: ald | ouatale bibladaseieet ch pace Te ean cae gp ofl ORR. Same ote VA te we NS [ 4 Al | g ww V'C Ss os aiibhiaed cL Pa del Mal Ld heheh ue TAS BT |i. Lbi Luba ldeee vef~\ a leet tyy wets wr = “weg CET Herat ag | woe lt ag ict ak ISU ee Wann ne saa sethantes vu re Nol te Ge pA pe wad wd i) PAN OAS tpiA) vy lg go OOH | A. ee eyed AAR 1 let ee Ale ey X | | | for oe AIT I hAR i | Ah he ete td Mw iY : | Vga werevvere Si UO dug gee | Syte ATs j) im” san SUE WOT. tang wit yy tds ed ° UTE BS 5 an? mel fs 2 fa See es Ly ; at WOR A BA Were~ Ls pri i Se aa by Pt hal Ay we a 3 Ps vhs ad | arene: ~ <5 co \ SSNS A j | ; y v po ‘A BS oe & i‘. 4, a Buf OT Ge MDPC CET M Destino Lili Titi evn tyoery ys a = od nog te Wyyl 4 ve'e SS ag Bean © | wa fag PT ie rORa~ 4 pew Ay} Fah le ec au vei qo -1F% VPSTUTVTD CAE EEE TLE Nt A Nelabeet bags “wrthicng Ah) bi ’ < ¥, ang des ST an ad A wo" ih fi Pd TT Tain NWA ewes, Ot ete a ex ) es a SLE ih i baletatialal folate tothe Mrattversecetosen Me “Waly TTT dhe) ppb de a lhe a hiss wwil TM vy eaue AR Pa - lala 4 h ha mee lll eieen™ Migeilel men eva oes CETL EP wry ) 'vr Peete URN eee tema e Ti Mal ieee | Pipe Telesyn: Mi PAO ld ar) rr eacetan : : ever" ye, Co SR SST sects he Wh at _A | VE “4 ~ a Wey Wee nce 2%, | mM owl aie fifty : Tome PbL as oe a ; ~~ Fare. ee eaeelee ‘ eRe Ame os. 4k Vd bephldl T oT A ne ¥ ~) ed ’ Voce SO hei le PT La bp yee hitler eT aie TUL TET cihabheehlaeha Al TTT TT iadeaabaia | MARR LL pi AA ol 4 d¥ 2 g ¥ may 4 z ert Ts : ga me ee ay rapa nl VO” "os ont ULC Webeseew tee egy tte da | @~ TTR Ps /layt ey af cere ererorerntet! Zee oueabucue LL hi "A peryes Me Wert wriewl anv: one AAP AAA ainetpme nacaaammme oii EL (UTIL! << fed w Nae - 5 ae ee Ad vw Vg we “eee A eT > ba AAJ ‘See SNE s 5 é Wut MT tet ads PAR at pee i h tg AS : " tay Nid feos. : oes NUN Why bia —— pal ia ¥ ee vy | ewes = *~ 2 ete — d : ae ore ‘ty a Ss > xp = View \eeyet ws pod Ml ge eu 4 1 Suv auvweeys Wey pay Tey i fe or : eee Await) i} 1 peer hee | yo? wie ow ‘\vvewes” Ey tl Pia eae As i ganna aay ¥ : ma eu OT et ts | ad saul geen wy ere | 28 weit a SE Sd wh | eee | AN. ~ : : ’ ww eae wm sme thi eit | ea Wy (AY : Woo. ~ = AY se With, Pha addi t w® dye EE 4 . een tle td ee eae ab mille Yl ayoe ~" eurce’ a + Sa 1 avty : o i VB Jee rece: ~ eed SG || a t | meee SAY 4 Bee | 4ug-< ‘©, ail e sities Jot s Foy!” : TELL EPL babi SSeS PLE i A Wye ¥ mer ahs eer ee , Ne “We WORT iL ale iqyuetattl eh ea” wil HP eu saniar tei nit whe. -% oseuil se sy sd Dated Pad ss 4¥Srpa 2 iw 6 ig cael wt. ee a As ~ ete VSS Teen tiagsuey a] SP. ve ( an re LL ao @ ti OW 4 Fons Oe J ae Seven Cet. Pb bare rin Wye PPE ewe | PANN, 2710 sae oT SSsEL ww - 1 : a- *4A ar VW lelle: FOURNAL 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 ALBERT JOHANNSEN C. R. VAN HISE Petrology Structural 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 Cornell University CHARLES BARROIS CLD WAL COdaE France U.S. Geological Survey ALBRECHT PENCK J. C. BRANNER Germany Stanford University HANS REUSCH W. B. CLARK Norway Johns Hopkins University GERARD DE GEER O. A. DERBY Sweden Brazil Wi ee DAVE) Australia Les Aiaian nti VOU MIE YaIrE The Gnibversity of Chicago Press CHICAGO, ILLINOIS Published February, March, May, June, August, September, November, December, 1911 Composed and Printed By The University of Chicago Press Chicago, Illinois, U.S.A. CONTENTS OF VOLUME XIX NUMBER I CLIMATE AND PHYSICAL CONDITIONS OF THE KEEWATIN. A. P. Coleman THE AGENCY OF MANGANESE IN THE SUPERFICIAL ALTERATION AND SECONDARY ENRICHMENT OF GOLD Deposits. William H. Emmons LocaL DECOMPOSITION OF ROCK BY THE CORROSIVE ACTION OF PRE- GuaciAL PEAt-Bocs. Edwin W. Humphreys and Alexis A. Julien THE Focus oF Post-GLaciAL Uptirr NorTH OF THE GREAT LAKES. J. W. Spencer : : 4 ‘ : A : : A GEOLOGICAL RouTE THROUGH CENTRAL AstA Minor. William T. M. Forbes , 3 s : ; - THE VARIATIONS OF GLACIERS. XV. Harry Fielding Reid REVIEWS NUMBER II THE SOUTHERLY EXTENSION OF THE ONONDAGA SEA IN THE ALLE- GHENY REGION. E. M. Kindle THE MISSISSIPPIAN-PENNSYLVANIAN UNCONFORMITY AND THE SHARON CONGLOMERATE. G. F. Lamb : ; THE WICHITA FORMATION OF NORTHERN Texas. C. H. Gordon, George H. Girty, and David White 5 : : NOTES ON THE OSTEOLOGY OF THE SKULL OF ParRioTIcHus. E. B. Branson HicH TERRACES AND ABANDONED VALLEYS IN WESTERN PENNSYLVANIA. Eugene Wesley Shaw REQUISITE CONDITIONS FOR THE FORMATION OF ICE RAMPARTS. William H. Hobbs 2 : : , THE TERMINAL MORAINE OF THE PUGET SOUND GLACIER. J. Harlen Bretz EDITORIAL: (hips SEEDING OF WoRLDS. I. C. C: ARTESIAN WATERS OF ARGENTINA. B. W. PETROGRAPHICAL ABSTRACTS AND REVIEWS REVIEWS 47 57 61 83 go 97 104 110 135 140 157 161 175 178 181 189 vl CONTENTS OF VOLUME XIX NUMBER III CERTAIN PHASES OF GLACIAL Erosion. Thomas C. Chamberlin and Rollin T. Chamberlin ; : ; , ; VALLEY Frinuinc By INTERMITTENT STREAMS. A. E. Parkins . ORIGINAL IcE STRUCTURES PRESERVED IN UNCONSOLIDATED SANDS. Charles P. Berkey and Jesse E. Hyde RESTORATION OF SEYMOURIA BAYLORENSIS BROILI, AN AMERICAN Cotytosaur. S. W. Williston GEOLOGIC AND PETROGRAPHIC NOTES ON THE REGION ABOUT CAICARA, VENEZUELA. T. A. Bendrat Tar AGE OF THE TYPE EXPOSURES OF THE LAFAYETTE FORMATION. Udward W. Berry Ti. RIPPLES OF THE BEDFORD AND BEREA FORMATIONS OF CENTRAL AND SOUTHERN OHIO, WITH NOTES ON THE PALEOGEOGRAPHY oF THAT Epocu. Jesse E. Hyde A PosstpLtE LIMITING EFFECT OF GROUND-WATER UPON EOLIAN Erosion. Joseph E. Pogue RECENTLY DISCOVERED Hot SPRINGS IN ARKANSAS. A. H. Purdue REVIEWS PETROLOGICAL ABSTRACTS AND REVIEWS NUMBER IV MacMAtic DIFFERENTIATION IN Hawai. Reginald A. Daly PETROGRAPHIC TERMS FOR FIELD Use. Albert Johannsen THE EvoLutIonN oF LIMESTONE AND Dotomite. I. Edward. Steidt. mann THE RECURRENCE OF TROPIDOLEPTUS CARINATUS IN THE CHEMUNG FAUNA OF VIRGINIA. E. M. Kindle FURTHER DATA ON THE STRATIGRAPHIC POSITION OF THE LANCE FORMATION (‘‘CERATOPS BEpDs’’). F. H. Knowlton . LARGE GLACIAL BowWLpERS. George D. Hubbard . REVIEWS NUMBER V SAMUEL CALVIN. H. Foster Bain . THE EVOLUTION OF LIMESTONE AND Dotomite. II. Edward Steidt- mann PAGE 193 217 223 232 238 249 Boy 270 272 276 283 289 317 323 346 358 Sil 381 385 392 CONTENTS OF VOLUME XIX THE DIFFERENTIATION OF KEWEENAWAN DIABASES IN THE VICINITY or LAKE Nipicon. E. S. Moore GENERA OF MISSISSIPPIAN Loop-BEARING BRACHIOPADO. Stuart Weller PHYSIOGRAPHIC STUDIES IN THE SAN JuAN District oF COLORODA. Wallace W. Atwood THE VARIATIONS OF GLACIERS. XVI. Harry Fielding Reid . PETROLOGICAL ABSTRACTS AND REVIEWS REVIEWS NUMBER VI PRELIMINARY STATEMENT CONCERNING A NEW SYSTEM OF QUATER- NARY LAKES IN THE MIssIssippI Basin. Eugene Wesley Shaw GRAVEL AS A RESISTANT Rock. John Lyon Rich . THE CRETACEOUS AND TERTIARY FORMATION OF WESTERN NORTH DAKOTA: AND EASTERN Montana. A. G. Leonard ON THE GENUS SYRINGOPLEURA SCHUCHERT. George H. Girty PRELIMINARY Notes on Some IcNEous ROCKS OF Japan. I. S. Koézu PRELIMINARY NoTES ON SoME IGNEOUS Rocks oF JAPAN. II. S. Kozu : d : : : : PRELIMINARY NOTES ON SOME IGNEOUS Rocks oF JAPAN. III. S. Kozu REVIEWS NUMBER VII THE Iowan Drirr. Samuel Calvin THE THEORY OF Isostasy. Harmon Lewis : : ; SPECULATIONS REGARDING THE GENESIS OF THE DIAMOND. Orville A. Derby : ; , : : : : : : PRELIMINARY NOTES ON SOME IGNEOUS Rocks OF JAPAN. IV. S. Kozu ; : : : FACTORS INFLUENCING THE ROUNDING OF SAND GRAINS. Victor Ziegler : : : 4 : THE UNCONFORMITY BETWEEN THE BEDFORD AND BEREA FORMATIONS OF NORTHERN Onto. Wilbur Greeley Burroughs Vil PAGE 429 439 449 454 462 469 481 492 597 548 555 561 566 576 Saal 603 627 632 645 655 Vill CONTENTS OF VOLUME XIX EDLLORTAL Gs CG. REVIEWS ; RECENT PUBLICATIONS . NUMBER VIII Tue BEARINGS OF RADIOACTIVITY ON GrEoLocy. T. C. Chamberlin THE WING-FINGER OF PTERODACTYLS, WITH RESTORATION OF Nycrosaurus. S. W. Williston THE TERRESTRIAL DEPOSITS OF OWENS VALLEY, CALIFORNIA. Arthur C. Trowbridge i ; : On CoRUNDUM-SYENITE (URALOSE) FROM Monrana. Austin F. Rogers A DrRaAwiInc-BoaRD WITH REVOLVING DISK FOR STEREOGRAPHIC Projection. Albert Johannsen REVIEWS i INDEX TO VoL. XIX ~ VOLUME. XIX NUMBER | THE JOURNAL or GEOLOGY A SEMI- QUARTERLY EDITED BY THOMAS C. CHAMBERLIN AND ROLLIN D. SALISBURY With the Active Collaboration of SAMUEL W. WILLISTON ALBERT JOHANNSEN WILLIAM H. EMMONS Vertebrate Paleontology Petrology Economic Geology STUART WELLER WALTER W. ATWOOD ROLLIN T. CHAMBERLIN Invertebrate Paleontology Physiography Dynamic Geology ASSOCIATE EDITORS SIR ARCHIBALD GEIKIE, Great Britain GROVE K. GILBERT, National Survey, Washington, D.C. HEINRICH ROSENBUSCH, Germany CHARLES D. WALCOTT, Smithsonian Institution THEODOR N, TSCHERNYSCHEW, Russia HENRY S. WILLIAMS, Cornell University CHARLES BARROIS, France JOSEPH P.IDDINGS, Washington, D.C. ALBRECHT PENCK, Germany JOHN C, BRANNER, Stanford University HANS REUSCH, Norway R. A. F. PENROSE, Philadelphia, Pa. GERARD DeEGEER, Sweden WILLIAM B. CLARK, Johns Hopkins University ORVILLE A. DERBY, Brazil WILLIAM H. HOBBS, University of Michigan T. W. E. DAVID, Australia FRANK D. ADAMS, McGill University BAILEY WILLIS, Argentine Republic CHARLES K,. LEITH, University of Wisconsin JANUARY -FEBRUARY, 1011 CONTENTS CLIMATE AND PHYSICAL CONDITIONS OF THE KEEWATIN - - - A. P. Coreman I THE AGENCY OF MANGANESE IN THE SUPERFICIAL ALTERATION AND SECOND- ARY ENRICHMENT OF GOLD DEPOSITS - - - - - - - Witiram H. Emmons = 15 LOCAL DECOMPOSITION OF ROCK BY THE CORROSIVE ACTION OF PRE-GLACIAL PEAT-BOGS - - - - - - - - - - -Epwin W. Humpureys anp ALExIs A. JULIEN 47 ON THE FOCUS OF POSTGLACIAL UPLIFT NORTH OF THE GREAT LAKES J. W. SPENCER 57 A GEOLOGICAL ROUTE THROUGH CENTRAL ASIA MINOR ~~ Wrtitam T.M. Forses 61 THE VARIATIONS OF GEACIERS. XV -- - -..- -,- = = - “HArRy FIeELpine REID “© .83 PLETE Sis g CS a 9 eS ASE Nite 9 he ages, ee eh cass Che University of Chicago Press CHICAGO, ILLINOIS CAMBRIDGE UNIVERSITY PRESS, Lonpon anp EpInpurRGH WILLIAM WESLEY & SON, Lonpon TH. STAUFFER, LErpzic Wh we print ore: y Advertisements in Our Journals re HIS journal is published with the object of promulgating the results of advanced study in the particular science which it represents. As a subscriber to the journal, you are, we assume, interested in the extent to which - this can be accomplished, as well as in the perfection of its mechanical presentation. Each of these lines of effort is attended by a large outlay of money. It follows that the greater the mcome of the journal, the more nearly it can approach an ideal attainment in both directions. A legitimate—and in some cases a preponderatingly large—source of income is advertising. We exercise a rigid censorship over advertising. We accept none that we do not believe to be thoroughly reliable. We consistently refuse any that might in the least degree be offensive or inappropriate to our clientéle of subscribers. | _ In view of this, our readers may properly consult their own interests and at the same time render substantial assistance to the journal by patronizing the advertisers whose announcements appear in these pages, and by mentioning the journal when so doing. The University of Chicago Press Chicago, Illinois ERRATA In the article by James H. Gardner, published in No. 8, 1910, of this Journal, the following errata should be noted: Plate I, between pages 702 and 703, belongs to the preceding paper, by Messrs. Ball and Shaler. Plate II, between pages 742 and 743 in the paper by Mr. Robin- son, belongs between pages 708 and 709. | Label for Fig. 3, p. 716, is for Fig. 7, p. 724. Label for Pigs 4,ip. 728. 1s tor Mig. a7 oO: Label for Fig: 5, ore under the conditions which generally prevail is very subordinate. Though the amount so contributed may facilitate the solution of gold, it is probably inadequate to form sufficient higher manganates or similar salts to suppress effectively the action of ferrous sul- phate. Under such conditions the gold could not travel to the reducing-zone below the water level, but would be precipitated practically at the place where it had been dissolved. 6. Concentration in the oxidized zone——The concentration of gold in the oxidized zone near the surface, where the waters t Bulletin No. 330, U.S. Geological Survey (1908). MANGANESE IN GOLD DEPOSITS AI remove the valueless elements more rapidly than gold, is fully treated by T. A. Rickard in his paper on the “‘Bonanzas in Gold Veins.”’* Undoubtedly this is an important process in lodes which do not contain manganese, or in manganiferous lodes in areas where the waters do not contain appreciable chloride. In the oxi- dized zone it is sometimes difficult to distinguish the ore which has been enriched by this process from ore which has been enriched lower down by the solution and precipitation of gold, and which, as a result of erosion, is now nearer the surface. It cannot be denied that fine gold migrates downward in suspension; but in all probability this process does not operate to an important extent in the deeper part of the oxidized zone. If the enrichment in gold is due simply to the removal of other constituents, it is important to consider the volume- and mass-relations before and after enrichment, and to compare them with the present values. In some cases, it can be shown that the enriched ore occupies in the lode about the same space as was occupied before oxidation. Let it be supposed that a pyritic gold ore has been altered to a limonite gold ore, and that gold has neither been removed nor added. Limonite (sp. gr. from 3.6 to 4), if it is pseudomorphic after pyrite (sp. gr. from 4.95 to 5.10) and if not more cellular, weighs about 75 per cent as much as the pyrite. In those speci- mens which I have broken, cellular spaces occupy in general about ro per cent of the volume of the pseudomorph. With no gold added, the ore should not be more than twice as rich as the primary ore, even if a large factor is introduced to allow for SiO, removed and for such cellular spaces. Rich bunches of ore are much more common in the oxidized zone than in the primary sulphides of such lodes. They are present in some lodes which carry little or no manganese in the gangue, and which below the water level show no deposition of gold by descending solutions. Some of them are doubtless residual pockets of rich ore which were richer than the main ore body when deposited as sulphides, but others are doubtless ores to which gold has been added in the process of oxidation near the water-table by the solution and precipitation of gold in the presence of the t Trans., XX XI, 198-220 (1901). 42 WILLIAM H. EMMONS small amount of manganese contributed by the country rock. In view of the relations shown by the chemical experiments it is probable that a very little manganese will accomplish the solution of gold, but that it requires considerably more manganese to form appreciable amounts of the higher manganese compounds which delay the deposition of gold, suppressing its precipitation by ferrous sulphate. In the absence of larger amounts of the higher manganese compounds, the gold would probably be pre- cipitated almost as soon as the solutions encountered the zone where any considerable amount of pyrite was exposed in the partly oxidized ore. From this it follows that deposits showing only traces of manganese, presumably supplied from the country rock, are not enriched far below the zone of oxidation. 7. Vertical relation of deep-seated enrichment in gold to chalco- citization.—In several of the great copper districts of the West gold is a by-product of considerable value. In another group of deposits, mainly of middle or late Tertiary age and younger than the copper deposits, silver and gold are the principal metals, and copper, when present, is only a by-product. . But in some of these precious-metal ores chalcocite is, nevertheless, the most abundant metallic mineral, often constituting 2 or 3 per cent of the vein matter. Frequently it forms a coating over pyrite or other minerals. Some of this ore, appearing in general not far below the water-table, is fractured, spongy quartz, coated with pulver- ulent chalcocite. It frequently contains good values in silver, and more gold than the oxidized ore or the deeper-seated sulphide ore. Clearly, the conditions which favor chalcocitization are favorable also to the precipitation of silver and gold. The exact chemical reaction which yields chalcocite is not known. At 100° C., according to Dr. H. N. Stokes,’ the reaction with pyrite is probably about as follows: 5Fes,-+-14CusO,- 12H O—7 Cus shiesO,- 12,50). In the cold, the reaction may differ in details, but: without doubt much ferrous and acid sulphate is set free. Attendant reactions Unpublished. MSS quoted by Lindgren in Professional Paper No. 43, U.S. Geological Survey, 183 (1905), and in Weed’s translation of Beck’s textbook. MANGANESE IN GOLD DEPOSITS 43 confirm this statement; for, if calcite is present, gypsum is formed by the reaction of H,SO, on lime carbonate; and, if the wall- rocks are sericitic, kaolin is formed by the acid reacting upon silicates, the potash going into solution as sulphate. The abun- dant ferrous sulphate must quickly drive the gold from solution, and it apparently follows that there may be no appreciable enrich- ment of gold below the zone where chalcocitization is the prevail- ing process. VI. REVIEW OF MINING DISTRICTS 1. If gold is more readily dissolved in manganiferous deposits, it would be supposed that placers form less readily from pyritic manganiferous lodes than from lodes containing no manganese. | If, in areas where the waters carry appreciable chlorine, placers have formed as extensively from such lodes as from lodes free from manganese, then the hypothesis fails. 2. The manganiferous lodes, in areas of chloride waters, as in the undrained areas of the Great Basin, should in general show less gold at the outcrop and in the upper portion of the oxidized zone than below. Im silver-gold deposits, however, silver, on account of the insolubility of the chloride, may remain, or be concentrated, in the oxidized manganiferous zone. Bunches of rich gold ore carrying oxidized manganese in the oxidized zone are not necesssarily fatal to the theory; for, as already stated, these . are probably residual from the zone of secondary enrichment. An extensive enrichment in gold of the oxidized manganiferous ores at the surface, which are shown not to be residual from the zone of secondary ores, would indicate that the selective processes lack quantitative value, if the waters carry chlorine, and if the primary ores, from which the manganiferous oxidized ores are derived, carry appreciable pyrite to supply sulphate. 3. If in certain lodes gold migrates below the water-table, it should be precipitated quickly by ferrous sulphate. But MnO, converts ferrous sulphate to ferric sulphate, which does not pre- cipitate gold. Hence, MnO, favors the solution of gold, and converting ‘the ferrous salt to ferric sulphate removes the pre- cipitant. Consequently, if auriferous lodes show enrichment in 44 WILLIAM H. EMMONS the deeper zone but related to the present surface of the country, the manganiferous lodes should, the other favorable conditions provided, show greater differences in values with respect to gold than lodes free from manganese. Gold provinces of the United States —As Lindgren’ pointed out in 1902, the principal gold deposits of the United States may be divided into four groups. ‘The deposits of each group belong mainly to one metallogenetic epoch, and certain relationships are clearly shown. ‘This classification, which has thrown much light on the genesis of the deposits, is useful as an instrument for study and for comparison of the deposits with respect to the problem of the migration of gold in them. t. The Appalachian gold deposits, and those of the Home- stake type in South Dakota, are the most important representa- tives of the oldest group. ‘These deposits generally yield placers, are usually low grade below the water level, and are singularly free from bonanzas. They are, in general, not greatly leached near the surface, and may have been enriched by the removal of other material more rapidly than gold. At only one of them, the Haile mine, in South Carolina, it is thought probable that gold has been carried below the water level. Judging from descriptions, practically all of these deposits are free from manganese. 2. The California gold veins and related deposits in Nevada (Silver Peak) and in Alaska (Treadwell, etc.) are younger than the Appalachian deposits, and were probably formed in the main in early Cretaceous times. These deposits, where physiographic conditions are favorable, have generally yielded rich placers. At many places, moreover, the ore is worked at the very surface, and, there is very little evidence of the migration of gold to the deeper zones. In the places where detailed work has been done, rhodo- chrosite is never a gangue mineral, although manganese oxide does occur in traces in the country rock, and rhodochrosite is found in a few places in veinlets in the mining districts but not associated with the gold veins. 3. The deposits of the third group are later than the early t“The Gold Production of North America,” Trans., XXXIII, 790-845 (1903); ““Metallogenetic Epochs,” Economic Geology, IV, No. 5, 409-20 (Aug., 1909). MANGANESE IN GOLD DEPOSITS 45 Cretaceous, and some of them are probably early Tertiary. They are extensively developed in Montana, Nevada, Utah, and Colorado. Mr. Lindgren calls this group the Central Belt. Many of its deposits have yielded considerable gold, and in certain other districts very closely related genetically (Butte, Georgetown silver-gold lodes, Cortez Nevada, Tintic, etc.) much gold has been obtained as a by-product to copper or silver mining. Some of these deposits have yielded placers and some have not. At Philipsburg and Neihart, Mont., Georgetown, Colo., and else- where, the deposits show a secondary enrichment of silver below the water-table. At Philipsburg, and probably at some other places, an enrichment in gold accompanies this concentration of silver. Some of the lodes of group 3 carry much manganese, and some carry none. Present data are meager for most of these dis- tricts. The determination of gold from the surface down in a large number of deposits would serve as a useful check to the con- clusions based upon the chemistry of the processes involved in its solution and precipitation. 4. Group 4 includes the most recent ore deposits in the United States. All of them are Tertiary, and most of them are Miocene or Pliocene. In general, they were formed relatively near the surface, and in some places it is highly probable that not more than a thousand feet of vein material has been removed by erosion since the ores were deposited. The majority of these deposits carry silver, and in many of them its value is greater than that of the gold; but they have supplied, notwithstanding, about 25 per cent of the gold production of North America. They are typi- cally developed in Nevada (Comstock, Tonopah, Goldfield, Tus- carora, Gold Circle); California (Bodie); Idaho (De Lamar); South Dakota (later than Homestake type); Colorado (Cripple Creek, Idaho Springs, Rosita Hills, San Juan, etc.); Montana (Little Rockies, Kendall, etc.). Many occurrences in Mexico should probably be placed here, also. The deposits of this group have not supplied much placer gold. Many of these deposits are in arid countries, where conditions for working placers are not favorable; but even those in well-watered districts supply rela- tively little placer gold. Manganese is abundant in some of 46 WILLIAM H. EMMONS these deposits (Comstock, Exposed Treasure, Tonopah); it is very sparingly present in others (Little Rockies); in still others (Goldfield) it is almost entirely absent. A few small placers are associated with the manganiferous lodes, although at some places they seem to have been derived from veins near by which are not manganiferous. Many of the Cali- fornia veins carry rich ore at the very surface, but the Tertiary gold veins are generally richer in gold a few feet below the surface than at the outcrop. Doubtless, many of them would have been overlooked if it had not been for the concentration of horn silver and argentiferous pyromorphite at the surface. It thus appears that practically all of the manganiferous gold deposits of the United States, so far as they have been described, may be included in groups 3 and 4; that nearly all described deposits where relations indicate a migration of gold belong to the same groups; that placers are much less abundantly devel- oped than in groups 1 and 2; and that outcrops less frequently supply gold; that secondary enrichment below the water-table, if carried on at all, proceeds with extreme slowness in groups 1 and 2, but may be more pronounced in deposits of groups 3 and 4. Not all deposits of 3 and 4 carry manganese, however, and those which do not carry it show relationships more nearly approxi- mating those which hold in the California gold veins. The migra- tion of gold in the more important auriferous deposits of the United States is discussed in some detail in Bull. 46, Amer. Inst. Mining Engineers, 817-37. LOCAL DECOMPOSITION OF ROCK BY THE CORROSIVE ACTION OF PRE-GLACIAL PEAT-BOGS EDWIN W. HUMPHREYS AND ALEXIS A. JULIEN While the layer of decayed rock which once overlay the region around New York City has been generally planed off by the conti- nental glacier, certain small isolated spots have been noted from time to time in which masses of rotten schist still remain. Their decay is commonly attributed to weathering action, and their escape from the glacial scour at such points, to their probable protection by projecting eminences of rock under whose lee they are supposed to lie. Excavation in schist—An unusually large occurrence of this kind has been recently exposed in an excavation for a cellar on the east side of the junction of Southern Boulevard and Westchester Avenue, Borough of the Bronx, New York City, whose general form and dimensions‘ are shown in Fig. 1. The gneissic schists here present the foliation with usual high angle, 70° to 90°; strike N. 23° E.and S. 23° W. Asmall anticlinal fold crosses the strata, as shown in the diagram (Fig. 1) whose axis runs N. 52° E. and S. 52° W. A small overthrow is shown in its cross-section at the northern end, and at its southern end it pitches to the southwest at an angle of about 30°. The rock con- sists chiefly of a fine granular aggregate of quartz, with much biotite in minute black scales, and more or less disseminated white feldspar. Throughout the western half of the excavation, however, many thin seams of pegmatitic gneiss and of gray quartz are intercalated, up to nine inches in thickness. Pegmatite dike-——A pegmatite dike, about five feet in width, nearly vertical, also cuts obliquely through the schists, with a course of N. 30° E. and S. 30° W. At many points, small projections or apophyses branch out into the schist along its course and are, 1 We wish to express our indebtedness for these data to Mr. C. S. Shumway, superintendent of the Construction Department of the American Real Estate Co. 47 48 EDWIN W. HUMPHREYS AND ALEXIS A. JULIEN apparently, connected with some of the pegmatitic seams inter- calated in the schist. The position of the dike, on the west side of the overthrow in the anticline at the north end, suggests that it has there acted as an obstacle against the northwestward thrust of the beds and so produced the westward distortion of the upper side of the fold. The pegmatite itself is an aggregate of grayish quartz, white feldspar, and very little mica, of the rather uniform medium texture usual in the dikes of the Bronx region, with grains rarely exceeding two or three inches. SOUTHERN BOULEVARD. Oy YY i] Ue AL al Dike Decomposed Schist Undecomposed Schist Fic. 1.—Relative positions of decomposed schist, undecomposed schist, and pegmatite dike. Decay of schist—In the eastern part of the excavation, the rock was hard and sound, and needed to be blasted for removal. In the western, the schist was thoroughly decomposed throughout to an undetermined depth, so soft that it was easily removed with pick and shovel, bluish to purple gray in color, and in texture passing from a gritty aggregate almost to a clay; the latter corresponded closely to the glacial clays of similar color commonly found about the city. The two tracts, fresh and decayed, were separated by an exceedingly sharp contact (the line A—B in the diagram), so that, DECOMPOSITION OF ROCK BY CORROSIVE ACTION 49 in the cross-section at the north end (at the point B), the hand placed across this line would rest on the left upon the decomposed schist, easily dislodged by the touch of a finger, and on the right upon the hard fresh rock. This contact is shown in Fig. 2, with the decayed schist on the left, and on the right the same rock in rugged, hard condition. The trend of this sharp division line was N. 28° E. and S. 28° W., approxi- mately parallel to the course of the dike. However, in the cross- section, a few seams of decayed rock were noticed to the east of this line, descending a yard or more into the solid schist. The same section showed that the upper eroded surface of the schist descended from a height of four- ‘teen feet at the point B along the northeast wall, to a height of seven and one-half feet, in a distance of fifty-four feet to the corner on Westchester Avenue. Decay of pegmatite.—A similar decay has atiected the pegmatite, Fic. 2.—Contact of decomposed and muchofwhosefeldsparhaspassed — ndecomposed schist. into a white kaolinic clay, so that this rock also was easily removed by means of the pick. Although it is even now much more tough and solid than the sur- rounding schist, it appears to have been planed off by the ice at about the same level, as shown near the bottom of the cross-section (Fig. 3) where the north end of the dike strikes the wall at West- chester Avenue. Above it lies a layer of till, and then a slab of granitic gneiss. It should be also noted that the decay above described is entirely exceptional in this region. For example, in another excavation in the schist, a few hundred feet to the north, the same schist was found practically undecomposed and sound. So also as to the numerous other pegmatite dikes in the Bronx, all we have observed are solid and show almost no decay. 50 EDWIN W. HUMPHREYS AND ALEXIS A. JULIEN Glacial deposits —The layer of glacial deposits, which overlies the schist at this locality, as shown in the following generalized cross-section along Westchester Avenue (Fig. 4), about fifteen feet from the street level to the greatest depth in the excavation, is yet to be considered. ee Fawn-colored micaceous sand with some trap bowlders. . . ves Slab: of pega titicsomelsstaare pees eee ae ee I Gray sandy gull swith staatedsbowldersem arc a ee 3 Slabyonjpegmia titleremeiss sane i sa ae oes ne I Gray till irich amma caceousiclay ay ees eset eee i-3 Slabt ok pesmatiticremeisses serie srneyra yen nie eae oie t-3 Gray bowlder: clawys.werrecsxs tye cs case canes Mata ete barb cere near 3-3 Slab sofpesmatiticremelssien. 00 ashen rc hae eer 3-23 Blue-etay bowldericlayincn ac, oe Note cleo Pee anes vee i Slabs of pegmatitic gneiss and Manhattan schist.......... If Blue-eray bowlderm clays 4 cone seeeie se ree eee I Decayed schist in place with vertical foliation, intercalated Withethinuseams Olspe cimatice ss. ci crn neers ars aera 6-83 The remarkable deposit of ground moraine, which here rests upon the upturned edges of the schist, is thus found to consist largely, in the two hundred feet of section exposed along the two avenues, of a succession of huge, overlapping sheets of granitic gneiss separated by layers of sand and till or bowlder clay. The gneiss slabs,-of which a series of from four to eight are shown in any particular part of the section, consist of a rather fine-grained granitoid gneiss of constitution similar to that of the pegmatite. Their dimensions in cross-section vary from about 3 to 35 feet in length, and in thickness from 1 to 30 inches or more. There was no opportunity to determine their real shape, but appar- ently they consisted of flat sheets, often thinning down toward the edges to an inch or less. Some show fracture and faulting in place, as by the effect of superincumbent pressure (Fig. 5), and occasion- ally the extension of such a slab toward its edge into a thin pliable sheet, one or two inches thick, reveals a marked curvature as by pressure from above (Fig. 6). Toward the bottom of the section, they may be accompanied by a few small sheets of fine biotitic schist, like that of the underlying rock in place. The granitoid gneiss in these slabs shows partial to thorough decay, so that they DECOMPOSITION OF ROCK BY CORROSIVE ACTION 51 mark the cross-section by a series of conspicuous, white, kaolinic, lenticular bands, contrasting with the intervening layers of dark till. The bowlder clay of these intervening layers is very dense and compact, some- times sandy, some- times rich in mica and clay, and con- tains few pebbles and occasional bowl- ders up to about two feet in diameter, which may show sharp, glacial striae. Fic. 3—Manner in which the pegmatite dike was These consist partly planed off by the glacier. of rocks of the vicin- ity, quartz from seams, granitoid and hornblende gneiss, etc., and partly of rocks from the Palisades on the west bank of the Hudson River, about five miles distant, viz., diabase, coarse red sandstone, indurated shale from the con- tact underlying the trap, etc. Nearly all these bowlders are hard and un- decomposed. At several points where such a trap bowlder rested im- Fic. 4.—Section along Westchester Avenue. mediately upon a granite slab, the latter was deeply indented, its folia separating and rising a little around the bowlder, above the upper level of the slab. Immediately 52 EDWIN W. HUMPHREYS AND ALEXIS A. JULIEN under the bowlder, the folia of the granite showed differentiation and deformation as by a crushing force from above; but the lower Fic. 5.—Fracture and faulting shown by one of the transported slabs. level: or mther sslalp rarely showed much if any depression in a direction below the bowlder (Figs. 7 and 8). . Cause of decay.— Inseeking to account for this peculiar de- composition in one tract of the schist, the action of the weather is barred out, on account of the absence of such decomposition in adjoining areas of schist, as well as in the other pegmatite dikes of the Bronx region, the absence of concentration of iron oxide from agencies of mere oxidation,’ and the sharp line of de- markation between this tract and the unchanged schist. All the facts point to some agency which could produce deep local corrosion, and the considerable leaching shown by the removal of iron oxide and by the Fic. 6—Curvature produced by pressure on a trans- ported slab. residues of white kaolin. The presence of the pegmatite dike across the middle of this tract, and its parallelism to the sharply t Stremme, Zis. f. prkt. Geol., XVI (1908), 128. i i i ; DECOMPOSITION OF ROCK BY CORROSIVE ACTION 53 defined border of the decay on the east, at once suggest its possible connection in some way with this chemical action. This might be referred to an attack of the schist by the magmatic vapors, “‘ the post-volcanic gas exhalations’’ of Weinschenk, accompanying the eruption of a dike of acid constitution, but for the entire absence of such effect in the vicinity of the tourmaline-bearing granite dikes which abound throughout this region. Taking all the facts here observed, we conclude that at this locality we find proofs, in the deep erosion, solution, and leaching, of the long-continued action of humus acids from peat-water, resulting in products which correspond to the ‘““Grauerde”’ of Ger- many, studied by Ramann, Wiist, Selle, Stremme, etc.* It seems probable that this deep local decay of both gneis- sic schists and the inclosed pegmatite records the continu- ous corrosion of an ancient pre-glacial peat-bog. The east- Fic. 7.—Showing how a bowlder was forced into one ern border of the bog of the transported slabs. appears to be marked by the sharply defined eastern limit of this decayed tract. The wall of impervious pegmatite may perhaps have formed a dam to confine the corrosive liquids, as in a vat, along this edge of the ancient bog; in such case the limit of corrosion would naturally lie parallel to the line of the dam. i With the prevalent tendency to attribute the formation of the original layer of: laterite over the northern part of our continent mainly or exclusively to weathering by meteoric agencies, there seems to have been little recognition of the view above suggested in explanation of the local instances of deeper decomposition of tH. Rosler, Zis. f. prkt. Geol., XVI (1908), 251-54. 54 EDWIN W. HUMPHREYS AND ALEXIS A. JULIEN crystalline rocks which have escaped the glacial scouring. It therefore may be added that we find abundant evidence of the wide distribution of tundra and peat-bogs all over this region, for a long period before the advance of the continental glacier as well as since its retreat. In the adjoining region, Westchester County, Mather recorded, sixty years ago, observations on peat-bogs, covering in the aggregate nearly 400 acres. Throughout the Bronx tract, in all directions around the locality we have described, we have noted many remnants of these, in street and house excavations, which have not yet been destroyed by the advances of the great city. ; These now vary widely in area and in depth. A few in- stances will be pre- sented to show that this form of chemical corrosion must have been here an active factor in degrading even the elevations of the rock-surface. Thus, along the low valley now oc- Frc. 8—Another bowlder that was forced into a slab. cupied by Morris Avenue, the depres- sion in the glaciated surface was formerly filled with peat, even now particularly well shown in the vicinity of 170th Street. In filling in this street with rock, the peat was forced up in places to a height of ten feet on each side, and its surface cracked in all direc- tions, revealing pockets of fresh-water shells. Thence the bog certainly stretched for a quarter of a mile, with a width of several hundred feet; while there is evidence of its former extension south- ward, probably as far as the Harlem River, and northward for an indeterminate, but long, distance. At 178th Street and Honeywell Avenue, a peat-bog yet remains and has recently been partially excavated, of which the original area, we estimate, must have occu- pied several hundred acres along the low valley. Its depth, as DECOMPOSITION OF ROCK BY CORROSIVE ACTION 55 proved by driven piles, reached, here, twenty-two feet. The bed of this formerly great swamp is now crossed by Daly Avenue, Honeywell Avenue, Southern Boulevard, Mapes Avenue, and Prospect Avenue. At Daly Avenue near Tremont Avenue, the depth of the peat was such that it was found necessary for foundations to drive piles forty-five feet in length. Not only were the lower grounds so filled up, especially the long valley depressions, such as those of the Bronx River, Eastchester Creek, Tibbit’s Creek, etc., but thin local sheets seem to have rested in the hollows among the rounded hummocks of the glaciated upland; these are in part still represented by little marshes, or the ponds in the various parks. It appears but a moderate estimate to assert that at least one-third of the surface of this region was once covered by an almost continuous sheet of fresh-water bog, out of which the higher elevations protruded as knobs of forest-covered rock. Along the adjoining coast at Hunt’s Point, Bartow, etc., these ancient bogs have been since overlaid, during the subsidence now in progress, by a sheet of salt meadow, surrounding a large num- ber of small scattered islets of now bare outcrops of gneiss and granite. Further evidence of the early and long activity of organic acids in solution, removal, concentration, and deposit of iron oxide from the surface of these rocks is afforded by numerous accumulations of bog iron ore once found throughout this region as well as over Manhattan Island. Though generally small, some of these were of sufficient volume to be of economic importance and use two hundred years ago. Escape of decayed schist from removal by the glacier.—There was here no knob or eminence on the northwest for the protection of the softened schists from the scour of the ice moving from that direction. On the contrary, a low valley lies on that side, which we presume was occupied by the peat-bog. The pegmatite itself, though softened, probably served long as the main protection of the schist, in connection with the pegmatitic branches and seams intercalated in the schist in this part of the tract. The next result- ing condition was apparently the erosion of this surface of the schist in an inclined plane, tending to lift the edge of the ice-sheet 56 EDWIN W. HUMPHREYS AND ALEXIS A. JULIEN up to the surface of the solid rock. The last phase appears to have been the plucking-up of huge thin slabs from a mass of thinly foliated granite, somewhere in the valley adjoining on the west, and their deposit as a ground-moraine over this inclined plane, with intervening sheets of bowlder clay, in a kind of natural masonry, for further protection of the underlying soft schists. Evidences and measure of the superincumbent pressure.—Soft as this granite is now found, it is obvious that it must have possessed ‘much strength and rigidity at the time of its transport by the ice in the form of slabs, mostly from a few inches up to a foot in thick- ness, although commonly ten to twenty feet or more in greatest extension. ‘The pressure upon them, as well as their rigidity, is shown by the frequent fractures and faulting, and the bending of thin edges. Still more significant is the crushing of the rock within the slabs at the contact with overlying bowlders of trap, which have been pressed down into pockets in the granite. In one case (Fig. 7) the bowlder appears to have been lifted subsequently some- what out of its pit and the clay forced in beneath it. In another (Fig. 8) the crushed granite rises around the imbedded bowlder, which was eighteen inches in diameter, as if the rock was almost plastic, either on account of the great pressure or of its own softened condition, or both. We had almost hoped to have found here a natural record of the weight of the superincumbent ice, and there- fore of its thickness, by estimating the volume of the granite crushed beneath the imbedded portion of the bowlder. This was found impracticable, from the impossibility of determining the crushing strength of the rock at the time of its penetration. However, we already possess some measure of the thickness of the ice-sheet in this region in the presence of glacial striae, often an inch in depth, at points 250 to 300 feet in elevation; e.g., on the edges of the gneiss over the summit of Inwood Heights, Manhattan Island, and on the trap along the edge of the Palisade escarpment, on the west side of the Hudson River. These imply a pressure which could hardly have been exerted by a sheet of ice less than 1,000 feet in thickness. ONDE, FOCUS, OF POSTGLACIAL UPLIFT NORTH OF THE GREAT LAKES J. W. SPENCER Washington, D.C. The first determination of the approximate location of the focus of postglacial uplift, based upon the amount of rise found in the beaches about Lake Ontario and Georgian Bay, appeared in a volume, to which access is difficult," and for this reason the passage may be cited, as the revision to be given below will be found in general conformity with the previous conclusions. If the axis of maximum elevation for the various triangles about Lake Ontario and Georgian Bay be produced, they meet near latitude 51° N., and longitude 743° W., a few miles west of Lake Mistassi and east of the southern end of James’ Bay. Although mainly radiating from the focus, the axes of maximum elevation for the different triangles are not uniform, and are locally modified, as along the western side of Lake Ontario, where there is found a secondary axis of uplift to the east. Combining the more western axes with those of the eastern end of the lake, another focus of uplift appears near the “Height of Land”’ between Lake Ontario and Hudson Bay, in about latitude 48° N., and longitude 76° W. From the double foci it may be inferred that the uplift reached its maximum along a line joining the foci, or that the axis of maximum regional uplift was meridional and located along the eastern end of Lake Ontario, increasing in amount until near the “Height of Land,” and thence with a diminishing ratio, or even depression, towards the north..... At any rate, it is in the region southeast of Hudson Bay that the maximum differential elevation of the earth’s crust, which involved the Iroquois beach, is to. be found.? ; Since that time (1889), Gilbert, De Geer, Taylor, and recently Goldthwaite have illustrated more or less fully the rise by isobars, which is only another mode of expressing the same phenomena, while Coleman has redetermined some of the triangles. Combining additional measurements obtained from Fairchild and Goldthwaite, I have recalculated the mean rise in the various triangles from the present heights of the beaches about Lake t Transactions of the Royal Society of Canada, VII, sec. iv, 189, read May 5, 1889. 2 J. W. Spencer, zbid., 189. 57 58 J. W. SPENCER Ontario and Georgian Bay, The postglacial bulge is like a sheet raised by an object thrust under it, the height increasing from out- ward from the focus of rise, but we can differentiate it in triangles Height 7] , of land Al FOCUS OR POST GLACIAL DEFORMATION as originally determined in 1887-88 and revised in |9|0 | ia Grand Bend Hamiltoq eal f°? lev and obtain the mean rate for each. These should cover the sur- veyed region, and be as nearly equilateral or right-angular as possible. Thus the following results have been obtained. The mean rate of rise from the lowest points in the triangles and the direction, based upon the height of the Iroquois beach, cover the region of Lake Ontario. iE FOCUS OF POSTGLACTAL UPLIFT 59 Triangles between Hamilton, Lewiston, and Scarboro,! 2 feet, N. 22° E. Lewiston, Rochester, Colborne, Ont., 2.5 feet, N. 17° E. Rochester, Sodus, Trenton, Ont., 3.6 feet, N. 10° E. Sodus, Rome, east of Watertown, N.Y., 5.5 feet, N. 3° E. These lines converge approximately in lat. N. 49° E., and long. 76° W. With the revised figures, the longitude is found to be the same, while the latitude is only 60 geographical miles north of the original determination, with the meridian of maximum uplift found to be just beyond the eastern end of Lake Ontario as origi- nally computed. Based upon the rise of the Algonquin beach east of Lake Huron, the mean rise in the triangles is found to be: Between Grand Bend, Southampton, and Rosedale, 1.3 feet per mile, N. 35° E. Holland Landing, Wyebridge, and Rosedale, 3.4 feet per mile, N. 23° E. Bradford, Owen Sound, Wyebridge, 3.1 feet, N. 27° E. Combining the former of these triangles with those about Lake Ontario, the lines of rise from Grand Bend, and from Holland Landing, converge to the same point as those from Lake Ontario, but if the mean rate for the third triangle (which takes in a more western equivalent) be used the focus will be in about lat. 48° N. Upham, Taylor, and Leverett have found the rate of rise in the region to the northwest to be of smaller amount, where Gold- thwaite suggests that the rise also somewhat coincides with the height of land as found by me farther east in 1888, where the maximum amount is computed at some 250 miles north of Ottawa City, or a few miles to the west of this meridian. The inferences to be drawn from these observations are: (1) that until a downward slope shall be found, we should conclude that the rise described continues to near the region indicated, beyond which the postglacial warping is downward; (2) that: eastward of the 76th meridian, for any assumed latitude, the post- glacial rise disappears and comes to be replaced by a downward slope. This is a question that has given the writer much solicitude, in an effort to determine the locus of downward warping in the 1 At a point 12 miles east of Toronto. Ii, in place of this, the elevation at Carl- ton (5 miles west of Toronto station) be taken, the line of rise is found to be N. 27° E. 60 J.W. SPENCER field. As the higher beaches have been found by all of us to indicate the greatest amount of warping, we should not expect to find a great amount in the altitudes of the Champlain marine deposits, but we see these recurring at so many points to about 500 feet above sea-level, that, upon examining their locations in the St. Lawrence Valley, it is noticeable that they occur along segments of the circle, roughly speaking, with the radii converging to the vicinity of the focus found; while in receding toward Gaspé, New Brunswick, and Nova Scotia the marine deposits rise only to lower and lower alti- tudes. These data are well known, so that an undue demand on the reader’s patience need not be made by their repetition here. So also with regard to most of the elevations on the Iroquois and Algonquin beaches. Again, the eastern equivalent of the warp- ing of the Iroquois, south of Lake Ontario, and in the Mohawk Valley, supports the hypothesis of declining postglacial warping in that direction, after passing the eastern end of Lake Ontario. The question of the location of the line of maximum post- glacial elevation radiating from the region mentioned raises several points in physical geography; one of these being the explanation of the cause of the rise, as due to the disappearance of the glaciers, for this locality was hardly the center of glacial dispersion. This idea, however, is here thrown out for others to consider. Since these notes were written, the admirable paper of Professor Goldthwaite on the “‘Isobases of the Algonquin and Iroquois Beaches”? has appeared. While the treatment of the postglacial warping by isobasic lines is scarcely other than a different mode from determining the mean rise in the various triangles, yet each has a significance of its own. The isobases indicate a regional rise toward the Laurentian axis. The triangles carried this rise to the “ Height of Land” and show the locus of maximum rise north of the Great Lakes. NOTE ON THE ACCOMPANYING Map The triangles about Lake Ontario are based on the instrumental measure- ments of the Iroquois beach, at the places on the map; those east of Lake Huron, on the measurements of the Algonquin beach. Height, at Rome and Sodus after Fairchild, of Holland Landing and Rosedale (in place of my Kirt- vill) after Goldthwaite. The other points are from my own surveys. A GEOLOGICAL ROUTE THROUGH CENTRAL ASIA MINOR FROM AFIUN KARA HISSAR VIA SIVRI HISSAR, ANGORA, SUNGURLU, AND THE MALYA TCHOL TO CAESAREA WILLIAM T. M. FORBES, PH.D. New Brunswick, N.J. The following paper is the summary of a series of notes taken in the summer of 1907, in connection with the Cornell Expedition to Asia Minor and the Assyro-Babylonian Orient. In a worked- over area like most of Europe and the United States, such a series of observations, probably somewhat inaccurate because of their hurried character, would add little to our knowledge. But in central Asia Minor the case is far different. Travel has been difficult and travelers are few. The men who have studied the geology of even a part of central Asia Minor could be numbered on one’s fingers, and but one or two of them had the advantage of being trained geologists used to the work and to the country, and with the leisure to stop and examine. For this reason it is that these notes represent new territory in practically their whole length. At certain points only did we cross (geologically) known territory. The observations were taken from horseback, as had to be done under the conditions of travel. It was rarely possible to stop to investigate a place or to visit again one that had been passed. This must have resulted in some errors, especially as the caravan could not carry any great weight of specimens. ' Frequently the rocks were fossiliferous, making their date certain, but unconformities were so frequent that it was not wholly safe to consider surrounding rocks as of the same date as the fossiliferous strata. Specimens were preserved wherever fossils were found (the majority were Eocene nummulites). These are deposited in the museum at Harvard and a study of them by 61 62 WILLIAM T. M. FORBES specialists would certainly result in a more accurate dating of many strata. Because of the peculiar conditions the description will usually follow the itinerary of the party, which was as follows: Afiun Kara Hissar, Phrygian Monuments, Aktash Kopri, Sivri Hissar, Gordium, Polatly, Hammam (Haimané), Giaour Kalesi, Angora (Enguri), Assi Yuzgad, Yakshy Khan, Izz-ed-Din, Sungurlu, Boghaz Koi—with a sidetrip to Eytik—Yuzgad, Medjidié, across the Malya Tchol to Bash; Hadji Bektash, Kara Burun, Avanos, Inje Su, and Caesarea (Kaisari). At this point the author was obliged to leave the party and hurry back to Constantinople. A few notes were taken from the train north of Afiun Kara Hissar which have been used to help out the map of that section. Distances are reckoned roughly in hours (at the rate of a walk- ing horse, three miles an hour), as that is the usual unit of measure- ment in the country. The principal geologist of Asia Minor was Tchihatcheff,? and his map remains the only one of much of the country. His work is now fifty years old and an experienced modern student would doubtless modify much of it. Still to the present time the man who has his books at his elbow will not miss very much of what is known of the geology of the eastern two-thirds of the territory. Yuzgad, and on an even grander scale, Caesarea, are in the center of regions which should prove exceedingly interesting. The complex resulting from several periods of igneous action makes a fascinating puzzle to disentangle. At these points I can add almost nothing to Tchihatcheff’s account, but can fully verify the existence of the confusion he reports. Because of the impossibility of determining the date of most tT have not been entirely consistent in my transliteration of native names of places, etc. There is no established method, and two books on the country will hardly agree in their methods. J and y (vowel), in particular, represent different sounds in Turkish, but they are not sharply defined and I should perhaps have used y more freely than I have. The distinction between g and & also generally represents a mere difference in Turkish spelling. Q might perhaps have been used more freely, following Arabic precedent. I follow the spelling of Tchihatcheff’s name as it appears on the title-page of his large work—in French. It is spelled differently in the German reports of his travels. ie GEOLOGICAL ROUTE THROUGH ASIA MINOR 63 of the rocks, they are classified on the maps rather from super- ficial characters. I have specially indicated the points where fossils have been found. The rocks may be grouped as igneous (of various sorts), metamorphic (largely Paleozoic where their relationships are known), obliquely stratified (Mesozoic and Eocene, especially the latter), and horizontally bedded sedimen- taries (Miocene and later as a rule). Thorough tracing-out of the relationships of strata and thorough collecting of the fossils can alone give a much more accurate knowledge of the dates of the various deposits of Asia Minor. In connection with the regular archaeological report of the expedition I expect to publish this matter in a less technical way and with reference rather to its interrelation with the various past peoples of Asia Minor and their culture. I wish to express my indebtedness to the members of the ex- pedition in many ways, and especially to Jesse E. Wrench, who did most of the topographic work; also to Professor J. B. Wood- worth, under whose direction and advice this report was prepared, and to the other authorities of Harvard University who have helped me in the matter of books, instruments, etc. MOUNTAINOUS PHRYGIA Comparatively few notes were taken in this district, and no specimens were collected. The substratum of the country is metamorphic, appearing as schists along the railroad cut between Ihsanié and Diiver (Deuyer), at the entrance to the mountainous section southwest of Ayaz In, and in smaller bands east of Yazili Kaya. There were also three outcrops in the Sakaria plain, one a considerable band at the eastern end of the Yazili Kaya lime- stones, and the others east and west of Aktash Kopri, as shown on the map. Quite as frequently the metamorphic rocks were limestones. This was the case along the railroad, north of the mapped area for a considerable distance, and also in a large area all about Yazili Kaya. Overlying the metamorphic rocks are everywhere igneous rocks, Neocene in date. These lie in horizontal beds, lavas, or tuffs, and are sometimes so rotted as to be indeterminable. Of the 64 WILLIAM T. M. FORBES same period also are several local deposits of sandstone and con- glomerate. These are cut by the railroad near Hammam and elsewhere; they form the basis of many of the sculptured rocks of the country. More frankly volcanic are the white tuffs of Ayaz In, and the lava mesas which make a dominant feature of the landscape east of the railroad, and all about Yazili Kaya. The lacustrine gravels of the Sakaria Valley approach quite close to Yazili Kaya on the east and mark the northeastern bound- ary of Mountainous Phrygia. As reported by Tchihatcheff and Hamilton the region south of the author’s route is of the same character, but the dominance of igneous rocks becomes less. The conspicuous volcanic necks of Afiun Kara Hissar are well described by Tchihatcheff and others. THE SIVRI HISSAR RANGE Passing over the lacustrine plains of the Sakaria River for the present, we reach the next point of interest in the Sivri Hissar Mountains, conspicuous among them Kodja Bel. At this place the stratified rocks would seem to belong to the same group as those in Phrygia, but the core of the range is a granite (a syenite in the popular sense of the word, as it is a fine-grained granite with little or no mica). East of Bala Hissar there is an area of limestone on the very top of the range, surrounded on both sides with the syenite, apparently lifted up on top of it. Near the city (Sivri Hissar) there is a complex of metamorphic rocks (schists and gneiss), through which the road passes on the east side of the mountains. Kodja Bel, a conspicuous peak southeast of the city, and the Kaimas peak to the northwest, seem to be similar. Tchihatcheff has reported on the northwestern part of the range; conditions are essentially the same, an alternation of syenite (granite) with various metamorphics. GORDIUM AND POLATLY The neighborhood of Polatly, unlike the preceding localities, has fossiliferous rocks, making it possible to fix this district as Eocene. Nummulites are the dominant feature, as elsewhere in Asia Minor eocenes. GEOLOGICAL ROUTE THROUGH ASIA MINOR 65 To the west of the Sakaria no Eocene rocks were found 7m sztu, but the cairns built by the shepherds are of fossiliferous limestone. / Fic. 1.—Sketch of the hills southeast of Gordium, interpreted as a laccolith. The stratified shales are indicated with fine oblique hatching; the lower trap is coarsely, and the upper trap finely, cross-hatched. The baked layer of the shales is shown solid black. Possible faults are indicated. Fic. 2.—Cross-section through ABC. The symbols are the same as in the pre- ceding figure. Probably outcrops of the rock occur. On the east side of the river the lacustrine plain is quite narrow, and is replaced by hilly country. This is composed, to the south of Polatly, largely of Eocene limestones, but farther north of light-colored (yellow or 66 WILLIAM T. M. FORBES greenish) shales, which have the appearance of hardly consolidated clays. There were also many outcrops of a dense igneous rock, necks southwest and east of Polatly, and sheets nearer to the village and to the northwest. ‘To the northwest, as the plan shows, the situation becomes quite complex; in the plan the strata are inter- preted as representing a laccolith, with overlying sedimentary rocks and flows, sloping away in at least three directions from its uncovered core. The eastern part was, unfortunately, passed over in the night, so that I cannot say whether the conditions were the same on that side or not. Overlying the sedimentary rocks was a sheet of lava. This had baked the clay red for the thick- ness of about a foot; making a very conspicuous layer. The red color was quite extensive toward the north, east of Gordium, so probably the trap sheet had once been much larger, but has been eroded off, leaving only the baked brick layer as a memento. At present, of course, these deposits can only be marked as “ prob- ably Eocene.” Hamilton reports similar mixtures of sedimentary and trap rocks north of Polatly, in the neighborhood of ‘‘ Begesch”’ oe djez or Beikos ?). HAIMANE Separated, at least in the line of our route, by a region of recent deposits, from the Polatly limestones and shales, there lies to the east the strikingly arranged Haimané district. In the immediate vicinity of Hammam no fossils have been found, though the rocks (shales) look promising enough. At Kaya Bashy there were plenty of shells in a limy conglomerate, apparently largely Anomias, but they were so much injured in transport that one can hardly determine whether the rocks belong with the eocenes of Polatly, or with the Jura and Lias which other authors have reported to the west of Angora. While Tchihatcheff, who has passed through the district at right angles to our route, considers them as probably Jurassic, I should incline rather to the other conclusion, especially as some of the Polatly nummulites were in quite similar-looking rocks. - At any rate, they are sedimentary deposits, obliquely banded, GEOLOGICAL ROUTE THROUGH ASIA MINOR 67 very brilliantly colored in red and yellow, and rarely obscured by plant cover, so that their bedding can be traced for long distances. No unconformities were noticed, but the point of transition between the shales immediately about Hammam and the calcareous con- glomerates and sandstones to the north was not seen. As reported by Tchihatcheff these deposits must be very extensive to the north of our route; in fact they and others similar dominate the formations of central Asia Minor. A gorge about an hour north of Hammam, near the village of Arif, passes through a mass of much denser limestone, which seemed to be conformable to the other deposits, but was very different in appearance. It is indicated by heavier hatching on the large map. Giaour Kalesi is close to the boundary between the series of sandstones just described and one of the great plains that make the type-landscape of central Anatolia. The boundary runs north- east and southwest, and was followed most of the way from Ham- mam Merkes to the Hohan Gol. The castle itself, however, is on a pinnacle of very different rock, more similar to those which wall the gorge at Arif and also to those at Angora and Assi Yuz- gad. It may then be of the same period as the surrounding rock, or with the Angora series, much older. There was no noticeable continuation of it through the surrounding rolling country. All the apparently earlier walls of- the castle were built of it. The stones were small and yet show no great signs of weathering, in marked contrast to the condition of Boghaz Koi, also built of its local marble during the same period of history. Less than an hour east of Giaour Kalesi is the village of Oyaja, built about the base of two trap necks, like’ a miniature Angora. These, or other similar outcrops, furnished the material for the later heaviest walls of Giaour Kalesi. Looking off from the top of Giaour Kalesi the hills seemed spotted with deep green, the characteristic mark a little farther east of the serpentines, and here doubtless due to the same cause. The spots seemed to have no regular arrangement and perhaps marked small volcanic necks, which, being soft, did not project above the general level. 68 WILLIAM T. M. FORBES THE ANGORA DISTRICT In the neighborhood of Angora I first came across the confused mass of rocks that seems to be typical of the igneous areas of Asia Minor. We stopped some time at Angora, and a day at Ortak6i, near by, giving rather more opportunity than usual to study the conditions. The series that leaves the strongest impression with one is a group of schists, extending roughly northeast and southwest, alternating between dark schists with hornblende or mica, sometimes very dense, and a very friable, whitish type, which seemed to have talc or sericite for its foundation (a snap- judgment, as there was of course no opportunity to go back). Neither of these types had the superficial appearance of stratified rock, but the relation of the two schists to each other and to the limestone of Elma Dagh convinces me that they were originally sedimentary. Tchihatcheff calls the whole system serpentine, and considers it igneous. They were apparently interrupted by a lava flow from Angora, southeast of the city where Tchihatcheff crossed the Elma Dagh, but the clay-slates, ‘‘Thonschiefer,’”’ on the road south from Angora, would seem to belong to the same bed; at least they have the same relation to the limestone. I crossed the entire width of the schists, going a few rods north, and three miles south, from Ortak6i. South of Angora the Thonschiefer were of about the same width. The marble was traversed in two places, and was also noted by Tchihatcheff about half-way between these two, giving a good idea of it. It seems to form the whole crest of the Elma Dagh and may extend quite a little farther at each end. Hamilton and Tchihatcheff’s notes would however seem to indicate that it is limited by Jurassic sandstones to the southwest, and apparently to the northeast also before reaching the Kyzyl Yrmak. Might the coalbeds reported near Kalejik, on the Kyzyl Yrmak, belong to the same system? To the north the schists were very soon cut off by igneous rocks of various kinds, but Hamilton reports both schists and limestones again north of them for some twenty miles northeast from Angora. The most typical of the igneous rocks are the necks which rise in Angora itself. These are of a reddish or purple trachytic GEOLOGICAL ROUTE THROUGH ASIA MINOR 69 DAMM GERGWYA : NX SS SSSUSSS RMX GSGCAAN X NY ASSSSs CESS Se y Ne LKR 55 e ey; Fic. 3 Wise SWS VEKC SS SEH EH NR RRS RS Sanne BRC SVs SS aN YS SS ia KOSS + WWOIQWZEy RROD RRQ co Kwong LOSERS XS ANOS Oe : SQ RRS SRS H SSK CRESS SARIS \ SSSSN AN SII SHUG XI N\ z Ey, ~ eee < < CBee GSA is Lr. ee \ JN = cas Nan I — (er Wwe a Gies u 7) = bed OG evee oa (Es ee ieiers = 2 eee Siac. [E. [BEN Ney Se SS SS a wie s VN SLO Oe See ee Sean eo oe ee Se SiS sass ce SF ESSE Cm KOT \ ro ~ ) SY | RQ SQA RGA WRK HMMA SVESSSS SSE SS TESS ESE EEN EE SES EH 70 WILLIAM T. M. FORBES porphyry.t The tuff which covers large areas east of the city, and makes the hill west of it, is made of the same volcanic rock. North of Ortakoi there isa long outcrop of the same type of rock, apparently here a lava flow, interbedded between two layers of late conglomerates, which in their turn have been made up of the schists, etc., of the region, and also of a fine-grained sandstone evidently not very old. Over the upper bed is a layer of white tuff, which one would naturally associate with the trachytes at Angora, and over this again a flow of very dark trap of indefinite extent. This trap would seem to make all the mountains to the north, at least for some distance. ‘There are also dykes of it cutting all the earlier beds, and necks of it north and northeast of Ortak6i. There is a small neck of the Angora trachyte also penetrating the limestone three miles south of Ortak6i. To sum up, there seem to have been the following periods of deposit: “(1) the system of limestones and schists which were probably metamorphosed and eroded before the next period; (2) the sandstone which formed an element of the conglomerates, and so must have had time to become consolidated before the date of the eruptions; (3) the eruptions of Angora trachyte, form- ing also the white, and porphyry tuffs (during lulls in this the con- glomerates north of Ortakoi were deposited by the precursor of the Enguri Su); (4) the period of the dark traps. Since the last there has been time for the whole lansdcape to be eroded down to its roots, leaving even the latest volcanic rocks as necks, and flows which have been tilted to decided angles. Almost a continuation of the Angora complex is the district about Kylyjlar. Just north of Assi Yuzgad there is a volcanic mass, apparently a sheet extending northward. Soon after reach- ing the hilltop the marbles about Assi Yuzgad, which have domi- nated since the last watershed west of the town, are in their turn replaced by an area of dense dark volcanic rock, mostly altered into serpentine, which extends, with various admixtures, almost * Bukowski has studied the igneous rocks of this district at some length. He finds the dominant rocks to be a variety of andesite, with quite a number of other igneous types, however. So probably the so-called trachyte and trap of the older geologists of Asia Minor should often be interpreted rather as andesite. He traces the extent of this igneous area to the north. See the Bibliography. PLATE I / 32° ; S] a /AFIUN KARA HISSAR 3 ANGORA. Based on FR. Kiepert’s Karle von Hleinasien, and Topographic Notes by Jesse E.Wrench. Geo- logical otes by W.7.M.Forbes FOR EXPLANATION OF SYMBOLS SEEAHE ANGORA-GASAREASHEET,/ EOCENE Y, Ye BORDIUM A Y, f IAS BR MOHAN GOL HAMMAM, EQCENE @) 3° Grotocy, Vor. XIX, No. 1 JouRNAL OF forth of Here. / NE METAMORPHIC Lt STONE METAMORPHIC Y LIMESTONE rs / 1 8 (2 regoooe. 2 OPCISGOG. LAR ER LE ‘3 A Prare I JURA, Se ae on oe TRACHYTE METAMORPHIC LIME STONE SYEMITE 4 MCUs Taine Canaan \ ee eH LACUSTRINE WA APerbee cal, ~~ 4 “AFIUN KARA HSSAR ANGORA. = / NS Based on PR. Kiepert’s Karte von Aleinasien, ane Topographic Notes by Jesse &. Wrencd, Geo- logical Motes by WIM Forbes, FOR EXPLANATION OF SYMBOLS SEE THE ANGORA*GAESAREA SHEET, ROCRNE OMAN a OL HAIMA vis MOCENE 5 Z Rae > 3 39 GEOLOGICAL ROUTE THROUGH ASIA MINOR He to the Kyzyl Yrmak. For the first couple of miles it is inter- rupted by several reappearances of the marble. Two of the hills south, also, are crowned by later horizontal beds. The dominance of the serpentines gives the whole country as far as one can see a deep green tint, varying in spots to pale green and to liver-red. After crossing a broad alluvial plain without outcrops, and then a narrow ridge of volcanic rock similar to the deposits to the west, we reach the immediate vicinity of Kylyjlar. (Up: YG Fic. 4.—Index map of Asia Minor. On the east slope of the last ridge west of Kylyjlar, there is a small outcrop of schist, dipping steeply to the northwest, and similar to that east of Angora. Underlying it is a small patch of syenite, not indicated on the map, and east of that again a flat- topped hill that dominates the whole valley. This hill seems to be formed of the serpentines, but is capped with a layer of the gray and white Angora limestone, apparently originally continu- ous with the beds east of the valley. East of Kylyjlar the domi-~ nant rock is still the same serpentine, but with the denser type less common, and more mixed than before with the greenish white tuffs (?). For a mile immediately east of the town, however, it 72 WILLIAM T. M. FORBES is clearly a tuff with large pieces of the dense volcanic rock as a foundation. Interbedded with this tuff there comes in quick succession a series all dipping northeast: in order going eastward, conglomerate, Angora limestone, tuff, limestone, tuff again. There is also a very conspicuous line of the pale-green rock running south behind Kyly- jlar, parallel to the valley. I did not find its contact with the stratified rocks, possibly a dyke of some kind. The marble appears in several other places between this district and the Kyzyl Yrmak, and also (but here the white crystalline type we found about Assi Yuzgad) in a prominent hill just across the Kyzyl Yrmak. Two miles east of Kylyjlar the serpentines east of the path are replaced by syenite, but they continue south of the syenite, and underlie the Neocene rocks for a distance farther. On leaving the syenites, now only a long mile from the river, we ourselves strike into the Neocene deposits that fill the bed of the river and extend indefinitely eastward. The neocenes are an alternation of conglomerate, sandstone, and a fine-grained rock: like pale- brown sugar (perhaps the “‘saccharine limestone” of Tchihatcheff). The latter is well exposed immediately around Yakshy Khan and seems to be the top layer of the series. THE SUNGURLU SERIES The section between the Kyzyl Yrmak and the Delidjé Yrmak is dominated by Eocene deposits. However, here and there igneous masses were seen, and they were probably numerous off of our route. Apparently serpentines form considerable outcrops south of Yakshy Khan, along the west bank of the Kyzyl Yrmak, and about a third way to Izz-ed-Din it again becomes probable that the distant rocks to the south are serpentines. An hour northeast of this last deposit there is a very conspicuous outcrop of syenite, cut from east to west by a crooked gorge which serves asaroad. Immediately north of Yaghly there is evidently another ‘igneous outcrop abruptly cut off to the south by a small brook, in a way that suggests the possibility of a fault running northwest and southeast. East of this line, almost as far as Eyiik, and from there south JoURNAL oF GEOLOGY, Eocene PLATE II Dn6esGog SANSA i LOCEOOM 36° | en ANGORA*C/ESAREA Based on R. Kiepert's "Harte yon Kleinasien," ana Topographic Noles by Jesse FE Wrench. Geo- 39 logical Notes by W.T.M Forbes. TRACHYTE EOCENE ie Si DOLERITE LACUS TRI(NE Ba Ie f ACA SAREA, TRACHYTE Journat or GroLocy, Vor. XIX, No. 1 Prare IT 40° os" a - j ANGORA*CASAREA Based on R.Kiepert’s “Karte yon e€inasien,” ana Topo Aa Notes by Jesse & Wranenaigees logical Notes by W.T.M Forbes. 36* METAMORP WIC LimesTgne fy GRANITE A. SERPENTINE a Wh DIORITE OIORITE LACUSTRINE / Vi JS y a, } , \ NS METAMORPHIC TRACHYTE y, LIMESTONE Vy, ~ \ BQCENE EOCENE ete, / DOLERITE TRACHYTE, NS / LACUSTRINE ME)\DIE re SG : YO TX CASAREA, Ws EOCENE He EXPLANATION ¢ SYMBOLS Superficial Deposits. Neoce nes. Wor/zontally Bedded). = (UM) Zt | Metamorphic Limestones Asst Yurgad type of Limestones. Schists. KARABURUN Toachy lesyard Similar Up A lgneédus Rocks. Y psi ZG Mixture of lgneous Racks, Tiffs,and Facene ana Neocene Deposits. A Mesozoic and Eocene LL) Deposits (Mastly cal- careous sandstones 5 ui with steep dips) lerpentine. NJ fossiliferous Strata . ; NG of thesame typo. Co Undetermined Deposits 2 -~-. Author's Route. . Ne IN THIS STYLE OF LETTER Deposits Noted by other Authors. IN THIS STYLE OF LETTER Jdontified Fossils farous Strata. Syenite ana Granite Je TRACHYTE LAcus TRINE ROCENE MINA tal, GEOLOGICAL ROUTE THROUGH ASIA MINOR 73 to Yuzgad, rocks of Eocene facies were constantly in sight, the immediate vicinity of Boghaz K6i making the only serious inter- ruption. In the immediate vicinity of Aktché K6yiin our route took us out of this Eocene area into the lacustrine plain which so often accompanies the larger rivers of Anatolia. The rocks of Eocene facies dip at various angles, and do not seem to be entirely conformable. However, as in unquestionably conformable series in that vicinity the dips and strikes often change very abruptly, I should not dare to say that more than one period is represented. One may take, for instance, the case sketched here. This is a frontal view of a bluff, past which the Fic. 5.—Frontal view of a bluff a short distance east of Alembeyli, showing very rapid change in dip of the strata. The stippled layers represent gypsum; the rest is sandstone. The road passes through the right-hand depression. (Somewhat dia- grammatic.) road ran (through the central gap, in fact). The entire part of the ridge shown in the figure was but a few rods long, and as one can see, the dip of the strata has changed considerably. Actually aside from this anticlinal structure the whole series dipped away moderately. At this particular point, near Alembeyli, there was some tendency for the dips to be moderate and easterly. South- west of Aktché K6yiin the dips of the nearer strata were about the same, but some strata were seen dipping at angles of over 45°. As already mentioned the two sets appeared unconformable but this may have been because of inability to see the intermediate rocks. At Sungurlu, at the east end of the village, there is a thin bed of plant remains, but they proved too fragile for transport. BOGHAZ KOI The dominant rock at Boghaz KGi is a marble breccia, with a bright red cementing material, making a striking pattern. Occa- sionally the marble is more massive, and then may appear either 74 WILLIAM T. M. FORBES like the Angora or the Assi Yuzgad types, showing that the latter two can well be of the same date. _ Mixed all through with these marbles is a serpentine, in which, wherever the outcrops are clear, the marble appears to be floating Serpentine RQY PateGreen S. ATA Tuff Morble \w Sandstone EE} Neocene. Ke Ne Route KYO ORR AR RRA, SELES : ——- Other Roads iy KK ‘ZB iY, is : best) > YAR? Ze, ZA Fic. 6 in large blocks. The serpentine is less in evidence immediately under the ruins at Boghaz K@i, but even there the excavations have exposed it enough to suggest that the arrangement is the same. At the point where the dominant limestones give way to ser- pentines, as one goes south from the village and the ruins, there «is an east-and-west bed of a dense siliceous rock that also appears GEOLOGICAL ROUTE THROUGH ASIA MINOR 75 a second time farther south, interbedded with sandstones. In both cases the dip is northeast. Farther east and up the main valley as far north as Devret the rocks and their arrangement are different. Here there are two bands of a dark tuff, with fragments of vesicular trap, imbedded in sandstone, and the whole dipping to the east. Where we went, Emirler. Fic. 7.—Sketch of a small area near Emirler, just north of Boghaz K6i, showing the relation between limestone and serpentine usual in that district. The hatched areas are limestone; the remainder is probably mostly or entirely serpentine, but largely covered up. : as shown on the map, it seems to disappear within a short distance under the serpentines. But it probably extends farther to the southeast and east, through the area left white on the map. North-northeast of Boghaz Koi, the mixture of serpentine and marble continues half-way to Eyiik, the serpentine dominating for the southern half and disappearing in the northern half. About at the point where the serpentine ceases to appear in any quantity there are several outcrops of denser igneous rocks. The space immediately south of Eyiik is occupied by fine-grained Neocene limestones, containing plant remains in poor condition. A little 76 WILLIAM T. M. FORBES farther west, and more directly north of Boghaz KGi, there is no sign of the limestone or serpentine, but the Eocene (?) rocks which extend west to Yaghly take their place. THE YUZGAD COMPLEX Five or six miles south of Boghaz K6i the serpentines and limestones again make way for the rocks of Eocene facies, but now they are much interrupted by igneous rocks in great variety. Our road, the new chaussée, ran at first southeast, over the divide. Then we entered the valley of a large stream which flows off to the southwest and is followed by the old chaussée. We went upstream (to the east) for a couple of miles along one of its tributaries, after coming down from the divide along another; and then we turned abruptly southeast to cross the very top of Kabak Tepé, the moun- tain just north of Yuzgad, by a very complex system of zigzags. From the height of land till we left the main valley to climb Kabak Tepé, and actually till we were fairly up on the slopes of Kabak Tepé, the supposed eocenes make up the mass of the rock. South- west of the road about a mile south of the divide there could be seen a flat mesa which has been used for ‘‘cliff-dwellings.”’ It is probably a tuff or soft trap like the beds so used elsewhere, and made a distinctly incongruous note among the other rocks of the district. In any case it is Neocene in date, and unconformable on the local bedrock. Northeast of the road in the same neighbor- hood there are several appearances of granite (probably more nearly of the date of the bedrock). At the point of forking of the old and new chaussées, where the road ceases to go south down one tributary and turns east up the other, there are a couple of trap dykes, both small, but perhaps outliers from more extensive intrusions to the north. A couple of miles south of the valley of the tributary running from east to west the supposed eocenes either disappear or else change their character entirely under the influence of the many igneous rocks, which now become dominant. Of this district one can only say that it is a practically inextricable tangle. It is com- posed, among other rocks, of granites, dark traps, schists, frag- ments of Eocene beds (some containing fossils according to Tchi- GEOLOGICAL ROUTE THROUGH ASIA MINOR Te hatcheff), tuffs, and Neocene sandstones and conglomerates. Im- mediately about the town there are several outcrops of tuff. Hamilton and Tchihatcheff report the same types as making the entire region west to near the Delidjé Yrmak, and northeast for an equal distance. To the southwest, however, after about ten miles they gradually give place to neocenes, which extend in the main to Hadji Bektash. Tchihatcheff discusses this complex with the dolerites. Getchi Kalesi, the mountain to the east of Medjidié, in the northern part of the Malya Tchol, is only the culminating point in a limestone range, cored with trap, which extends from west of Medjidié southeast for a dozen miles. The limestones at several points contain Eocene nummulites in quantity. At the point where the road traverses this series, south of Medjidié, the eocenes are not conspicuous, and the traps are locally interrupted, but behind the city, the traps stand up in a series of prominent and ragged hills. Even here I have a feeling that the igneous dyke is not entirely continuous. A little farther south, and on the other side of the path, the igneous rocks appear again in a couple of amorphous masses (as seen from a distance), the eocenes remain- ing inconspicuous, but Getchi Kalesi itself is of a somewhat differ- ent structure. The dyke here seems to be fairly continuous, and in general makes the crest of the ridge. Leaning against the west side of this is a long series of Eocene limestones, etc., all with dips of about 60° to the north. Apparently on the east side of the dyke the same facing occurs, and the very highest point of all is formed by one of these strata which is continuous across the top of the dyke. This very topmost point furnished one of my num- mulite specimens, through the kindness of Mr. Wrench. A little farther south the village of Mahmatly is situated in a very striking gorge, which marks the boundary between a lower and a higher level of the Malya Tchél. On the steep sides of this gorge, as well as the escarpments that lead up to its mouth, the neocenes are interrupted, laying bare the substratum of the Tchél, which is evidently of the same system as Getchi Kalesi mountain. Half a day’s journey farther south there is a long hill a moderate distance west of the road, which again shows the 78 WILLIAM T. M. FORBES steeply dipping strata of the Eocene series. At one place Num- mulites levigata (Lutétien period of the Eocene) was picked up, but not 7m situ. + 2 e, O << < o< = <= vehi 3 yr Castle §& Oe < SSS KKK — Sx SS an es YS Fic. 8.—Kara Burun and immediate vicinity. The trap is indicated by hori- zontal lines, the granite by coarse cross-hatching. Gardens have been shown in stipple, indicating the position of springs. There are also some smaller igneous outcrops in the neighbor- hood of Medjidié, which do not seem to belong to the Getchi Kalesi range, in particular a neck of very coarse porphyry some five miles northeast of that town and on the other side of the river. GEOLOGICAL ROUTE THROUGH ASIA MINOR 79 The district between the Malya Tchél and the Kyzyl Yrmak valley to the south is again apparently eocene in date, resembling closely in appearance the Haimané and the vicinity of Sungurlu. The very top of the divide south of Hadji Bektash showed no out- crops, but the pebbles brought down and the appearance of the distant hills would imply that it also had a volcanic core. It isa very much more insignificant ridge than Kiepert’s map would . suggest. KARA BURUN The village of Kara Burun is located on the east slope of a mesa capped with a sheet of hard black trap. This sheet disappears abruptly at the north end of the village against a steep bluff of much-rotted granite, which in its turn is capped with a second sheet of trap exactly similar to the first, but on a higher level. The lower level trap, like the upper, seems underlaid with granite. Kast of the upper level, the granite is laid bare in several places, but east of the lower level there are no near outcrops. On the south boundary of the granite outcrop there is a line of springs marked by gardens and villages, of which the first is Kara Burun itself. The whole, with the line between the upper and the lower Kara Burun traps and the southern boundary of the Eocene deposits farther to the east, forms a line nearly parallel to the Kyzyl Yrmak river which I have interpreted as possibly a fault. East of Kara Burun the Kyzyl Yrmak valley, as far at least as Avanos, is filled with a series of almost horizontally bedded neo- cenes, more or less tufaceous, which gradually rise as one goes east. They are cut off to the south by the valley of the river, and seem on the north to end abruptly against the eocenes. Farther to the east, as one approaches Avanos, the eocenes appear from below and the later deposits make only a narrow cornice against the bluff. Some of this series of beds are more or less water-worn conglomer- ates, while others are fine-grained tuffs of very even texture. The latter especially have been much used by the troglodytes for excavating houses, churches, and tombs. South of the river one can see a great confusion of lava-sheets, the spaces between which are taken up by vast masses of tuff. Occasionally the tufaceous matter would become less noticeable, 80 WILLIAM T. M. FORBES and they would grade into the usual Neocene conglomerates. The trap-sheets hardly appear north of the river, except at Kara Burun. Several miles west of Inje Su there is a perfectly flat plain, formed by the vesicular surface of one of the trap-sheets. Nearer to Inje Su itself a stream has cut a deep gorge in this bed, exposing the underlying tuff. The district between Avanos and Inje Su is the famous troglo- dyte country, which also extends a long distance to the south, to the west of Mount Argaeus. In the neighborhood of Urgiib there had evidently been at one time a thick layer of fine homogeneous tuff, capped with a thin trap-sheet, which though harder than the tuff was itself easily weathered and cracked into blocks. Erosion has cut this whole district into a mass of cones of tuff, the higher ones of which are still capped with small blocks of trap. Between these higher ones there are a vast number of shorter cones, whose lava caps have fallen off, and which are fast being eroded away. When the cap falls off it sometimes finds new lodgment at a lower level, and becomes the nucleus of a new shorter cone. Hundreds of the cones have been used by the troglodytes for excavating houses, and many of these are still in use. Where the country is at a little higher level, at Urgiib village, the country is not broken up into separate cones, but there is a large mass of tuff, crowned by a continuous sheet, and terminated to the north with a con- tinuous cliff. The village was originally a system of troglodyte houses excavated in the face of this cliff, but most of the houses have added built facades in more recent times. It is still distinctly a troglodyte village nevertheless. Beyond Inje Su notes were not taken, but the general character of the country does not change. Tchihatcheff spent considerable time in this district, and gives a long and interesting account of it in the section ‘‘trachytes”’ of his geology of Asia Minor. THE LACUSTRINES In this survey I have passed over several sections of the route with hardly a word. These are occupied by the characteristic Neocene (lacustrine) deposits which seem to cover nearly half the surface of Anatolia. They are in general horizontally bedded or GEOLOGICAL ROUTE THROUGH ASIA MINOR SI nearly so, sometimes fossiliferous—then generally Pliocene—often formed of the same materials as the eocenes of their district. Still more often they show a more or less characteristic appear- ance, and may usually be distinguished by their horizontal bedding. In every place where they come in contact with the trachyte (andesite) deposits, they grade into their tuffs, and are evidently in a general way of the same period. This shows conspicuously at Yuzgad and along the Kyzyl Yrmak near Caesarea. Here is a list of the places where I found deposits of this type to predominate: The Sakaria valley near Aktash, and from there to Sivri Hissar. The Sakaria valley from Sivri Hissar east to the river at Gordium. The country east of Hammam Merkes and Giaour Kalesi. The Kyzyl Yrmak valley from Yakshy Khan to Yaghly. The region about Eyiik. The region beginning just south of Yuzgad and extending the entire length of the Malya Tchol almost to Hadji Bektash. BIBLIOGRAPHY Congres géologique internationale. Compte rendu de la [X® session, Vienne, 1903. This contains two important papers on the Bibliography of Asia Minor, so complete that it seems unnecessary to give a detailed bibliography here. These are: Touta, Franz. Der gegenwirtige Stand der geologischen Erforschung der Balkan-Halbinsel und des Orients, p. 175; followed by— Touta, Franz. Ubersicht iiber die geologische Literatur der Balkanhalbinsel mit Morea, des Archipels mit Creta und Cypern, der Halbinsel Anatolien, Syrien und Paliastinas, pp. 185 to 330. This is a bibliography of over 1,300 titles, arranged chronologically. von Bukowski. GerjzA: Neuere Fortschritte in der Kenntnis der Strati- graphie von Kleinasien. Loc. cit., p. 393. A bibliography arranged by authors. Hamitton, WILLIAM JOHN. Researches in Asia Minor. 2 vols. London, 1842. A book of travels with frequent geological notes. The geological part, as well as that of the other older authors, has been summarized by Tchihatcheff, and harmonized with his own observations. | TCHIHATCHEFF, PAuL DE. Asie Mineure. Paris, 1866 to 1869. Paléontologie by A. d’Archiac, P. Fischer, and E. de Verneuil, in one vol., with atlas. Asie Mineure. Géologie, in 3 vols., with a geological map of Asia Minor, and one of the Bosporus. The classic. Asie Mineure. There are numerous other papers by Tchihatcheff, for which one may consult the bibliographies cited above. 82 WILLIAM T. M. FORBES SCHAFFER, F. Cilicia. In Petermanns Mittheilungen, Erganzungsheft 141, 1903. It contains a geological map of Cilicia and the neighboring district, which adjoins the route of our studies on the south. LAPPARENT, A. DE. ‘Traité de géologie. Ve. édition, 1906. Summarizes the stratigraphical knowledge of Asia Minor along with the rest of the world. Lronuarp, R. Geologische Skizze des galatischen Andesitgebiets nordlich von Angora. Neues Jahrbuch fiir Mineralogie, etc. Beilageband XVI, 99 to 109, with a sketch map. p’Arcutac, A AND J. Hamme. Description des animaux fossiles du groupe num- mulitique de l’Inde, precédée d’un résumé géologique et d’une monographie des nummulites. Paris, 1853. A useful reference book on the nummulites. Witson, Str CHARLES. Handbook for travelers in Asia Minor, Transcaucasia, Persia, etc. (Murray’s Handbook). The most convenient book for geographic information. Kierert, RicHArD. Karte von Kleinasien. 1:400,000. Berlin, published in sheets. The standard map of the country. THE VARIATIONS OF GLACIERS. XV" HARRY FIELDING REID Johns Hopkins University The following is a summary of the Fourteenth Annual Report of the International Committee on Glaciers.” REPORT OF GLACIERS FOR 1908 Swiss Alps.—Of the ninety glaciers which were measured in 1908, fifty-three are in undoubted or probable retreat, one is certainly advancing, and thirteen are possibly advancing. The retreat, therefore, is general. Certain small glaciers have for some years shown signs, more or less definite, of advance. Short glaciers respond more quickly than long ones to the changes in snow-fall, and may make a number of small variations which are not indicated by large glaciers.’ Eastern Alps.—A large part of the observations on the glaciers of the Eastern Alps were carried out under the auspices and at the expense of the German and Austrian Alpine Club. The general retreat was dominant between 1907 and 1908, as it has been for several years past. Only a single glacier, the Wansee- ferner, in the Oztetal, has advanced; its advance amounted to 15 meters. The other glaciers showed retreats amounting in some cases tO 22, meters:? Italian Alps.—The observations of the Italian glaciers were all the results of private enterprise. All the glaciers observed on the south side of the Alps were apparently in retreat, except possibly a few in the Maritime Alps, which, seen from a distance, had appar- ently enlarged slightly; but this observation is doubtful.s French Alps.—Many observations on the snow-fall and varia- t The earlier reports appeared in the Journal of Geology, Vols. III-XVII. 2 Zeitschrift fiir Gletscherkunde (1910), IV, 161-76. . 3 Report of Professor Forel and M. Muret. 4 Report of Professor Briickner. 5 Report of Professor Marinelli. 83 84 HARRY FIELDING REID tions of glaciers were made under the direction of the Minister of Agriculture. During the winter of 1907-8, the amount of snow-fall was regis- tered in twenty-seven stations in Savoy; although the snow-fall was less than in the previous year, still the total amount was fairly large. The maximum does not increase with the altitude, as the altitude becomes high; but the observations along this line are still incomplete. On Mont Blanc the Glacier de Bionnassay, which between 1906 and 1907 had advanced 38 meters, in 1907-8 advanced 17.5 meters farther. The three other glaciers observed in this region had all retreated.’ The retreat of the Glacier du Tour amounted to 52.5 meters. In the Maurienne three small contiguous glaciers made a slight ad- vance; a fourth glacier, not far distant, had retreated markedly." Pyrenees.—A number of glaciers observed in this region showed a tendency to a slight advance.’ Swedish Alps—The retreat of the glaciers of Lapland, noticed since 1900, has been confirmed. During the summer of 1908 sig- nals were placed near many glaciers which will lead to more definite results in the future. Norwegian Alps.—The glaciers of the Jotenheim are in marked retreat, whereas those of the Jostedal and Folgenfon, nearer the coast, show an equally marked advance, the variations between 1907 and 1908 amounting in some cases to about 30 meters.’ Canada.—The [Illecillewaet Glacier retreated about 4o feet between 1907 and 1908.3 Himalaya.—Although no new observations have been made, there are indications that since the survey of 1872-75 the glaciers of Garhwal and Kashmir have retreated considerably.! REPORT OF THE GLACIERS OF THE UNITED STATES FOR 1909° Hallett Glacier, Colorado, shows no change between 1908 and tgo09 (Mills). ™ Report of M. Rabot. 3 Report of Mr. Vaux. 2 Report of M. Oyen. 4 Report of Mr. Freshfield. 5 A synopsis of this report will appear in the Fifteenth Annual Report of the Inter- national Committee. The report on the glaciers of the United States for the year 1908 was given in this Journal (XVII, No. 7, pp. 667-71). THE VARIATIONS OF GLACIERS 85 The glaciers of Mt. Hood, Oregon, show a marked recession since 1906 and they have also decreased in thickness. The White Glacier has receded about 400 meters; Sandy Glacier 100 to 200 meters; Reid Glacier 50 to 100 meters, and Zig-Zag is hardly more than an ice-bank. The glaciers on the north side of the mountain, as seen from the summit, also seemed reduced in size (Montgomery). Lyman Glacier, near Lake Chelan in central Washington, is still diminishing (Rusk). Mt. Baker, Washington, the most northerly of the great vol- canic cones which rise above the Cascade Range, was surveyed during the summer of 1909 by the United Stated Geological Sur- vey, the party being under the direction of Mr. J. E. Blackburn. Mt. Baker, 10,745 feet high, is covered with ice and snow above an altitude-of 5,000 feet, divided into a few separate masses by narrow ridges of rock. At the lower levels glacier tongues develop, some of which extend to as low an altitude as 3,500 feet. Seven glaciers have been given names, the Roosevelt, Mazama, Wells Creek, Sholes, Park, Boulder Creek, and Nooksack. The last has not been fully explored and may later be divided and receive several names. The ice of these glaciers is especially broken up by large and numer- ous crevasses. The crater of the mountain, from which steam is still escaping, is about 1,000 feet below the flat snow-covered sum- mit. The moraines in front of the glaciers’ ends and the polished and grooved rock along their sides show clearly that they are retreating and d:minishing in volume. Fifteen miles northeast of Mt. Baker rises the spire of Mt. Shuksan, 9,038 feet; a glacier on ‘ts wes:ern side breaks over a cliff and the ice collects to form a reconstructed glacier at a lower level; this glacier then falls over a second cliff and forms a second reconstructed glacier still lower down (Blackburn). The United States Coast and Geodetic Survey has published a map of Glacier Bay, and the surrounding area, on a scale of 1/160,- ooo, from surveys made in 1907.’ It is interesting to compare this map with the earlier one published by the survey in 1899.2, The t Tt is numbered 8306 and was published in January, rgro. 2 No. 3095. 86 HARRY FIELDING REID latter was based on surveys made by Reid in 1892, with additions taken from the surveys of the United States Coast and Geodetic Survey and of the Canadian Boundary Commission between 1884 and 1895. At the first glance one is struck by the smaller area covered by the ice, and the correspondingly greater area of bare rock; for instance, the ridge between Casement and McBride glaciers was broken in the earlier map by many arms of ice con- necting the two glaciers; the later map shows this ridge as continuous and much broadened. Similar changes are noted in other parts of the map. This indicates not merely the melting of small connecting arms of ice, but also a general lowering of the whole surface of the ice. Dying Glacier at the head of Tidal Inlet has entirely disappeared, and Dirt Glacier, immediately east of Muir Inlet, is not represented on the later map. I am inclined to think that this glacier has not completely melted, but that its very thick covering of moraine has masked its character. Large areas of rock are free of the ice which covered them in 1892. The ends of the tide-water glaciers have receded greatly (as noted in earlier reports of this series) and allowed the inlets to penetrate farther into the land. The end of Muir Glacier has receded and divided into two parts, separated by the rocky island which appeared as two dis- tinct nunataks in 1892 about 3 miles from the ice-front. To the north the glacier has receded 8.5 miles. The ice surrounds the water on three sides; bergs are discharged most actively at the northern end of the inlet. To the east the glacier has receded 3 miles and ends in a sloping surface just reaching- the water. (Since 1907 this portion has receded still farther and now rests on a sandy beach, where it is forming a terminal moraine Dunann.) The total increase in the area of the inlet between 1892 and 1907 was Ig square miles. Carroll Glacier does not seem to have receded, but the Rendu has retreated about half a mile. Grand Pacific Glacier has receded 73 miles, almost as much as the Muir, and its inlet has increased by 14 square miles. Johns Hopkins has receded 3 miles, increasing its inlet by 55 square miles, and separating from one of its southern THE VARIATIONS OF GLACIERS 87 tributaries, which becomes an independent tide-water glacier. Reid Glacier seems to have receded about 3 mile. In 1892 Hugh Miller Glacier attained tide-level at two termini; one, on the north, barely reached the water and had a sloping surface; this has retreated about half a mile. The other terminus, on the east, was divided by a rocky mass, north of which the ice re- sembled the northern terminus but south of which it ended in a cliff discharging bergs. ‘The northern part of this terminus has receded about one mile and has uncovered much rock, about 1% square miles; the southern part has receded about 13 miles and the inlet has increased by about 2 square miles. Charpentier Glacier has receded about 13 miles and its inlet has increased by one square mile. Geikie Glacier has receded about ? mile and Wood Glacier has greatly diminished in size, though it still seems to reach tide- water as in 1892 without an ice cliff. The total increase in the area of Glacier Bay, as the result of the recession of the glaciers, amounts to about 50 square miles. Professor Ralph S. Tarr has published a detailed account of the Yakutat Bay Glaciers, with many illustrations and maps, which includes all information regarding these glaciers available at the end of 1906.1. The remarkable advance of some of these glaciers in the interval between Professor Tarr’s visits to them in 1905 and 1906 are carefully considered and ascribed to extraordi- nary supplies of snow shaken down from the mountains by earth- quakes in 1899.2 This very excellent monograph can receive only a cursory notice here. Professors Tarr and Lawrence Martin organized an expedition under the auspices of the National Geo- graphic Society to revisit Yakutat Bay and Prince William Sound in 1909. Professor Martin has sent me the following outlines of the results of this expedition: The National Geographic Society’s Alaskan Expedition of 1909 in charge of R. S. Tarr and Lawrence Martin observed the following variations of glaciers. In Yakutat Bay Hubbard Glacier seemed to be beginning to advance more ™“The Yakutat Bay Region, Alaska,” U.S. Geological Survey, Professional Paper No. 64, Washington, 1909. 2 Mentioned in an earlier report of this series (this Journal [1908], XVI, 54-55). 88 HARRY FIELDING REID rapidly; Lucia Glacier was advancing rapidly and overriding a nunatak after semi-stagnation since before 1890; Hidden Glacier had advanced 3 kms. in less than 3 years and had returned to semi-stagnation; Nunatak Glacier was continuing the retreat in progress since 1890, having retreated over { km. since 1906 or nearly 53 kms. since 1895; Turner Glacier had advanced slightly since 1906. The Variegated, Haenke, Atrevida, and the Marvine lobe of Malaspina Glacier had ceased the spasmodic advance which Tarr observed in 1906 and explained, not as climatic, but as part of a glacier flood due to earthquake avalanching. Haenke Glacier, which advanced and became tidal between September, 1905 and June, 1906, had retreated before 1909 so that it no longer discharged icebergs, being fronted by a low gravel cliff. It was once more mantled with ablation moraine, as were large parts of Varie- gated and Atrevida glaciers and the Marvine lobe of Malaspina Glacier. Our party easily crossed Variegated and Atrevida Glaciers in 1909 in the parts most impassably crevassed in 1906. The advance of three additional glaciers between 1906 and 1909 and the quick return to semi-stagnation in 1909 of the four that were rapidly advancing in 1906 gives additional proof of the earthquake-avalanche hypothesis for certain variations of mountain glaciers. On the lower Copper River the Miles, Childs, and Baird glaciers were, in r90g, in about the same conditions as when they were seen by Abercrombie in 1884, by Allen in 1885, by Hayes in 1891, and by Schrader in 1900. Parts of Miles and Baird glaciers have been stagnant and forest-covered for at least twenty-five years. Five miles of railway track has been laid on Baird Glacier. Childs Glacier seems to be advancing and forcing Copper River eastward, according to Johnson. The rate of movement near its northern margin in July, 1909 was about 4 feet a day. During the last half of July, 1900, abla- tion lowered the surface of Childs Glacier at the rate of 7 inches a day. In eastern Prince William Sound, Valdez Glacier is retreating, as it has been since 1898 excepting the slight advance between 1905 and 1908 recorded by Grant. Shoup Glacier has been retreating since 1898 except for a slight advance, perhaps, in the spring of 1909. Columbia Glacier was continuing the advance observed by Grant in 1908 and early in July, 1909. The eastern margin had advanced, before August, 1909, making a decided lobation, but not reaching the forest along the whole margin. The western margin had advanced more than 800 feet up to the forest of Gilbert’s maximum of 1892, as was also the case at Heather Island where the middle of the glacier was destroying the forest in August, 1909. The United States Geological Survey has published a bulle- tin’ containing a short account of the glaciers of the Wrangell «F, H. Moffit and Adolph Knopf, ‘‘The Mineral Resources of the Nabesna- White River District, Alaska, with a Section on the Quaternary by S. R. Capps,” U.S. Geol. Survey, Bull. No. 417, Washington, 1910. THE VARIATIONS OF GLACIERS 89 Mountains, Alaska, and a topographic map of the region; and from it we draw the following information: A very important feature of the Wrangell Mountains is the great ice cap ‘that occupies the crest of the range and that has its greatest development in the region around Mount Wrangell. From the periphery of this great feed- ing- ground valley glaciers extend in all directions down the more important drainage lines. The Nabesna and the Chisana are by far the largest of these glaciers. The former is about 55 miles long and has an area of about 400 square miles. The latter is 30 miles long with an area of 135 square miles. There are many smaller ice tongues, and even small glaciers independent of the main ice cap. The St. Elias Mountains, south of White River, are snow-capped in much the same way as the Wrangell Mountains. Most of the mountain range is unexplored, however, and the extent and area of the ice field is unknown. All the more important tributary valleys to the north are occupied by valley glaciers, the largest and best known of which is the Russell Glacier, at the head of White River. The main lobe of ice in the head of the White Valley is between 6 and 7 miles long and about 23 miles wide, and most of the ice moves in a northeast direction. A small crescentic lobe, however, moves westward into the head of Skolai Creek. Formerly the glaciation was much more extensive, but very little information is available to determine what changes are taking place at present. In 1891 the western terminus of Russell Glacier was a smooth slope, but in 1909 it was a wall of ice from 25 to 75 feet high. This certainly indicates an advance of the ice, but at the northeastern terminus the ice passes into the moraine without a clear line of demarkation, indicating a slow, gradual retreat. The Nizana Glacier was formerly crossed by prospectors going to the White River region, but it has become so crevassed as to be practically impassable, which suggests an- advance of the ice. REVIEWS Testing for Metallurgical Processes. By James A. Barr. San Francisco: The Mining and Scientific Press; London: The Mining Magazine, 1910. Pp. 216. $2.00 delivered. This book, which is based on a course of lectures given by Mr. Barr at the Michigan College of Mines, is a laboratory manual for the student of metallurgy and for the mining engineer. The treatment differs from that of the textbooks on metallurgy in that the methods for testing are fully treated and minute details for many of the operations are given. It is designed not to take the place of the textbooks on metallurgy but to supplement them. The subjects treated include amalgamation, chlorination, cyaniding, concentration, smelting, calculation of lead and copper slags, cost data, etc. The treatment, while condensed, is exceptionally clear. The work should be appreciated by students, mining chemists, and engineers. W. H. E. Economic Theory with Special Reference to the United States. By HerinricH Ries. 3d ed. New York: Macmillan, 1910. Pp. 589. The third edition of this work is revised and greatly amplified. The treatment of the non-metallic minerals, which covers about 300 pages, is well arranged, and the data are clearly presented. The coal fields of the United States are described in considerable detail and the occurrences of other hydrocarbons are mentioned or briefly described. Chapters are devoted to building stones, clays, limes and cements, salines, gypsum, fertilizers, abrasives, minor non-metallic minerals, and underground waters. The illustrations and text figures are well chosen and clearly executed. The references are numerous, but are placed at the end of each chapter, a practice which, though saving space, renders them less accessible to the reader or student. The treatment of the metals is superior to that of previous editions. Although the book is intended primarily as a text, it should serve a useful purpose as a work of reference to the engineer or geologist who wishes general information regarding the occurrence and uses of certain minerals and the literature of the subject. W. Hz. E. go REVIEWS QI The Geology of New Zealand. By JAMES PARK, Professor of Mining and Mining Geology in the University of Otago. Pp. 488, with 145 illustrations, 27 plates, and a colored geological map. London: Whitcombe & Tombs, Limited, 1910. This new work is welcome to the geologic reader because it gives in organized, systematic, and relatively brief form a general view of the geology of a country whose geologic literature is otherwise scattered and to most geologists not readily accessible. It must also be acceptable to the teachers and students of New Zealand in that it gives them a view of geological history founded on the formational record of their own land. The work combines some of the features of a synoptic governmental report with those of a textbook. It was written originally for the Department of Mines, but only a part of it was published by the government—a fact which probably accounts for a seemingly dis- proportionate treatment of certain topics as compared with others, and also some lack of continuous progression under the control of a well- chosen scheme. Detailed descriptions of the various formations comprise the first portion of the work. Fach series is discussed first under the head of distribution, thickness, and age; then the faunas and floras are taken up, followed by the economic minerals and the igneous activity of the time. As might not unnaturally be expected in a country where even today the glaciers are such splendid spectacles, the Glacial Period has received much fuller treatment proportionately than the other periods. In an interesting discussion upon the excavating power of glaciers, the assertion is made that it is certain that ice can only excavate its bed when the pressure of its mass exceeds the ultimate crushing strength of the bed rock, and that the pigmy valley glaciers of today are incapable of excavating their beds. That glaciérs, even those of the small valley types, may be active eroding agents seems to find much less favor with the English school of geologists than with the American. The last portion of the book is devoted to economic geology. Natu- rally the greatest emphasis and fullest treatment are given to the very extensive coal fields and the important gold deposits, both of which have long attracted notice. A very welcome feature of the book is the closing chapter, which presents a complete bibliography of the geology of New Zealand cover- ing 56 pages. This book places the principal facts of New Zealand geology at the disposal of any geologist who reads English. Rew e. 92 REVIEWS Topologie. Etude du terrain. Par le G&NERAL BERTHAUT. 2 vols., quarto, pp. 674; 265 full-page topographic maps; 65 text figures. Paris: Service Géographique de l’Armée, 1909. This title covers a masterly philosophical treatise upon the evolution of land forms. The presentation is founded upon a thorough analysis of the geologic agencies which co-operate to form and to alter the surface features. The different classes of topographic features are described in the light of the various deformative and physiographic processes to which they owe their origin. These processes are taken up successively, and as the peculiarities and characteristics of the resulting topography are minutely described, they are vividly illustrated by the introduction of topographic maps. The subject is further developed from a discussion of these maps, which are so numerous as to constitute one of the leading attractions of the work. Most of the maps are selected from the topo- graphic surveys of France and the French possessions in North Africa, with occasional sheets from the Swiss Alps, Norway, and the United States. Re leis La sécurité dans les mines. Etude pratique des causes des accidents dans les mines et des moyens employés pour les prévenir. By H. ScuHMERBER. Paris: Ch. Béranger, éditeur, ro10. Pp. 659; figs. 589. Now that the people of this country have been awakened to the ned of greater safety in coal mining and efforts are being made to better the mining conditions, this new work on the engineering phase of the problem is very timely. It should be understood, however, that the geological and strictly scientific aspects of the problem of mine explosions scarcely enter at all into the author’s treatment and hence the book contains little of interest to geologists as such. But as an engineering work, which in truth is all that it attempts to be, it is an admirable treatise. RC: Leading American Men of Science. Edited by DAvip STARR Jorpan. New York: Henry & Holt Co., 1910. Pp. 471, with 17 portraits. This volume is made up of biographical sketches of seventeen men of the past selected as leaders in American science by a zodlogist of eminence. The selection embraces an astronomer, a chemist, a geolo- REVIEWS 93 gist, four zodlogists, two ornithologists, two paleontologists, one anatomist, one botanist—ten out of the seventeen from the biological group—and four physicists. The individuals chosen and the authors of the essays are as follows: Benjamin Thompson, Count Rumford, Physicist. By Edwin E. Slosson. Alexander Wilson, Ornithologist. By Witmer Stone. John James Audubon, Ornithologist. By Witmer Stone. Benjamin Silliman, Chemist. By Daniel Coit Gilman. Joseph Henry, Physicist. By Simon Newcomb. : Louis Agassiz, Zodlogist. By Charles Frederick Holder. Jeffries Wyman, Anatomist. By Burt G. Wilder. Asa Gray, Botanist. By John M. Coulter. James Dwight Dana, Geologist. By William North Rice. Spencer Fullerton Baird, Zodlogist. By Charles Frederick Holder. Othniel Charles Marsh, Paleontologist. By George Bird Grinnell. Edward Drinker Cope, Paleontologist. By Marcus Benjamin. Josiah Willard Gibbs, Physicist. By Edwin E. Slosson. Simon Newcomb, Astronomer. By Marcus Benjamin. George Brown Goode, Zodlogist. By David Starr Jordan. Henry Augustus Rowland, Physicist. By Ira Remsen. William Keith Brooks, Zodlogist. By E. A. Andrews. Students of geology will be most interested in the lives of Dana, Marsh, and Cope, the leading events of whose fruitful scientific careers are clearly set forth. Resa: Geology and Ore Deposits of Republic Mining District. By JOSEPH B. UmpLtesy. Washington Geological Survey, Bulletin No. 1. Epos.) fgsis; pl 12). Olympia, 1010: Physiographically the -Republic mining district in northeastern Washington appears to be an extension of the Interior Plateau of British Columbia and to be allied in Tertiary history with it. At the same time it seems to belong to a different physiographic unit from the central Cascades. The oldest rocks exposed in the Republic district are metamorphic, and are provisionally assigned to the Carboniferous. In early or middle Mesozoic times there occurred great batholithic intrusions of granodiorité. Following these came a great period of erosion lasting until the middle of the Tertiary. During this time there was developed an Eocene peneplain which was lifted and trenched before the end of 04 REVIEWS the Oligocene. The rocks next in order are dacite flows of Oligocene age. The remaining Tertiary history is written in several periods of igneous activity—andesite flows, intrusive latite porphyries, and a basaltic eruption during the Pleistocene... The total bullion production of the camp since the first discovery of ore in 1896 has been about $2,000,000, of which approximately 90 per cent has been gold and the remainder silver. The veins are thought to be genetically related to the latite porphyry intrusion and are made up of quartz, chalcedony, opal, calcite, and adularia, carrying incon- spicuous amounts of pyrite and possibly gold, in association with anti- mony, sulphur, and selenium. Though the deposits at Republic are not altogether like any others known in the United States, they most closely resemble the lodes of the Great Basin province. Their striking feature is the great amount of selenium in the ores, and they are thus best correlated with Tonopah and Goldfield, the only other camps in the United States known to pro- duce selenium ores. The report closes with a detailed description of the principal mines of the district, of which the New Republic mine is easily the leader. } aes 1 Ba Ce Notes on Explosive Mine Gases and Dusts with Special Rejerence to Explosions in the Monongah, Darr, and Naomi Coal Mines. By Rortitrin THomAs CHAMBERLIN. U.S. Geol. Surv. Bulletin 383. The results of a series of experiments carried out by the author throw new light on the nature of the explosive material and on the conditions governing explosions in coal mines, and should be of great practical, as well as scientific, value. As soon as possible after the explosions in the mines mentioned, samples of the mine atmosphere were collected and analyzed. Another series of experiments was carried out to determine the probable condition of the gas in the coal, whether (1) imprisoned in*minute cavities, (2) occluded or dissolved in the substance of the coal, or (3) the result of slowly operating chemical processes. ‘This was done by studying the rate of liberation of gas (1) from coal bottled in vacuum, (2) from crushing the coal, and (3) from heating the coal. A careful study was also made of the position and nature of the dust in passage-ways and on timbers in the mines after the explosions. REVIEWS 95 It is concluded that if methane were the sole explosive gas, only local explosions near the face of the coal could result. Coal dust is present, however, in large quantities and can under proper conditions become explosive. The chief restraining agent on dust explosion is dampness, and the presence of a high proportion of non-combustible shale dust. A great reduction of the moisture in mine atmospheres results from the incoming of cold air at the beginning of winter, and it is observed that most of the great explosions have been at that time. It is a general belief that old dust exposed for a long time to the air is more dangerous than fresh dust, but the author shows by experiment that this belief is erroneous, and that fresh dust is the-more explosive. Reve: Reconnaissance of the Book Cliffs Coal Field between Grand River, Colorado, and Sunnyside, Utah. By G. B. RicHarpson. U.S. Geol. Surv. Bulletin 371. The field forms a part of the south rim of the Uinta basin, around whose margin the outcrops of coal-bearing rocks can be traced for more than five hundred miles. Three formations of Cretaceous rocks are mapped: the Dakota sandstone lying unconformably on Morrison beds, the Mancos shale of Colorado and Montana age, and the Mesaverde formation which is overlain unconformably by Wasatch beds. The Mesaverde is partly marine and partly non-marine, the marine part showing close similarity to the upper Mancos shale and the non-marine to the Laramie. The age is placed as pre-Laramie, the Laramie epoch being supposedly represented by the unconformity above. Coal of good quality occurs in the lower part of the Mesaverde forma- tion in some localities. Several beds are present, but no single bed has been traced for more than a few miles. The coal of the region is little developed. hy 18s IW Cenozoic Mammal Horizons of Western North America. By HENRY FAIRFIELD OsBorn, with Faunal Lists of the Tertiary Mammalia of the West by WittiAM DitLER Matruew. U.S. Geol. Surv. Bulletin 361. This report is primarily a correlation of the mammal-bearing horizons of the Cenozoic with one another and with those of Europe, with a brief characterization of each horizon. In the Tertiary, six faunal phases are 96 REVIEWS recognized, containing eighteen subdivisions, while a seventh phase belongs to the Pleistocene. Three faunal phases containing seven subdivisions belong to the Eocene, the fourth phase, containing seven subdivisions, extends through the lower Miocene, the fifth phase extends through the middle and upper Miocene, and the sixth through the Pliocene. The conclusion is that North America promises to give a nearly complete and unbroken history of the Tertiary in certain regions, though much work still remains to be done. The chief remaining gap is now in the Pliocene stratigraphy. Je; Reale Keep The Dust Down In schoolrooms and gymnasiums it is of the greatest importance that everything be done to benefit the health even in the smallest degree. The harmful effects of too ‘much dust in the air are well known. The constant shuffling of feet in the schoolroom, and the more violent exercise in the gymnasium, stir up dust and circulate it in dangerous quantities. It is of the @ greatest importance to general health that the amount of this floating dust should be reduced. | mea \ \ ™D SS \\ holds down all dust that settles, and prevents its circulation in the air. Standard Floor Dressing is a special preparation, and vegetable an and animal germs cannot find subsistence in it. They are held oA === down and swept away at the end of the day. Many schools, \cymnasiums and stores, where the importance of reducing dust was |recognized, have been quick to avail themselves of the properties Standard Floor Dressing. - Illustrated booklet sent free—A booklet on “Dust anger and How to Avoid It” will be mailed to you free immediately pon receipt of your request. It contains much valuable information nd is a book you should have. | y =—-_ . Not intended for household use. Standard Oil Company (Incorporated ) Impure air and sick- ness are caused by OIL and GAS stoves, faulty fur- naces and dry steam heat. In every liv- ing room keep an open vessel contain- ing water and Chiorides | The Odoriess Disinfectant. It is an odorless, colorless liquid disinfectant. and deodorizer which instantly destroys foul odors and germs. Sold everywhere by druggists and high-class grocers. Write to Henry B. Platt, 42 Cliff St., New York, for new illustrated book free. 2 TO KEEP IN GOOD TRIM MUST LOOK WELL TO THE CONDITION OF THE SKIN” TO THIS END THE BATH SHOULD have been established over 60 YEARS. By our system of payments every family in moderate cir- cumstances can own a VOS€ piano. We take old instruments in exchange and deliver the new piano in your home free of expense, Write for Catalogue D and explanations, vose & SONS PIANO CO., Boston, Mass, nal 6... ee. VOS VOLUME XIX NUMBER 2 LTE JOURNAL or GEOLOGY A SEMI- QUARTERLY EDITED BY THOMAS C. CHAMBERLIN AND ROLLIN D. SALISBURY . With the Active Collaboration of SAMUEL W. WILLISTON ALBERT JOHANNSEN WILLIAM H. EMMONS Vertebrate Paleontology Petrology Economic Geology STUART WELLER WALTER W, ATWOOD ROLLIN T. CHAMBERLIN Invertebrate Paleontology Physiography Dynamic Geology ASSOCIATE EDITORS SIR ARCHIBALD GEIKIE, Great Britain GROVE K. GILBERT, National Survey, Washington, D.C. HEINRICH ROSENBUSCH, Germany CHARLES D. WALCOTT, Smithsonian Institution THEODOR N. TSCHERNYSCHEW, Russia HENRY S. WILLIAMS, Cornell University CHARLES BARROIS, France JOSEPH P.IDDINGS, Washington, D.C. ALBRECHT PENCK, Germany JOHN C., BRANNER, Stanford University HANS REUSCH, Norway 5 R, A. F. PENROSE, Philadelphia, Pa. GERARD DEGEER, Sweden p WILLIAM B. CLARK, Johns Hopkins University ORVILLE A. DERBY, Brazil WILLIAM H. HOBBS, University of Michigan T. W. E. DAVID, Australia FRANK D. ADAMS, McGill University BAILEY WILLIS, Argentine Republic CHARLES K. LEITH, University of Wisconsin FEBRUARY -MARCH, 1911 CONTENTS THE SOUTHERLY fe LIES wee ce pee OINOR EEGs Ee IN THE ALLEGHENY REGION - >> - aiong ees CINDER ah 107 THE ae er Ee ie oy CONFORMITY ge aoa SHARON CON- GLOMERATE mies tice = Sl hy Gar CANBY STO: THE WICHITA FORMATION OF NORTHERN TEXAS C. H. Gorpon, Grorce H. Girty, AnD Davi WHITE NOTES ON THE OSTEOLOGY OF THE SKULL OF‘ PARIOTICHUS'- E. B. Branson HIGH TERRACES AND ABANDONED VALLEYS IN WESTERN PENNSYLVANIA EUGENE WESLEY SHAW REQUISITE “CONDITIONS FOR THE FORMATION OF ICE RAMPARTS Witt1am H. Hosss THE TERMINAL MORAINE OF THE PUGET SOUND GLACIER J. Harten Brevz EDITORIAL: PES BEDING «ORY WORLDS rect acer i cular hint wdc, Rog wae TEESE Dane ee lat CoG ARTESIAN WATERS OF ARGENTINA - - - - A Si ade ee aie ne od eee WW PE EROGRAPHICAL) ABSTRACTS, AND REVIEWS = - >= => 0-02 2 ase ee ROB; WET Sistah, Oc (= apenas amt em lens Ree) GE eR USN Eh Pe toh ah del a) ie a ge Che Aniversity of Chicago press CHICAGO, ILLINOIS AGENTS: CAMBRIDGE UNIVERSITY PRESS, London anp EpINBURGH) WILLIAM WESLEY & SON, Lonpon TH. STAUFFER, LEtIpzic PUBLICATIONS IN THE BIOLOGICAL SCIENCES BOTANY Morphology of Gymnosperms. By JOHN M. CouLTER AND CHARLES J. CHAMBERLAIN. 470 pp., 462 illustrations, 8vo, cloth; post- paid, $4.22. Methods in Plant Histology. By CHARLEs J. CHAMBERLAIN. Second edition. x-+-262 pp., illustrated, 8vo, cloth; net, $2.25; postpaid, $2.39. Mitosis in Pellia. By CHARLES J. CHAMBER- LAIN. With three lithographic plates. 18 pp., 4to, paper; net, 50 cents; postpaid, 53 cents. The Ecological Relations of the Vegetation on the Sand Dunes of Lake Michigan. By HENRY C. COWLES. cents; postpaid, 81 cents. Oogenesis in Saprolegnia. By BRADLEY M. Davis. With twolithographic plates. 34 pp., 4to, paper; net, 75 cents ; postpaid, 79 cents. The Phylogeny of Angiosperms. By JoHN M. COULTER. 6pp., 4to, paper; net, 25 cents; postpaid, 27 cents. The Life History of Polysiphonia Violacea. By SHIGEO YAMANOUCHI. 54 pp., 10 plates, 8vo, paper; net, $1.00; postpaid, $1.05. ZOOLOGY Animal Micrology: Practical Exercises in Micro- scopical Methods. By MICHAEL F. GUYER. 250 pp., with 71 illustrations, 8vo, cloth; net, $1.75; postpaid, $1.88. The Development of Colors and Color Patterns of Coleoptera, with Observations on the Development of Colors in Other Orders of Insects. By Witiiam L. Tower. With three colored lithographic plates. 40 pp., 4to, paper; net, $1.00; postpaid, $1.04. The Animal Ecology of the Cold Spring Sand Spit, with Remarks on the Theory of Adap- tation. By CHARLES B, DAVENPORT. 22 pp., 4to, paper; net, 50 cents; postpaid, 53 cents. Laboratory Outlines for the Study of the Em- bryology of the Chick and the Pig. By FRANK R. LILLIE. 48 pp., paper; net, 25 cents; postpaid, 27 cents. BACTERIOLOGY A Laboratory Guide in Bacteriology. By PAuL G. HEINEMANN. xvi-+144 pp., illustrated, 12mo, cloth; net, $1.50; postpaid, $1.61. The Self-Purification of Streams. By EDWIN O. JORDAN. With two maps. 12 pp., 4to, paper; net, 25 cents; postpaid, 27 cents. 118 pp., paper; net, 75 PHYSIOLOGY Studies in General Physiology. By JACQUES LorB. In two parts. 806 pp., 8vo, cloth; net, $7.50; postpaid, $7.91. Physical Chemistry in the Service of the Sciences. By Jacospus H. VAN ’r HOFF. Translated by ALEXANDER SMITH. xviii -- 126 pp., illustrated, 8vo, cloth; net, $1.50; postpaid, $1.62. “Lucid, terse, concentrated.”’—Kuowledge and Sczen- tific News (London). “‘ The volume is an unusually elegant one, which makes a strong appeal to the book-lover as well as to the chemist.”” —Journal of American Chemical Soczety. “This is an extremely readable book.”—T7echnzcal World, “‘ Die Uebersetzung ist ausgezeichnet.”—Zeztschrift fur phystkalische Chemte. NEUROLOGY Neurological Technique. By Irvinc Har- DESTY. xii-+ 184 pp., 8vo, cloth; net, $1.75; postpaid, $1.87. : ‘¢ We do not know any other book of its size that seems quite as complete and useful.’’—Journal of American Medical Assoctation. ‘** As a whole, we know of no similar book which will be as valuable to the student of neurological technique.”— American Journal of Insanity. 3 ‘As a succinct, but sufficiently comprehensive intro- duction and laboratory guide to the subject, the book may be warmly recommended,’’—Brztzsh Medzcal Fournal., ‘The Finer Structure of the Neurones in the Nervous System of the White Rat. By SHINKISHI HATAI. With four colored plates. 14 pp. 4to, paper; net, 75 cents; postpaid, 78 cents. ANATOMY A Description of the Brains and Spinal Cords of Two Brothers, Dead of Hereditary Ataxia, of the Series in the Family De- scribed by Dr. Sanger Brown. By LEWELLYS F. BARKER. With three heliotype plates and forty figures. 50 pp., 4to, paper; net, $2.00; postpaid, $2.06. ‘¢'The article is extremely interesting to neurologists and medical men. It shows a profound insight and knowledge of the disease treated.” —Kzowledge. ‘¢ Altogether it is an elaborate and well-executed essay.” —Medical Record. The Distribution of Blood-Vessels in the Labyrinth of the Ear of Sus Scrofa Domes- ticus. By Grorce E. SHAMBAUGH. With eight colored plates. 20 pp., 4to, paper; net, $1.25; postpaid, $1.29. The Structure of the Glands of Brunner. By ROBERT R. BENSLEY. With five plates. 50 pp-, 4to, paper; net, $1.00; postpaid, $1.05. THE UNIVERSITY OF CHICAGO PRESS Address Dept. P. CHICAGO, ILLINOIS THE JOURNAL OF GEOLOGY Uae Bix ALR VIVA RC, TOL: THE SOUTHERLY EXTENSION OF THE ONONDAGA SEA IN THE ALLEGHENY REGION*™ E. M. KINDLE It is proposed in this paper to present some of the evidence which calls for a distinct modification of the current conception of the extent of the Onondaga sea in the eastern part of the United States. Before submitting the new data the reader’s attention will be invited to certain features of the previously recorded faunal and lithologic facts relating to the Onondaga sediments which, in the writer’s opinion, have led to some misconceptions regarding the character and extent of the Onondaga sea in the eastern states. The Onondaga fauna as developed in the states of New York, Ohio, Indiana, and Kentucky was one of the first of our Paleozoic faunas to be studied and described. The reports of the state sur- veys of these states, supplemented by numerous unofficial papers in which this fauna has been recorded and illustrated, have made it one of the best known of the Paleozoic faunas. It is a noteworthy fact, however, that all of the various contributions to our knowl- edge of this fauna have dealt with a nearly pure limestone fauna. If one were to seek a comprehensive idea of the character of the Onondaga sea and its sediments from the published descriptions of the fauna and the limestones holding it, he would get the con- ception of a sea in which only limestones were deposited. To any- t Published by permission of the Director of the United States Geological Survey. Vol. XIX, No. 2 O7 98 E. M. KINDLE one who admits that the factors controlling marine sedimentation were essentially the same in Paleozoic and recent times, a Devonian sea in which only calcareous sediments accumulated is a manifest absurdity. We know of no continental or other seas in which there are not a variety of types of sediment accumulating simultane- ously. Papers which have undertaken to deal with this fauna in a large way and weld its evidence into the new science of paleo- geography have naturally been influenced by the fact that the only faunas described from the Onondaga sea were limestone faunas. Translated into the form of a paleogeographic map this class of evidence taken alone gives us a sea whose outlines inclose only limestone sediments. This was a serious defect in Professor Charles Schuchert’s first map of the Onondaga sea.t. The shore- lines given by it for the Onondaga sea in the central states inclosed a sea from 100 to 300 miles in width. All of the known Onondaga deposits included by the shorelines of the map are limestones. The recently published map of the middle Onondaga by Professor Schuchert? shows improvement in this respect, since it includes the shales and argillaceous limestone bands holding the Onondaga fauna which was discovered in central Pennsylvania by Charles Butts and determined by the writer. The later map, however, still gives us a conception of the Onondaga sea far from that which the writer’s recent studies in the Allegheny region appear to demand. The writer’s criticism, it may be stated here, is directed primarily, not to Professor Schuchert’s map, which incorporated all of the positive evidence available at the time of its preparation, but at the incompleteness of the evidence in a region where it might be expected to be fairly complete. In order to ascertain to what extent recorded evidence and opinion will enable us to reconstruct the shorelines of the Onondaga sea within the limits of the eastern states so that they will appear consistent and rational with reference to the character of the known deposits of that sea, we may consider briefly the principal sources * Charles Schuchert, ‘On the Faunal Provinces of the Middle Devonic of America and the Devonic Coral sub-Provinces of Russia, with Two Paleographic maps,” Am. Geol. (1903), XXXII, 137-62, Pl. 20. 2 Charles Schuchert, ‘‘Paleogeography of North America,” Bull. Geol. Soc. Am., XX (1910), 75. THE ONONDAGA SEA IN THE ALLEGHENY REGION 99 of its sediments. The comparatively thin mass of sediment which accumulated during the whole of the Devonian in the central states affords satisfactory evidence that the land area adjacent to the Devonian sea on the west had slight relief, and furnished compara- tively little sediment at any time during the Devonian. On the east side of the Devonian sea, however, physiographic conditions were very different. Willis' has shown that during much of the Devonian period there lay immediately southeast of the Alle- 'gheny region the highlands of Appalachia. This old land area furnished to the interior Devonian sea of the Appalachian region, between the beginning of the Hamilton epoch and the close of the Devonian, a mass of sediments which, if restored upon a sea-level plain of Appalachia, ‘would constitute a mountain range closely resembling in height, extent, and mass the Sierra Nevada of California.’ According to the prevailing view: this fertile source of Devonian sediments was elevated at the close of the Oriskany to such an extent that throughout Onondaga time the Allegheny region was a land area. Such elevation, if it occurred, must have resulted in accelerated erosion in the Devonian highlands, and in an in- creased volume of sediments in the Onondaga sea. If this hypo- thetical uplift occurred, it could not have failed to have been registered by a great thickness of coarse clastic sediments in the narrow Onondaga sea which, as outlined by Schuchert’s map, extended as a narrow belt across the adjacent portions of the present states of Kentucky, Indiana, and Ohio. Instead of such coarse clastics we find in these states, as previously noted, only limestones. representing. sedimentation near the eastern shore of the Onondaga sea as outlined by Schuchert.4 The utter impossi- bility of harmonizing the pure limestone deposits representing the Onondaga in the Ohio valley with this currently accepted theory of diastrophism in the Allegheny region would appear to be a suff- cient reason for discarding it. If, however, we assume that Appa- t Md. Geol. Survey, Special Publication, Vol. IV, Pt. I, pp. 61-62. elbtd eps 62: 3 Charles Schuchert, ‘‘Paleogeography of North America,” Bull. Geol. Soc. Am., XX (1910), 492. siibtds. Rls 7s. 100 E. M. KINDLE lachia was not elevated and the Devonian shoreline was not pushed westward at the initiation of Onondaga time, we would still expect as a probability non-calcareous sediments to predominate in the eastern portion of the Onondaga sea. That portion of the Onon- daga sea adjacent to the land area which furnished 10,000 feet of non-calcareous Devonian sediments in post-Onondaga time would be likely to acquire chiefly non-calcareous sediments even in an epoch so favorable to calcareous sedimentation as the Onondaga. A considerable mass of paleontologic and stratigraphic data which has been gathered by the writer shows that Onondaga sedi- ments are present in the Allegheny region and are mainly of this non-calcareous type, as might have been expected from theoretical considerations. The recent discovery of an Onondaga fauna in the Allegheny region which occurs in a series of drab or dark shales and thin interbedded argillaceous limestones thus very materially supplements the hitherto one-sided character of the available data relating to the nature of the fauna and sediments of the Onondaga sea. The sediments holding this fauna are of such a character as we might have expected to be accumulating on some portion of the Onondaga sea floor if we may judge by analogy with the processes of sedimentation now in operation in the largest continental seas. Since this fauna will be described and figured in a forthcoming bulletin of the United States Geological Survey, only the most general facts regarding it will be presented here. The fauna comprises more than one hundred species. The correlation of this Allegheny fauna with the New York Onondaga fauna is based primarily upon the presence in it of such well-known species as Ano plotheca acutiplicata, Rhipidomella vanuxemt1, Spirifer acuminatus, and Odontocephalus aegeria. The great abundance and general distribution of the first named of these species is a con- spicuous characteristic of the fauna. In point of abundance and wide distribution in this argillaceous facies of the Onondaga, Ano plotheca acutiplicata is as prominent as is Spirifer acuminatus in the well-known calcareous facies. It is interesting to note in this connection that while Anoplotheca acutiplicata is a familiar species in the Onondaga limestone of eastern New York comparatively near the region under discussion, it is unknown in the more westerly areas THE ONONDAGA SEA IN THE ALLEGHENY REGION 101 of the limestones of Onondaga age in Ohio, Indiana, and Illinois. Its occurrence in typical Onondaga limestone only in an area which is nearly adjacent to the region of the shaly facies of the formation suggests that the latter type of sediments furnished its normal and most congenial habitat. Spzrifer acuminatus, on the other hand, does not extend very far to the southward into the region occupied by the argillaceous facies of the Onondaga. Other Onondaga species, however, like Odontocephalus aegeria, appear to be equally adapted and distributed in both types of sediment. Some of the stratigraphic data relating to this fauna may be very briefly summarized as follows: The calcareous shales holding this fauna are generally preceded in the sections by the Oriskany sandstone and always followed by the dark fissile and comparatively barren shales of the Marcel- lus. These two limiting formations exhibit in general essentially the same lithologic characters throughout Pennsylvania, Mary- land, West Virginia, and much of Virginia as in New York. Both are, however, much thicker in this more southerly region than in the type region of the Onondaga limestone in New York. In the Helderberg mountain region the Onondaga and the Hamilton faunas are separated by 300 feet of comparatively barren dark Marcellus shale, and in western New York by about half this thick- ness, while in Pennsylvania and southward these shales often have a thickness of more than 500 feet. While the succession from the Onondaga fauna to the Marcellus fauna above is a uniform one throughout most of the Allegheny region, as it is in New York, the succession at the base of the fauna is not everywhere precisely the same. In most of the territory the Onondaga beds rest upon the Oriskany, but in some of the Penn- sylvania sections they immediately follow beds representing the Esopus shale. In respect to its underlying formation, however, the Onondaga shows less variation than in New York, where, in different areas, it is found to follow the Manlius, Oriskany, Esopus, and Schoharie. Thus, we find that this fauna occupies in the Allegheny region the same relative position in the succession of faunas as the Onondaga fauna does in the standard sections of New York. The stratigraphic evidence, therefore, coincides with the paleontologic 102 hk. M. KINDLE evidence already briefly cited in pointing to the Onondaga age of the fauna. We may now consider the bearing of the data which have been cited on the modification of the current conception of the eastern shoreline of the Onondaga sea in the eastern United States. The Onondaga formation extends scarcely south of the Dela- ware River according to most of the papers dealing with the stratig- raphy of the Devonian in the Allegheny region, thus giving it a north-south extension of scarcely 200 miles. This comparatively insignificant southerly extension of a fauna which is so persistent in a westerly direction seems more surprising when it is recalled that all of the other faunas characterizing the major divisions of the New York Devonian section have with one or two exceptions been traced southward from New York entirely across Pennsyl- vania. Thus itis seen that the prevailing conception of the Onondaga formation and fauna, which presumes their absence south of New York, gives to it an anomalous position as compared with the other important formations of the Devonian section of New York. The evidence which the writer has gathered during three seasons of field work in the Allegheny region indicates that the southerly extension of the Onondaga fauna is quite comparable in distance with its westerly extension. The field studies of the writer have shown that the Onondaga fauna in the Allegheny region extends far to the southward of the area in which nearly pure limestones were deposited during Onondaga time into a region where shale-forming sediments partially or completely dominated those of calcareous type. This fauna has been found in nearly all the sections studied from New York to Tennessee. The direct bearing of these new data on the paleogeography of Onondaga time is obvious. Its incorporation involves the exten- sion of the eastern shoreline of the Onondaga sea in a southwesterly direction from southeastern New York to the eastward of the Allegheny region instead of far to the westward of it, as now drawn, across the states of Ohio, Indiana, and Kentucky. In the light of this new evidence it appears that the eastern shoreline of the Onondaga sea trended southwesterly across north-central New Jersey and southeastern Pennsylvania. It probably traversed the THE ONONDAGA SEA IN THE ALLEGHENY REGION 103 states of Maryland and Virginia near the present axis of the Blue ~Ridge Mountains. From southwestern Virginia this shoreline appears to have trended westerly not far from the Kentucky- Tennessee line as far as the valley of the Tennessee River where it resumed its southerly trend. This revision of the shorelines of the Onondaga sea gives, instead of the Cincinnati peninsula of Schuchert’s map, a Cincinnati island. This, and probably other smaller islands, interrupted the continuity of the Onondaga sea, which, in the region of the Ohio valley, reached a maximum width of about 500 miles from northwest to southeast. THE MISSISSIPPIAN-PENNSYLVANIAN UNCON-— FORMITY AND THE SHARON CONGLOMERATE" G. F. LAMB Mount Union College There exists in northern Ohio a well-defined boundary between the strata of the Mississippian and Pennsylvanian ages, a boundary marked by a pronounced unconformity. The upper limit of the Mississippian is the top of the well-known Cuyahoga formation, and the lower limit of the Pennsylvanian is the bottom of the equally well-known Sharon conglomerate. So far as the writer is aware the Sharon has been generally regarded as a formation of general extent around the northern and northwestern border of the Appalachian coal basin, and resting upon the Mississippian in a continuous sheet except where removed by erosion. Field work the past summer in Mahoning, Trumbull, Portage, Summit, and Geauga counties has revealed some facts that lead the writer to believe that the Sharon conglomerate is not the simple formation that it has been thought to be, and that it has a setting of unusual interest. Following its outcrop from place to place, the formation is found to change in structure quickly, to disappear suddenly, and to be absent over considerable areas, letting later rocks form the contact with the Cuyahoga. Where its development is greatest, it les in troughs of the Cuyahoga. Further, it is found to occur in belts having a more or less north-and-south direction, and these belts, in places at least, are not now and never have been connected from east to west. This is due, in part at least, to the fact that the conglomerate lies between ridges of the Cuyahoga, and not alone to post-Pennsylvanian erosion. t Published by permission of Dr. J. A. Bownocker, state geologist of Ohio. Pre- sented at the twentieth meeting of the Ohio Academy of Science, Akron, November 25, I9IO. IO4 MISSISSIPPIAN-PENNSYLVANIAN UNCONFORMITY 105 This manner of occurrence calls attention to the surface upon which the Pennsylvanian rests. Whatever may be the case else- where, the writer believes that greater erosion of the upper Missis- sippian occurred in northern Ohio than is generally known. Instead of the Sharon resting upon a nearly uniform plane, it is found that the surface of the Cuyahoga has a relief of nearly 200 feet, and it is significant that where the depressions are greatest, the Sharon is also thickest. The regional or belt-like occurrence of the con- glomerate, and its apparent relationship to depressions in the Cuyahoga, along with the structure and variability of the stratum, have led the writer to the conclusion that these depressions are creek and river valleys, and that the conglomerate is a deposit of stream gravels, and that the overlying sandstones of the Potts- ville are, to a greater or less extent, river and delta deposits. Some of the data on which this view is based are added. Val- leys in the Cuyahoga formation are of general occurrence. The most conspicuous and deepest one so far found may be noted in some detail. This valley lies in the eastern edge of Portage and Geauga counties, about half-way between Akron and the state line, and its course is roughly north and south. At Akron, the top of the Cuyahoga formation lies about 940 above sea; due east, at Mineral Ridge, west of Youngstown, at 962; at Newton Falls, between these two points, and 5 miles north of the Akron-Mineral Ridge line, it lies below 850, or about 100 feet lower than to the east or west. If such a line be drawn from east to west half-way between Akron and Cleveland, the same depression in the Cuya- hoga is again found. At Brandywine Falls, 15 miles north of Akron, the top of the Cuyahoga formation lies at 1,040; near Howland Springs, due east of Brandywine, at 1,044; and at Nelson Ledges between these two points at 956, or again nearly 100 feet lower. Another east and west comparison may be cited. At Burton, due east of Cleveland, the top of the Cuyahoga lies at 1,090 and due east on the state line at 1,190, or 100 feet higher. These three middle points—Newton Falls, Nelson Ledges, and Burton—are in line, roughly, north and south, and are clearly in a depression of the Cuyahoga formation, since rock of this formation lies higher both to the east and west. Further, this depression cannot be 106 G. F. LAMB assigned to a syncline, as is proven by the nearly horizontal posi- tion of the Berea in the same direction. It is worthy of note that the Sharon and overlying sandstones in the line of this old valley reach their greatest development in Ohio, and form a great body of conglomerate and sandrock extending southward from southern Lake County, through Geauga and Portage counties, at least as far south as northern Stark County. ‘The evidence is strong that the conglomerate and overlying sandstones in this great ridge are stream deposits, and will be discussed later. It may be objected that the distances involved in the three lines across this supposed valley are of such length as to be of doubtful value. Data are at hand, however, which confirm fully what the three lines of elevation show. At Brandywine the Sharon base is 210 feet above the Berea; due east at Nelson Ledges only about 75 feet; near Newton Falls only about 75 feet; but on the state-line nearly due east of Nelson Ledges nearly 300. The meaning of these figures is clear, and shows deep erosion, which is still further confirmed by the presence of hills of the Cuyahoga in the very region in which the erosion was greatest. As stated above, the top of the Cuyahoga near Newton Falls lies below 850 and in 2+ miles north rises to 1,040 above sea-level. It therefore forms a hill at least 190 feet high, with no trace of the Sharon or overlying sandstones. Within 35 miles to the northwest from this hill, and in a direction opposed to the dip, the surface sinks to g19 feet at least. At Nelson Ledges the conglomerate is about 75 feet thick, and one solid mass from bottom to top. It appears to the observer that it may be expected to continue for miles to the north, but instead it thins out quickly on the steep slope of another Cuyahoga hill, which rises from 956, at the base of the Ledge:. to i1o7,7a) ise, oh 15m Teet) ant emilee NVae mm 2 miles to the northwest from this hill, the surface drops again to gogo, or 117 feet, as seen in the Parkman gorge. From this point the surface rises again to the northeast, 180 feet in 24 miles, then falls toward the northwest. At Newton Falls, there is a like rise toward the northeast from below 850 to 941, in about 3 miles. Now all points which show these old hills are on or near the eastern margin of this rock ridge, and in every case MISSISSIPPIAN-PENNSYLVANIAN UNCONFORMITY 107 bear evidence of a more or less westerly slope toward the ridge. They are clearly hills bordering a valley, and are conclusive evi- dence of former dissection to a depth of nearly 200 feet. This same hill and valley topography of the Cuyahoga is found all through eastern Trumbull and northern Mahoning counties, with the conglomerate often absent, and with the Sharon coal lying close above the Cuyahoga. One of the finest exposures of the unconformity occurs in Mineral Ridge, south of Niles, and near the Mahoning-Trumbull line. A deep east-and-west ravine cuts through a north-and-south Cuyahoga ridge finely exposing the contact, showing the horizontal shale and flaggy layers of the Cuyahoga, overlain by the steeply inclined strata of the Pennsylvanian. ‘The slope of the Cuyahoga is toward the east, and at an angle of about 25°, is ragged or stair- step like, and is directly overlain by 2 or 3 feet of crude, mixed sandstone, without lamination or bedding planes, which grades quickly into a bluish shale, then to a carbonaceous shale which carries the well-known and formerly much-worked bed of iron ore. The ore is a highly ferruginous limestone, which is certainly the Lowellville limestone. Directly above the ore is a bed of coal— the Mineral Ridge coal—which lies only 8 feet above the Cuyahoga. The sandstone, shale, ore, and coal all lie at the same steep angle above the Cuyahoga. I have stated above that the Sharon conglomerate bears evi- dence of being a stream deposit. This appears from its position, its constitution, and its structure. In some places it is little else than a mass of quartz pebbles which range in size from coarse sand to half the size of the fist. (Commonly the stratum is an alterna- tion of sand beds and pebble layers, of constant variation both horizontally and vertically. Bottom-set, fore-set, and top-set beds are common. The sudden change from sand to gravel, and the very variable structure of the sand beds, all of which may be repeated several times in a single rock face, can be accounted for only by stream action. There is not any feature of the con- glomerate that stream action does not produce. On the other hand, the writer is unable to conceive of any other agency capable of producing a like stratum. 108 (GSE EANEB It will be interesting to note the most exaggerated conglomeratic development found. It occurs at the base of the Sharon as exposed in the gorge at the village of Parkman, Geauga County. Lying directly upon the Cuyahoga, and representing a stream velocity of probably 3 miles an hour, is a 3-foot bed composed not of pebbles alone, but of cobble stones, or pieces of flagstone from the Cuyahoga, some angular, some rounded and flat and well worn, 2 to 3 inches thick and more than a foot in diameter, standing and lying in all positions and mixed with sand and pebbles. It is a veritable picture of the stones and gravel and sand all mixed that we have all seen many times on the inside curve of streams. A more con- vincing evidence of stream deposit in former ages can hardly be found. It is worthy of note here that two distinct stages in the deposi- tion of the Sharon are displayed in this gorge. At to or 12 feet above the base the conglomeratic character is entirely absent, a rather fine soft sandstone occurs, the top of which is quite undulat- ing, as if eroded. Resting directly upon the undulating surface, with a sharp line, is the massive conglomeratic rock characteristic of the Sharon. The transition is sudden and very conspicuous and is well shown at a number of points in the gorge. At Nelson Ledges the base of the conglomerate lies at 956, and 3 mile west conglomerate is found at 1,160 above sea. This whole thickness of 204 feet is not to be assigned to the Sharon, however. Overlying sandstones are conglomeratic in this locality and suffi- ciently so to be mistaken easily for the Sharon itself. Two miles south of this point and about $ mile south of Nelson village at Ledge Haven Mill conglomerate rock is found on Tinker Creek. There are clearly two stages of conglomerate formation here. The bed of the creek below the fall is conglomerate of unknown thick- ness. It is directly overlain by 5 to 6 feet of dark gray sandy shale and this is overlain in turn by 30 to 4o feet of conglomerate. The top of the lower stratum lies at 952, as seen at the foot of the fall beside the mill. The shale stratum is strongly suggestive of the horizon of the Sharon coal. It also strongly suggests relationship to the two-stage phase of the conglomerate observed in the Parkman gorge. At the latter place this transition occurs at a level of about MISSISSIPPIAN-PENNSYLVANIAN UNCONFORMITY 109 1,000 feet above sea, and is nearly 5 miles north of the above mill, and when dip is taken into account the probability is very strong that the phenomena seen at the two places belong to the same horizon. A quite singular feature occurs in this shale at the fall. Near its middle, and imbedded in it, lies a lenticular mass of conglomerate a foot thick and probably weighing nearly a ton. It contains large quartz pebbles, much pyrites of iron, and an impression of a cala- mite. How was it transported to this place where only fine sedi- ments were being deposited? Where did it come from, and from what rock formation was it detached? For the conglomerate beneath must have been only a stratum of sand and gravel when it was deposited. In central Ohio three other formations intervene between the Cuyahoga and the base of the Pennsylvanian, the lower one of which—the Black Hand—is known to be conglom- eratic in part. Is this conglomerate block imbedded in this shale a remnant of the Black Hand which once may have overlain the Cuyahoga in northern Ohio, and was completely removed by erosion before the close of the Mississippian? ‘These are ques- tions to which only further study may reveal the answer. THE WICHITA FORMATION OF NORTHERN TEXAS! C. H. GORDON University of Tennessee, Knoxville, Tenn. With discussions of the Fauna and Flora by GEORGE H. GIRTY ann DAVID WHITE INTRODUCTION The geology of the “Red Beds” area of northern Texas has long been recognized as one of the perplexing problems of North American geology. The interest aroused by the discovery in these beds of a fauna which was regarded by Cope, C. A. White, and others as Permian has brought forth a number of papers bearing on this region, most of which are based on transient visits In search of fossils, generally with scant attention to the detail of stratig- raphy. This paper is based upon investigations made in connection with the study of underground water conditions for the United States Geological Survey during the field seasons of 1906 and 1907. The collections of invertebrate fossils made in the course of the investigations were submitted to Dr. George H. Girty of the Survey, who also had for study additional materials collected by E. O. Ulrich in former years. STRATIGRAPHY OF THE REGION The ‘‘Red Beds”’ area.—The area occupied by the “Red Beds” in northwestern Texas is bounded on the west by the eastern escarp- ment of the Llano Estacado, and extends eastward along the Red River to Montague County, where the formations pass from sight beneath the basal beds of the Cretaceous. From this point the eastern boundary of the “‘Red Beds” bears south and then west- ward, following approximately the lines between Jack and Clay, and Young and Archer counties as far west as the Salt Fork of the Brazos. From this point it bears southwestward to the south- « Published by permission of the Director of the United States Geological Survey. IIo WICHITA FORMATION OF NORTHERN TEXAS III eastern corner of Haskell County, thence irregularly south until it meets the Cretaceous again in Concho County. As thus outlined, the “Red Beds” occupy an area of irregular shape 80 to roo miles in width in the southern portion, while at { 4 Mi ATV Pana Quaternary [__] PERMIAN : Recent Seymour Undifferentiated Wichita Deposits Gravel Clear Fork and Formation Double Mountain PENNSYLVANIAN ‘wiih Cisco Canyon Strawn Formation Limestone Formation the north they extend eastward fully twice that distance along the south side of Red River. If a line be drawn from a point on Red River near the mouth of Pease River southwestward through Seymour to the northeastern corner of Haskell County and thence southward, it will mark approximately the eastern boundary of a series of red clays and red sandy shales containing gypsum in vary- ing amounts, to which the names Clear Fork and Double Mountain IRIED) C. H. GORDON were applied by Cummins in reports of the Texas Geological Survey. These are evidently the equivalents of the beds included by Gould? in the formations to which he applied the names Greer and Quarter- master. As these beds have no connection with the problem in hand, they may be dismissed from further consideration. It is to that portion of the “Red Beds” area adjoining the Red River and extending eastward from the line above indicated that most of the discussions concerning the Texas Permian apply. This is the type area of the Wichita formation of Texas. ‘The western part of this area is characterized by the occurrence of beds of limestone and blue shale interbedded with red clays and sandstones, while the eastern part is notable for the entire absence of limestones and the very limited development of blue shale and clay. If a line be drawn from a point where the Salt Fork of the Brazos crosses the boundary between Throckmorton and Young counties, a little east of north to Red River, it will mark approximately the boundary between the areas thus lithologically distinguished. According to Cummins’ earlier writings? most of the rocks of this western area were assigned to the Clear Fork formation, while the strata occur- ring toward the east constitute his original Wichita division. Many of the fossils on which his conclusions regarding the Permian age of the beds were based, however, appear to have come from the basal portion of the limestone series in eastern Baylor County. In the earlier reports the Wichita formation is described as hav- ing no surface development south of the point where the “Red Beds” boundary meets the South Fork of the Brazos River in the northeastern corner of Throckmorton County. From that point southward the Clear Fork formation is said to rest directly upon the “Albany,” considered to be the highest division of the “Coal Measures” in that region. This peculiar relation of the Wichita formation was conceived to be due to overlap, and hence it was believed that an unconformity marked the relations of these beds to the “‘Coal Measures.” In a later paper,? read before the Texas Charles N. Gould, Water-Supply and Irrigation Paper No. rg9t (1907), 14-19. 2 Geological Survey of Texas, II (1890), 401. See map facing p. 552. 3.W.C. Cummins, Transactions of the Texas Academy of Science (1897), I1, 93-907. WICHITA FORMATION OF NORTHERN TEXAS 113 Academy of Science, Cummins announced the discovery of evidence showing that the limestones of eastern Baylor County are the same as those of the “‘Albany.”’ In this paper the beds of Baylor County are said to constitute the upper part of the Wichita. Owing to the discontinuance of the Texas Survey the report on this area prepared for the Fifth Annual Report has not appeared. Rocks of the Wichita area.—KEast of Baylor County the rocks consist for the most part of red concretionary clays, red sandstones and sandy shales with occasional beds of blue shales, and bluish to grayish-white sandstones. Limestones are conspicuously absent. Occasional impure nodular layers occur, however, which contain considerable calcareous matter, but these do not constitute strata of limestone. The sandstones are usually soft and distinctly cross- bedded. In some places they are shaly, in others massive. Some layers are streaked and specked with grains of black iron oxide, while others contain nodular masses and concretions of iron ore. The clays are mostly deep red or red mottled with bluish-white and drab colors. ‘The red clays contain an abundance of nodular concretions of irregular shape varying in size from that of a pea to masses 4 or 5 inches in diameter. They consist of clay, iron, and lime, and at times are hollow or with the interior filled with cal- careous clay or lime carbonate. In some cases they are flattened and stand in vertical position in the clays, suggesting their origin through the filling of fissures. Occasionally a bed is met with consisting of rounded lumps of hardened clay cemented together by ferruginous matter, repre- senting what Cummins called “‘a peculiar conglomerate.” This formation is believed to have had its origin in the breaking-up of a bed of clay by running water or wave action. In places the bluish clays are copper bearing. Efforts to mine these deposits, however, have not been profitable. The ore occurs in the form of small nodules in the clays and also as a replacement of wood." In the sandstones occasional traces of plants occur, and some- times remains capable of identification are found. White reports Taeniopteris from the sandstones near Fulda. The stratification ry. F. Cummins, First Annual Report, Geological Survey (Texas, 1889), 188-96. II4 C. H. GORDON of the beds is very irregular. The sandstones, shales, and clays grade into each other both vertically and horizontally. Moreover there is a monotonous similarity in the sandstones and _ shales respectively throughout the area, which, taken in connection with the absence of any persistent easily recognizable stratum, renders the stratigraphic correlation of the beds, except within very narrow limits, practically impossible. In eastern Clay and Montague counties, the beds, considered Cisco, show a greater development of sandstones some of which are conglomeratic. In the western part of the area, however, no true conglomerates were observed. As to the thickness of the Wichita, no definite statement can be made. Certain of the beds may be traced for a limited distance sufficient to indicate a general westward dipping of the strata. Cummins estimates it to be 35 feet per mile, which is probably too high. The width of the outcrop in an east-west direction is about 50 miles, which, assuming a regular inclination of 25 feet per mile, would give a thickness of 1,250 feet for the beds out- cropping in this portion of the field. How much of this should be referred to the Cisco is conjectural, but probably not less than half. A well put down for oil at Electra, which is located near the top of the formation, passes through 1,790 feet of red clays with some sandstone and red sandy shales. At Petrolia, which is near the middle of the outcrop, the oil wells are for the most part about 400 feet deep, chiefly in red clays and shales. Drilling has extended to a depth of 800 feet in some instances and indicates an increase in the proportion of blue shales below, but no reliable record could be obtained of the lower formations passed through. At Archer City a well 737 feet deep shows red clays and reddish sandstones predominating to a depth of 670 feet. Below this the drill revealed similar deposits but in diminished proportion, as compared with the light-colored sands and bluish clays. Since thé upper beds of the Cisco in this region are prevailingly red, how- ever, no reliable conclusion can be drawn from well records as to the plane of division between the formations. In the bluffs of the Wichita River in the northwestern corner of Archer County some beds of limestones aggregating 4 feet in WICHITA FORMATION OF NORTHERN TEXAS 5 thickness appear at the top of the escarpment on the west side of Horseshoe Lake, and outcrops of these appear at intervals along the boundary of Archer and Baylor counties. This limestone is earthy, very hard, dark blue where fresh, and weathers to dark brown or black. It is underlain by 4 feet of blue clay. The remainder of the section to the base of the hill, about 100 feet, consists of red concretion-bearing clays with a limited development of red and white shaly sandstone. From this point westward the stratification becomes more regular, consisting of the blue shales alternating with the red, the red being predominant, with an occa- sional bed of dark earthy limestone containing usually an abun- dance of poorly preserved fossils. At the Bar-X ranch on the Wichita River in the northeast cor- ner of Baylor County near the Old Military Crossing, several ledges of hard limestone appear in the river bluffs separated by varying thicknesses of blue shale, alternating with red clay. The beds dip to the westward at inclinations estimated at 20 to 30 feet per mile. Proceeding up the river from this point, limestones appear at intervals in increasing development, the best outcrops occurring about 2 miles east of where the Seymour-Vernon road crosses the river. Here an escarpment go feet in height has the lower two- thirds composed of red and blue shales alternating with beds of limestone. The middle of the section consists of red and concre- tionary clays and sandstones. Some of the ledges of limestone are massive, but others are thin-bedded and shaly, and separated by varying thicknesses of bluish clay. Locally the thin-bedded limestones and their included shale grade horizontally into more massively bedded limestones. Fossils are not plentiful in this locality. The same beds are exposed again northward in the banks of Beaver Creek. At Seymour the limestones are well exposed in the banks of the river where they are quarried to some extent and furnish a stone that is well adapted to ordinary uses. The beds are here transected by the Salt Fork of the Brazos River, which flows in a relatively narrow valley between steep bluffs 200 feet high, made up of interbedded red and blue clays, and lime- stones. The limestones of Baylor County area are generally fossil- 116 C. H. GORDON iferous. Owing to the hardness of the rock, however, good speci- mens are difficult to obtain. Toward the south there is an increase in the development of blue shale and limestone, while the red clays and sand show a corresponding diminution. In a recent paper’ Case has endeavored to correlate certain of the sandstones occur- ring throughout the area, one of which he calls Fulda, from a little station by that name in eastern Baylor County. With this sand- stone he correlates others which outcrop as far east as Wichita Falls, a distance of 37 miles. With this conclusion the writer is not in accord. In the first place, the sandstones at Fulda are underlain by some thin limestones which outcrop toward the north- east in the northwestern part of Archer County. It is quite apparent that the sandstones in eastern Archer and Wichita counties represent horizons below these limestone beds. Assum- ing the general westward dip of the strata to be no more than 20 to 25 feet per mile, there must be a descent of not less than 650 to 800 feet to which must be added the rise of the plateau surface which is about 200 feet, making a total of 850 to 1,000 feet between the horizon represented at Wichita Falls and that at Fulda and rendering untenable the correlations suggested. Albany area.—The eastern boundary of the Clear Fork and Double Mountain formation in eastern Jones County is marked approximately by the Clear Fork River. The region to the east of this point to the limits of the Cretaceous in western Parker and Wise counties, a distance of over too miles, known as the Brazos Coal Field, is occupied by rocks of Carboniferous age. These beds, which have a thickness of nearly 7,000 feet, present lithological, stratigraphic, and faunal characteristics, which permit their separation into four well-marked divisions, known as the Strawn, Canyon, Cisco, and “Albany” divisions.? Southward in the Colorado Coal Field the equivalent rocks were first studied by Tarr,’ who rE. C. Case, Bulletin of the American Museum of Natural History, XXIII (1907), 659-064. 2 These names appear first in the First Annual Report of the Geological Survey of Texas in the State Geologist’s ‘‘Report of Progress,” pp. Ixv—lxvii. Hill, however, credits them to Cummins (Twenty-first Annual Report, U.S. Geological Survey, Part VII, 97). 3R.S. Tarr, First Annual Report of the Geological Survey of Texas (1889), 201-16. WICHITA FORMATION OF NORTHERN TEXAS in, subdivided them into five divisions as follows: Richland, Milburn, ’ Brownwood, Waldrip, and Coleman. Later the Milburn was included in the Brownwood.t The relations of these rocks as now recognized are as follows: ? Colorado Field (Tarr) Brazos Field (Cummins) Thicknesses in Feet Coleman “Albany” 1,200 Waldrip Cisco 800 Brownwood } : - Brownwood Canyon 800 Milburn \ _ Richland Strawn 4,100 The beds dip to the west at a low inclination estimated by Cummins to be 30 feet per mile for the “Albany” and 75 for the Canyon. : Limestones constitute the dominant characteristics of the “Albany” and Canyon formations, while sandy shales and sand- stones, with some conglomerates, make up the larger part of the Strawn and Cisco formations. It is with the two uppermost of these, the ‘“Albany’’ and Cisco, that the ““Red Beds” problem is concerned. The “Albany.”’—The “‘Albany,’’ named from the county seat of Shackelford County, consists of blue, gray, and yellowish lime- stones, alternating with beds of blue and dark-gray shales. The upper 500 feet are characterized by massive beds of hard blue limestone, with partings of blue shale, while the lower portion shows a greater development of shale, the limestone being for the most part thin-bedded and shaly. The heavy ledges of limestone appear at the surface in a succession of terraces which extend in sinuous curves from north to south. Sandstones and conglomerates are almost entirely lacking. The formation contains an abundant marine fauna, which, taken in connection with the notable devel- ‘opment of limestones, indicates deep seas and quiet conditions of deposition. Above, the formation grades rather abruptly into red gypsiferous clays and red sandy shales and sandstones. The base of ~R. T. Hill, Twenty-first Annual Report of the U.S. Geological Survey, Part VIL (1899, Ig00), 98. 2 The thicknesses cited are those given by Drake, ‘“‘ Report of the Colorado Coal Field of Texas,” Fourth Annual Report, Texas Geological Survey (1892), 371-446. 118 C. H. GORDON the formation is placed just below the main limestone and the blue shale series, the line marking the boundary with the Cisco coincid- ing approximately with the east line of Shackelford County. The Cisco.—Below the “Albany,” and outcropping to the east of that formation, is the Cisco, which is composed of sandstones and shales, with some conglomerates and two or three beds of coal. Occasional beds of limestones occur in the lower part of the forma- tion and again near the top. Coal outcrops along the Salt Fork of the Brazos River west of Graham in Young County, and else- where to the northeast and southwest. Some of the beds of coal are associated with limestones, in one case a thickness of two or three feet of limestone resting directly upon a bed of coal. The conglomerates consist of sub-angular fragments of flinty blue lime- stone and chert cemented together by a ferruginous sand. Nodules and hollow concretions of limonitic iron ore are common. These conglomerates have been recognized at two different horizons and in widely separated localities. Their exact relations, however, have not been clearly defined. In Stevens County the clays are mostly blue and yellow. Limestones appear at intervals, but these thin out northward, while the clays show a corresponding increase in development. Relation of the‘ Albany” to the Wichita.—When traced northward, the limestones of both the “‘Albany”’ and Cisco formations diminish in thickness, while there is a corresponding increase in the inter- vening beds of shale. In the case of the ‘“‘Albany”’ the limestones show also a change, becoming more earthy and irregular in their texture, and some of the beds passing into gray indurated clays. The few limestones in the upper part of the Cisco formation dis- appear entirely in the northern part of Young County. Along with this change there is an increasing development of red clay, alternating with the blue. The massive beds of limestones con- stituting the upper part of the “Albany” along the Clear Fork in northwestern Shackelford County and in western Throckmorton County were traced northward as far as Beaver Creek in eastern Wilbarger County. They appear in more or less continuous exposures as far north as Seymour, north of which they are covered, but are again exposed, greatly diminished in thickness on Big WICHITA FORMATION OF NORTHERN TEXAS 119 Wichita River and Beaver Creek in the line of their strike north- ward. Greater difficulty is encountered in the effort to trace the lower beds of the “Albany,” owing to the greater proportions of clays and sands and the disturbed condition of sedimentation, both conditions becoming more pronounced as the beds are followed northward. Certain of the limestone beds, however, are persistent, although showing changes in their physical character, and by means of these the eastern boundary of the formation was ascertained with a fair degree of accuracy. At Fane Mountain, a low eleva- tion in the southeastern corner of Throckmorton County, is an outcropping of limestone characterized by an abundance of M yalina permiana. These beds occur at intervals northward in eastern Throckmorton County, and at Spring Creek in the northwestern corner of Young County they outcrop in the bank of the river about a mile from the post-office. Here the beds show locally a gradation into sandstone suggesting near-shore conditions of sedimentation. On Godwin’s Creek, in the western part of Archer County, the diminished representatives of these, or possibly somewhat higher, beds appear, as also farther north on the Big Wichita River. The limestone which outcrops on the Big Wichita north of Fulda, referred to on p. 116, is evidently one of the lowermost beds. The most northerly appearance of presumably the equivalents of these beds was noted in the vicinity of Electra in the western part of Wichita County, where occasional plates of limestone appear over the surface apparently as a result of the weathering out of lenses of limestone in the clays. In the case of the Cisco formation the changes which these undergo toward the North have not had care- ful study. The limestone, however, appears to thin out entirely in the northern part of Young County, there being no representa- tives of these formations in the “‘Red Beds” area except it be the impure, calcareous nodular beds described above. Nowhere in the southern area so far as observed are there any indications of unconformity. Notwithstanding the lithological and faunal characteristics which distinguish the “‘Albany,”’ these beds appear perfectly conformable with the Cisco below and the Clear Fork above, nor is there within the formation any indication of stratigraphic discordance. The change in the lithological character 120 C. H. GORDON of the beds toward the north is evidently the result of differences in the conditions of sedimentation. The character of this part of the formation suggests very strongly its origin on a coastal plain, or river delta, to the south and west of which lay the sea in which were deposited the marine “Albany” sediments. The inter- relations of the two kinds of sediments suggest oscillation of the shoreline upon a relatively wide coastal plain. These changes may be explained as the result of oscillation of the land surface or, possibly better, by the slow but intermittent sinking of the coastal region. As suggested by Case,’ Beede,”? and others, the materials of the “Red Beds”’ were evidently derived from a land mass on the north, of which the Wichita and Arbuckle mountains are the remnants. The following quotation from Beede’s paper is especially pertinent: The Arbuckle and Wichita mountains are probably the source of much of the red sediment in which they are partially buried, and the former mountains are directly responsible for the eastern extension of these beds in central Oklahoma. The extent to which the lighter colored sediments of Kansas and Texas are replaced by red sediments in Oklahoma and near it represents in a rough way the limits of the influence of these mountains on the deposits of the time by the spread of their sediments. By the time the deposition of the light colored sediments had ceased the conditions had become such that nearly all the sedi- ments derived from the land surrounding the basin were red. FAUNAL RELATIONS In the course of the field work collections of fossils were made at many localities, chiefly in the region occupied by the “Albany” beds. At the close of this paper is given a list of the invertebrate fossils obtained from the Albany and Wichita areas. ‘The list includes the collection made by the author, and those made several years since by Mr. E. O. Ulrich. The localities are indicated on the map by corresponding numbers. ‘These remains indicate, accord- ing to Dr. Girty, a marked identity in the invertebrate faunas of the Albany and Wichita areas. In the collections several different faunas can be discriminated. One of these has the brachiopod ~E. C. Case, Bulletin of the American Museum of Natural History, XXIII (1907), 659-64. 2 J. W. Beede, Journal of Geology, XVII (1909), 714. WICHITA FORMATION OF NORTHERN TEXAS 27 element fairly well represented, Derbya cymbula being generally present, and the pelecypod Myalina deltoidea rather abundant. Another contrasting fauna has, as a rule, brachiopods absent or greatly diminished, but is plentifully supplied with large nautiloids. The faunas appear to have been contemporaneous, both occurring throughout the formation, but in different localities. The nauti- loid facies, however, is more prominent in the upper series of beds. The invertebrate remains of this region were studied by C. A. White, who considered them to be Permian. A map on which the localities were shown was prepared for the Fifth Annual Report of the Texas Geological Survey, but never published.? The collections of vertebrates, which in past years have attracted so much attention, were made in the adjoining portions of Baylor and Archer counties. Cope, who first studied them, considered them to be of Permian age. A description of the localities where these remains were discovered has only recently appeared in print. From this description, which is not accompanied by a map, it ap- pears that no fossils were obtained east of the middle of Archer County. In late years interest in the vertebrate remains of the Wichita formation has been renewed and much new material has been obtained, more particularly through the labors of Williston and Case. The results of their investigations have appeared in varlous papers. The plant remains from this region have been studied by Fontaine and White! and by David White. The last named spent several days in the field in rg09 and collected considerable material from two near-by localities, one, two and one-half miles south of Fulda, and the other four miles southeast of that place. As pro- visionally identified this material is as follows: tC. A. White, U.S. Geological Survey Bulletin 77 (1891). 2 Transactions of the Texas Academy of Science (1897), 95. 3W. C. Cummins, Journal of Geology, XVI (1908), 737-45. 41. C. White, Bulletin of the Geological Society of America, III (1892), 217-18. Study based on identifications by W. N. Fontaine. 5 No. 1: Cassil Hollow, two and one-half miles south of Fulda, Texas. No. 2: Breaks of the Little Wichita, one-half mile south of the river, and four miles southeast of Fulda, Tex. The beds are just over the bone-bearing limestone. The species in bold-faced type are characteristic of the Permian. 122 C. H. GORDON Locality No. 1 Pecopteris arborescens Pecopteris hemitelioides Pecopteris densifolia ? Pecopteris grandifolia Pecopteris mertensioides ? Gigantopteris sp. (cf. nicotianifolia) Neuropteris (cf. lindahli) Aphlebia sp. Taeniopteris multinervis Annularia spicata Sphenophyllum ? sp. Sigillariostrobus hastatus Walchia schneideri ? Gomphostrobus bifidus Cardiocarpon n. sp. Carpolithes sp. Pelecypods Estheria and fish scales Locality No. 2 Pecopteris hemitelioides Pecopteris grandifolia Pecopteris candolleana Pecopteris tenuinervis Diplothmema ? sp. Odontopteris fischeri ? Odontopteris neuropteroides Neuropteris cordata Taeniopteris coriacea ? Taeniopteris abnormis Taeniopteris n. sp. Sphenophyllum obovatum Sigillaria sp. (leaf) Gomphostrobus ? sp. Cordaites principalis Poacordaites cf. tenuifolius Walchia piniformis Aspidiopsis sp. Araucarites n. sp. Cardiocarpon n. sp. Insect wings Estheria Anthracosia - Ostracods and fish scales CORRELATIONS That the limestone series of Baylor County is the equivalent of the “Albany” formation of the southern area is fully established by both the stratigraphic and the faunal evidence. The beds in the northern area, which include the limestones, shales, and sandstones of Baylor County and the sandstones and shales of Archer and Wichita counties, constitute the Wichita formation. Our investigations therefore fully support the conclusions of Cummins‘? and Adams? as to the equivalency of the “‘Albany” and Wichita formations. =W. C. Cummins, Transactions of the Texas Academy of Science, 1 (1897), 93-907. ? George I. Adams, Bulletin of the Geological Society of America, XIV (1903), TQI—200. 3 Along with the limestones of northeastern Baylor County which Cummins has designated as the top of the Wichita the writer would include the overlying beds of shale and limestone mapped by him as Clear Fork, which outcrop in the banks of the Big Wichita about a mile east of the Seymour-Vernon road and northward on Beaver Creek. WICHITA FORMATION OF NORTHERN TEXAS 123 Gould! correlated the Clear Fork with the Enid, Blaine, and Woodward formations of Oklahoma. In making this correlation, he evidently followed Cummins’ earlier writings, in which the beds of Baylor County were considered to be Clear Fork. Williston states? that the Enid formation of Gould is identical with the beds of Baylor County. NOMENCLATURE In the paper cited, Adams has contended that the terms Wichita, Clear Fork, and Double Mountain should be discarded as having no stratigraphical significance. In his latest papers, Cummins recommends the abandonment of the term Albany and the use of the term Wichita for the formation. In view of the conflicting statements that have been made as to the relations of the beds called Wichita we were at first inclined to agree with the first- named writer in recommending the abandonment of the term Wichita. Further consideration, however, leads us to conclude that with a revised definition it will be best to retain the name Wichita for the formation overlying the Cisco, which it is now gen- erally agreed should be regarded as of lower Permian age, and to abandon the name “‘ Albany.”’ The series of red clays and sandstones with their included gypsum deposits which in Texas overlie the Wichita formation and to which the names Clear Fork and Double Mountain have been given have not as yet received much study. With the limited amount of knowledge available the attempt to subdivide these beds seems to the author unwarranted, and they are, therefore, here mapped as “undifferentiated Clear Fork and Double Mountain.” CLASSIFICATIONS The Permian age of the beds to which the name of Wichita was originally applied has been accepted quite generally, though there are not wanting those who regard the evidence as unsatisfactory. It was based chiefly upon the vertebrate and plant remains. In the southward, or ‘‘Albany,” area the beds are wholly marine and ™C.N. Gould, Water-Supply Paper No. 154, U.S. Geological Survey (1906), 17. 2 Letter to the author dated August 6, 1g09. 124 C. H. GORDON destitute of both plants and vertebrates, though abounding in the remains of invertebrates. The Pennsylvanian aspect of this fauna has strongly impressed some investigators, including the author of this paper, and doubt was entertained as to whether the plane of separation between the Pennsylvanian and the Permian should be drawn at the base or at the top of the formation. The studies of David White, Beede, and others have contributed much in recent years to a knowledge of the Permian in American and in the main support the view of the Permian age of the Wichita formation. In a recent paper Beedet has ably discussed the Permian of Kansas, with which he correlates the ‘““Red Beds” of Texas. Cummins correlates the beds of eastern Baylor County which he regards as the top of the Wichita formation with the Fort Riley limestone of the Chase group of Kansas. ‘“‘It is quite certain that the Fort Riley horizon is the same as the Wichita of Texas and is at the very top of the division.’ The top boundary of the Wichita formation was drawn by Cummins? at the top of a stratum of red clay over- lain by thin beds of limestone and blue shales at a point on the Big Wichita four miles west of the east boundary of Baylor County. However, as we have shown, beds which are undoubtedly the same as those which appear at Seymour and southward in Throckmorton County appear in the banks of the Big Wichita River some eight to ten miles west of this point. The thickness of the strata included here, which overlie Cummins’ topmost beds, and are here included with them in the Wichita formation, is estimated to be 250 to 300 feet. The whole limestone and shale series of Baylor County, thus included as the upper division of the Wichita formation, is provisionally placed at 450 to 500 feet, and consists, as shown elsewhere, of limestone beds of varying thicknesses separated by varying but usually great thicknesses of shale. How much of this is to be correlated with the Fort Riley lime- stones can be determined only by more detailed stratigraphic and paleontologic studies. Cummins evidently intended to include tJ. N. Beede, Journal of Geology, XVII (1909),-710-29; Kansas University Science Bulletin, 1V, No. 3 (1907). 2W. F. Cummins, Transactions of the Texas Academy of Science, II (1897), 98. 3 Second Annual Report, Texas Geological Survey (1891), 402, 403; also Fourth Annual Report (1893), 224. WICHITA FORMATION OF NORTHERN TEXAS 125 the lower beds only in his correlation. It may be that further studies will show that the overlying beds of the Winfield limestones of Kansas are represented here. DISCUSSION BY GEORGE H. GIRTY The equivalence in a general way of the fossiliferous late Carboniferous beds of Kansas and Texas has long been recognized and in both cases they have very generally been cited as Permian. Cummins,’ partly on stratigraphic and partly on paleontologic evidence, reached the conclusion that the Fort Riley limestone of Kansas occupies a position at the top of the Wichita formation of Texas. The Fort Riley is the middle formation of the Chase group, the lowest group of the Kansas Permian, so that the bottom of the Wichita may well be as low as the base of the Permian of Kansas. This correlation of Cummins is probably the most pre- cise and the best sustained of any, and it is also in accord with some recent paleobotanic evidence. Mr. White states in the present paper in discussing the fossil plants which he obtained from the Wichita formation that the latter is probably referable to the Chase group of Kansas. Not until recently, it seems to me, has adequate evidence been adduced either for distinguishing the Permian of Kansas and that of Texas sharply from the underlying Pennsylvanian or for cor- relating them with the Permian of Europe. C. A. White found the Wichita fauna to have essentially a Pennsylvanian (‘Coal Measures’’) facies, in which, however, certain characteristic Permian Ammonites occur. A similar conclusion seems to be demanded by the evidence of the present collections. In all, 75 species have been discriminated in the Wichita collec- tions which I have studied, the local distribution of which is shown in the table prepared by Mr. Gordon accompanying the present paper. The identifications naturally vary in precision and refine- ment. In many cases it has been possible to name only the genus to which a species belongs. This is sometimes due to the fact that the species is undescribed. In a few instances species have been cited by comparison with others, e.g., Bellerophon aff. harrodt. « Trans. Texas Acad. Sci., II (1897), 98. 126 GEORGE H. GIRTY If such citations are included as species identified, 48 species of the fauna are identified and 27 are unidentified. Of the 48 species identified, 37 are known to occur in the Pennsylvanian rocks of the Mississippi Valley. Most of them are cited by Dr. Beede in his table showing the Pennsylvanian faunas of Kansas. The large percentage of indeterminata introduces a considerable possibility of error in the inference that 75 per cent of the fauna of the Wichita formation consists of well-known Pennsylvanian types, but it is undoubtedly true that in the main this fauna has a Pennsylvanian facies. One or two new forms at present excluded from the identi- fied species would somewhat decrease this percentage. On the other hand, of the 25 per cent which is not known to occur in the Pennsylvanian of the Mississippi Valley, relatively few species are characteristic of the Permian of that area; still fewer, if any, are characteristic of the Permian of Europe. Some of them occur in western faunas, probably contemporaneous with the eastern Pennsylvanian. Bellerophon subpapillosus is one of these. Twenty- five in a hundred, therefore, far overstates the percentage of char- acteristic Permian species. Such percentage, however, might be considerably increased by the inclusion of certain species known to occur in the Wichita formation but not represented in the Survey collections. I refer especially to the Ammonite forms described by C. A. White from the Military Crossing of the Wichita. These are by all means the most diagnostic Permian types of the fauna. How little characteristic of it they really are, however, is shown by the fact that later collections made at the same place fail to contain them, although a special search was made to secure addi- tional representatives. Mr. White finds that.about 50 per cent of the Wichita flora consists of species characteristic of the Permian, while most of the remainder are known to occur in rocks regarded as of Permian age. If we omit the fauna of the. Kansas Permian, to include which would be a sort of circulus vitiosus, no condition comparable to this has been demonstrated by the invertebrate fossils and, in so far as I have seen the evidence, no such condition exists. Iam, therefore, accepting the Permian age of the Kansas and Texas beds, but at present strictly on the paleobotanic evidence. WICHITA FORMATION OF NORTHERN TEXAS Toy, If the upper part of the Carboniferous section of Texas is to be discriminated as Permian, the line of division, as indicated also by the paleobotanic evidence, would probably best be taken at the base of the Wichita. An inspection of the faunas collected from the strata immediately concerned in this report shows a rather noteworthy change of facies between the Wichita and the Cisco—a change, however, which is more or less progressive and has its beginning in earlier beds. This shows itself rather in a limitation than in a change of fauna and in the prominence of certain groups more rare below. Thus the brachiopods, pelecypods, gasteropods, etc., are much less in evidence in the Wichita than in the Cisco, but, as already pointed out by C. A. White," they are essentially the same as those of the normal Pennsylvanian fauna. In the Wichita, however, we have ~a remarkable development of the Cephalopoda, which in the earlier sediments are rare. Just what significance faunal changes of this sort possess it is difficult to say. It seems to be a change comparable to that which is more strikingly illustrated when a thin calcareous sheet with a marine fauna occurs in the middle of a coal deposit. Here, of course, there is an absolute change from the animal life of the cal- careous stratum to the plant life of the coal and roof shale, but in this case the significance is not ambiguous and it is clearly not stratigraphic. So I think the faunal change marked by a substitu- tion of one predominating animal type for another may often be more safely interpreted as environmental than as stratigraphic in itsimport. At the same time the stratigraphic significance may be present also, which would appear to be the case with the Wichita auna, as indicated by the fossil plants. Nevertheless, this change, as marking the evolution from one geologic period to another, would be more impressive if the molluscan and molluscoidean groups were continued into the Wichita and with a difference of facies such as is usually found when the faunas of other systems are con- trasted. U.S. Geol. Surv., Bull. 77 (1891), 30-39. 2 T mean of course that there is usually no time break and no appreciable change of fauna in the general region accompanying the phenomenon. 128 GEORGE H. GIRTY In connection with the correlation of the Wichita formation with the Permian of Europe, it may be well once again to consider the use and definition of the term Permian. As is well known, Murchison correlated with the English ‘‘ Mill- stone grit”? a series of sandy beds which underlies the typical Russian Permian, and therefore this series, to which the name Arta beds or Artinskian was subsequently given, was distinctly excluded from the original or typical Permian. It has since been recognized that the Arta beds are not the equivalent of the ‘ Mill- stone grit,’’ and that the fossils which they contain show affinities with both the “Upper Carboniferous”? below and the Permian above. The Artinskian therefore came to be called also “ Permo- Carboniferous,” and by many writers it is included with the other -under the name Permian. While the typical Permian is usually underlain by the sandstones of the Artinskian, over a considerable and well-defined area a heavy series of limestones and dolomites has been found to intervene. This apparently lenticular mass has been called the Kungur-stufe, and on paleontologic evidence has been by Tschernyschew united with the Artinsk and included under the term “ Permo-Carbonit- erous,’ which, therefore, comprises two divisions, the-Arta beds below and the Kungur beds above. Now, the propriety of including the original Permian and ‘“‘Permo-Carboniferous”’ in a single group is, of course, a question quite apart from the nomenclature which should be used, and it is a question with regard to which one who has not studied the rocks and fossils in the typical region can hardly render an authoritative opinion. There seems to be European authority both for exclud- ing the ‘‘Permo-Carboniferous”’ from the Permian and for includ- ing it with it, the greater number of writers, it may be, adopting the latter course. As for the plants, Mr. White states that “from the paleobotanical standpoint the Artinsk stage of Russia is clearly Permian.” My own knowledge of the facts is only that of the library, but I should judge that the faunal break was greater between the Gschelian and the ‘‘Permo-Carboniferous” than between the ‘““Permo-Carboniferous” and the original Permian. ‘That is, of the WICHITA FORMATION OF NORTHERN TEXAS 1209 very varied brachiopod fauna described by Tschernyschew from the Gschel but a small number of species appear to pass over into the Artinsk, and I infer that much the same is true of other groups. Both for this reason and because the Artinsk seems to introduce a new ‘‘cycle of deposition,” I would be disposed to group the ‘‘Permo-Carboniferous”’ with the beds above rather than the beds below, not feeling, however, that my opinion on this point deserves much weight. Now, while there may be diversity of opinion about grouping together the ‘‘ Permo-Carboniferous” and Permian, all must agree that it is bad usage to employ the name Permian in two different senses, especially for the whole and at the same time for a part. Although the question is international as well as national, the proposition to remedy the present unfortunate condition would come with greater force and propriety from European writers. To me, personally, it is naturally a matter of indifference whether the term Permian is used for the series and a new name introduced for the beds above the ‘‘Permo-Carboniferous,” or used for the beds above the “‘Permo-Carboniferous” and a new name intro- duced for the series.. The former alternative has in its favor the fact of perhaps greater usage; the latter, that it is the original and authoritative usage. I cannot believe that the unscientific procedure of employing the term in two senses will continue indefinitely, and consequently whatever we now do, short of the fundamental courses just named, must be more or less of a make- shift. It does not, perhaps, make much difference which method is adopted in this provisional manner, but as the main object is to be clear and exact, it has to me seemed the better plan to use Permian in the original and authentic sense. It seems to me obvious that the Artinskian and Permian should be assembled under one division or separated into several, entirely as the sum of the evidence from all sources dictates. I have not the personal acquaintance with the beds, their faunas and floras, their field relations, etc., which would entitle me to an opinion of my own as to how they should be classified. It seems to be a moot t Possibly some older name could be revived for the Permo-Carboniferous and Permian, such as Dyas, as suggested by David White. 130 DAVID WHITE point whether the Arta beds should be regarded as a separate division or included with the Permian, and it matters little for purposes of correlation whether an American writer follows one group of authorities rather than the other. Personally, I am quite willing to include them both in a single division of the time scale, and although believing that propriety would be better served by retaining Permian for only the upper division, I am willing to extend that term to cover the entire series because of the usage which it has received in this sense, but I am not willing, for reasons which must be obvious, to call the whole Permian and the upper part - also Permian, and for the sake of precision I have been temporarily calling the upper beds Permian, the lower beds ‘“ Permo-Carbon- iferous,” and the whole ‘‘ Permo-Carboniferous”’ and Permian. If, in my Guadalupian report and elsewhere, I restricted the term Permian to the supra-Artinskian beds, it was done as a matter of procedure in nomenclature. I had no opinion of my own as to the classification of the beds to express or defend, although, if I had, excellent authority could be named in support of my position. DISCUSSION BY DAVID WHITE The plant material collected by myself from the breaks of the Little Wichita River near Fulda, Tex., is derived from two near-by localities, both near the middle of the Wichita formation. The fossil plants previously listed by Fontaine and White from two other localities, and recorded’ by them as Permian, appear to repre- sent a mixed flora, one of the localities being under suspicion of Pennsylvanian age. Neither of the latter two localities was visited by me on account of the lack of time; but on the basis of informa- tion received, I am disposed to believe that the stratigraphically lower beds at Antelope are probably Pennsylvanian. The identifications given on p. 122 are provisional. Later it is hoped, when the material will have been increased both geographi- cally and stratigraphically, a formal report covering the floras of the ‘Red Beds” will be prepared. The species printed bold-face in the lists on p. 122 are characteristic of the Permian. They point somewhat distinctly to the Rothliegende age of the beds. t Bull. Geol. Soc. Amer., III (1892), 217. WICHITA FORMATION OF NORTHERN TEXAS 131 All the Old World species in the lists occur in the Permian of - western Europe, and of the remaining species apparently every one which is not new is found in the Permian of Kansas. Taeniop- teris, in simple fronds, is represented by several species character- istically lower Permian. Other types proper to the Permian are the Odontopteris form, the genus Gomphostrobus, Annularia spicata, the Sphenophyllum forms, one of which approaches S. stowken- bergit, and the scales provisionally referred to Araucarites, while the presence of Walchia assures a horizon as high as the highest ‘Coal Measures.” The presence of Gigantopteris, abundant at locality No. 2, is particularly notable since the genus is not definitely known except from the coal fields of central and southern China, where it occurs in beds associated with the coals overlying other terranes which, on the evidence of their contained invertebrates, have been referred by the French geologists to the lower Permian. The genus is certainly close to, if not actually identical with, a form described from several small fragments from the Permian of the Ural region. In accordance with the paleobotanical standards of western Europe, I refer the plants of the Little Wichita in Texas to the lower Permian, the terranes being probably referable to the Chase group in Kansas. In this connection it should be observed, how- ever, that the Artinskian flora of the Urals is essentially Permian, and that paleobotanists universally agree with the general usage of the geologists of western Europe in referring the Artinsk to the Permian. DESCRIPTION OF LOCALITIES Notre.—The number at the left is the locality number as given at the head of the list and indicated on the map. The first numbers following the description of locality are the Survey permanent record numbers, the second the temporary or field numbers. t. Bar-X Crossing, Big Wichita River, three miles north of Fulda Station, 5247 (Gt. 67). 2. Bluff of Wichita River, one mile west of Bar-X ranch house, 5243 (Rt. 20). 3. One mile east of Old Military Crossing, Wichita River, 7025. 4. Two miles north of Wichita River, near Old Military Crossing. 70250. DAVID WHITE 132 xx xX xX xX ie ae wmnpunjoid wnq{fAydoydoT POO OD G00 05 76 646 050-0 6 ds peices BIIVUOGIVI “je [eee eo oaks ny30 “ye BNsULy oiReginicelsetahiehvass © elustie/ alavh stlelse ds viyipsoda’y Getetene oc: Os Ov0.0-c oy Om. os ds V][OSOUI | 0 OO roegeo. Ura OO nt bao O8o ds eviodinysty one die liaelisleaniefiokselachtsta site tle 0-0 ds eulynsn yy ae ee snsoyidedqns snwoydnyy eidepesrejiene hehielhelsemenviienhssle ds Aen saployfnes ae aa snsopousod snyeyduiony Sites Daten prim nah (UOSMEP PLIOASH| Ore 350 FR0 he genes ea CSeSTULO TUG Sint afiel atvellehoitepetlelielscifelie: eerie, av ds eile latte “ene coemeliovebie) oe, eynquiAd vAqio(y ov OW eel op iete letaiel esietre liek salts. ds wni[eyuaqd Berry sl[vquapios0 usqded07][9q ie i SIE Gr as 9H dIV[NQO[S Ssv1Id0[OD) pigeulenteriefowiseltavist aie eyyIqns eyisodwio;) oh sey Te\leriosia devreuteielisivelistteliabint a ds soyouoy) as ee ha) ake vueysou sisdouvong el elsecamsbiatia aries evese® alissiaie ds SO ees (porey “ye SE Cute Si snssvio uoydoraypag eto tcerateelahistishetvecis teil) strahy/s\%al ds d VI[[OMOYV Rea ca mace BOLIJUIOUOD PI[IIILISY SROs Da en De cbatltea ds fre: eet efielreltes.e eynovied ee is sisuaysviqou vuurdinouay “ o-oo tTro0-0 Otho 60-0 ds stqaoitdg atone SlUvpRoIo}UL vUTOpoIeRYyds sae Reece sljetiestq viodo}das ia emer S sisuoyodo} BIyIMaSpaS A tyod Sf ppd c oso) Ove Gen. O ds Sateen nace moran tea cas snssoiduio9 snpoztyas * saploipuspopride] viodoquioy yy Pane ge SNYR[NITJOITUIS Be ici Grete rar ica IO) snyONporg Omit eto oO Orosdetna ted ioc ds Sit aaa snyeounoqns foe apes tunoyyeo snioydoins[g at ia aea vuuvo “ye vydAsorseylg alee] is, seigeliemeite) ter cemehenie)terte, ds viodoyAyd sree aes lo[quinp seiad0oRY g “+ WINUBITAOJJUOUL WNTISOT[PEI9q DS boo oto Do yoput epodAsapag Oo ant, 0-0: OF0 010 0 0.02050 ds svi99004}IO DOGO =0)lo Of 0c0-0-0 WyIIAVq “ye BTNONN eitewaePteptondTiepteleqieteshisikelaeitel-tetl ovis ds snyien CtstoD 0 OD Goo 1 oONd 3 ds sisdooleN 0 00-0 Gs080 0) 050-00 vurtuiod naire Soup eae - paployap Pe eae eyenus}}esod vulpes JL eliecie\to Selle) altel: sb (0) sale 0, « ds sors + BIgaIa] "Ye eTuosIyoINyy area BYLISOIRINS LI[OYII eles Misiculeelle(veuielrelie (‘ds *AQS) BUIIUOXO'T 134 DAVID WHITE . Four to five miles northwest of Old Military Crossing, Wichita River, 70256. . Near Old Military Crossing, Wichita River, 7025d. Three miles northwest of Fulda Station, near No. 2, 7026. 8. Four miles northwest of Mabelle Station (section house), 7028. 9. Eight miles southeast of Seymour, 5998 Deep Creek. to. Head of Godwin Creek in eastern Baylor County, 7031, 70314. t1. Godwin Creek near county line on Seymour-Archer city road, 5242 (Raazo) 12. One to two miles northeast of Spring Creek, Young County (Rt. 30). 13. One mile west of Spring Creek, 7035 (Gr. 12, Gr. 14). 14. Five miles south of Spring Creek, in Butte, 5216 (Gr. 11). r5. Seven miles south of Spring Creek, Young County, 5217 (Gr. 15). 16. Rocky Ford, Salt Fork, southeast corner of Baylor County, 5218 (Gr. 9). 17. Quarry bank of Salt Fork, Seymour, 5220 (Gr. 7). 18. West bank of Salt Fork, eight miles south of Seymour, eee (Gr. 35). tg. Nine and one-half miles south of Seymour, Miller’s Creek, 5221, 5224 (Gr. 34). 20. Buttes, near wagon road, half-way between Throckmorton and Sey- mour, 7036 (Gt. 32). 21. Three miles north of Throckmorton, 5215, 5227 (Gr. 30). 22. Five miles west of Woodson, Throckmorton County, 5219, 5223, 52220 (Gr: 25, Gr. 26, 'Gr.27)). 23. Fane Mountain, three miles southwest of Murray P.O., 5226 (Gr. 24). 24. Paint Creek, southeast corner of Haskell County, 5245 (Rr. 32). 25. Clear Fork, near southeast corner of Haskell County, 5231, 5232, 5240 ( Birn6, Bin 7 bis): 26. Round Mountain on the Clear Fork, near 25, 5237, 5237@ (Rt. 34). NOLES SON THE OSTEOLOGY OF THE SKULE, OF PARIOTICHUS E. B. BRANSON The University of Missouri In the summer of tg10 Dr. S. W. Williston asked the writer to study the Pariotichus skulls in Walker Geological Museum of the University of Chicago to see if they would throw any light on some of the undecided points concerning the osteology of that genus. The material was fragmentary with the exception of one remark- able specimen of Pariotichus laticeps Williston, a skull of Parzoti- chus aguti Cope ?, and the base of a skull of an unidentified species. Some of the undecided questions were: Are squamosal and pro- squamosal both present? Is there a distinct quadratojugal ? What are the homologies of the tabulare, if such a bone is present ? What are the homologies of the so-called epiotics, quadratojugals of Case? Is a presphenoid present ? and What is the arrangement of the bones in the base of the skull ? The writer’s thanks are due Dr. Williston for the use of the specimens and for discussions during the investigation. In a paper published in 1878" Cope gave the name Parvotichus brachyops to an imperfect skull from the Permian of Texas, and later in the same paper described a more perfect skull as Ecto- cynodon ordinatus. As he supposed that the former had the roof of the skull unsculptured he referred the specimens to different genera. In 1882 he described Ectocynodon aguti? and in 1888 Ecto- cynodon incisivus3 In 1896 he referred all of the Ectocynodonts to Pariotichus and named two more species, P. aduncus* and P.isolomus.s In the paper where he named the latter he described t Proc. Am. Philos. Soc., XVII, 508. 2 Tbid., XX, 290. 3 Trans. Am. Philos. Soc., XVI, 290. 4 Proc. Am. Philos. Soc., XX XV, 135. 5 Tbid., XXXIV, 145. 135 136 E. B. BRANSON Captorhinus angusticeps, which has recently been referred to Pariotichus by Broom.’ In 1909 Williston described and figured Pariotichus laticeps.? Roof of skull.—No separation of the squamosal into two bones was observed in either Pariotichus or the closely related genus Labidosaurus. Williston first called attention to this in Labido- saurus’ and Broom shows no separation in his figures of Pariotichus.4 ‘A quadratojugal is present in its normal position in the temporal region and this bone is also present in Labidosaurus. Its distinct- ness is not apparent in the type specimen of Pariotichus laticeps Williston, and was first noted in a specimen of Labidosaurus recently acquired by Walker Geological Museum, and corroborated by examination of other specimens. Dr. Case calls an element in the base of the skull the quadratojugal, but it seems to be a part of the squamosal. This part of the squamosal is indicated by the numeral “2” in Fig. 3. In a specimen of Labidosaurus figured by Williston, this part of the bone seems to be separate, but in all other specimens examined there is no evidence of separation. Dr. Williston worked over all of the skulls of Pariotichus and Labidosaurus in the Walker Geological Museum to see if we agreed -on this point and we are now in accord in saying that this is prob- ably not a separate element. In 1883 in describing Pariotichus megalops, since referred to Isodectes, Cope said: ‘‘At the extreme posterior angle is a very small element in contact with the supraoccipital which may be the true intercalare.’’> In 1896 he figured this bone in Pariotichus aguit Cope,° and Case? and Broom! figure it in Pariotichus angus- ticeps Cope. It is present in the form figured in this paper; and in one or two other specimens of Pariotichus examined by the writer t Bull. Am. Mus. Nat. Hist., XXVIII, 218: 2 Biol. Bull., XVII, 241-55. 3 Am. Jour. Anat., X (1910), 74. 4 Bull. Am. Mus. Nat. Hist., XXVIII, 218. 5 Proc. Am. Philos. Soc., XX, 630. 6 Am. Naturalist, XXX, Pl. VII. 7 Bull. Am. Mus. Nat. Hist., XXVIII, 194. 8 [bid., XXVIII, 218. OSTEOLQGY OF THE SKULL OF PARIOTICHUS 137 it is distinctly separated from the parietal, but there is no indica- tion of it in Labidosaurus. Cope also applied the name tabulare to the element and recently Broom has suggested the name post- temporal. ‘There seems to be no valid objection to tabulare and it has the advantage of priority over Broom’s name. All writers seem to be agreed about the rest of the bones in the roof of the shull. Base of skull.—The bases of several skulls examined during the investigation were fairly well preserved and the one from which Fig. 3 was made is almost perfect. This shows the post-parietals in the same position as figured by Williston in Labidosaurus' and by Case in Edaphosaurus? and Pariotichus. (Case calls them epiotics in Edaphosaurus. ) The exoccipitals are large and articulate with the squamosals after passing in front of the inturned edge of the latter, the quadra- tojugals of Case. The stapes, tympanic of Broom, articulates at - its distal end with the lower inner end of the quadrate. In the drawing it is not shown distinctly separated from the exoccipital, the sutures not having been determined. ‘The separation in this form is probably as shown by Williston in Labidosaurus. The position of the quadrate is almost vertical with a broad bladelike process above and a heavy expanded portion below. The bladelike portion projects forward almost parallel with the median line of the skull, and the posterior end of the pterygoid rests against it. Its upper end comes in contact with the squamosal and the outer side of the base touches the quadratojugal. Floor of skull—The pterygoids extend from near the posterior end of the skull almost to the anterior end. They meet in the median line and are not separated by the basisphenoid as shown by Broom in Pariotichus angusticeps Cope. The sutures between the long slender palatines and the pterygoids were made out in one specimen from the anterior end to near the posterior end, as shown by solid lines in the drawing. There are strong indications of a transverse as shown by broken lines in Fig. 4, but the evidence t Amer. Jour. Anat., X, Pl. III, Fig. 4. 2 Revision of the Pelycosauria of North America, 1907, p. 153. 3 Bull. Am. Mus. Nat. Hist., XXVIII (1910), 218. 138 E. B. BRANSON is not entirely convincing. The presphenoid is perfectly preserved in one specimen in Walker Geological Museum but is lost in all others examined. It is slender and extends about half the distance from the basisphenoid to the anterior end of the skull. ‘The sutures between the vomers and palatines are not evident in any of the specimens studied. In the specimen shown in Figs. 1 and 2, which is probably Partoti- chus aguti Cope, though it has only two teeth on the premaxillaries, the teeth on the maxillae are in one row to behind the fifth where a second row begins inside the first, and behind this two other rather indistinctly defined rows appear. Nearly all of the teeth are sub- circular in cross-section near the base, but some of the posterior ones are more or less compressed laterally. The distinction between Pariotichus and Labidosaurus made by Cope, that the latter had the teeth of the maxillae in one row, breaks down in the Walker Geological Museum specimens. In two specimens examined during the present investigation a second row of teeth is evident and in other specimens the preservation is not such as to show whether there is more than one row. Mandible.—The dentary makes up the outer part of the anterior half of the mandible. Just behind the dentary there is a short coronoid which occupies about one-third of the width of the jaw and sends upward a very large coronoid process almost equal in width to the rest of the mandible. Behind the coronoid the angular makes up most of the outer part of the posterior half of the jaw, and also sends forward a slender process between the dentary and the splenial, which reaches almost to the tip of the mandible. Above the angular and separated from it by a suture that runs diagonally across the jaw and passes to the posterior inferior corner, there seems to be a surangular. The splenial is not well preserved in any specimen observed and all that can be determined is that it is a broad flat bone on the inside of the jaw. The articular is imper- fect in all of the specimens, but in a perfect specimen of a closely allied form, found completely separated from the other bones of the jaw, it is heavy at the posterior end and sends a long slender process forward. ie emai foe URN oct PLATE [ JOURNAL OF GEOLOGY, VoL. XIX, No. 2 OSTEOLOGY OF THE SKULL OF PARIOTICHUS 139 EXPLANATION OF PLATE (The figures on this plate are natural size) Fic. 1.—Top view of skull of Pariotichus aguti Cope. Fic. 2.—Lateral view of skull of Pariotichus aguti Cope. Fic. 3.—Base of skull of Pariotichus, species unidentified. Fic. 4.—Floor of skull of Pariotichus restored from three specimens. I,parietal; 2,squamosal; 3, postorbital; 4, quadratojugal; 5, jugal; 6, frontal; 7, prefrontal; 8, nasal; 9, maxilla; 10, premaxillae; 11, dentary; 12, angular; 13, coronoid; 14, surangular; 15, tabulare; 16, vomer; 17, pterygoids; 18, palatines; 19, transverse; 20, quadrate; 21, basisphenoid; 22, basioccipital; 23, stapes; 24, exoccipital; 25, supraoccipital; 26, postparietal. (As the surface sculpture was not well preserved in any of the specimens no attempt was made to reproduce it exactly in the drawings.) Lined areas are restored. HIGH TERRACES AND ABANDONED VALLEYS IN WESTERN PENNSYLVANIA’ EUGENE WESLEY SHAW U.S. Geological Survey, Washington, D.C. The terraces with which this paper has to do are the well-known gravel-covered rock shelves found along the Allegheny, Mononga- hela, and other large streams of western Pennsylvania, about 200 feet above present stream channels. The abandoned parts of valleys are closely associated with the terraces, being found at the same elevation, and in many places the two are connected. Fig. 1 shows the principal areas of high terrace. The region includes all the Ohio River basin above New Martinsville, where there was formerly a divide. There are, however, terraces and abandoned parts of valleys of the same age on the Kanawha, Guyandot, Big Sandy, Kentucky, and other streams. The impressiveness of these features is attested by the long list of names of eminent men who have studied and described parts of them. This list includes Stevenson, Leslie, Jilson, Chance, Wright, Chamberlin, Gilbert, I. C. White, Tight, Canaalell E. H. Williams, Leverett, and others. Some of the earliest workers believed that the terraces were due to a submergence and marine erosion. Stevenson in 1879 (Proc. Am. Phil. Soc., XVIII, 289-316) called attention to benches along the valley of the Monongahela and its tributaries. He divided them into a higher series of twenty benches, and a lower - one of five. The higher series he attributed to marine action. They are probably entirely above those under discussion, and later work on them has shown that they are obscure and are probably due to hard layers of rock. The lower series of Steven- son seems to include those under discussion, and he refers them to stream action, without going into details of development. « Published by permission of the Director of the United States Geological Survey, Washington, D.C. 140 Fic. 1.—Principal areas of high terrace. Black areas have glacial gravel; those in outline have local gravel only. 142 EUGENE WESLEY SHAW In 1883 Professor G. F. Wright presented evidence of a large ice dam at Cincinnati, and shortly thereafter Professor I. C. White, in a paper before the American Association for the Advancement of Science, referred the terrace deposits of the Monongahela to that dam. Chamberlin, in 1890 (Bull. U.S. Geol. Surv. No. 58, 13-38), showed that the upper series described by Stevenson could not be ascribed to the ice dam, because of their great range in altitude. He also pointed out certain characters of Stevenson’s lower series which indicated that they were of fluviatile, not lacustrine, origin. These characters were: (1) the terraces slope with the present streams; (2) the material capping the terraces is distinctly fluvial; (3) they are rock platforms; (4) the form and distribution of the terraces is of fluvial, not lacustrine, order; (5) the abandoned channels must have been of stream origin. In 1896 Professor White expressed himself (Am. Geol., XVIII, December, 1896, 368-79) as still convinced that the glacial lake, Monongahela, did exist and was responsible for the terrace deposits, but that the ice dam was probably not at Cincinnati, but in the vicinity-of Beaver, Pa. In the Masontown-Uniontown folio, published in 1902, M. R. Campbell advances the theory that the deposits and abandoned channels are due to local ice dams which formed in Kansan time. He points out the fact that it is an extremely difficult and slow process for a stream to cut off any of its meander in a rugged region like Pennsylvania, and that it is impossible for a stream to establish a totally new course unless the conditions under which it operates are very different from those which normally affect the develop- ment of streams. Again, as an objection to the view of Professor White, Mr. Campbell states that while it would be possible for a stream to change its course by superimposition if it were first caused to silt up its valley and then permitted to cut down again, he finds that part of the Carmichaels abandoned channel was not so silted up, and he therefore concludes that the change of course was not due to silting up and superimposition, but to local causes. Mr. Camp- bell’s idea is that ice jams formed in glacial time and that these TERRACES AND VALLEYS IN WESTERN PENNSYLVANIA 143 grew until they formed huge dams too or more feet high, and that they persisted until deposits over 100 feet thick accumulated above them. In many cases these dams not only gave rise to terraces but caused the rivers to abandon their old valleys and cut new ones. In the Amity folio Frederick G. Clapp expresses the belief that Professor White’s theory—that of ponded waters throughout much of western Pennsylvania—will best account for the phenomena. He states that the upper limit of the stream deposits in all the val- leys of southwestern Pennsylvania and parts of adjacent states has a vertical range of but little over 100 feet, but since Mr. Clapp’s work was published the gravel has been found to lie at an elevation of over 1,200 feet on Clarion River, making the vertical range more than 200 feet. . The data gathered by the present writer, instead of lending support to any one of these views, seem rather to indicate that the high terraces and abandoned channels on all the rivers developed as a unit, through the overloading of the Allegheny in early glacial time. The terraces may be divided into two groups, which have certain essential differences. Those of the first group are capped with glacial gravel, and are found along the Allegheny and Ohio. Those of the second bear material of local derivation, and are found on streams tributary to the Allegheny and Ohio. There are other differences which will be brought out later. In this connection it should be stated that there are a few remnants of older gravels, which lie at various elevations above the main high terrace forma- tion, and in some places have been let down by erosion, so that they seem to connect with the much more extensive deposit below, but the older gravels have very slight extent. TERRACES OF THE ALLEGHENY AND OHIO The terraces of the Allegheny and Ohio are almost continuous from the mouth of the Clarion to Pittsburgh, and on down the Ohio. The gravel deposits on them are thin or absent where crossed by lateral streams; in other words, where erosion has been most severe; but enough remains to indicate clearly the position of the 144 EUGENE WESLEY SHAW original upper surface. At over one hundred places the upper limit of gravel has been determined by level, and that limit is in all cases very nearly 300 feet above present low water. ‘The eleva- tion of the rock floor beneath the deposits has also been determined at many points, and is found to be a little less than 200 feet above the present position of the rivers. Thus, the upper limit of gravel TABLE SHOWING ELEVATIONS OF HIGH TERRACES IN WESTERN PENNSYLVANIA : 2 Upper Limit Place sale from Wap Tent eck oon | Beets ec ee ent Stream. Foxburg quadrangle *One mile north of Callens- Ueto nual saree IIo 1,180 1,160+ 970 230 > Rurnipholese a nee 108 1,170 I,120— 930 240 1,160 Mouth of Clarion River.... 102 I,150 1,035 846 304 Mouth of Bear Run ....... 99 1,145 LOZ Gini moAO 305 Montene yas: ion wa steiner: 96 1,140 THOUS) |e O32 308 Kittanning quadrangle Redbank san cms sanieicnes 81 . 1,100 950 810 290 sHorde@ityaiGe ssc voce ee 538 1,025-+ 885 763 262+ and 980 New Kensington quadrangle Wharentum Ge snr ane arS 30 1,000-+- 075 725 Onis a Carnegie quadrangle | | INIFANS N55 obo ob abu oo sos 22 T,000 896 698.4 300-+ Beaver quadrangle BeaVviersna city ie ° 978 goo 672 306 Latrobe quadrangle *One mile northeast of | Blairsville eee 80+ | 1,060 sim goo 160 Burgettstown quadrangle *One and one-half miles northeast of Burgettstown 28+ 1,028 1,015 047 81 *Gravel of local derivation (not glacial). falls regularly from 1,145 feet at Foxburg to 1,010 feet at Pitts- burgh; the rock floor beneath the gravel from 1,015 to about 880 feet, and the river from 845 to’7oo feet. Here, then, are three approximately parallel planes, each of which slopes about 140 feet in 80 miles. In other words, the gravel formation holds its thickness of about 125 feet, and slopes in the direction of present stream flow. See table. The pebbles are well rounded, and lie in a matrix of sand and TERRACES AND VALLEYS IN WESTERN PENNSYLVANIA 145 clay, though in some places there is so little fine material that the gravel is dug from pits and used without further washing. In such places beds of gravel are separated by lenses of clay, but on the whole the formation is homogeneous. That the deposit is of fluviatile and not lacustrine origin seems to be shown decisively by the characters to which Chamberlin has called attention: The deposit slopes regularly with the present streams throughout their winding courses. A lake deposit would be horizontal unless affected by crustal deformation, and in that case the slope would not change direction at just the places where the course of the river changes. Second, the material is distinctly fluvial, consisting of irregularly bedded gravel which contains lenticular masses of silt and clay. A lake deposit in a valley might have deltas containing some coarse material, but in no way could coarse glacial débris, poured into the end of a narrow lake too or more miles long, be evenly distributed so that the resulting forma- tion throughout its length would be homogeneous and of uniform thickness. A ‘There seems to be good evidence also, as Leverett has pointed out, that in pre-Glacial time the Clarion was the headwater portion of the Allegheny, a divide crossing the present course of the latter stream just above the mouth of the former, and that the glacier, by. cutting off the outlets of the drainage of the area to the north, forced the water to cut across the divide to the old Lower Allegheny, thus thrusting greatness upon the Allegheny basin. Through the new cut were discharged great volumes of glacial outwash—too great for the Allegheny to transport—and_ the coarsest part of the débris was spread along the bottom of the valley, forming a typical valley train which had a nearly uniform thick- ness throughout its length. The bodies of gravel on the high terraces of the Allegheny and Ohio, then, are the remnants of this old valley train. ‘The overloaded condition of the Allegheny was probably due to several causes, among which the following may be mentioned as being more or less effective: First, an actual increase in load derived from (a) material fed more or less directly to the streams by the glacier; (b) débris from the cutting of new gorges across 146 EUGENE WESLEY SHAW old divides; (c) material brought after the ice melted, by tributaries as they cut new valleys. Second, a decrease in velocity and carry- ing power, produced by (a) the attraction of the ice mass; within a degree of the ice front this may so have changed water level that in a stream flowing away from the ice a gradient of 1? feet per mile might have been reduced to 13 or 14 feet per mile; (0) crustal deformation, due to the weight of the ice; (c) the divides crossed; each of these would check the velocity and cause deposits for a short distance upstream; and ice jams operate in a similar way. Third, a possible but not probable decrease in volume, arising from a change in climate. It is probable that during Kansan time the river had a larger volume than now because it was carrying the run-off from a much larger territory. TERRACES OF TRIBUTARY RIVERS The second group of high-terrace deposits is found on streams tributary to the Allegheny and Ohio. Those along the Clarion River may be taken as typical and described in detail.‘ At Fox- burg the high gravels of the Clarion connect and mingle with those of the Allegheny, both the rock floors and the upper surfaces of the deposits connecting, without interruption (see Fig. 3). The Clarion gravels are much like those of the Allegheny, but differ from them in the following respects: First, the material is of local, not glacial, origin. Second, the thickness decreases upstream. Third, the gravels are as a whole much finer, only the base being as coarse as the glacial gravels. There are some minor points of dissimilarity, but these are the important ones in the present discussion. In present distribution the gravels are as continuous as those of the Allegheny. There is scarcely a half-mile of the lower part of the valley where they are absent or even approximately so. That the Clarion terraces are of stream origin is shown by char- acters similar to.and as decisive as those of the Allegheny terraces mentioned above, and certain important features indicate the imme- diate cause of the accumulation of gravel. First, at the confluence of the two rivers the high-terrace gravels correspond exactly in elevation and thickness. Second, the Clarion gravel rises and * See Foxburg-Clarion folio, U.S. Geol. Surv. (in press). TERRACES AND VALLEYS IN WESTERN PENNSYLVANIA 147 becomes thinner and narrower upstream, and at a distance of 20 miles from the Allegheny the formation has the width, thinness, and coarseness of an ordinary flood-plain deposit. These facts suggest at once that the Clarion terraces owe their existence to conditions on the river into which that stream dis- charged. When the Allegheny began to aggrade, the effect was that of a gradually growing dam across the mouth of the Clarion. This caused the latter stream to drop the coarsest part of its load. The dam did not grow so rapidly as to produce a pond in the river above, but aggradation kept pace with the growth of the dam. In other words, at its mouth the Clarion built up as rapidly as the Allegheny. This is shown by the even downstream dip of the Clarion gravels and by their coarseness. If any ponded stage had existed, the deposit would have been coarse only at the upper end of the pond and would have taken the form of a delta. But of course the dam did not affect the Clarion throughout its length. On the contrary, when the dam began to grow, its influence was felt only in that part of the stream immediately above. As it grew the area affected by it extended farther and farther from the Allegheny and the river built up to a new gradient, over which it was just able to carry its normal load. The coarser part of the gravel was dropped where the gradient changed from the old to the new. This point gradually moved upstream and the extended coarse deposit became the basal coarse part of the forma- tion. The Clarion then silted up because its master stream, the Allegheny, was aggrading, and the elevation of its outlet was being raised. The Allegheny aggraded because of great increase in load, the Clarion because of decrease of gradient. The absolute load of the latter stream has not changed materially since the dawn of the Quaternary period. Space will not permit of complete description of all the high terraces, but the work of the Clarion may be taken as a type of the work of those streams which discharged into the overladen Allegheny and Ohio. Redbank Creek, the Conemaugh, Kiski- menitas, Youghiogheny, and Monongahela show similar characters. On all except the smallest of the tributaries of the Allegheny, there are deposits connecting with the early glacial valley train, such 148 EUGENE WESLEY SHAW deposits rising, thinning,.and narrowing upstream, and consisting of mixed coarse and fine material of local origin, the proportion of fine being somewhat greater. than in the valley train. The larger the tributary the more gradually does the deposit rise and thin, for the larger streams have less fall, and there is less difference between the old gradient, with which the streams were more than able to carry their loads, and the adjusted gradients, with which the streams did neither cut nor fill. To illustrate, certain facts indicate that in pre-Glacial time the lower 50 miles of the Monongahela had a fall of about one foot Fic. 2.—Longitudinal section of deposit on a stream tributary to one which is overloaded with glacial débris. A distinguishing character of such deposits is that they are definitely limited upstream by the convergence of the old profile in use when the stream was cutting down, and the adjusted profile with which the stream is just able to carry the load delivered to it by headwaters and side streams. per milé, that the adjusted gradient was about g inches per mile, and that the valley train athwart the mouth of the river was about t1o feet thick. At this point the Monongahela fill should have been tio feet thick, and this thickness should have decreased upstream by 1 foot minus 9 inches, or 3 inches per mile, and at 50 miles the formation should have been thinner by 130 inches, or 125 feet. The results obtained by actual observation in the field accord very closely with these figures. The deposit is nearly too feet thick at the West Virginia line. As another example, the Clarion gravels thin almost 50 per cent in 1o miles. Originally, as indicated by the base of the gravels, the stream had a fall of about 125 feet in the lower 10 miles of its course. The adjusting of the gradient reduced this to 60 feet. The difference between these figures, or 65 feet, plus the thickness of the deposit at the upper end of the ro miles, or 50 feet, is 115 feet, which is the thickness of the deposit at the lower end of the valley. TERRACES AND VALLEYS IN WESTERN PENNSYLVANIA 149 To sum up, the inferred history of the terraces reads about as follows. The Allegheny was overloaded at a certain point. The material was spread out evenly from the place of overloading. On each tributary stream deposits accumulated, first at the point of junction with the overloaded one, then farther and farther upstream. ‘These processes continued until the load and gradient of the Allegheny were so adjusted that the river was able to carry its load. Later, probably on account of an elevation of the land, the stream has cut through its deposit and 200 feet below the level of the old rock floor. ‘““ABANDONED CHANNELS”’ In close association with the high terraces are the many so-called abandoned channels or side tracks to the main lines of drainage. Examples are found not only in western Pennsylvania, but along the Ohio, Mississippi, and a large number of tributary streams. Genetically these features seem to be similar, though some devel- oped early in the Quaternary period and others later. The abandoned part of the Monongahela valley at Carmichaels, Pa., referred to on p. 142, has been described as containing evidence of a huge local dam of ice, but to the present writer the evidence did not seem to indicate a local barrier for the following reasons: (1) The deposit thins at the position of the supposed dam not abruptly, but irregularly, and a mile or more below considerable thicknesses are found. (2) Just below the place of thinning, the formation extends up the valley side to the altitude of the upper limit of gravel, and a little farther away are extensive bodies of the deposits, fully roo feet thick. (3) The thinner parts are found at a place where erosion has been very severe—where the gravel has been dissected by a good-sized tributary. It appears, therefore, that the thin part of the deposit is simply a result of irregular clear- ing-out of the old valley by the tributary and is a feature to be expected. The stream seems to have cut down quickly through the silt and gravel, but when it came to hard rock it hesitated, mean- dered a little, and then cut down farther, leaving the shelf covered with pebbles and bowlders concentrated from the original deposit. The fact that just below the site of the dam the formation is found 150 EUGENE WESLEY SHAW today extending up on the side of the valley and a few miles away the full thickness of over too feet is present, is evidence that the deposit was formerly roo feet deep here as it is elsewhere. There could scarcely be any other possibility except that the valley-side deposit represents an older fill, and there is no foundation for such an assumption. A theoretical consideration of the question of local ice dams yields interesting results. The possibility of an initial ice jam is not to be questioned. Moreover, the supposition that such a jam might be large in a northward or iceward flowing stream in a subglacial climate is reasonable and is supported by known con- ditions on the McKenzie and other streams which work under somewhat similar circumstances. But the ice dams in this case must have been several times as high as the highest known and must have persisted through many summers warm enough to melt back the thousands of feet of ice in a continental ice sheet. Indeed if we assume that the Monon- gahela carried the same amount of suspended matter which it carries today (in all probability it did not carry so much), that all its load of undissolved matter was dropped, and that immediately after the reservoir became filled the dam went out, we get a minimum estimate for the life of the dam of about 1,000 years. If only a quarter of the material were dropped the time would be 4,000 years. During this time the run-off of the basin must have passed over the dam, for if the dam had suddenly risen above the height of cols in near-by divides, the lake immediately behind the dam would not have been silted up. Moreover, considerable coarse material is found just above the supposed dam, indicating that there were strong currents and that only a small fraction of the suspended matter was dropped. The hypothesis of an ice dam, therefore, involves the assump- tion that the Monongahela, which since early glacial time has, with a very low gradient, removed rock material to a depth of 200 feet for more than too miles, was for centuries unable to cut through or undermine these blocks of ice over which its gradient and eroding power must have been that of a cascade or waterfall. The assumed floor of the valley below the site of the dam is 60 feet below TERRACES AND VALLEYS IN WESTERN PENNSYLVANIA 151 the top of the fill above the dam. The drop in water level must have been as great or greater, and yet the dam must have withstood the pressure and the wear year after year for thousands of years. Parker oxbow.—One of the most famous of the abandoned val- leys is the old oxbow at Parker’s Landing (see Fig. 3). It was first described in detail by Chance (Second Geol. Surv. of Pa., Rept. VV, 1880, 17-22). He calls attention to the disproportion- ate size and breadth of the valleys of the two small streams which now flow from the oxbow, and also to the fact that between the heads of the streams there is low swampy ground. Glacial gravels of probable Kansan age are found almost continuously around the loop and in some places the deposit is over 50 feet thick. Chance inferred that at the time of the earliest ice advance this oxbow was occupied by the Allegheny River, and at a subse- quent time the neck was severed. G. F. Wright held that this channel was formed and abandoned before glaciation, and that the glacial material now found in the oxbow was deposited there at a time when the Allegheny, being overloaded with Kansan outwash, aggraded up to a position some- what above the oxbow; that the gravel was carried into the ends of the loop, but the river never reoccupied the entire loop. Wright has long advocated the idea that the Allegheny was cut to about 50 feet below its present channel in pre-Glacial time, and that the glacial valley train was thus about 350 feet thick, filling the inner gorge and part of the broad valley above. Chamberlin and Gilbert studied the problem in 1889, and their conclusions agree essentially with those of Chance, and are found in Bulletin U.S. Geological Survey No. 58, 31. In 1894 Wright again published a paper (Am. Jour. Sci., 3d ser., XLVII, 173-75) in which he holds to his previous conclusions. In 1900 E. H. Williams presented a paper at the Albany meeting of the Geological Society of America (Bull. G.S.A., XII, 1900, 463) in which he agreed with Wright that the river has not occupied the oxbow since the beginning of the Glacial period, but he went so far as to hold that the river never did flow around the so-called oxbow. He ascribes the feature to the work of two small streams which “rise on opposite sides of a low col and de- MW ws z S$ 20 Wa) 6 an aS) (5) Ars o ee Ta S op No =i Neh; (o} re see a Ay \O Bives 2 sp = 2 aK a6 a es z 8 A ls o> “oO & 3 Ss 8b oO toa fey m 8 Tah Sp ede ah IES Fie | 3 8. ae) ors ace) 3 5 ) vey Sur TERRACES AND VALLEYS IN WESTERN PENNSYLVANIA 153 bouche into the Allegheny gorge within a mile of one another, and in Glacial time these two valleys were filled by overwash deposits mingled with material from the immediately adjacent slopes.” He states also that the rock floor of the abandoned channel is not level, but falls down rapidly toward the river. He does not, however, explain the fact that the col between the heads of the two streams is low and swampy, whereas there is not a case of two large streams rising in an area of swampy ground in the whole unglaciated area of western Pennsylvania, and he says nothing about the broad steep-walled valley through which the small streams flow. Frank Leverett (U.S. Geol. Surv. Mon. 41, 242) considers EK. H. Williams’ view ‘“‘more consistent with the features than the one presented by Chance,” and says further (apparently misinter- preting Williams’ view): “It refers the opening of the double channel, resembling the forks of an oxbow, to a shifting of a smaller tributary of the Allegheny from one side to the other of a low hill that stood nearly opposite the point at which the tributary entered the valley.” The data gathered by the writer indicate, first, that the so-called Parker oxbow is an abandoned channel of Allegheny River, and so is properly called an oxbow. ‘The characters which force such a conclusion are: (a) the depression has the size and shape of the Allegheny valley, having a comparatively uniform width of about a mile, and bounding walls from roo to 300 feet high; (6) the shape is a broad smooth curve with the side of the valley inside the loop gently sloping, and that outside high and steep like the present valley around curves of the river (it resembles, for example, the curve of the Clarion 1 mile south of Turniphole, Foxburg quad- rangle); (c) a current with something like the strength of a river must have flowed around the bend, for pebbles up to 6 inches in diameter are found at the most extreme part of the loop. Second: The abandoned channel was occupied in a part of Kansan time. The presence of Kansan outwash on the floor, which is at nearly the same elevation as the floor under Kansan material near by, indicates that the last great event before the abandonment of the oxbow was the advance of the Kansan ice 154 EUGENE WESLEY SHAW sheet. The abandonment took place before the stream began again to cut down, for deposits are found around the loop almost as high as the highest gravel. The broad valley around the oxbow was cut previous to this time. One can only conjecture how long a period of time was necessary for this. There is some evidence that the rock floor of the east end of the loop is higher than the Parker strath. If this be true the oxbow must have been developed either in pre-Kansan time, before the stream had cut as low as the Parker strath, or after the Allegheny had aggraded until it was high enough to take this route. How- ever this may be, the close association of the abandoned channel with the high terrace, and the occurrence of Kansan material in the channel, show that whenever it was formed, it was occupied and abandoned in Kansan time. Third: The length, depth, and narrowness of the rock channel through which the river now flows across the neck of the oxbow suggests that the oxbow was not cut off in the way that streams ordinarily cut off their meander, but points rather to superimposi- tion. The present valley across the neck of the abandoned channel is a narrow rock gorge over a mile long, and the top of the gorge extends up to the level of the highest part of the old channel. Another abandoned valley which is thought to show the method of development very well is found on the Allegheny, a few miles northeast of Pittsburgh, and opposite Verona. The topography suggests at once that this feature is a cut-off loop of the Allegheny, and it is found on a level with the high terraces. The width is nearly as great as that of the old valley of the Allegheny, and glacial gravels are found in it. But on closer inspection it is found that the width of the valley and the thickness of the deposit decrease rapidly away from the present course of the river, and this through a rise in the rock floor. Also there is an impressive amount of fine material and a scarcity of bowlders. Finally, at the extreme end of the loop the old valley, if such it be, is very narrow, and the deposit but a few feet thick. The meaning of these features seems quite evident. At the time the Allegheny began to aggrade, the position of this loop was occu- pied by two small tributary streams. The divide between them TERRACES AND VALLEYS IN WESTERN PENNSYLVANIA 155 was, at one place, a little less than 100 feet above the river-valley floor. As the river rose it dropped some coarse materials in the ends of the tributary valleys, but a mile away the ponded water was quiet and the deposit fine. This process continued until the Allegheny reached the elevation of the lowest point in the divide between the streams, about 3 miles away. Then the river current was separated and a part flowed slowly up one tributary and down the other, carrying some coarse and much fine material (varying from season to season) and shaping the col into the form of a valley. Finally the river cut down again, abandoning the course which it had occupied temporarily for the much shorter original course. . A part of the history of the loop is reflected also in its present drainage. When the river left it, the run-off was naturally in the direction in which the river had flowed, out one arm and back down the other. But a new small tributary in the position of the upper one of the old ones is now cutting back into the upper end of the abandoned valley, driving the head of the other stream back and annexing a part of its unnatural drainage basin. All of the other abandoned parts of the valleys of this region have been examined carefully and seem to have been developed in the same way—by silting up and redissection—and the process is the same whether the case be on the Allegheny or on one of the tributary streams. In many cases the new courses made available by the silting-up of the old channels were about as direct as the old, and in certain of such cases the stream cut down in its old course, while in others it assumed a new course. Thus some of the aban- doned valleys mark courses temporarily occupied by the rivers, while others show old and long-used courses. It is a significant fact that the changes were, in nearly every case, from a longer route to a shorter one. This would scarcely have been true had the rivers been driven from their courses by ice dams. ‘There are some cols which stand just a few feet higher than the highest gravel, and these were, of course, never crossed by the rivers. Indeed, if aggradation had proceeded 50 or 100 per cent farther there would have been an amazing network of long and devious “‘abandoned”’ valleys. 156 EUGENE WESLEY SHAW SUMMARY Summarizing, the method of development of the high terraces and abandoned parts of valleys of western Pennsylvania seems to be as follows: (1) The development of a valley train over 100 feet thick, along the Allegheny and Ohio; (2) from the beginning the aggradation of this stream produced an effect felt on every tributary, and a portion of each, beginning at its mouth and extend- ing gradually upstream, became silted up. (The lower end of each tributary valley thus took on a form resembling the half- filled character of the valley of the master stream.) (3) As the rivers built up they found themselves flowing at the height of one after another of the lowest places in near-by divides, and at such times and places the currents were divided and the cols were occu- pied. This overloaded condition of the streams lasted a long time and there were many fluctuations, for at some places, as at Pitts- burgh and Belle Vernon, there are two or three well-developed valleys side by side. (4) When final redissection began, the rivers chose the channels momentarily most desirable. In most cases the short route was the principal factor in the choice, but in others the largest current at the time and other comparatively trivial conditions determined the courses of the streams. As in all cases of superimposition, the resistance of underlying rock played no part in their location, and at many places the rivers soon found themselves sawing into hard rock where near by were courses through unconsolidated materials. REQUISITE CONDITIONS FOR THE FORMATION OF ICE RAMPARTS WILLIAM H. HOBBS University of Michigan In a recent paper’ Mr. J.B. Tyrrell, late of the Geological Sur- vey of Canada, has made the assertion that though he has now for many winters made observations on and about the Canadian Lakes, he has never detected any evidence of ice push against shores as a result of expansion. He thus discredits the accepted explanation of ice ramparts. To one who has in other localities seen the ramparts 7m process of formation from this cause, it seems important to supply an explanation for the failure of such an expe- rienced and careful observer as Mr. Tyrrell to observe the same phenomenon. The quite obvious fact is that ice ramparts are greatly restricted in their occurrence, a number of special conditions being essential to their formation. Mr. Tyrrell’s paper fortunately shows that some of these conditions were lacking in the districts which he studied. In order that these requisite conditions may clearly be under- stood, it will be necessary to give in brief outline the theory of formation of normal ice ramparts through ice expansion. The initial ice cover of the winter season on our northern lakes usually forms with only moderately cold air temperatures. These may be assumed to be but a few degrees below the freezing point, and the cover, once formed to a thickness of an inch, grows quite slowly from the under surface. After it has acquired a considerable thickness, the arrival of one of the ‘‘cold waves” contracts the ice cover by lowering its temperature through contact with the colder air layers. Under this contraction fissures open in the ice to the accompaniment of loud rumblings, water rises to fill them and is 1J. B. Tyrrell, “Ice on Canadian Lakes,” Trans. Can. Inst. (1910), IX, 1-9 (reprint), pls. 1-6. 157 158 - WILLIAM H. HOBBS quickly frozen in the prevailing low temperature so as to form intercalated “‘planks” of younger ice. The lake cover is thus again completed at a low temperature, so that a ‘warm wave,” if it can quickly communicate its temperature to the ice, causes an expansion which according to Tyrrell amounts to one to three inches per mile per degree Fahrenheit. Thus expanded the ice cover is too large, and a push is exerted against the shore zf the cover 1s a structure competent to transmit the stresses induced in tt. The range of action of this push, and the consequent size of the ridge raised upon the shore will depend upon the number of times the process is repeated; for each alternation of “cold”? and “warm” wave introduces a new series of wedges into the ice cover and correspondingly extends its margins. To recapitulate: (1) there must be a wide and probably also a relatively sudden alternation of lower and higher air tempera- tures over the lake: (2) these temperature changes must be | promptly communicated to the ice; (3) the ice cover regarded as a girder must be competent to transmit the stresses to the shore; and (4) for large effects the alternations of temperature must be several times repeated. Obviously, also, the shores of the lake must be of such form and materials as to be subject to movement under stresses below the crushing strength of the ice itself. The first and last conditions are meteorological and can be determined for any given district. Not only is a severe winter climate essential, but there must be an alternating occurrence of cold and warm waves. The second and third conditions are crucial. In Buckley’s studies of ice ramparts at Madison, Wisconsin, the most thorough that have been made,’ it was found that ramparts seldom formed during seasons when the lakes were snow covered. The probable explanation of this is that snow blankets the ice and prevents a quick communication to it of the air temperatures above the snow surface. We have here emphasized the element of time, for the reason that studies in Greenland show that air temperatures are slowly communicated downward through snow blankets to very tE. R. Buckley, “Ice Ramparts,” Trans. Wis. Acad. (1901), XIII, 141-62; pls. 1-18 (discussion by C. R. Van Hise). THE FORMATION OF ICE RAMPARTS 159 considerable depths. It is well known from studies of the “fatigue” of materials under stress that they often yield to slowly acting stresses that would be transmitted undiminished in intensity if quickly applied. Snow blanketing of the ice, from the evidence in Mr. Tyrrell’s paper, would appear to be very general within the districts which he studied. Further limitations upon the formation of ice ramparts are imposed by the third condition—the incompetency of the ice cover as a transmitter of stresses. With the ice serving as a strut, Fic. 1.—Sketch map showing the position of ice ramparts and of buckled ice ridge formed on Lake Mendota at Madison, Wisconsin (based on Buckley’s Map). its push can be transmitted effectively only when the cover 1s main- tained as a plane surface. Lack of homogeneity or of absolute uniformity in strength, and variation in form of the surface at which stress is applied, will with increasing length of beam intro- duce an important stress component tending to buckle the beam and dissipate the energy transmitted by it—the competency of a strut to transmit stresses is inversely as its length. Experience shows that lakes or arms of lakes which are much over a mile and a half across do not develop important ice ramparts. On Lake Mendota at Madison, the best ramparts are found upon the shores of University Bay, which is about three-fourths of a mile across. Outside this bay the lake ice is raised each winter into a sharp 160 WILLIAM H. HOBBS ridge extending from the outer margin of the bay (the peninsula of Picnic Point) across the wide portion of the lake to the opposite shore, and about this section no ramparts are developed (see Fig. 1). Ice ramparts can thus form only on shores of lakes which have relatively small size or on small bays of larger lakes, though a width of at least half a mile is probably necessary in order to secure sufficient dilatation of the ice cover to make ramparts of appreciable size. Anything which tends to deform the ice cover from a perfect plane will effectively destroy its competency as a girder, and then no ramparts will form. Mr. Tyrrell has shown in his valuable paper that young lake ice will support, without bending, less than its own thickness of dry snow, and that the ice on Canadian lakes is bowed down under its load of snow to such an extent that water comes to the surface through cracks and further increases the bending. To sum up, the heavy snow cover alone would by blanketing the ice, but probably even more by bending it, effectually prevent the formation of normal ice ramparts. As already stated, such ramparts may actually be seen in process of formation during a warm wave in any favorable winter about Lake Mendota at Madison, Wisconsin. It is fully realized that rafts of floating ice drifted by the winds at the time of the spring “break up” do also produce small bowlder ridges on shores which bear a close resemblance to some of the types of normal ice ramparts. THE TERMINAL MORAINE OF THE PUGET SOUND GLACIER J. HARLEN BRETZ I. GENERAL CHARACTER OF THE COUNTRY SOUTH OF PUGET SOUND The region of Puget Sound, inclosed between the Olympic and Cascade ranges, is a heavily drift-covered lowland. The drift is deeply incised by broad valleys of meridional trend, some occu- pied by arms of the Sound, some by lakes, and others by streams. The summits of the plateaus and hills of drift accord in a general level so that, seen from overlooking mountain peaks, the region appears to be a vast plain, interrupted only by a few rock hills, remnants of the preglacial topography rising above the drift. Immediately south of the Sound the drift assumes a different | facies. The trough valleys disappear and the plain becomes continuous and is widely covered with gravel outwash. It is still a part of the great Puget Sound drift plain. It is diversified with morainic eminences occasionally, of a character different from that of the drift hills lying between the valleys of Puget Sound farther north. The most southerly extended of these hills is a belt which constitutes a part of the terminal moraine of the Puget Sound glacier. Beyond the southern portion of the Puget Sound depression is an abrupt transition in the topography. Rock hills of pre- glacial sculpture, lying beyond the limit of glaciation, begin here and continue:southward past the Columbia. In western Washing- ton no other such area of low rock surface occurs as must here exist beneath the heavy drift mantle of Puget Sound. On the southeast, the area is overlooked by the magnificent Tertiary volcano Rainier. On the south is a region of rock hills bearing no group name. They are drained by the upper Des Chutes and the Skookum Chuck rivers and have a maximum altitude of about 2,000 feet. Farther west lie the Black Hills, whose highest altitude is probably not greater than 2,000 feet. 161 162 J. HARLEN BRETZ These hills are bounded on the east and northwest sides by low, wide valleys which constituted the two chief routes of glacial water discharge from the basin of Puget Sound: The Olympic foothills rise farther northwest in the main Olympic Range. At the time of greatest extent of the ice, the northern slope of the Black Hills was overridden and a lobe extended down on either side, the hills determining a broad re-entrant in the ice front. Il. PREVIOUS WORK ON THE MORAINE In a general way, the glacial drift in Puget Sound has long been known to terminate some distance south of Olympia, and the gravel plains have been commonly recognized as outwash deposits from the ice. No detailed work, however, has been done in the region except by Willis and Smith on the Tacoma quad- rangle.’ Here the contact between Pleistocene deposits from the glaciers of the Cascades and Mt. Rainier and the Puget Sound drift has been traced along the northwest flank of the volcano to the southern edge of the quadrangle. Warren Upham has described,? from a hasty reconnaissance, what he believed to be the terminal moraine lying between the base of Mt. Rainier and the Black Hills. He interpreted the remarkable gravel mounds of the outwash plains of the region as morainic topography of peculiar type. No observer, so far as the writer is aware, has previously noted the existence of the western lobe of the glacier, lying between the Olympic Mountains and the Black Hills. . III. MORAINE COURSE ACROSS THE GEOSYNCLINE The westernmost geosyncline of North America is regarded by stratigraphers as finding its representative on the Washington coast in the Puget Sound depression. Were it not for the accident of glaciation, this structural valley would today embrace a broad inland sea, but the thick drift deposit constitutes a filling sufficient to maintain most of the surface above sea-level. The terminal t Bailey. Willis and G. O. Smith, ‘“‘Tacoma Folio, No. 54,” U.S. Geol. Survey. 2 Warren Upham, “Glacial and Modified Drift in Seattle, Tacoma and Olympia,”’ American Geologist, XXIV, No. 4. TERMINAL MORAINE OF PUGET SOUND GLACIER 163 moraine is built far enough to the south to lie in general over rock surfaces above sea-level; indeed, the moraine forms a fairly con- ? MOUNTAINS GLACIATION OF PUGET SOUND Morrzontal Ruling ---- Extent of Vashon G/aciation Ver(Gal fusing: — —= = On Fast = Cascade Weve and /¢e On West —Olympic Neve and /ce Dotled Areas =---—-—fytra-moraimc. Vashon Outwash 0 10 20 30 stant boundary between the depressed region to the north, now drift filled, and the rugged, stream-carved rock hills southward. The field work on which this paper is based has been done under some difficulty because of the undeveloped nature of the country, 164 J. HARLEN BRETZ and its incompleteness must largely be charged to the same reason. Large tracts about Puget Sound are yet covered with virgin forest whose density is such that passage for any considerable distance is next to impossible. The lack of roads, trails, and inhabitants over many square miles forces the investigator to shoulder his pack of blankets and food, and travel the country on foot. De- tailed work is not practicable under these circumstances. It will probably be years before the moraine can be mapped with accuracy, since the task must wait on the agricultural development of the country. On the eastern margin of the Tacoma quadrangle, Willis and Smith? found a broad sheet of till spread by a piedmont glacier from the Cascades. The contact between this drift sheet, named the Osceola till, and the Vashon or youngest till sheet of the Puget Sound glacier was found to be marked by a belt of hummocky topography of morainic aspect, considerably different from that of the ground moraine on either side. No definite marginal or interlobate moraines were found, the phenomena being apparently referable to subglacial accumulation. Short eskers were a notable feature. This study has taken up the continuation of the contact between local glacial deposits, and those far traveled down the Puget Sound depression from the north, on the south edge of the Tacoma quadrangle in a densely wooded country traversed by a few second- ary roads and one highway, the Mt. Rainier automobile road. The geological map of the Tacoma folio maps the Rainier Pleistocene drift on the south edge of the quadrangle, farther west than the writer has found it. The automobile road follows a north to south course for 5 or 6 miles, parallel to and about 6 miles west of Lake Kapowsin, and for this whole distance traverses the till plain of the Puget Sound Vashon glacier. Northward toward Tacoma are extensive areas of outwash gravel deposited during the recession of the Puget Sound ice. The till plain rises southward from the outwash with an abrupt morainic slope, ascending 200 feet in one mile. The slope is thrown into several successive ridges of till trending east to west, on the south sides of some of which were « Bailey Willis and G: O. Smith, of. cit. TERMINAL MORAINE OF PUGET SOUND GLACIER 165 distinct kames. The till is the characteristic blue-gray arenaceous material, with laminae and rounded cobbles, which is identified throughout the Puget Sound country as Vashon. ‘The presence of numerous varieties of rolled granite cobbles in the moraine and in the plain southward is a safe criterion for the identification of the till as the Vashon rather than the Osceola till of the Cascades. The moraine ridges on its northern flank and broad till plain lying southward are topographic features in accord with this interpreta- tion. Though the boundary between Vashon and Osceola till was not located, it obviously lies between the Mt. Rainier high- way and Lake Kapowsin, the lake lying at the base of the foothills of Mt. Rainier. The dominance of Puget Sound ice at the western base of the Rainier foothill country is proved conclusively by the common occurrence of bowlders and cobbles of several granitic types char- acteristic of the drift of Puget Sound and unknown to the adjacent Cascades. The postglacial gorge of Nisqually River, 300 feet in maximum depth and with vertical and even overhanging walls, is two miles ~ long and occurs where the river enters the area of Puget Sound drift. A 40-foot section of outwash, containing frequent Vashon drift materials, overlies the rock floor in which the canyon is cut at LeGrande. Farther up the canyon no drift was found. A trail crosses the divide between the Nisqually and Des Chutes rivers just south of the canyon noted, entering the latter stream at the headwaters. Scattered granitic bowlders of Vashon drift were found up to an altitude of 1,220 feet on the Nisqually side, but no traces of drift were found in the remaining 200 feet of ascent or in the valley of the Des Chutes on the other side until the altitude of 1,200 feet was reached, a few miles down from the headwaters. Here scattered erratics occur on the hillsides, and at goo feet is a level terrace composed of fine material with inter- . spersed pebbles, probably a lacustrine deposit caused by the ice entering the lower valley and blocking the drainage. Two miles below this terrace, whose soil has determined the loca- tion of several small farms in the wilderness, is found the terminal moraine of the Puget Sound Vashon glacier. The surface is exceed- 166 J. HARLEN BRETZ ingly bowldery, granite is very abundant, kettles containing lakelets and bogs are common, and the subsoil is typical Vashon till. The margin of this bowldery drift may be traced about the west and north of the Bald Hills from the Des Chutes to the Nisqually River and is in places thrown into sharply defined ridges. Occasionally the forest seems growing on one gigantic bowlder heap. A preglacial valley descending to the northwest has been dammed, giving rise to Little Bald Hill Lake, a picturesque body of water in the heart of the wilderness. Another such valley has three morainic ridges thrown across it at descending altitudes, a marsh or alluvial flat lying behind each ridge. Pronounced relief of the moraine on the north slope of the Bald Hills was found, but the unbroken forest prevented satisfactory examination. The same difficulty of examination is presented by most of the country from the Bald Hills west to Tenino. In general, the drift- covered area bears the farms and roads, the region immediately beyond the ice limit rising in rocky hills which constitute the divide between the Des Chutes River and the Skookum Chuck. A trav- erse across this divide found the moraine disposed in bowldery ridges along the base of the hills with a marginal drainage channel separating the frontal ridge from the bold rock hill slope. No erratic material or evidence of ice action was found on the ascent to the divide crest, the glacier of Puget Sound having succeeded in barely reaching the northern base of the hill region. The town of Tenino is situated on an area of gravel outwash lying immediately south of the moraine. The rock hills die away toward the west just south of the town and glacial drainage escaped southward to the lower Skookum Chuck through a broad, gravel- filled valley. Glacial outwash was also carried westward from Tenino toward Grand Mound and Gate to join the extensive areas there outspread. The Skookum Chuck bears a train of glacial gravel which entered it somewhere in the unsurveyed region of the Huckleberry Moun- tains, presumably from Mt. Rainier’s Pleistocene glaciers.- But careful search revealed absolutely no granite or sedimentary meta- morphics in this gravel for a distance of 6 miles along its course. Only when the western limits of the rock hills were approached, TERMINAL MORAINE OF PUGET SOUND GLACIER 1607 and below a low pass across the divide to the Des Chutes River, was Puget Sound glacial gravel found in the Skookum Chuck valley. Clear Lake, at McIntosh station, 4 miles east of Tenino, lies in a marginal drainage channel discharging westward into the outwash gravel area at Tenino. The terminal moraine lies imme- diately north of this lake. North of Tenino, the moraine is of a character considerably changed from that in the Bald Hill region. It has here become a single massive till ridge on the plain, and sur- face bowlders are not sufficiently numerous to attract attention. It is two miles wide and 250 feet above its base on both north and south sides, the highest point examined reaching 550 feet A.T. On each side, it is flanked by an outwash gravel plain bearing peculiar tumuli. The till mass appears to cover several rock knobs and hills, whose existence may have in some measure determined its location and relief. Both east and west of Tenino, quarries in sandstone have been opened on the slopes which rise farther to the north in the moraine. ‘The road north from Tenino to Olympia cuts into decayed shale strata im situ at the summit of its grade across the moraine at about one-half the maximum height of the moraine, and at McIntosh rock outcrops occur on the south base of the moraine. The hills which rise south of Tenino were carefully examined for drift materials. Three distinct terraces of outwash gravel were found, occasionally showing forests beds descending southward toward the Skookum Chuck. The highest gravel lies 360 feet A.T., and above it drift abruptly ceases. Flanking the frontal margin of the moraine from Tenino west to Black River is an extensive area of outwash gravel, known as the Grand Mound Prairie. It is entirely barren of forest growth and almost useless for any agricultural purpose because of the coarseness and depth of the gravel. At the contact between moraine and outwash examined no apron structure was found. The gravel plain apparently was built by outwash occurring through breaks in the moraine ridge and not by outflow from the ice edge when standing at its maximum limit. ) The whole region south of Puget Sound bears much outwash, both 168 J. HARLEN BRETZ extra-morainic in position and lying back of the ice limit. These areas are all alike in being natural prairies because of the coarseness of the soil and in bearing a surface deposit of black silt of variable thickness. Many of them exhibit a very interesting surficial development into mounds of fairly uniform size and distribution composed of mingled gravel and silt without stratification. Where typically developed, they resemble a field of closely spaced hay- cocks. Their origin is not clear. Grand Mound Prairie bears these tumuli over a considerable portion of its extent. Some distance back from the frontal edge of the terminal moraine between Tenino and Little Rock a new railroad grade affords frequent exposures of the Vashon till overlying drift of much greater age and with bedrock often appearing beneath the drift. Hills of the moraine occur on the east side of Black River a mile south of Little Rock, while across the river on the west, a morainic tract of low relief occurs about a mile wide. In this tract is a splendid exposure of Vashon till highly charged with rounded gravel which is doubtless overridden and incorporated outwash material. Mima Prairie, southwest of Little Rock, is another part of the outwash gravel plain and forms a sharp re-entrant angle in the surface till exposures, though hardly recording such an ice margin form, the till being probably buried beneath this northward angle of the outwash. Between Mima Prairie and the Black Hills, unweathered Vashon till was observed in a gravel pit with a thick- ness of three feet overlying a very red and decayed till of undeter- mined depth. Small pebbles of the latter were often easily cut in two with a knife, while those of the overlying Vashon were firm and unweathered. No drift is found back in the Black Hills except a sprinkling of pebbles in re-arranged residual material on the slopes which face the broad drift plain eastward. The region is exceedingly difficult to examine, the forest being almost impassable. Entrance into the hill region is gained on a logging railroad and on various trails. One road crosses near the northern part of the hills, passing west from Olympia close to Summit Lake. Drift has been found near this lake on the north slope of the hills up to an altitude of 1,460 a TERMINAL MORAINE OF PUGET SOUND GLACIER 169 feet, falling short a few tens of feet of reaching the summit. No till has been found in the valleys of any of the south-flowing streams of the region. f Summit Lake lies in the upper part of a preglacial valley, the lower southern portion of which bears a drift filling. The ice sheet certainly overrode the divide at the northeast of Summit ‘Lake but it brought over no drift. Farther south, however, the valley opens into a larger one trending east and west, and from both directions in this, till was carried into the Black Hills. Again the relation of agriculture to the drift is illustrated in the occurrence of several small farms on the broadened valley floor produced by drift filling while elsewhere the region is covered with primeval forest or the waste of logged-off land. At least two distinct valley trains cross the western part of the Black Hills to the Chehalis River, the larger of these being a filling so complete that several rock hills rise like nunataks from the gravel plain. This enters the Chehalis valley at Elma, in the vicinity of which it is deeply incised by creeks, its structure being thus plainly revealed. A feature of the gravel is the prevailing reddish color, fairly uniform throughout the mass. The freshness of the pebbles and the youthfulness of drainage on the plain, however, show this staining to be due to some other cause than age. The Vashon till near the head of this valley train is also deeply red while its pebbles are fresh. The explanation is thought to be found in the incorporation of residual material from the basalt rocks of the Black Hills. The country lying between these hills and the Olympic Moun- tains is practically a great gravelly waste. The forest is thin over large areas and open prairies occur in the region south of Hood’s Canal. The moraine hills when found are often largely buried in outwash and the extreme limit of the ice as mapped is consequently only approximate, being based on the occurrence of till outcrops above the gravel plain. No definite ridging tangential to the ice margin was observed in the till hills seen, though their occurrence forms a zone a few miles wide, whose outer margin has been indi- cated as the limit of Puget Sound ice to the west. The character of the till, where exposed in railroad cuts and 170 J. HARLEN BREDZ stream valleys, appears identical with that shown in the vicinity of Seattle, on the slope of the Bald Hills, and in other widely separated regions. The matrix is somewhat sandy, the pebbles and bowlders are rounded, and large erratics are rare. Granite of various kinds is abundant, though granite is not known in the neighboring Olympics. The till is seen to overlie fresh gravel in a few sections with a thickness of about three feet. Its altitude probably does — not reach much above 450 feet A.T. Lake Nahwatzel lies in a decidedly morainic area, the monoto- nous gravel plain giving place to rolling hills of till which rise 50 feet above the lake surface. These morainic hills lie probably over the lowest preglacial rock surface between the Black Hills uplift and the Olympic foothills, and in such a situation we may find an explanation of the more pronounced morainic expression. The till along the margin from Matlock to the Black Hills often shows a large proportion of deep red clayey material inter- mingled with fresh pebbles. The presence of such material, doubtless from the incorporation of the residual soil of basalt of which there are frequent outcrops, is to be expected near the ice margin providing the ice was overriding a region previously unglaciated. The approximate moraine course from Matlock northward bends abruptly back toward Hood’s Canal, the greater length of which is closely bordered by the Olympic Mountains on the west. The extent to which Puget Sound drift penetrated into the valleys of these mountains is known in but one case, that of the Skokomish River. Rock along this stream’s course is practically absent below Lake Cushman, while the mountain walls rise almost from the lake shores on the upstream side. Puget Sound drift of Vashon age composes an extensive plateau 400-800 feet above Hood’s Canal, extending back from Lilliwaup Creek directly west to Lake Cushman and also southward to the broad, pre-Vashon lower Skokomish valley. One large rock hill rises through this till plateau just south of the Lilliwaup, otherwise the surface is of rolling ground moraine with occasional shallow kettles. Across the Skokomish to the west are foothills with little orno drift. To the south of the great bend of this stream, extensive TERMINAL MORAINE OF PUGET SOUND GLACIER 171 outwash gravels begin, continuing to Shelton in one direction and across the Puget Sound divide to the Satsop in another. In this latter direction, the outwash largely buries the moraine near Mat- lock and becomes extra-morainic in its further extent. On the east side of Lake Cushman, the till plateau becomes ridged and kettley, though a dense forest prevents satisfactory examination. The morainic character is best seen along the trail from the head of the lake to Lilliwaup. The material on the lake- ward face of these ridged drift hills nowhere contains granite, though two very careful examinations were made. In but one place are granite pebbles found on the shore or in the immediate vicinity of the lake, this being in the bed and delta of the largest stream entering the lake from the northeast. Yet a mile back from the lake, to the east, granite bowlders are found lying on the surface, becoming very numerous two or three miles farther east. The limit of the Puget Sound drift is thus seen to lie close to the lower end of Lake Cushman, the basin of which is caused by the damming of the Skokomish River valley. The inner slope of the drift dam is probably faced with the terminal deposits of the Skokomish valley glacier, which was unable to advance farther in the face of the overwhelming mass of the Vashon glacier. It may have earlier deployed farther out on the plain, but if so the deposits are buried beneath the Vashon drift. That a valley glacier must have existed back of the drift dam of Lake Cushman when the Puget Sound ice was at its maximum is evident, else the lake basin would have filled with outwash. A till with very angular débris, none characteristic of Puget Sound drift, lies back of the drift dam on the slope of Mt. Ellinor, immediately north of the lake. It is estimated to reach 500 feet higher than the lake surface. As shown on the map, the western margin of the Puget Sound glacier north of Lake Cushman is approximate only. .The moun- tains rise close to Hood’s Canal throughout the remaining distance included in the accompanying map, and in all probability there existed no embayment of Puget Sound ice in the other river valleys entering the Canal comparable to that of the Skokomish valley. 172 J. HARLEN BRETZ IV. GENERAL CONSIDERATIONS Considering the altitudes of the terminal moraine only where facing driftless country to the south, its crest is found to have no great range in elevation above the sea. On the north slope of the Bald Hills, near the headwaters of the Des Chutes River, the moraine crest 1s probably nowhere more than goo feet A.T., though erratics occur 320 feet higher. Near Tenino, where the moraine is most typically developed on the plain, the crest is probably less than 600 feet in altitude. The existence of buried rock hills in the moraine in this region has been noted. At Little Rock, the moraine surface on the west side of Black River can hardly have been lowered by erosion of escaping glacial water or subsequent stream action, and is approximately 150 feet above the sea, the lowest altitude in the moraine. From this altitude is a descending slope southward, on which the ice ceased to advance. The oppos- ing northern flanks of the Black Hills, deeply cut by valleys, did not permit assumption of the moraine form. Drift, however, has its upper limit in the re-entrant angle which they produced, at an altitude of 1,460 feet. The flattened lobe northwest of these hills has its moraine hills about Lake Nahwatzel at 450 feet A.T. Puget Sound and Olympic drift damming Lake Cushman reaches observed heights of 950 feet above the sea. The data available for an estimate of the thickness of the ice and its frontal slope are meager. Three miles from Little Rock, the glacier left its till at the eastern foot of the Black Hills at an altitude of about 150 feet. From here it is 10 miles north’to the upper drift limit near Summit Lake, at 1,460 feet A.T. The slope in this instance is approximately 130 feet per mile. Fifteen miles east of Seattle rises the peak of Mt. Issaquah, about 3,000 feet A.T., whose frost-riven summit bears no residual soil comparable to that found on hills of much the same lava rock beyond the limit of the drift: Scattered erratic pebbles were found on the summit, their number increasing on the lower slopes. With the maximum depth of the Sound near Seattle at 964 feet, we may conclude that in the latitude of Seattle the glacier attained a thickness of 4,000 feet, allowing very little for central surface convexity, which would increase the estimate an unknown amount. TERMINAL MORAINE OF PUGET SOUND GLACIER 7.8 Evidence of the lack of vigorous movement near the frontal margin of the glacier is shown in the occurrence of deeply decayed material overridden by the ice. Shale strata, profoundly decom- posed, are exposed east of Little Rock. Though slightly crumpled and in one case bearing an intruded arm of the till, this incoherent and rotted shale has been but little eroded by the ice, though it is two or three miles back from the moraine front. West of Little Rock, where Vashon till is found at its farthest southern extent along the Black Hills, a knob of old red till is exposed beneath it. Depth of weathering and staining are the same on the slopes as on the summit of this knob, hence the inference that no erosion of the projecting softened till was produced by Vashon ice. The accompanying map indicates only the extra-morainic out- wash. Great areas lie within the moraine limits of essentially the same character and age. In the case of all outwash deposits, the discharging water was received by the Chehalis valley largely on the east or west side of the Black Hills. Extensive tracts are rendered as worthless for agriculture by these outwash plains as though in an arid country. For example, the road through the sparse forest extending from Lake Nahwatzel to Shelton crosses but one stream bed and this carries water only during the very rainy winters and no valley has been cut. As already noted, the moraine across the low area between the Black Hills and the Olympic foot- hills has been partially buried in the flood of gravel and its relief much reduced. The question of contribution from valley glaciers in the border- ing Cascades and Olympics cannot be adequately treated in our present state of knowledge. Valley glaciers in these mountains on the Soundward slopes debouched into a great mass practically filling the depression from rim to rim. That they would perform much erosion under such conditions is not to be expected. Willis has found the till sheet of a Cascade piedmont glacier on the eastern part of the Tacoma quadrangle, a part of which is indicated on the accompanying map. The relative insignificance of the Skoko- mish glacier whose lower extremity occupied the basin of Lake Cushman has been shown. No evidence has yet been found that tributary glaciers north of these two produced any perceptible 174 J. HARLEN BRETZ effect on the mass of the course of the great Vashon glacier, whose volume and thickness was of course greater northward. Definite recessional moraines are yet unknown in the Puget Soun~ cocntry. Between the terminal moraine and the southern arms of the Sound are occasional moraine hills and ridges which will probably resolve themselves into linear arrangement when carefully studied and will constitute recessional moraine deposits. But in the larger area of longitudinally ridged drift among the arms of the Sound, there is little of morainic origin beyond scattered lodge moraine hillocks in the valleys. Russell’ first noted that there are two till sheets in Fuget Sound basin, recording two glaciations. Willis? has named these the Admiralty and Vashon, with the latter of which we have had to do. The frequently weathered condition of the Admiralty till or of its superposed outwash has been pointed out by Willis as evidence of long exposure before the Vashon glaciation. The freshness and slight erosion of the Vashon till sheet and moraine evince an age comparable to that of the Wisconsin drift. A notable feature of the Puget Sound glaciation, shown by the failure of constant careful search to find older till beyond the moraine, is that the last glaciation of the region, doubtless Wiscon- sin in age, was the most extensive. Frequent incorporation of residual soil in the Vashon till is the best evidence which might be secured, in the absence of deep sections, that it overlies areas never previously glaciated. ' Bailey Willis, ‘‘ Drift Phenomena of Puget Sound,” Bull. Geol. Soc. Am., IX. 2 Willis and Smith, ‘“‘Tacoma Folio No. 54,” U.S. Geol. Survey. 2) DIRO RIAL THE SEEDING OF WORLDS As a sort of initiation stunt precedent to admission into the fraternity of agencies of good and regular standing, every new agent that is brought into view by the ongoings of science is likely to be set to the task of solving some large part of the outstanding puzzles that still vex the wise men of our craft. “Light pressure”’ is one of the latest novitiates on trial, and has been set to the stunt of seeding the habitable but not inhabited worlds by spores from some previous spore-growing world. The seeding of the first world is mercifully not made a part of the stunt. So too, to help out the novitiate somewhat, the hazards of the cold of space are mitigated by bringing to bear certain novel tenets about endurance of extreme cold, and by cutting the time by the great speed of the trip from world to world under the new pressure. ‘The stunt still remains a stiff one and is interesting, but the fraternity seems to be missing the best part, the getting home to the new world; no doubt because it is so far off. The start of the spore from the spore-growing planet is not without its little difficulties; for the seed, be it even so light as the airy fluff of the puffball, must yet not only get out to the very top of the air, but it must be pushed off by the pressure of the light at a speed of some 5 or 6 miles a second to be able to get away from the pull of the parent world, if that world be a body like our familiar acquaintance, the earth. A Krakatoan blast, however, can no doubt give the spore a lift, if need be. But the getting away is not the interesting part of the stunt; it is the landing. If ‘“‘light pressure’? has once pushed the spore out of the clutches of the parent world and got it well under way, all is likely to go well till the bounds of the sun’s sphere of control are reached and the border of the domain of the other sun is entered, for that sun is likely to push back as much as the parent sun pushed out. In matters of this sort one sun seems unwilling to be the dumping- aed 175 176 EDITORIAL ground of another sun. So now, between the opposing pushes of the rival suns, comes the real trial of skill or luck in landing the spore. If the seed be duly planted, the fraternity door should surely open for the candidate magna cum laude. On leaving the domain of the old sun and entering the field of new suns, care or luck in hitting on a sun that shines less bright than the one that has pushed the spore out is surely needed, or else the back-push of the brighter sun will grow in time to be stronger than the on-push of the old sun and the spore will be stopped or turned aside. If someone churlishly remarks that the seeding of new worlds can thus only go down the scale of solar radiance, let that pass; it is enough to seed at long distance any world. Hitting upon a sun of duly lesser radiance, the spore must shoot straight for it, quite straight, center to center, for if the backward push of the sun ahead is a little awry at the front, the spore will be pushed aside and out of line, and once off the line it will be turned more and more away and surely go astray. Nor must the chosen sun move out of line while the spore is coming toward it, or else the front push will surely turn the spore away. No sun must be hit upon but one that will stand still, if such there be, while the spore is getting home to the new planet, or, if no sun stands still, a sun must be hit upon that is coming toward or else is going straight away from the advancing seed. All ill luck in hitting the right path or in hitting on a sun moy- ing straight toward or straight away from the speeding spore once duly escaped, the larger perils are past, but not all; there are perils of side pushes. In hitting upon a star of proper weakness of radiance and coming or going or standing still duly, the spore may chance to pass some brighter star off the line and its side push may turn the spore off its course; or stars may be thicker or brighter on one side or another and the spore be put off its course by their united pushes. Where, then, it may ayain be churlishly asked, is a spore to go if all the suns push it away? Well, it is not a part of this stunt to chase up lost spores; still, there are ‘“‘dark lanes” and ‘“‘coal sacs”? and ‘‘openings”’ leading out into room “outside the universe.” Then too there are perils of planets as well as perils of suns. EDITORIAL 177 As the spore pushes down against the radiance of the defendant sun, one of whose planets, near enough to it to keep duly warm, is to be seeded for a new life kingdom, a planet just at the right spot must be hit upon. Luck must here stand the spore in good stead, for the chances are not the best. If the planets of the chosen sun circle round it cross-ways, in any but the minutest degree, they will never be in the center-to-center line of the spore’s path, for, as we have seen, the spore must keep true to line or the back- ward push of the light pressure in front, striking aslant, will turn the spore off. There is a chance indeed that a spore will get down to just the right point and then be turned off just so as to strike a planet that is off line, but it is not a chance to stake much on. To have any fair chance of getting home to a planet while the spore keeps straight on toward the repellent sun, under the superior inertia it got from the sun it left, the planet must circle round the sun in a path that cuts this line. And then, too, the planet must be there at just the right time. The spore must no doubt cross the spot in the wink of an eye, or less, and the new world must be there on exact time if it is to be seeded. It is not unfair that it should be made to be there on time as its part of the stunt, for the spore has come far to do its part. . Now if all has gone well thus far there is only the landing left. Ii the spore was pushed out from the old sun too fast, it may plunge so swiftly into the air of the new world as to strike fire and burn or brown itself fatally. But if pushed out just right at the start and pushed back just right on the road, it may land with little more than the speed forced by the pull of the new earth, a matter of a few miles a second, it may be. When the speed of the spore is stopped and it floats in the outer air of the new earth it may perchance from being too hot come quickly to be too cold and the change from warmth to chill may try its salamandrine powers before it sinks to the warm air low down or to the ground in which it is to grow. The luck of the spore must stay by it a little farther in its lighting. All may be lost if it falls on polar snow, or mountain peak, or desert plain, or perchance in the ocean midst, if it is not 178 EDITORIAL a salt-water spore. It must fall in a spot where it can grow, where its family, as it comes to have one, may live and multiply and grow into a kingdom, for if it fails in this last, the kingdom will not be won. The stunt may be perilous; but it is easy to see how easy it is to do if done just right. Light is the great foster-farmer of the earth, the truly great farmer; and we now see how clearly and truly ‘‘light pressure” is the long-distance seed-planter of the worlds. Te Cace ARTESIAN WATERS OF ARGENTINA The climate of a part of Argentina is semi-arid, and the geo- logical formations which are regarded as Quaternary and Later Tertiary are, in the western and central districts of the country, saline to a degree which indicates prolonged duration of aridity. The region of the Pampas which covers the province of Buenos Aires and stretches northward west of the Parana does not exhibit this characteristic, having apparently long enjoyed a more humid climate, as it does now. The foothills of the Andes are also well watered. But with the exception of these last-named regions, a great part of the country suffers from lack of good water. This condition may, however, be in some measure relieved by proper development of artesian supplies. Many wells have been sunk already, but without adequate geological investigation. In the Pampas, water is found at a general depth of 20 meters more or less, and is pumped to the surface by windmills. It may be said that the development of the livestock industry of Argentina would be impossible were it not for this supply which comes from eolian, alluvial deposits of Quaternary and Tertiary age. A different geological condition exists from the Rio Colorado southward in what may be best described as northern Patagonia. In that region there are local elevations occupying a middle position between the Atlantic and Pacific, composed of granites and older rocks possibly of Paleozoic age, and rising to altitudes of 300 to 1,000 meters. These mountains are not represented upon any map and their distribution is not known, but they have been de- scribed by Moreno and other explorers. Upon their flanks there EDITORIAL 179 is an extensive formation of gray sandstone which attains a thick- ness of several hundred feet and is very porous. It slopes gently toward the Atlantic and pure water flows from it in outcrops near the coast. The head of water in these strata is unknown. Farther south in Patagonia the central sierra is replaced by plateau country and in Comodoro Rivadavia, in latitude 46 near the coast, wells which were sunk by the government in search of water developed petroleum. There is a large area in this region in which the geologic structure and the possibilities of artesian water need to be developed. In the great plains east of the Andes there are glacial deposits which may furnish superficial supplies like those of the Dakotas, and the marine Tertiary and Mesozoic strata afford con- ditions not unlike those of southern California. Here as well as in the valleys among the spurs of the Andes from Patagonia to Bolivia the geological structure is complicated and the problem of artesian water is one of peculiar difficulty as well as of great interest. Our present knowledge of these conditions rests upon recon- naissance work and the stratigraphic and paleontologic observa- tions of the Geological Survey of Argentina. No work based upon topographic maps and systematic structure has as yet been undertaken. The problem is therefore one whose elements are as yet to be developed. The Argentine government is using every means to encourage settlement and development of the rich agri- cultural regions which lie in the zone of sufficient rainfall east of the Andes, and also the vast grazing district of Patagonia.. In order to afford ready communication it is building railroads at great national expense and operating them. Zhe need of pure water for locomotive use as well as for other purposes has thus been made critically evident, and the minister of public works, Senor Ramos Mexia, has adopted a plan for making surveys for the determina- tion of artesian water conditions along the lines of national rail- ways. He contemplates topographical and geological surveys of a character similar to those executed by the United States Geological Survey, from which he derived the initial suggestion. He last summer applied to the United States government for the services of a geologist and such assistants as he might need, and our govern- 180 EDITORIAL ment has responded cordially to that request. Mr. Bailey Willis has accordingly entered into a contract for the term of two years, to execute topographical and geological surveys for the specific purpose of ascertaining artesian water possibilities in those districts which the minister may designate. With him are associated Mr. Chester W. Washburne of the United States Survey, Mr. J. R. Pemberton of Stanford University, and Mr. Wellington D. Jones of the University of Chicago, as geologists, and Mr. C. L. Nelson and Mr. W. B. Lewis as topographers, and the party has recently sailed for Argentina to enter upon the work. While these surveys have a specific purpose, their possibilities of usefulness in develop- ing the natural resources and encouraging settlement in the regions surveyed will not be overlooked, and the work will be founded on those scientific studies upon which alone practical conclusions can safely rest. Thus it is hoped that a definite contribution to knowledge in geography and geology may be made. It is desirable to point out that the Argentine government has a geological survey which has been in existence since 1903 in its present organization and which dates back half a century as a bureau of mines. It is under the direction of Senor E. M. Her- mitte, who is assisted by Messrs. Bodenbender, Keidel, and Schil- ler, three German geologists who have done excellent stratigraphic and paleontologic work, particularly in districts of the central ‘Argentine Andes. They have unfortunately not been supplied with maps. The established Bureau of Mines, Geology, and Hydrology is under the Minister of Agriculture. The surveys which are about to be made are undertaken by the Minister of Public Works. The two operations are thus officially distinct, but it is hoped and anticipated that they may be mutually helpful. B. W. PETROGRAPHICAL ABSTRACTS AND REVIEWS Epitrep By ALBERT JOHANNSEN! BENEDICKS, CARL, AND TENOW, OtoF. ‘“‘A Simple Method for Photographing Large Preparations in Polarized Light,” Bull. Geol. Inst. Univ. Upsala, 1X (1910), 21-23. For the description of the comparatively simple apparatus used, reference must be made to the original paper. W. T. SCHALLER Bow tes, OLiver. Tables for the Determination of Common Rocks. New York: Van Nostrand, 1910. 16mo, pp. 64+84 advs. 50 cents net. Cui bono ? Written, as this book is, for “‘beginners in lithology,” it is especially unfortunate that the author’s statements are often very misleading. For example, in the chapter on “Rock Classification”? the statement is made that “igneous rocks ... . represent the original solid crust of the earth,” and that “sediments . . . . are but modifications, or recon- structed phases, of this primary type.’ A short chapter on the deter- mination of the rock-forming minerals is followed by 18 pages of tables for the determination of the common rocks. The methods of identifica- tion are given in extremely brief form, but would a “beginner,” or any- one else, classify andesite, quartz porphyry, felsite, or phonolite as “ashy, and often with a few phenocrysts, mostly cellular”’ ? The book ends with a ten-page chapter on “ Building Stones” and a seven-page glossary. The volume is No. 125 of Van Nostrand’s Science Series and is uniform in size and binding with the remainder of the set. ALBERT JOHANNSEN Bowman, H. L., anpD CLarkE, H. E. ‘On the Structure and Composition of the Chandakapur Meteoric Stone,” Min. Mag., XV (1910), 350-76. Pls. 2, and analyses. A full description, with extensive chemical work, ona large piece of the meteoric stone which fell at or near Chandakapur, India, on June 6, t Authors’ abstracts will be welcomed and may be sent to Albert Johannsen, Walker Geological Museum, The University of Chicago, Chicago, IIl. 181 182 PETROGRAPHICAL ABSTRACTS AND REVIEWS 1838. It is an intermediate chondrite, with olivine and pyroxene as the most important constituents. Metallic iron and nickel form nearly 6 per cent, and combined iron and nickel, 5 per cent. W. T. SCHALLER Date, T. Netson. ‘‘The Cambrian Conglomerate of Ripton in. Vermont,” Am. Jour. Sct., XXX (1910), 267-70. Figs. 3. A conglomerate formed of pre-Cambrian pebbles generally held together in a highly metamorphosed “muscovite-quartz schist with more or less magnetite.” The pebbles are a beach formation and are of local origin as is shown by their large size and by their similarity to adjacent rocks. ALBERT JOHANNSEN Duparc, WUNDER, AND Sapot. ‘“‘Les minéraux des pegmatites des environs d’Antsirabé 4 Madagascar,” Mém. Soc. Phys. et d’Hist. Nat. Genéve, XXXVI (1910), fasc. 3, 283-410. The geology of Madagascar is briefly described and then, in detail, are described the rocks around Antsirabé. These include basalts, granites, quartz diorites, pegmatites, cipolines, quartzites, and mica schists. The localities of the pegmatites are then given in detail. The pegmatites occur chiefly in the cipoline and are formed principally of microcline and quartz, or plagioclase (near albite) and quartz. Mica, tourmaline, beryl, garnet, and pyroxene are also present. In the second part of the paper are mineralogical descriptions of microcline, amazonite, lepidolite, lithionite (zinnwaldite), beryl (rose- pink and aquamarine), tourmaline, spodumene, spessartite, garnet, and cordierite from the mica schist of Mount Ibity. W. T. SCHALLER GraBHam, G. W. ‘An Improved Form of Petrological Micro- scope; with Some General Notes on the Illumination of Micro- scopic Objects,” Min. Mag., XV (1910), 335-49. Figs. 5; Dit Suggests several improvements on a Dick microscope, namely, a better adjustment for the condenser system, a triple nose-piece, iris diaphragm, and a slot for introducing screens below the stage. The graduated circle is placed below the ocular. Several other suggested improvements have already been used on other microscopes. Several PETROGRAPHICAL ABSTRACTS AND REVIEWS 183 pages are devoted to the “Illumination of the Object.” An explanation of the ‘“‘white-line effect”’ (Becke’s line) is given for parallel light where the contact plane of the two minerals tn question is at various inclinations. W. T. SCHALLER GRAYSON, H. J. “Modern Improvements in Rock Section Cutting “Apparatus,” Proc. Roy. Soc. Victoria, XXIII (1910), 65-8t. Risin: Describes an apparatus, constructed for the University of Melbourne, with which the writer is able to slice, grind, and mount thin sections of about an inch in diameter and of a thickness of less than 0.001 inch, from rocks of the hardness of granite, in not more than ten minutes. Using two cuts with a diamond saw for each slide, the cost per section is about one shilling. A mechanical device for doing the rough grinding would be an im- provement. With a number of laps running simultaneously, the greater length of time required for each section would be no drawback, and there would be a considerable reduction in cost since it would not be necessary to use diamond dust. ALBERT JOHANNSEN Grout, FRANK F. ‘“‘The Composition of Some Minnesota Rocks and Minerals,”’ Science, XXXII (1910), 312-15. A preliminary statement regarding the composition of certain Minne- sota rocks. There are given analyses of seven rocks and fourteen minerals. Two or three types of granite occur in laccoliths of considerable size in the Keewatin schists and are considered by the author as probably of that age. These granites are intersected by diabase, quartz diabase, and quartz porphyry dikes, and there occur a few masses of gabbro. Most of the Minnesota effusive rocks belong to three types of diabase which, chemically, are classed as Hessose, Bandose, and Auvergnose. The country rock was tested for copper. The common theory of the origin of the Lake Superior copper deposits is that of lateral secretion from the diabases. In the present tests it was found that copper occurs in all the main types of rock, and, so far as could be judged from the ten samples tested, the fresher the rock, the larger the amount of copper. It varied in amount from 0.029 to o.o12 per cent. ALBERT JOHANNSEN 184 PETROGRAPHICAL ABSTRACTS AND REVIEWS Hoécsom, A. G. ‘‘Ueber einen Eisenmeteorit von Muonionalusta im nérdlichsten Schweden,” Bull. Geol. Inst. Univ. Upsala, TX (i010) 220538) elem This is a description of the first iron meteorite found in Sweden. The essential constituents are the iron-nickel kamazite, taenite, and plessite. Troilite and daubréelite form a minor part. Chemically, the meteorite contains or per cent Fe and 8 per cent Ni. W. T. SCHALLER Dre LapPpARENT, JACQuES. ‘“‘Les gabbros et diorites de Saint- Quay-Portrieux et leur liaison avec les pegmatites qui les traversent,”’ Bull. de la Soc. Francaise de Minéralogie, XX XIII Goro), 254-70: Near Saint-Quay-Portrieux on the coast of Brittany, intrusive in mica schists, there is a mass of rather coarse hypersthene-gabbro with a periphery of dioritic facies. Both gabbro and diorite contain inclusions of a finer-grained hypersthene-bearing rock with the structure of beer- bachite. These rocks are cut by dikes of aplite essentially composed of labradorite and quartz. The diorite and the marginal, but not the central, part of the gabbro are cut also by small dikes of pegmatite com- posed essentially of microcline, albite, quartz, and a little biotite, with local muscovite and tourmaline. The albite has crystallized before the microcline. The principal types are represented by five analyses. The microscope shows the hypersthene of the gabbro in process of replacement by a mixture of biotite and quartz, and the augite more or less uralitized. In the peripheral “‘diorite’”’ both alterations are much more advanced; the augite is almost completely uralitized, and the hypersthene wholly replaced by biotite and quartz. The author ascribes these changes to the agency of the pegmatite and believes them to have been effected before the gabbro was fully consolidated. He considers for reasons not fully stated that the first phase was the production of soda-lime feldspar by the reaction with the femic magma of siliceous alkaline vapors, rich at first in soda. He supposes the vapors subse- quently to have become more abundant and richer in potash, water, and boric acid. The quartz and biotite, it is pointed out, would be formed by combination of the constituents of hypersthene with those of potash- feldspar; there is evidence that this reaction took place in the central gabbro before the hypersthene was completely crystallized, and in the PETROGRAPHICAL ABSTRACTS AND REVIEWS 185 peripheral “diorite”’ even before that mineral was individualized. The transformation of augite to amphibole, accompanied by crystallization of quartz, is considered to have been the final reaction, effected mainly by the water and boric acid in which the vapors became relatively richer as the consolidation of alkalies and silica progressed. M. de Lapparent believes that the action of the kind here described is common, and especially, that it has occurred in certain American rocks. F. C. CALKINS MicHEL-LEvy, ALBERT. ‘“‘Les terrains primaires du Morvan et de la Loire,” chap. v, ‘‘ Etude pétrographique et chimique des roches éruptives du faisceau synclinal du Morvan,” Bulletin des Services de la Carte Géologique de la France, XVIII (1908), 209-68. The area described is part of the central plateau of France, made classic by the thorough studies of the elder Michel-Lévy and others. Its rocks furnished the basis for some important principles of the science, and some of them are illustrated in the beautiful plates that accompany the ‘‘ Minéralogie Micrographique.” . 2p. 170. The most important part of this paper is a report on the Gold Hill Copper District by F. B. Laney (pp. 20-55). This district is located in the south-central part of the state just west of the Yadkin River. The rocks are slates and igneous rocks of various kinds, and of different periods of intrusion. The ores are (1) auriferous pyrite and chal- copyrite in a quartz gangue and (2) slightly auriferous bornite and chalcocite in a quartz epidote gangue. No attempt is made to corre- late the period or periods of ore deposition with a period of igneous activity or to determine the age of the ores. The remainder of the paper is chiefly statistical. Ee Rese Paleontology of the Coalinga District, Fresno and Kings Counties, California. By RAtepH ARNotp (U.S. Geol. Surv. Bull. 396). _ Pp. ror and plates 30. The district forms a strip roughly fifty miles long by fifteen miles wide along the border between the Coast ranges and the San Joaquin valley. The eastern slope of the mountains is formed by a great thick- ness of strata dipping toward the valley, successively younger formations being exposed to the east. The rocks of the district range in age from the Franciscan formation, which is probably Jurassic, to rocks of recent age, with an unconformity at the base of almost every formation. A description of the formations with faunal lists is followed by descrip- tion of forms from the Tejon formation (Eocene), the Vaqueros, the Jacalitos, and the Etchegoin formations (Miocene), and the Tulare formation (Freshwater Pliocene). a Ree: | Floors im free from Dust OF; Hygienic Schoolroom Floors Hygienic conditions in schools and in rooms of all public : buildings should be maintained with the most scrupulous care, for a dust-laden atmosphere is a constant menace to health. Continuous activity on the part of pupils stirs up the dust from the floor and keeps it in circulation. Proper ventilation will assist materially in keeping dust at a minimum, but the only solution of this problem is to eliminate the dust entirely. “This can be successfully accomplished by treating floors with - STANDARD — FLOOR DRESSING Actual use has proved beyond question its effectiveness as a dust-exterminator—the danger from disease contagion from dust being reduced almost one hundred per cent. Standard Floor Dressing is also a remarkable preservative for floors. It not only keeps the floors from splintering and cracking but actually lessens the labor in caring for them. Dealers everywhere sell Standard Floor Dressing in barrels and cans. Apply three or four times a year for best results. “WE WILL PROVE the remarkable efficiency of Standard Floor Dressing at our expense. On request we will apply it to the floor of one schoolroom or corridor free of all charge. To localities far removed from our agencies, we will send_free sample with full directions for applying. Every Board of Education, every School Superintendent, Principal and ‘Teacher should know how to guard the health of children under their care. Learn what eminent medical authorities have to say on the subject. You will find interesting reading in our free booklet ‘‘ Dust and Its Dangers.’’ Write for a copy. STANDARD OIL COMPANY (Incorporated) Ordinary dusting scatters but does not remove dust and germs. Use cheese-cloth dampened with tepid water to which a little Platt’s Seated Chlorides has been added. Wring out Us Paton, till dry so that it will not streak the furniture, etc. ) Platts Chlorides, The Qdorless Disinfectant. Alcolorless liquid, safe and economical. It does not cover one odor with another, but removes the cause. : 3 are now using a soap that not | makes you clean, but makes you 7 tell your friends about it, that the VOLUME XIX | NUMBER 3 THE JOURNAL of GEOLOGY A SEMI-QUARTERLY EDITED BY THOMAS C; CHAMBERLIN AND ROLLIN D. SALISBURY With the Active Collaboration of SAMUEL W. WILLISTON ALBERT JOHANNSEN WILLIAM H. EMMONS Vertebrate Paleontology Petrology Economic Geology STUART WELLER WALTER W. ATWOOD ROLLIN T. CHAMBERLIN Invertebrate Paleontology Physiography Dynamic Geology : ASSOCIATE EDITORS ‘ SIR ARCHIBALD GEIKIE, Great Britain GROVE K. GILBERT, National Survey, Washington, D.C. HEINRICH ROSENBUSCH, Germany CHARLES D. WALCOTT, Smithsonian Institution THEODOR N. TSCHERNYSCHEW, Russia HENRY S. WILLIAMS, Cornell University CHARLES BARROIS, France JOSEPH P.IDDINGS, Washington, D.C, ALBRECHT PENCK, Germany JOHN C, BRANNER, Stanford University HANS REUSCH, Norway RALPH A. F. PENROSE, Philadelphia, Pa. GERARD DEGEER, Sweden * WILLIAM B, CLARK, Johns Hopkins University ORVILLE A. DERBY, Brazil - WILLIAM H. HOBBS, University of Michigan _T. W. EDGEWORTH DAVID, Australia FRANK D. ADAMS, McGill University BAILEY WILLIS, Argentine Republic : CHARLES K. LEITH, eae of Wisconsin APRIL-MAY, 1911 CONTENTS CERTAIN PHASES OF GLACIAL. EROSION THOMAS C. CHAMBERLIN AND ROLLIN T. CHAMBERLIN 193 VALLEY FILLING BY INTERMITTENT STREAMS - - - = - - - A. E. Parkins 217 ORIGINAL ICE STRUCTURES PRESERVED IN UNCONSOLIDATED SANDS é CHARLES P. BERKEY AND Jesse E. HypE 223 RESTORATION OF SENIORS Pe SLORENSIS PROUD, aN AMERICAN COTY— ELOSAUR. =. =\ =. - - - S. W. WILLISTON 232 GEOLOGIC AND HSNO ele iB Swide IN: ole EB GON ABOUT CATCARA, WV OEINGE ZU Ne ee one See IRS A. BENDRAT 238 THE «AGE OF THE TYPE EXPOSURES OF THE LAFAYETTE FORMATION EDWARD W. BERRY 249 SRE GRIPPER Se Ok THE BEDFORD. AND BEREA FORMATIONS OF CENTRAL AND SOUTHERN OHIO, WITH NOTES ON THE PALEOGEOGRAPHY OF THAT EPOCH JEssE E. HypE 257 A POSSIBLE LIMITING EFFECT OF GROUND-WATER EON EOLIAN EROSION JosEpH E. PoGUE 270 RECENTLY DISCOVERED HOT SPRINGS IN ARKANSAS - = - - “A. H. Purpue 272 SETTER LL RS A AR a eS i ly I alae a sears NPS hE Lee 97 PORPOLOCHKGAL ABSTRACTS SANDY REVIEWS... oe Mee) ee ee ee Re Che Uuibersity of Chicago press CHICAGO, ILLINOIS AGENTS: CAMBRIDGE UNIVERSITY PRESS, Lonpon anp EpINBURGH WILLIAM WESLEY & SON, London TH. STAUFFER, Letpzic GOODE’S BASE MAPS a series of outline maps for all classes of work in applied sciences and the various fields of research Prepared by J. PAUL GOODE | Associate Professor of Geography in the University of Chicago This series is designed to meet the needs of teachers and students ina wide variety of work with maps. The maps are adapted to the use of classes of every grade, from the university to the common school: in geography, including commercial or economic geography, in physiography, geology, botany, zodlogy, anthropology and ethnology, sociology, econom- ics, politics and history. The maps have been prepared by being first drawn on a large scale, to insure accuracy of detail, and then greatly reduced in the engraving. In the quality of the drawing they are superior to most maps used in books and magazines. In two Sizes 8 x 103 in., 1 cent each 15 x 103 in., 3 cents each ~ 2 ° H . The World: on Mercator’s projection. No. 14. The British Isles: conic projection. No. 2. North America: onanequal-area pro- No. 16. Europe, Western and Southern: conic jection. projection. No. 3. South America: Sanson’s projection. No. 17. France: conic projection. No. 4. Europe: conic projection. No. 18. The Spanish Peninsula: conic projec- No. 5. Asia: Lambert’s equal area. _ tion. No. 6. Africa: Sanson’s projection. No. 19. Italy: conic projection. No. 7. Australasia: Mercator’s projection. No. 20. Central Europe: conic projection.* No. go. America (U.S.): conic projection* No, 21. The German Empire: conic projec- No. 10. America (U.S.): state outlines only; ton. conic projection. No. 24. The Levant: conic projection.* *.* Double size, not yet ready. When ordering the small or standard size, reference to the map desired should be by the appro- priate number in the above list. When the double-size map is desired, the symbol “A” should follow the number. Thus, No. 4 means the 8x1o% inch map of Europe; No. 4A means the 15 x 103 inch map of Europe. i i Map of America (U.S.) Showing All Counties 15 X Ios in., 3 cents each 21X15 in., 5 cents each THE UNIVERSITY OF CHICAGO PRESS CHICAGO, ILLINOIS BR Bags Riseee pie ees Oe ee ee tr Re eg Ps heaped Bie, Shy ah cay ae THE FOURNAL OF GEOLOGY APRII-MAY, rori CERTAIN PHASES OF GLACIAL EROSION THOMAS C. CHAMBERLIN anp ROLLIN T. CHAMBERLIN The University of Chicago Oscillation, more or less rhythmic, seems to be a phenomenon of the intellectual, as well as of the physical world. The doctrine of glacial erosion has its ups and downs in quite typical undulatory fashion. It seems that even individuals at times ride on the crest of the wave of advocacy and at other times sink into the hollows of doubt. These moods are apt so to distribute them- selves that while some workers are on the crest others are in the trough. The crest-riders have recently been much the most in view, but just now voices from the hollows of doubt are heard. The president of the British Association for the Advancement of Science, speaking from official vantage ground, voices a cautious skepticism as to the glacial parentage of certain kinds of configurda- tions that are held by others to be the erosive offspring of glaciers.* Professor Garwood goes beyond the measured skepticism of Dr. Bonney and gives a critical analysis of his grounds of doubt and laudably matches his destructive criticism with constructive interpretations. In these interpretations, he marshals topographic phenomena in support of the view that protection? is the character- istic effect of glaciers rather than erosion. tT. G. Bonney, Presidential Address before B.A.A.S. (Sheffield, 1910), Science, XXXII (1910), 321-36, 353-03. 2. J. Garwood, ‘‘Features of Alpine Scenery Due to Glacial Protection,” Geog. Jour. (September, 1910), pp. 310-39. Vol. XIX, No. 3 193 194 THOMAS C. CHAMBERLIN AND ROLLIN T. CHAMBERLIN So, too, among those who believe in the efficiency of glacial erosion, there has long been some doubt as to the truth, or at least as to the adequacy, of some of the processes to which the erosion has been attributed. It seems worth while, therefore, to add to the growing mass of matter some notes suggested by phenomena recently seen by us, without presuming that much is new either in the observations or in the suggestions. | I. THE CRITICAL STAGE FROM WHICH CERTAIN EROSION TYPES START It has seemed to us advantageous to study the initial stages of erosion to see, if possible, precisely what action gives the start to the type of erosion which thereafter controls the configuration, for it is the initial turn that most delicately measures the balance between the opposing tendencies. The contours that spring from ordinary wear and weathering are well known and may be restored from remnants when the greater part has been lost. Even when there has been no change in the agent and only a slight change in its mode of action, the old configuration can be distinguished from the new; as, for familiar example, the remnants of a peneplain are commonly made out with confidence after most of the plain has been cut away by the rejuvenation of the very drainage system that formed it. Much more clearly can remnants of contours be rebuilt into their originals when some new agency intervenes, especially a new agency whose habit of sculpture is distinctively at variance with that of the previous agency. As surface configurations are traced from regions dominated wholly by ordinary wear and weathering into regions that have been affected by local glaciation, it is usual to find the lower slopes of the unglaciated region and, in the main, the brows and tops of its hills and higher elevations, up to a certain limit,‘marked by contours of the familiar wear-and-weather type whose interpreta- tion is clear and whose restoration, when mutilated, may be made with great confidence. As such contours are traced into higher latitudes or higher altitudes where local glaciation has entered CERTAIN PHASES OF GLACIAL EROSION 195 sparsely as a modifying factor, it is usual to find the flowing con- tours of the wear-and-weather type replaced in certain spots by a type that may be said to be unconformable to the prevalent one, a type in which concavity replaces convexity, a type in which the surface has been broadly scooped out locally rather than rounded off generally or narrowly incised. ‘The broad scoop-like mode of excavation, as distinguished from the gully-form mode of narrow incision, is held to be distinctive in that it implies an agency that deployed its effects laterally rather than one which concentrated its action on axial lines. This, it is to be noted by way of precau- tion, is a distinction that applies chiefly to the initial stage of the two modes of erosion. They remain distinguished throughout but are not so declaredly diverse in later stages. The lodgment of snow, which is the primary factor in glacial work and determines its initial deployment, is controlled by the wind to an exceptional degree, and wind action is chiefly horizon- tal in its effects and is thus distinguished from rainfall and run-off, whose dominant actions are vertical. While the very first phases of this difference of action are not very important in themselves, they are believed to be significant as the initial factors in the localization as well as the deployment of the two classes of erosion. The relative locations of greatest rain-work and greatest snow- work respectively.—Precipitation is intimately dependent on the ascent of air so well laden with moisture that it reaches saturation by reason of the expansion and cooling caused by the ascent. It is for this reason that the ascent of moist air caused by rising over the windward face of any marked relief of the topography deter- mines precipitation on or near that face. As is well known the windward sides of mountain chains thus receive more precipitation than the leeward sides, as a rule. This holds true of snow-pre- cipitation as well as rain, though the snowfall is less prompt and less well localized. Where mountain ranges are broad and com- plex the snow caught on the windward side is usually greater than that which lodges on the leeward side, and the glaciers on the windward sides of mountain ranges are usually larger than those on the leeward sides. But such general community of distribution does not hold in detail, for the wind comes in as a 196 THOMAS C. CHAMBERLIN AND ROLLIN T. CHAMBERLIN local differentiating agency. Acting on rain, wind increases the amount per unit area that strikes the surface of an eminence on its windward side; it also somewhat increases the force of the impact on that side. On the other hand, wind tends to drive falling and fallen snow around the wind-swept side of the eminence into its lee and to heap it up in the eddies there, and on the areas protected from the wind. Thus the snowfall that, in the absence Fic. 1.—Snow lodgment on the side of the summit ridge of Mt. Victoria, Canadian Rockies. This ridge forms the continental divide. The snow has lodged on the Eastern or Albertan side in the lee of the crest. Photo. by R. T. C. of wind, would come to rest on the windward and lateral slopes of an eminence and later must drain away on these slopes is, under the action of wind, concentrated notably in patches in the lee. Considered therefore in detail, rain action is somewhat intensified on the windward side of prominences, while snow lodgment, leading on toward glacial action, is more markedly concentrated on their leeward slopes. The field use of this distinctive localization of rain-work and of snow-work respectively is qualified by the fact that, while the CERTAIN PHASES OF GLACIAL EROSION 197 prevalent air movement of a region may be nearly constant in general, the cyclonic movements that are the immediate agents that bring on precipitation introduce variation in the particular direction from which the wind blows at the critical time when the storm is on and the distinctive work in question is done. In the mid-latitudes of the northern hemispheres, the general air movement is toward the east but at the times of storms the wind not uncom- monly comes from the eastward. However, the general law that snow lodgment is most abundant on the prevailingly leeward sides of prominences seems to hold good. This is greatly aided by the shifting that takes place in the intervals between storms. The fact that the eddies formed in the lee of crests, domes, and knobs are the common spots of lodgment carries as a corollary the observation that the forms of the snowfields are usually broad, or ovoid. The windward edge is usually arched, and is often thickened near its upper border. Not unfrequently the thickened snow mass is wider transversely than in the line of slope. Often, too, it must be noted, the lodgment is concentrated in ravines and valleys that were shaped previously by drainage erosion, and in such cases the localization is less distinctive. The case best suited to a discriminative study is a broad or transversely elongate lodgment of snow in the lee of a well-rounded eminence from which the normal run-off is divergent. So long as such a snow mass lies passively where it lodged, there can be little doubt that it is protective rather than erosive, when compared with normal surface action. So long, too, as the later action is confined to a slow annual melting of the snow and a quiet run-off of the resulting water, the snow and water combined perhaps do less erosive work, on the whole, than would be done by the more forceful impact and the more prompt run-off of the equivalent rain, though qualifying conditions must be recognized on both sides. . The case of snow vs. rain, under these conditions, is not more than debatable at most and the modes of erosion in the two cases are essentially identical. But when the snow accumulates perennially so as to move as snow-ice in glacier fashion, the modes of erosion become diverse, 198 THOMAS C. CHAMBERLIN AND ROLLIN T. CHAMBERLIN and the configuration of the eroded surface is the test of the domi- nance of the one or the other type. It is obvious that the least eroded part of the eminence must come to stand forth and the Frc. 2.—Diagram to illustrate the effect of erosion upon a hill, on the assumption that the capping of ice, SC, is protective. The dotted line represents the original outline of the hill; the solid line, the contour resulting from erosion. most eroded part must retire toward the center. If the snow- covered flank or brow is indeed a protected area, it must gradually come to stand forth from the retiring wear-and-weather contours Fic. 3.—The same hill as in Fig. 2, eroded according to the hypothesis that ice is a superior eroding agent. SC represents the original snow bank which comes to occupy a basin as erosion goes on. adjacent, as a rather definite embossment, as illustrated in Fig. 2. As time goes on, the summit of the hill should migrate toward this protected area and it should tend to become the summit, while the snow-cap in turn migrates into its lee. A marked asym- metry should gradually develop. On the other hand, if the snow mass, accumulating from year CERTAIN PHASES OF GLACIAL EROSION 199 to year, comes to take on motion as a glacial body, the erosion to which its motion gives rise must take a form coincident with the moving part of the snow-ice mass. The erosion is assumed to be due to the adhesion of the snow-ice mass to the ground on which it rests—to the soil and loose rock at the start, to the progressively loosened and ground rock below later. A broad patch of soil and loose rock coincident in form with the moving part of the snow mass is first dragged away and the configuration of the scar is distinctive of the action. If erosion beneath the moving glacier mass continues the excavation will in time come to have the form shown in Fig. 3. Such excavations are to be looked upon as embryo cirques. They are found on the lee crests, brows, and slopes of round-topped mountains known to have been subjected to local glaciation. Less typical initial cirques are formed in ravines where snow lodgment gives rise to glaciers. If absolute certainty that there has never been any previous glacial action in a given region is regarded as a prerequisite to an irreproachable illustration of this class of actions, such a case is difficult to demonstrate because the configurations left by the older glaciations have often been so largely lost in the subsequent sculpturing of the common wear-and-weather type that the absence of previous glacial work is hard to prove in regions likely to have been glaciated recently, but this is only a question between the work of different glacial stages, not between glacial and aqueous methods. But though a region has been subjected to previous general glaciation, even rather recently, geologically speaking, the typical effects of local glaciation on rounded contours are dis- cernible much as in wholly unglaciated regions, for the con- tours shaped by the general ice movement conform to the domi- nant horizontality or the low inclination of such general ice move- ments, while the lines of local ice movement are decisively down- ward in conformity to the local slope. These considerations are here put in the theoretical form, but they suggested themselves almost as inductions during a summer trip along the coast of Norway in 1909. ‘They arose naturally from the abundant and instructive phenomena of that region, where former glacial action merges into present action. The configurations 200 THOMAS C. CHAMBERLIN AND ROLLIN T. CHAMBERLIN wrought by the older general glaciations do not seriously mask the distinctive work of the local glaciation that has followed and is, in some part, in action still. Broad excavations of the initial cirque type are common on the brows and slopes of the rounded mountains and on the islands that fringe this coast and on the mainland itself. They seemed to us clearly to be more common on the eastward sides of the islands than on the westward. The initial types are chiefly the products of modern action; indeed in Fic. 4.—Basins hollowed in a hillside by tiny glaciers. From the coast of Norway. Photo: by ReL. C. many cases the basins are still occupied by the snow-ice mass to which their shaping is due. The whole series taken together show various stages of the work of snow accumulation and earth excavation. Small, relatively wide basins, scooped broadly from hillsides, are variously occupied or empty according to altitude, latitude, or other condition favoring snow accretion or snow wastage. Their dimensions range downward to hollows not unlike pits on the brows of drift hills and upward to mountain cirques of typical form and magnitude. ‘They also range from mere cirque heads to cirque heads with short glacial appendages and thence on to longer and longer glacial tails until the peculiarities of the CERTAIN PHASES OF GLACIAL EROSION 201 head-work in the cirques are lost in the more familiar body-work and tail-work of the more accessible parts. Various stages and transitions are shown in the accompanying photographs. Fig. 4 shows five well-developed basins escalloped in a hillside. The two hollows on the left are round and wide and terminate below in well-defined platforms or steps at nearly the same level. Fic. 5.—A concave scallop on the brow of a projecting embossment. Apparently this is the work of sapping by the ice at the base of the cliff. Note the rounded convex glacier-polished outlines of the rest of the embossment. In the background is the Lyskamm, central Pennine Alps. Photo. by R. T. C. They are approximately as wide as they are long, showing that the ice which accumulated there has eaten its way in a distinctly broad fashion into the rock slope on which it lay. The vertical distance which any given part of the ice has moved its rock is small relative to the total amount of transportation accomplished. The work has been done very locally compared with the longitudinal movement of typical water action. The next two basins continue down the slopes to points much nearer to the sea. There has been 202 THOMAS C. CHAMBERLIN AND ROLLIN T. CHAMBERLIN more advance movement of the ice here. The basin on the right has become a glacial valley in an embryonic stage and the work of water erosion seems to form a larger factor. Sculpturing of similar sort is illustrated by Fig. 5, from the Swiss Alps. The brow of a long spur descending from the Zwil- linge has been scooped and hollowed in concave fashion by the sapping action of glacier ice. Occurring in the midst of a still Fic. 6.—The Glacier des Grandes Jorasses on the Italian side of the chain of Mont Blanc. The ice has sunk its bed into the rocky mountain wall and worked backward as implied by the distinct bench upon which it rests. Photo. by R. T. C. strongly glaciated area, this case is interesting for the reason that such sculpturing has been at work here for a comparatively short time only. ‘The rounded rock contours below and to the right of the hollow excavation have at no distant date been scraped and polished by the larger glaciers descending from the peaks above. The ordinary abrasive action of a moving body of ice is here illustrated. But the much smaller mass of snow and ice at the base of the cliff in the hollow appears to have operated in the very different and more potent manner of basal sapping at the schrund line. CERTAIN PHASES OF GLACIAL EROSION 203 Fig. 6, from the chain of Mont Blanc, represents the Glacier des Grandes Jorasses on the Italian side of the rugged mountain mass of the same name. Other similar glaciers to the left and right have etched their basins into the upper slopes of this great mountain rampart. These glacier-filled basins are deeply sunken and are as broad or broader near the base of their cirque walls than they are farther down toward the ends of the present ice tongues. At their heads they are terminated by precipitous rock walls. Extremely precipitous cliffs come down to the Glacier de Rochfort from the Aiguille du Géant and the col between it and the Aiguilles Marbrées. From these rock walls behind the ice there is a very decided change in slope to the gentle incline of the glacier floor below. In just the same way there is a very abrupt change of slope from the precipitous rocks of the Grandes Jorasses and Mont Mallet to the very moderately inclined surface of the Glacier des Grandes Jorasses at the foot of these steep cliffs. It is at the point where these cliffs join the less inclined basin floor beneath the glacier that the greatest cutting has occurred. Such a profile of cliff and floor coming together at a sharp angle is quite _unlike any gully erosion developed by ordinary running water in mountains of massive crystalline rocks. The greatest cutting has been beneath the glacier in the neighborhood of the bergschrund and directed backward into the mountain. II. CERTAIN SIGNIFICANT POSITIONS OF CIRQUES In our sketch of the initiation of cirques, we gave preference to cases located on leeward aspects of eminences favorable for snow lodgment but unfavorable for the concentration of running water. We noted that if the lee brow were protected by its snow covering, the crest should slowly shift toward the protected spot and the protecting snow-cap should shift in turn to its lee and thus combine to shape forth an asymmetrical mountain horn. On the other hand, if the snow mass becomes a superior erosive agent when it begins motion, and digs out a broad basin which in turn adds to the catchment of snow, and if at the same time the embry- onic glacier stopes headward, it, in its way, moves toward a summit position. It is clear that rainfall does not concentrate toward 204. THOMAS C. CHAMBERLIN AND ROLLIN T. CHAMBERLIN the summit of a rounded eminence in this way. Its trenches do advance headward, but they take the form of ravines, gulches, and gullies eating sharply, not broadly, backward. ‘The positions of cirques that are fully developed may be studied for evidence confirmatory of these deductions. In such study perhaps the most striking illustration of summit-ward creep is found in the crater cirques, a form that has attracted the attention of observant travelers but has not played as large a part in glacial literature as Fic. 7.—The Rendalstind on the west side of the Lyngenfjord, Norway. The summit has become crater-shaped by ice sculpturing. Photo. by R. T. C. perhaps it should. Mountains with crater-like summits are quite common along the Norwegian coast above the Arctic Circle, and they are likewise frequent enough in the Lofoten Islands to give characteristic profiles to the views obtained there from passing steamers. The Rendalstind, on the west side of the Lyngenfjord (Fig. 7), is an illustration of the type in a not very advanced stage of develop- ment. Glacier action in the summit basin is today actively in progress. Other mountains of the region reveal much more pronounced sculpturing of this sort where the action has either been more prolonged or more intense. Such a case is illustrated by CERTAIN PHASES OF GLACIAL EROSION 205 Fig. 8. This crater mountain, which happens to be crossed by the 7oth parallel, comprises the north end of Kaagé Island. Origi- nally it appears clearly to have been a more or less rounded dome or knob. A cirque starting with snow lodgment high up on the northeast slope of this eminence appears to have worked back Fic. 8.—A crater-like mountain top in a more advanced stage of erosion. The outer slopes of the conical mass show the familiar abrasive action of past general glaciation together with the lines of ordinary meteoric erosion. The steep walls of crater-like cirque are due to sapping by localized glaciers of late date. Kaagé Island, coast of Norway. Photo. by R. T.C. toward the summit by a stoping process until the cirque pit has come to occupy a sub-summit position. The mountain top has been hollowed out and now only a shell remains in place of the former flat-topped mass. It is like a volcanic crater broken down on one side. The inner walls are steep and cirque-like and the crater portion is filled with deep snow. Water erosion is not adapted to this sort of sculpturing. The central basin with cir- cular cirque cliffs gives every appearance of having resulted from THOMAS C. CHAMBERLIN AND ROLLIN T. CHAMBERLIN 206 ‘OL a Aq ‘oqoyd . “woreurI0F YI VY} Jo souvIsISOI oY} UI SadUDIByIp Aq poururiojzop yred ur Ayqeqoad osye sem JoDvIS ay} Jo Apog 94} Yyeeueq Surdo}js Jo yNser oy} ApIed o[TYM uIseq oY} Jo s[pprum 9y} ur doqs ydnaqe oy, “WH 9}eII0S ay] Jo SsoUMOIILU ay} pue ulseq enbaio oy} jo yYpeoiq 9Yy} Iv oJ0U Jo sjulod ayy, ‘a[suerpenb oprnqjay, ‘e8pry vrydog 3g Jo yurod Suneurm -[ND oY} WoIZ JSvayINOs SUIxOo] Useg ‘opeIO[OD Jo sureyunoy uenf ues dy} ur onbs1 yrurums snoredvs Alaa W—'6 ‘O1g CERTAIN PHASES OF GLACIAL EROSION 207 continued sapping and quarrying by glaciers eating their way backward much as they have in the Alpine cases cited. Fig. 9, from the San Juan Mountains of southwestern Colorado, is a striking example of a much more capacious summit cirque. The basin here is very broad, with a nearly level floor, while the cirque rim has been undercut till only a narrow horseshoe-shaped serrate ridge remains. Its jagged crest varies in altitude from 13,000 to 13,200 feet, while the mean elevation of the broad floor is about 12,800 feet. The breadth of the basin and the steepness and thinness of the amphitheater walls that skirt it show that this type of action has here gone about as far as it well could as a single stoping operation. The whole constitutes a signal case at its very climax. In the light of these illustrations, particularly Figs. 8 and 9, there seems no ground to doubt that the erosion suffered by the non-glaciated parts in such situations at least falls greatly behind that suffered by the parts covered by ice. III. THE DISTINCTIVE WORKING FACTOR The decided superiority of moving ice over moving water as an erosive agent lies chiefly, we think, in the rigid hold of the ice on matter set in its base or sides. This is sharply contrasted with the adhesion of water which is so feeble as to scarcely warrant the term “hold” at all. Water action finds some compensation, indeed, in the higher velocity it usually gives to the matter it carries, but near the point of origin of action—and this is the location chiefly under discussion here—the water is so distributed as not to be able to acquire much efficiency from concentration. The ice mass, on the contrary, is rigidly unified; its velocity is indeed low, but its mass and fixed coherence are high. It is in respect to this coherence that exaggerated views of the viscousness of ice are perhaps most misleading. Rigidity of grasp and mechani- cal firmness of action are specifically implied in the groovings, gougings, and crushings that distinguish the action of glaciers. Effectiveness of corrasive action is further implied in the chemical and physical nature of the rock-flour and fragments which these groovings, grindings, and crushings contribute to the glacial till 208 THOMAS C. CHAMBERLIN AND ROLLIN T. CHAMBERLIN and to the wash-products immediately derived from it through glacio-fluvial assortment.t The graving of a glacier’s bed by rock fragments set in its base or sides may be cited as specific evidence of an essentially rigid hold on the graving tools, and of an internally rigid, rather than fluent, motion of the mass holding the tools. The glacial grindings that are borne out with the subglacial waters and give milkiness to glacial streams seem to us irrefutable evidence of effective rasping and grooving of the rigid type, not of simple viscous overcreep. The very marked contrast between the turbid waters that flow from beneath glaciers and the relatively clear waters that flow down adjacent ungla- ciated valleys is very impressive and spectacular evidence of the superior erosive powers of glaciers. Closely allied to this lesson from grindings in transit is a less obtrusive one drawn from the contrast in the points where coarser matter which only strong transporting agents can handle is con- centrated respectively in glaciated and in non-glaciated valleys in regions of the same general type. The upper parts of non- glaciated mountain valleys in cold regions are usually burdened with heavy talus and large loose masses which the drainage is unable to carry away, while the glaciated parts of similar valleys are usually well scoured out and the moutonnéed sides and bottoms of U-shaped troughs take the place of the craggy outliers of V- shaped trenches in unglaciated valleys. But in the lower portions of the glaciated valleys below the reach of recent glacial action, aggradation very generally prevails, while in similar non-glaciated valleys degradation generally prevails, if not absolutely, at least relatively. Students of Alpine regions will recall multitudes of illustrations. A similar lesson is even more impressively enforced on the borders of the late Pleistocene glacial areas. In strong contrast to the state of the valleys of the adjacent driftless regions, the great glacio-fluvial valley trains with their thick heads next to the ice border, as well as the frontal aprons, show very conclusively the overladen condition of the glacial waters and_ their marked incompetency to fully carry away their burdens. = “Hillocks of Angular Gravel and Disturbed Stratification,” Am. Jour. Sci., XXVII (May, 1884), 378-90. CERTAIN PHASES OF GLACIAL EROSION 209 Now the regional precipitation is much the same for like areas and like situations in the glaciated and in the non-glaciated val- leys. Such differences as there may be appear to favor a greater run-off in the glaciated than in the non-glaciated basins, for the former are likely to be cooler and hence better condensers and the concentration of snow by wind action is there more effective. If so, the advantage in absolute carrying power lies with the waters of the glaciated valleys. If therefore the passing of a part of the precipitation through the glacial form and the moving of this part, so far as 1t moves as Ice, is protective, the débris should tend to remain in the upper glaciated sections thus protected; while the waters of the valley below the glacier having some excess in volume and having less burden to carry should tend to degrade the lower reaches of the valley more effectively than if the glacier were absent. That the facts are precisely the opposite seems good additional evidence that the glacial form of water compared with the aqueous form increases notably the corrasion and the trans- porting power. And so it seems to us that the fringing outwash aprons and the thick-headed valley trains of the Pleistocene join with the aggraded states of the lower stretches of present glacial valleys and with their turbid glacial streams, their mouton- néed walls, and their glacial scorings to testify to the superior erosive efficiency of glaciers. Traced back analytically to the properties that gave rise to them, these corrasive products point to a glacier’s power to take firm hold on rock fragments imbedded in its base and sides and move them on while it uses them as graving tools. Dhe-amountvor erosion is not great in either case, but in the opinion of the writer the occurrence of an unconformity in this region at appar- STRATIGRAPHIC POSITION OF LANCE FORMATION 367 ently the same horizon as that in the Dakotas tends to show that the break has more than local significance.” The observations recorded by Mr. Calvert in the eastern portion of this South Dakota field were supplemented and extended to the western part of the state by Mr. V. H. Barnett in 1910, while making a hasty reconnaissance trip across the country. For instance, on the south bank of the Moreau River, near Govert P.O., in Sec. 21, Tr5N, R8E, South Dakota, Mr. Barnett found what is perhaps the most marked evidence yet recorded of unconform- able relations between beds thought to be Fox Hills and the Lance formation. Mr. Barnett traced the Lance formation continuously from the central part of the state to the point mentioned above, where it was found practically horizontal, while the underlying beds dip to the northwest at an angle of about 10°. These under- lying beds appear to be in the stratigraphic position of, and litho- logically similar to, beds resting immediately on Pierre shale, at Hoover, about 1o miles southwest, and there is no reasonable doubt regarding their age, but no paleontologic evidence was secured— or sought—at this locality. If the beds are not of Fox Hills age they must be older, which would indicate an unconformity of even greater magnitude than is presumed. Mr. Barnett secured a photograph of this section which he has kindly permitted me to reproduce here as Fig. 1. The full section of the supposed Fox Hills is not exposed at this point, but some distance west, at Castle Rock Butte (T12N., R5E), the following section was measured; Pierre 50 feet; Fox Hills 125 feet; Lance formation 140 feet, the latter overlain by higher Tertiary. The unconformity spoken of above by Calvert as occurring in Sec. 32, T7N, RO61rE, in Custer County, Montana, is on the west side of the anticline extending southeast from Glendive. It is clearly shown in the accompanying Figs. 2, 3, the negatives of which were made by Mr. C. F. Bowen, by whose consent they are included here. The Fox Hills with a thickness of about 70 feet dips at an angle of 5°, while the overlying Lance formation is horizontal. Mr. Calvert’s observations concerning the occurrence of the marine Fox Hills invertebrates in the basal members of the Lance 368 F, H. KNOWLTON formation may be briefly alluded to. It appears that in the hundreds of localities throughout the Dakotas, Wyoming, and Montana at which the contact between Fox Hills and the Lance Fics. 2, 3.—Eastern part of Custer County, Montana, showing erosional uncon- formity between Fox Hills and Lance. Photographs by Bowen. formation has been examined, only five localities, all within a limited area in South Dakota, have been noted in which the marine Fox Hills invertebrates occur above the acknowledged top of the Fox Hills, where they are often found commingled with certain STRATIGRAPHIC POSITION OF LANCE FORMATION 369 brackish-water forms. It does not appear that they have ever been found at a greater distance than 12 or 15 feet above the top of the Fox Hills, and since it further appears that in none of the four sections given’ does the Fox Hills exceed 115 feet in thickness, there is every probability that they were re-deposited in the chan- neled upper surface of the Fox Hills and that they did not live in association with the brackish-water forms with which they are now found entombed. The plant collections obtained from the Lance formation by Mr. Calvert and the members of the several parties under his charge show conclusively that the relation of this flora is unmis- takably with the Fort Union. In fact with the information at hand regarding distribution it is practically impossible without stratigraphic data to distinguish between the flora of the Lance formation and that of the acknowledged Fort Union. The lists of these collections follow: [5437]. NW ¢ Sec. 5, T2N, R88W, S. Dakota. North bank of Cannonball River, at McCord coal-bank, 150 feet above base of beds. Sequoia nordenskiéldi Heer Thuya interrupta Newb. Glyptostrobus europaeus Unger Populus speciosa Newb. Populus amblyrhyncha Ward Paliurus colombi? Heer Sapindus grandifoliolus Ward Celastrus alnifolia ? Ward Juglans sp. ? 2 new forms, gen. ? [5443]. SW % Sec. 23, T23N, Ra1E, Black Hills Meridian, 150 feet above base of beds. Sequoia nordenskidldi Heer Leguminosites arachnioides Lesq. [5444]. Near { cor. E. side Sec. 13, T22N, R22E. Black Hills Meridian. 125 feet above base of beds. 2 or 3 of same species as unnamed forms from the ‘‘somber beds”’ at Glendive, Montana. [5430]. Rattlesnake Butte, Cheyenne Indian Reservation, S. Dakota. 100 feet above base of beds. Glyptostrobus europaeus Newb. Taxodium occidentale Newb. tAm. Jour. Sct., XXX (1910), 174-77. [5422]. [5431]. [5432]. [5433]. [5434]. [5436]. F. H. KNOWLTON Viburnum marginatum Lesq. Cornus newberryi Hollick Salix sp. Quercus sp. 3 or 4 forms that are identical with unnamed species from Glendive, Montana. Near SE + Sec. 20, T14N, R1oE, Black Hills Meridian, S. Dakota. 100 feet or less above base of beds. Sequoia nordenskidldi ? Heer Sequoia langsdorfii Heer Platanus platanoides ? (Lesq.) Kn. Viburnum sp. (same as new species from Lance of Converse County, Wyoming). Lauraceous leaf (same as form found in “somber beds” at Glendive, Montana). SW i Sec. 4, TroN, R18E, S. Dakota. 300 feet above base of beds. Thuya interrupta Newb. Populus amblyrhyncha ? Ward Viburnum elongatum Ward Viburnum sp. ? Grewiopsis whitei ? Ward SE } Sec. 25, T20oN, R18E,S. Dakota. 300 feet above base of beds. Sequoia nordenskidldi ? Heer Zizyphus cf. Z. hyperboreus Heer Populus ? sp. Platanus ? sp. Sec. 33, T20N, R20E, S. Dakota. 150 feet above base of beds. Ginkgo adiantoides Heer Platanus raynoldsii ? Newb. Sapindus grandifoliolus ? Ward Viburnum (apparently same as unnamed species from Lance of Converse County, Wyoming). SE + Sec. 12, Tr9oN, R24E, S. Dakota. Base of beds. 3 fragmentary leaves, apparently same horizon as No. 5436. NE cor. Sec. 7, T17N, R24E, S. Dakota. Base of beds. Platanus haydenii Newb. Viburnum elongatum Ward Viburnum marginatum ? Lesq. Sapindus grandifoliolus ? Ward Dombeyopsis sp. Polygonum ? sp. Lauraceous leaf like that of Glendive, Montana 2 species same as unnamed form from Glendive STRATIGRAPHIC POSITION OF LANCE FORMATION 371 [5423]. South of Moreau River about 7 miles above Thunder Butte P.O., S. Dakota. Sec. 35, T14N, R1oE. Lower 4 feet of Lance formation. Thuya interrupta Newb. Sequoia nordenskidldi Heer Sequoia acuminata ? Lesq. Populus cuneata Newb. Viburnum marginatum ? Lesq. Leguminosites ? n. sp. Cyperacites sp. Monocotyledon—new It needs but a glance at the above lists to show how preponder- ating is the Fort Union facies. CONVERSE COUNTY, WYOMING Although Converse County, Wyoming, is the type locality for the Lance formation, and has been visited again and again by geologists and paleontologists, it is still a perennial source of dis- cussion and difference of opinion. From the first, difficulty has been experienced in drawing the line between the highest marine formation—the Fox Hills—and the overlying dinosaur-bearing beds. The Fox Hills was estimated by Hatcher to have a thickness of 500 feet, and consists of an alternating series of sandstones and shales, with massive sandstones at the top which contain numer- ous large concretions and a rich marine fauna of characteristic Fox Hills species. The upper line was drawn somewhat arbitrarily at a six-inch band of hard sandstone which was thought to separate the fossil-bearing Fox Hills sandstone below from similar but sup- posedly non-fossiliferous sandstones above. When Dr. Stanton and I visited this region in 1896 we failed to secure evidence for changing the top line of the Fox Hills as established by Hatcher, though we did find four species of brackish-water invertebrates in clays above a forty-foot bed of massive sandstone over 400 feet above the highest fossiliferous Fox Hills horizon in that particular section. So the question rested until 1909, when Messrs. M. R. Camp- bell, T. W. Stanton, and R. W. Stone spent nearly a week in the region. Their principal contribution to the knowledge of the 272 F. H. KNOWLTON TCG; stratigraphy of the area was, according to Stanton,’ “the discovery that the marine Fox Hills deposits extend about 4oo feet higher than had previously been determined, and that non-marine coal- forming conditions were temporarily inaugurated here before the close of Fox Hills time.’ If Hatcher’s estimate of the thick- ness of the beds assigned by him to the Fox Hills was anywhere near correct this ‘discovery’? would seem to increase the total thickness to about goo feet, yet nowhere in the paper mentioned is a thickness greater than 400 or 500 feet claimed for it. This appears difficult to explain unless the lower as well as the upper limit of the formation has been changed. A number of sections are given by Dr. Stanton, in one of which at least, namely that on Buck Creek, the top of the Fox Hills appears to have been fixed by the presence of the plant Halymenttes major. ‘The thickness of the Fox Hills in this section is given as 505 feet, though the highest horizon at which marine Fox Hills invertebrates occur is about 180 feet below the top. In the section made on the divide between Lance and Buck Creeks the Fox Hills is said to have a thickness of 445 feet, though the lower member of the section only (30 feet above the Pierre shale) is indicated as containing a Fox Hills fauna. The section made on the south side of the Cheyenne River at the mouth of Lance Creek shows a thickness of 405 feet of Fox Hills above the Pierre, but the highest point in the section at which marine Fox Hills invertebrates were found is over too feet below the top. It further appears from this section that the upper four members, aggregating 115 feet in thickness, contain carbonaceous and lignitic shales as well as fragments of dinosaur bone and brackish-water invertebrates, certain of which are the same as those found in, and there said to indicate the Laramie age of, the 400 feet of beds already mentioned as reported by Stanton and Knowl- ton above the typical marine Fox Hills.2, To the writer it seems altogether more probable that the four upper members of this section belong to the Lance formation and not to the Fox Hills, and it appears that this was the view at first entertained by Dr. t Am. Jour. Sci., XXX (1910), 184. 2 Bull. Geol. Soc. Am., VIII (1897), 130. STRATIGRAPHIC POSITION OF LANCE FORMATION 373 Stanton, who says,’ ‘‘When studying the section it was believed that the upper four members belong to the Lance formation, but afterward when comparison was made with sections of the south end of the field it seemed more possible that all the beds examined here belong to the Fox Hills.”’ If this portion of the section is placed in the Lance formation, where it certainly appears to belong, the thickness of the Fox Hills in the section is reduced to 285 feet, or but little more than half of the maximum thickness assigned to beds of this age in the Converse County region. While this evidence may not be considered conclusive, it must at least be admitted that it strongly suggests the possibility that even here, as in the areas already discussed in the Dakotas and Montana, the Fox Hills is of variable thickness, due to the erosion of the upper portions before the deposition of the Lance formation. It is to be admitted, however, that all who have studied the Converse County areas have had, and still have, difficulty in fixing the upper line of the Fox Hills, but in this connection it is to be pointed out that while many students have visited or collected in the region, it still awaits the careful, painstaking study that has been given other fields, such, for instance, as the areas in the Dakotas and Eastern Montana, which have been described by Mr. Calvert. And in this connection it may be mentioned that although in Converse County the exact location and extent of the unconformity between Fox Hills and Lance is not definitely known, the time interval is undoubtedly indicated, since 150 miles to the southeast (i.e., opposite the-mouth of the Medicine Bow River) the same dinosaur-bearing beds are above an unconformity which separates them from 6,000 feet of unquestioned “‘Laramie,”’ while 100 miles to the east in the Dakotas, the Lance formation rests on an uneven surface which in some cases represents the removal of practically the whole thickness of the Fox Hills of the region. As a possible explanation of the difficulty experienced in detect- ing the presence of the unconformity between the Fox Hills and overlying Lance formation in this area, the following facts may be offered: the localities in Eastern Montana and Western South Dakota where the examples of the distinct angular and erosional tAm. Jour. Sci., XXX (1910), 185. 374 F. H. KNOWLTON unconformity are so well exhibited are all adjacent to the anti- clinal uplift which Calvert has shown extends southeast from the vicinity of Glendive, Montana, to the western line of the Dakotas. Here the uplift tilted the beds and accelerated the erosion, while in the flat country to the westward in Converse County and adja- cent areas, the erosion of the Fox Hills was relatively uniform, and when the Lance formation was later laid down over this surface the unconformable relations are difficult of detection. But as Cross long ago stated: ‘‘The visible conformity between the Cera- tops beds and Fox Hills in Converse County cannot be accepted, contrary to other evidence, as proving the former to have been deposited in the epoch next succeeding the Fox Hills.”’ UPPER LIMIT OF.THE LANCE FORMATION In my original paper on the Lance formation (‘‘Ceratops beds’’) I stated that everywhere throughout the vast region studied it was found conformably overlain by the acknowledged ‘‘yellow”’ Fort Union, adding that ‘‘of the many workers who have observed the field relations at hundreds of points, not one, so far as known to the writer, has recorded the presence of unconformity between them.’ Field work during the past two seasons has confirmed this statement in every particular, and there is yet to be observed a single locality at which unconformable relations have been even suspected. Hence it seems to have been demonstrated that sedimentation from one to the other was continuous and unin- terrupted. : At the time the original paper was published it was thought that the Lance formation and the acknowledged Fort Union (the lower and upper members of the Fort Union as they were there called) might usually be separated on lithologic grounds, the lower being generally dark and somber-colored and the upper usually yellow. Subsequent investigation, however, has failed to confirm this, for while in individual sections, or even within limited areas, a provisional lithologic separation may often be made, when regional studies were undertaken it was found that the lithologic difference was so variable within short distances as to be wholly t U.S. Geol. Survey, Mon. 27 (18096), 236. STRATIGRAPHIC POSITION OF LANCE FORMATION 375 unreliable. For instance when a coal-bed that occurred near the top of the so-called somber-colored Lance formation was traced accurately for only a few miles it was found that the position of the dark-colored and the yellow beds varied as much as 300 feet, that is, at one point, the coal-bed might be 150 feet down in the somber-colored portion, and at another, an equal distance up in the yellow beds. It may therefore be confidently stated that the Lance formation and acknowledged Fort Union cannot be sepa- rated formationally on either structural or lithologic grounds, though in general the lower beds are on the whole prevailingly somber in color, while the upper beds are prevailingly yellow. MAGNITUDE OF UNCONFORMITY AND BOUNDARY BETWEEN CRETACEOUS AND TERTIARY Having demonstrated the existence of unconformable relations between the Lance formation and the underlying formations, the question naturally arises as to the magnitude of this discordance. By some it is claimed that it is merely local and is not more impor- tant than other breaks said to occur at various intervals in the Lance formation, and the doubt is expressed whether, even if the unconformity is present, any great amount of erosion is indicated. The wide area over which its existence has now been demon- strated certainly removes it from the category of ‘local’? happen- ings, and the uniformity of its occurrence beneath the Lance forma- tion is sufficient indication of its Importance over any that have been thus far even apparently indicated within the formation. Now as to its magnitude. It has been shown that in Carbon County, Wyoming, the Lance formation is not only above the full thickness of the ‘‘Laramie”’ (6,000 feet) but is separated from it by an unconformity that Veatch states is fully 20,000 feet, and moreover this unconformity is in the same position as regards the Laramie as that in the Denver Basin of Colorado, which, accord- "ing to Cross, has involved the removal of from 12,000 to 15,000 feet of strata between the Laramie and overlying formations. It is possible that the figures given by Cross and Veatch may be too high, but even so, the unconformity is undoubtedly one of importance, and this would seem to dispose of the contention that 376 F, H. KNOWLTON the Lance, Arapahoe and Denver formations may be mere “phases of the Laramie.’’ Whether the Laramie and various post-Laramie beds were deposited and later removed throughout the Dakotas, Montana, and Wyoming, is not at present known, but certain it is that the unconformity at the base of the Lance formation represents the time interval during which in other areas they were laid down and subsequently removed in whole or in part. There- fore, in the opinion of the writer, this unconformity is an impor- tant one and must be so recognized in American geology. Since it has been demonstrated that the Lance formation is so inseparably associated with the Fort Union—that is, without a trace of an unconformity—and is separated from the underlying formations by an unconformity of such extent, this point becomes more clearly than ever the logical point at which to draw the line between Cretaceous and Tertiary. In establishing this line the stratigraphic, lithologic, and paleobotanical criteria are believed to be more competent than any other evidence thus far brought forward. LARGE GLACIAL BOWLDERS GEORGE D. HUBBARD Oberlin College Mention of large glacial bowlders is not uncommon. In fact most localities glaciated have their ‘“‘largest in the state.’”’ Some lie so as to reveal easily the fact that they have been transported. Others are more or less concealed, and some care is needed to determine whether the rock is really a transported mass or country rock in place. A mass of limestone in Ohio covering about three-quarters of an acre, and in places sixteen feet or more in thickness, was men- tioned by Orton in one of the older reports of Ohio geology and cited by Dana.’ In the Alps was found a mass containing about 200,000 cubic feet of rock or enough to cover a fourth of an acre twenty feet deep.?, Sardeson’ reports a block of limestone moved a short distance whose width was over roo feet, thickness several feet, and length unknown. Limestone bowlders, often large masses, are quite common in parts of Illinois, specifically in western Livingston County, in northern McLean, and in parts of Cham- paign, Ford, and Vermilion counties. Following is a detailed description of several masses or “‘pockets”’ of this rock which have been studied. On the south side of the Champaign-Ford county line one and one-half miles east of the northwest corner of Ludlow township are the remains of a large “pocket.” H. H. Atwood of Paxton who owns the farm says several loads of the rock have been drawn away for building purposes, but enough remains to mark the place distinctly. Near Saybrook, McLean County, are a number of localities where limestone is found at the surface. On the farm of Mr. Riggs, tJ. D. Dana, Manual of Geology, 5th ed. (1895), 960. 2 Thid., 248. 3 Jour. Geol. (1905), XIII, 351-57. 377 378 GEORGE D. HUBBARD one mile north and one and one-half miles west of Saybrook, lime was burned forty or fifty years ago. A small kiln was built and operated several years with rock from this deposit. A half-mile east of this kiln, past the schoolhouse, another ‘‘pocket”’ was opened and several loads drawn some thirty-five years ago. At present but few know of these limestone pits, for they have been entirely dug out and the holes are plowed over. Portions of the kilns and fragments of waste alone remain. Two miles north and one mile west from Saybrook are a number of slabs resembling flagging. These are quite numerous on one farm. On a farm ten miles west of Saybrook lime was burned for the local market, but at present the rock is apparently exhausted. In this locality, a good many loads for foundations and well curbs have also been drawn away. According to a boring for Mr. H. Cheney of Saybrook, bed rock was struck here at a depth of 236 feet. It is recorded that a five-foot limestone bowlder was struck in a gravel bed at a depth of 150 feet. A well digger here in con- versation said that in digging wells he frequently encountered limestone bowlders of various sizes, and noted several localities where the bowlder weighed from ten to twenty tons. A number of wells in the vicinity have been walled with the rock taken out in digging, supplemented with more found near by. “In fact,” he says, ‘‘there is lots of limestone scattered all over the country.” No bed rock, however, has ever been found about Saybrook except at considerable depths as in the well cited, 236 feet.1 With such thickness of drift as this, these masses of limestone cannot be in place. The largest drift mass of limestone is in Livingston County, about a mile and a half southwest of Fairbury, where Dr. Brewer has been taking out a great deal of limestone. The mass is along a small stream where the water divides, flowing around a little island. On the north bank of the south division and on both banks of the north division, rock is found; but on the extreme south bank no rock is known, nor is rock struck in any wells on the south side of the stream. Along the stream on the north side for t Frank Leverett, U.S.G.S. Mon. 38, 695, reports a boring for coal here reaching rock at 247 feet. LARGE GLACIAL BOWLDERS 370 a half-mile or more, and back from the stream a half-mile, all wells strike rock at some twelve to sixteen feet. Below the rock at the quarry is Clay, a soft sticky yellow body, called by the quarrymen ‘‘soap- stone.” Examination showed it to be glacial drift. No large pieces of rock can be obtained in the quarry, for the whole mass is shattered. The pieces vary in size from ten or fifteen to two hundred and fifty pounds, rarely larger than can be handled by one man. At the quarry the rock is from ten to fifteen feet thick, and two or three nearby wells are reported passing through it, one finding sixteen feet of rock. The rock seems to be almost exactly horizontal in the quarry, and it is struck at quite uniform depths in the neighboring wells. Inquiry for this stratum in the coal shafts, two in number, at Fairbury, failed to reveal its presence. One about a mile distant encountered a piece of rock at a depth of forty feet, but below it was more clay. The other about one and one-quarter miles distant found no rock for at least ninety feet. At McDowell a little quarry is operated in rock which has almost precisely the same characters as the one at Fairbury, but it is of less extent—ten or twelve feet thick, shattered and local. West and south of McDowell about two miles from Ocoya there are two or three little quarries opened. One near a little stream is operated by two men who have taken out over a hundred cords of rock in a single summer. The rock is eighteen feet thick at a maximum, but in places only five or six feet thick. Some parts of it are shelly or shattered, but toward the bottom, this mass is firmer than any other yet considered. Sometimes pieces twelve to sixteen inches thick and six to eight feet long are removed, but no blasting is done. The near proximity to the stream caused some trouble with water seepage, so a sumpf was dug through the rock and a pump put in. A crowbar was thrust down easily © in the bottom of this sumpf two or three feet. The quarrymen say the substratum is ‘“‘soapstone of variable character,” but it seems to be a well-packed, blue, pebbly clay with a greasy feel. That it is not one of the soft argillaceous layers of the Coal Measure rocks is shown by its pebbles. The edge of the rock is known in two directions. The edge along the stream is slanting, the other, 380 GEORGE D. HUBBARD nearly at right angles thereto and on the east end of the quarry, is perpendicular and very regular. Rock is struck in but one well in the vicinity. Rock has been taken out from similar, though smaller local pockets, in two other localities within 80 rods. The county surveyor of Livingston County says there are a good many local deposits along the Vermilion River, slabs, bowlders, and irregular pieces, but it is not continuous, and the layers are variously tilted. Usually these large masses are along morainal ridges. Some- times they are found along stream beds where they have been exposed by erosion. They cover areas varying from a few rods to over a hundred acres in extent, and differ in thickness from six or eight feet to eighteen or twenty feet. They are always in a shattered condition; often very much broken up, but sometines requiring some blasting to get out the rock. What seems the most surprising thing is that there is rarely much dip. The bedding in all the larger masses is almost horizontal. During early days when transportation was expensive, these masses of limestone were much used by the settlers, who made lime from some of them, and from others drew building material. The rock was more workable, and hence more desirable, than the granite bowlders. Their presence in the drift, and their distribution mostly in the large recessional moraines, seems to point to a glacial origin for them. Since most if not all the masses mentioned are of Carboniferous rock, as shown by their fossils, the sources could not have been more than fifty to seventy-five miles north, for beyond that limit there is no Carboniferous rock, from which they could have come. While no specific places have been found from which it is thought these large bowlders were plucked, it is believed that they may have come readily from the bluffs of a valley, or from hills a moderate distance to the north. REVIEWS Tron Ores, Fuels and Fluxes of the Birmingham District, Alabama. By Ernest F. BURCHARD AND CHARLES Butts. With Chap- ters on the “Origin of the Ores,” by Epwin C.Ecket. Bull. U.S. Geol. Surv. No. 400. Pp. 204. The Birmingham District, as here considered, extends as a long, narrow belt, about seventy-five miles in length by ten in width. The iron ores of the district lie in the broad, anticlinal Birmingham Valley which is structurally a part of the Appalachian Valley. An outline of the geology of the district shows rocks belonging to all the periods from the Cambrian to the Pennsylvanian, with unconformities separating all the systems except the Cambrian and the Ordovician, where the transition is within the Knox Dolomite, which here attains a thickness of 3,300 feet. An unconformity is found within the Ordovician. Within the area are extensive deposits of red hematite and brown ore, and important beds of coking coal and fluxing limestones. The red hematite or Clinton ore is found in the Clinton or Rock- wood formation which, in Alabama, consists of lenticular beds of sand- stone and shale with four well-marked ore horizons. The ores occur in lenticular beds analogous to strata, interbedded with limestone, sandstone, and shale. Three opposing theories have been advanced _to account for the origin of the Clinton ores: (1) original deposition; (2) residual enrichment by weathering; (3) replacement by percolating waters. Mr. Eckels shows that both the second and third theories are untenable, and that observations support the theory of primary sedi- mentary deposition. The brown ores or ores of the hydrous iron oxides belong to a type common in southeastern United States, occurring as irregular masses associated with residual clays, and underlain by limestones of Cambrian and Cambro-Ordovician age. Mr. Eckel points out very forcibly that the decay of a limestone carrying disseminated iron material would not of itself yield such a deposit of ore, but that some factor must be found whereby the iron is concentrated. In his opinion, this concen- tration usually took place in the limestone itself. The coke used in the blast furnaces of the district is made from coal mined in the Warrior coal field which lies to the northwest of the valley. Re: 381 82 REVIEWS Oo Annual Report of the Geological Survey of Western Australia for the Year tg09. By A. GipB MAITLAND, Government Geolo- gist. Pp. 32, maps 4, andtigs.. 3: The report contains a summary of the work done and the results obtained by each of the fifteen officers employed by the survey. Three bulletins were issued by the survey during the year tgo9: Bull. 33, ‘Geological Investigation in the Country Lying between 21 deg. 30 min. and 25 deg. 30 min. S. Lat. and 113 deg. 30 min. and 118 deg. 30 min. E. Long., Embracing Parts of the Gascoyne, Ashburton, and West Pilbara Goldfields’; Bull. 35, “‘ Geological Report upon the Gold and Copper Deposits of the Philips River Goldfield”; Bull. 37, “The Geological Features of the Country Lying along the Route of the Pro- posed Transcontinental Railway in Western Australia.”’ E.R. L. “The Dakota-Permian Contact in Kansas.” By F. C. GREENE. Kansas University Science Bulletin, Vol. V, No. 1 (October, 1909), pp. 1-8. The paper presents a summary of the relations of the Permian and the Cretaceous in Kansas, north of the Smoky Hill River. eR les Annual Report on the Mineral Production of Virginia during the Calendar Year 1908. Virginia Geological Survey Bull. No. I-A. By THomas LEONARD WATSON. Pp. 138. Virginia possesses an abundance and variety of mineral materials, about forty varieties of which are now exploited, many of them on a large scale. A table of the mineral production in 1908 shows a total value of nearly $18,000,000, of which iron makes up over $6,000,0c0. Under the heading Preliminary Generalities, the author presents a brief and interesting review of the physiography and general geology of the state, including several generalized sections from various parts of the state. The parts devoted to the various mineral deposits are chiefly descriptive and statistical. A valuable feature of the report is a series of maps showing the distribution in the state of a number of the most important of the mineral deposits. Ee Ree: REVIEWS 383 Annual Report of the State Geologist, Geological Survey of New Jersey, t909. By Henry B. Ktmmet, State Geologist. EVO), ey. Besides the administrative report this volume contains the following papers: ‘Report upon the Development of the Passaic Watershed by Small Storage Reservoirs,” by C. C. Vermeule; ‘Records of Wells in New Jersey, 1905-9,” by Henry B. Kiimmel and Howard M. Poland; “Notes on the Mineral Industry,” by Henry B. Kiimmel. E.R. L. ‘““A Proposed Classification of Petroleum and Natural Gas Fields Based on Structure.”” By FREDERICK G. CLAPP. Economic Geology, Vol. V, No. 6 (September, 1910), pp. 503-21. The classification proposed by the author of this paper is based on the “anticlinal” or ‘structural’? theory, which is called into use to explain the segregation of oil, water, and gas from a primary disseminated condition. Depending on the structures which have segregated and localized the pools, seven classes of oil and gas accumulations have been distinguished by the author: I, Where anticlinal and synclinal structure exists; II, Domes or quaquaversal structures; III, Sealed faults; IV, Oil and gas sealed in by asphaltic deposits; V, Contact of sedimentary and crystalline rocks; VI, Joint cracks; VII, Surrounding volcanic vents. Class I embraces most of the known oil fields and is subdivided into five subclasses to distinguish the various relations of the pools with anticlines and synclines. BeR. EL: ‘Outline Introduction to the Mineral Resources of Tennessee.”’ Extract (A) from Bulletin No. 2, Preliminary Papers on the Mineral Resources of Tennessee, State Geological Survey. By GEORGE H. ASHLEY, State Geologist. Pp. 65. This pamphlet contains a brief survey of the surface features of the state, the geological formations, and the geological history; and a list of the minerals of the state with a brief notice of their occurrence, use, etc. Bulletin No. 2, of which this is the first part to be published, is the first scientific publication of the newly established state survey, and is not intended as an original contribution but as a brief statement of facts already published, and is designed to meet the demand for immediate information on the mineral resources of the state. HR: 384 REVIEWS Summary Report of the Geological Survey Branch of the Department of Mines, Canada, for the Calendar Year 1909. By R. W. Brock, Directorsa-Pps 3078 Besides the administrative report of the director of the survey, there is included in this volume a short summary report by each of the geologists and officers of the survey, of the work carried out during the year. Almost all of the work at present being undertaken is along economic lines. i Regs “The Tectonic Lines of the Northern Part of the North American Cordillera.’ By W. Jorrc. Bull. Am. Geog. Soc., XLII (1910), 161-79. With map. This paper pictures the tectonic lines of the North American Cor- dillera from the 4oth parallel to Bering Sea. Though the author has based his work in part upon the reports of the geological surveys of the United States and Canada, he has confessedly followed Suess, in the main, both in subject-matter and in mode of treatment. The chief purpose of this paper is to consider in their larger relations the individual ranges and groups of ranges which go to make up this complex system. The interrelations of these mountain chains are discussed in a condensed synoptical form. The axes of the many separate, individual ranges are located on the map by heavy black-tectonic lines which show graphi- cally the distribution and direction of deformative movements. A prominent place is given to the mountain systems of Alaska. In conclusion the author suggests the subdivision of the North American Cordillera from Bering Sea to the Isthmus of Tehuantepec into three major divisions: (1) Northern Cordillera, or Alaskides; (2) Central Cordillera; (3) Southern Cordillera, or Lower California and the Mexican Highland. The boundary between the first and second divisions would be the zone of coalescence near the Alaskan boundary, that between the second and third the depression along Salton Sink, the Gila, and the Rio Grande. The decided Asiatic structure of the Alaskides is the reason given for recognizing them as an independent major subdivision of the Cordillera. Rowe e: This floor was never treated except with water sprinkled before sweeping. The boards are splintered and joints and nails sprung; it is old before its time. - ~The air above this floor was filled with dust and everything in the room was dust-covered. Do YOUR customers ; || Sold by dealers every- think of your store as a ~ where. dusty one? , | Not intended for house- Are YOUR floors dis- || hold use. integrating ? ; : Send for free booklet, “‘ Dust, and Its Dangers.”’ This floor,in a popular and successful store, was regularly treated with Standard Floor Dressing for years. It has worn down smoothly and evenly. All the dust that settles is held down, and the air iscleanand clear. Goods and fixtures remain bright and unsoiled by dust when a floor is Treated with Standard Floor Dressing EOFOROFOIOI OF OF OR OS OFOIOFOFOFOFOFOZOFOFOLOFOFOFOFOFOIOFOLOIG} The Delicious Flavor of Baker’s Breakfast Do not keep both food and germs in the refrigerator. To prevent musty smells Appeals to the strong and vigorous as well as to those who seek health and strength. and keep air of refrigerator pure and sweet, place a bowl containing sponge sprinkled with Platt’s Chlorides where food is kept. Wash sponge occasionally with Platts Chlorides, Walter Baker & Co. Ltd. The Odorless Disinfectant. Established 1780. Dorchester, Mass. A colorless tuquid, sate and economical. It does not SOT OTOIOLOTOIOLOIOLOLOLCLOICIOTOICHOICHOTOFOFOTOLOFOIO OF cover one odor with another, but removes the cause. IT IS A PERFECT FOOD DRINK 53 Highest Awards Poueeran U.S. Pat. od FOLOEOLOLOZOFOFOFOFOFOFOFOFOFOLOF OFOFOFOLOLO FOF OF OF OtOFOtOsOsOzOtotofzOtofzOfzOtotrofofopozojoso} }OLOPORFOFOTOROFOFOSOFOFOFOLOF OFOFOFOt OF OFopotototofzotosotopototototopotototorepotototototo} | ATHLETES | | TO KEEP IN GOOD TRIM MUST LOOK WELL — TO THE CONDITION OF THE SKIN. TO THIS END THE BATH SHOULD | BE HAD WITH have been established over 60 YEARS. By ¢ p system of payments every family in moderate ¢ cumstances can own a VOS€ piano. We take @ instruments in exchange and deliver the new pi in your home free of expense, Write for Catalogue D and explanations, vose & SONS PIANO CO., Boston, Mass. VOS VOLUME XIX é : ‘NUMBER 5 THE JOURNAL or GEOLOGY A SEMI- QUARTERLY EDITED BY THOMAS C. CHAMBERLIN AND ROLLIN D. SALISBURY With the Active Collaboration of SAMUEL W. WILLISTON ALBERT JOHANNSEN WILLIAM H. EMMONS Vertebrate Paleontology Petrology i Economic Geology STUART WELLER WALLACE W. ATWOOD ROLLIN T. CHAMBERLIN Invertebrate Paleontology Physiography Dynamic Geology ASSOCIATE EDITORS SIR ARCHIBALD GEIKIE, Great Britain Z GROVE K. GILBERT, National Survey, Washington, D.C. HEINRICH ROSENBUSCH, Germany CHARLES D. WALCOTT, Smithsonian Institution THEODOR N. TSCHERNYSCHEW, Russia HENRY S. WILLIAMS, Cornell University CHARLES BARROIS, France JOSEPH P.IDDINGS, Washington, D.C. ALBRECHT PENCK, Germany JOHN C. BRANNER, Stanford University HANS REUSCH, Norway RICHARD A. F. PENROSE, Jr., Philadelphia, Pa. GERARD DEGEER, Sweden WILLIAM B. CLARK, Johns Hopkins University ORVILLE A. DERBY, Brazil : WILLIAM H. HOBBS, University of Michigan T. W. EDGEWORTH DAVID, Australia FRANK D. ADAMS, McGill University BAILEY WILLIS, Argentine Republic CHARLES K. LEITH, University of Wisconsin s JULY-AUGUST, 1911 CONTENTS SAMUEL CALVIN - hea eae Pea a oak ee eee lt POSTER BAIN aes THE EVOLUTION OF LIMESTONE AND DOLOMITE. II - - Epwarp STeEmtTmMann 392 DIFFERENTIATION OF KEWEENAWAN DIABASES IN THE VICINITY OF LAKE NIPIGON - - - - - - - - = Bee a ease as ic shee ete Oe LOORIAAZO GENERA OF MISSISSIPPIAN LOOP-BEARING BRACHIOPODA = - - Stuart WELLER 439 PHYSIOGRAPHIC STUDIES IN THE SAN JUAN DISTRICT OF COLORADO WALLACE W. ATWOOD 449 THE VARIATIONS OF GLACIERS. KVI - - - - - -.- - Harry Freipinc Rem 454 BE GEOnOGIOAL ABSERACKS AND BEVIRWS 9-02 20 POO as en ee TRIS DENNY Sth go he teil et SY ls cates a ce ec aeRO COL Na pees aria Ryall Onin ara Che Aniversity of Chicago press CHICAGO, ILLINOIS AGENTS: THE CAMBRIDGE UNIVERSITY PRESS, Lonpon anp EDINBURGH { WILLIAM WESLEY & SON, Lonpon TH. STAUFFER, Leipzic THE MARUZEN-KABUSHIKI-KAISHA, Tokyo, Osaxa, Kyoto OUTLINES OF GEOLOGIC HISTORY WITH ESPECIAL REFERENCE 10 NORTH AMERICA A Series of Essays Involving a Discussion of Geologic Correla- tion, Originally Presented before Section E of the American Association for the Advancement of Science J* J c# EOLOGISTS and all readers of geologic literature will welcome & the publication, in book form, of an important series of essays and discussions on the subject of geologic correlation under the title, OUTLINES OF GEOLOGIC HIsTORY wiTH ESPECIAL REFERENCE TO NortH AmeERiIcA. The symposium was organized by Bailey Willis, and the papers were originally presented before Section E of the American Association for the Advancement of Science at Baltimore in December, 1908. They were first published by the Yournal of Geology and are now brought out in book form under the editorship of Rollin D. Salisbury. — The series as a whole represents the successful execution of the plan on which. all the monographs were based—namely, to formulate the principles of correlation as applied to the formations of the various geologic periods. The evolution of floras and faunas has been traced with especial attention to environment and correlation. As originally presented, the papers excited much interest and discussion. They embody the present state of knowledge and opinion concerning many of the fundamental problems of North American geology, and form an admirable supplement to earlier treatises and manuals. The value of the book is greatly enhanced by the fifteen paleo- geographic maps by Bailey Willis which accompany the papers. 316 pages, 8vo, cloth; $1.66 postpaid ADDRESS DEPARTMENT P THE UNIVERSITY OF CHICAGO PRESS CHICAGO, ILLINOIS THE JOURNAL OF GEOLOGY Ve LNe AUGUST, LOU, SAMUEL CALVIN H. FOSTER BAIN Samuel Calvin, who died at Iowa City, Iowa, April 17, 1911, was born in Wigtonshire, Scotland, February 2, 1840. He passed the first eleven years of his life in the little town of Wigton, living the simple, hardy life common in Scotch families of moderate means. As a boy, after school hours, he often paused at the edge of the cliff to look down upon Wigton Bay in which lay ships and schoon- ers from all parts of the world. At that period trade between America and the various little ports on Solway Firth was active, and the longing for travel, common to all boys, was greatly stimu- lated in young Calvin by the sight of the vessels that came and went. Among his school companions was James Wilson, now Secretary of Agriculture, and a friendship was there formed that continued through life. The Calvin and Wilson families both emigrated to the United States in 1851, the Calvins going direct to Iowa and the Wilsons following after a short sojourn in Connecticut. Thomas Calvin, the father of Samuel Calvin, took up land south of Man- chester, Iowa. At that time neighbors were few and far between in eastern Iowa. The first residents came from Indiana and, being accustomed to a wooded country, feared to venture upon the open prairie. They accordingly settled near the streams on what proved to be the poorer land. The Calvins and their asso- ciates, evidently thinking that having ventured so much in coming from Scotland they might as well venture more and save themselves Vol. XIX, No. 5 385 3286 H. FOSTER BAIN the work of clearing off timber, chose prairie land. It thus hap- pened that from early boyhood Samuel Calvin was familiar with the great granite bowlders that mark the fertile prairies covered by the Iowan drift. Between farm work, school, and the usual country sports, his time was fully occupied until about 1861, when he entered Lenox College, at Hopkinton, near by. Lenox College, was, and is, an excellent example of the small denominational colleges that the pioneers of the Middle West founded so prolifi- cally and supported with so much sacrifice. Without the equip- ment of a present-day university, or a staff of world-famous pro- fessors, it was still an excellent place for a young man desirous of getting at the fundamentals of the simple college curriculum of a half-century ago. Here Calvin remained and studied until near the close of the great Civil War, when, in company with most of the instructors and students who had not already gone to the front, he enlisted in 1864 in one of the Iowa regiments. Fortunately the war was nearly over. His military service was therefore neither long nor was it distinguished, in the sense of taking him into great battles. For the most part it was a period of dull routine, of guard duty and of marching, of occasional small skirmishes with the enemy, and a continual private skirmish for acceptable food and some comfort. He learned the rudiments of a soldier’s life and the routine of camping—the latter much the more valuable to him. At the close of the war, Calvin, with many others no longer young, went back to college to finish his studies. The college, however, had been practically wrecked. The call for men had taken both instructors and students, and while the buildings were still there, the life of the institution had been nearly broken up. After the ensuing reorganization Calvin found himself in the ranks of the instructors rather than among the students, and the inter- rupted college course was only completed as a result of much hard private study. At Lenox College, among other members of the new faculty, was Thomas H. Macbride, a graduate of Monmouth College, and a man who had had the advantage, then unusual for teachers in small western colleges, of study at the University of Bonn. Mac- bride’s interests centered in botany, and Calvin, while broadly SAMUEL CALVIN 387 concerned with the whole of natural history, was already beginning _ to specialize in paleontology and geology. The two men became intimate friends and close companions. ‘Together they explored the neighborhood and later, in the long vacations, the more distant parts of Iowa. With team, covered wagon, and simple camp outfit, trips were made as far west as the Missouri River, and collections of various sorts were brought back to enrich museum and class- room instruction. In the course of this work Calvin came into contact with C. A. White, at first state geologist and later professor in the State University at Iowa City. When White went East Calvin succeeded him at the university, and when the Iowa Geo- logical Survey was re-established in 1892, he also followed him in the position of state geologist. The change to Iowa City, which took place in 1874, was an agreeable one, since the university, having larger resources, offered a larger opportunity for work; and for work Calvin always was greedy. The State University of Iowa in 1874 was not a large institution and the professors found plenty to do. Calvin occupied what has been aptly described as the “settee rather than chair” of natural history, and, as he once whimsically phrased it, he was ever after ‘“‘shedding professor- ships.” As rapidly as funds would permit he divided the work and called other men to him. Among the first was Macbride, then Nutting, and others in succession until in the closing years of his work it was not only possible for him to confine his work to geology, but to have the aid of an able corps of assistants in that. His work in the other sciences, however, was more than time- serving. His interest in animal morphology was especially keen, and in organizing and conducting zodlogical explorations he did work of real value. The results of the Bahama expedition, which brought back such unheard-of wealth of specimens of living crinoids, were due in no small part to him. As a geologist Calvin’s name and fame are principally bound up with that of the lowa Geological Survey, which, except for a brief interim, he directed from the day of its organization, in 1892, to the day of his death. As a former member of the staff of this Survey I may be over-partial, but it is none the less my conviction that, considering time, place, and means, the Iowa Survey is and has 388 H. FOSTER BAIN been one of the best-conducted institutions of its kind in America. It serves the people and the state of Iowa well, meeting their peculiar needs. In other states different methods are necessary, as they will become in time in Iowa. The Iowa Survey owes much to Calvin. It also owes a not-to- be-forgotten debt to Charles R. Keyes, the principal assistant at the time the Survey was organized and through the period when plans were being formulated. Im brief the Survey has served two main purposes: (1) It has furnished handbooks, consisting of maps and reports on the geology of the individual counties, written in simple language adapted to the understanding of students and intelligent laymen; (2) scientific and technical reports have been issued as occasion offered, covering all phases of the natural history of the state and the economic development of its resources. According to the report for the year ending December 31, 1909, all but nine counties had been surveyed and reports prepared and issued. More detailed surveys, based on topographic maps made in co-operation with the United States Geological Survey, had already been begun in areas of especial economic importance, and, as conditions permit, this system of more refined mapping will doubtless be extended over the state. In the meantime the county reports serve extremely useful purposes and afford a sure basis for broad general studies. The special reports on scientific and economic subjects have been numerous and valuable, as is attested by the list of publications of the Survey. The water, the coal, clay, stone, lead, zinc, gypsum, and minor mineral resources have all been studied by specialists, and information essential to their economical development set forth. The mineral industry has responded to the stimulus and the total value of the mineral output of the state is now many times what it was when the Survey was organized. Towa is primarily an agricultural state and its mineral resources are relatively small; the educational phases of the Survey work have therefore always, and properly, been emphasized. On the narrow material basis, however, the Survey has made good return for the support of the state and it is deservedly popular. As an administrator, therefore, Calvin’s work shows him to have been successful. ee) SAMUEL CALVIN 389 In scientific research, Calvin’s personal interest lay mainly in paleontology and, in later years, in the study of the Pleistocene deposits. Aside from contributions to the paleontology of the Paleozoic invertebrates, Calvin will probably be best remembered for the discovery of the fauna of fish remains in the Devonian beds near Iowa City, and for his studies of the vertebrate remains of the Aftonian deposits. The fish remains have been investigated by C. R. Eastman, who found in them much of interest. The Afto- nian bones are important because they permit fixing the age of a widely scattered series of puzzling deposits. In his last published administrative report! Calvin speaks of them as follows: This remarkable interglacial fauna is not altogether new to science. The individual species, and to some extent the fauna as a whole, have for some time been known to students of paleontology. It has also been known that some of the species were at one er more times inhabitants of Iowa. But, so far as concerns Iowa, it was not known that the few discovered forms, which heretofore have been represented by isolated finds, were contemporaneous; and outside of Iowa, in territory ranging from Texas to western Nebraska, the exact age of the beds in which the remains of this assemblage of extinct mammals were found was not definitely fixed. The fauna as a whole is markedly different from that familiar to the pioneer settlers of this State, very different from that known to the pioneers in any part of America. True horses were represented by at least two species, both quite distinct from our domestic species; there were three species of elephant, one of imperial size, and there were two mastodons, making in all five great proboscidians; there was at least one species of camel, an extinct bison, a gigantic stag, and two ponderous, awkward, clumsy ground sloths. The smallest of the three ele- phants seems to be identical with the hairy elephant or northern mammoth of Europe and Asia; it furnishes to this unique fauna a distinctly boreal element. The two great sloths, on the other hand, contribute an element distinctly South American. As found in Iowa, the age of the fauna is definite and clear. The beds in which the remains occur belong to the Aftonian stage; these animals lived, and the beds in which their remains were buried were laid down, in an interval of comparatively mild climate between the first and second stages of Pleistocene glaciation. The quotation illustrates at once Calvin’s broad scientific interest and the singular clearness of his writings, which makes it a pleasure to readers and a fit model for beginners. Calvin’s active interest in the Pleistocene deposits dates from t Towa Geol. Surv., XX, xxii. 390 H. FOSTER BAIN 1896. In the spring of that year he undertook to put into shape long-accumulated notes on Johnson, the county in which Iowa City is situated. It chances that two lobes of Iowan drift reach across the northern border of this county. South of these lobes are the loess hills characteristic of much of the Iowan border and, in turn, the broad loess-Kansan plains. At that time the southern boundary of the Iowan had not been traced. From the known presence of two drifts at Afton and elsewhere in southern Iowa, it had been inferred that the Iowan boundary was much farther south than now appears true. The battle for two ice-sheets has been but too recently won to encourage belief in many. Calvin, however, intimately familiar with the typical Iowan drift plane of: Buchanan and Delaware counties, recognized at once that the phe- nomena of southern Johnson County required a new interpretation, and shortly thereafter hit upon the clue. The rest of the staff, inspired by his enthusiasm, started out like crusaders to over- turn and rebuild the Pleistocene column. The field was particu- larly favorable since in Iowa the different drift sheets are mainly deployed rather than superimposed, and since, also, nearly the whole sequence is represented. With the kindly counsel of T. C. Chamberlin, with friendly visits from R. D. Salisbury, J. E. Todd, Albrecht Penck, Frank Leverett, and others, the work went rapidly. There were many field conferences, and the winter meet- ings of the Iowa Academy of Science became notable for the dis- cussion of current Pleistocene problems. Naturally there were differences of opinion and later work has shown the need of some revision of first-stated conclusions. Out of it all, however, has come the recognition of the independence of the pre-Kansan, the thorough establishment of the Aftonian, and the concept of the Iowan ice sheet. As to the latter, especially, there has been, and still is, much difference of opinion. The Iowan drift is so peculiar, it is so local, and the phenomena are so puzzling, that some find them- selves unable to accept the evidence of its existence. It is not my purpose to review the proofs. That has already been done most excellently by Calvin himself. It is sufficient to repeat here a remark made by him at the close of the first field season devoted to this study: ‘‘The Iowan ice sheet did so many queer things that SAMUEL CALVIN 301 we will never blame anyone not familiar with it in the field for denying its existence; but, whatever the explanation, there is a con- sistent and co-ordinate set of phenomena that demands explanation, and that is best interpreted by the hypothesis of a separate Iowan icersheet. Aside from his work at the university and on the Iowa Survey Calvin did his full share in the general work of his chosen profes- sion. While he was not especially interested in economic geology, his advice was sought and valued by a number of Western mining companies; he was a member of the first Conservation Conference at the White House; he served on the National Advisory Board on Fuels and Structural Materials; he was an active member and officer of numerous professional societies, including the Geological Society of America, of which he was president in 1908; he was one of the founders of the American Geologist, and of the Associa- tion of State Geologists; and in many other positions he made his influence felt. He was a charming writer, a popular lecturer, and a most inspiring teacher. His personal influence was strong and deep, and the thousands of students who came into contact with him are today better men and women because he lived. He left a stainless record as a man and citizen, and an inspiring example to young men of the profession. THE EVOLUTION OF LIMESTONE AND DOLOMITE. II (Concluded) EDWARD STEIDTMANN University of Wisconsin PART If. CALCIUM AND MAGNESIUM IN THE PRODUCTS OF METAMORPHISM The products of metamorphism of the rocks may be classed as solids and solutes, or residuals and losses. Residuals, the mate- rials which remain in situ after a rock has suffered chemical change; losses, the materials which are dissolved, transported in solution, and redeposited elsewhere. A study of the fate of calcium and magnesium in rocks subjected to metamorphism shows that there is nearly always a marked tendency for a greater percentage loss of calcium than of magnesium. Magnesium tends to remain with the residuals to a greater degree than calcium. It will be shown that in the movement and redisposition of the residuals and losses of rock alteration, and in the reworking of these products by the same processes, again and again throughout geologic time, lies the history of a progressive enrichment of the lands in calcium , and their progressive depletion in magnesium. The evidence for the selective splitting off of calcium from the parent rocks in response to metamorphic processes and the accumulation of magnesium in the residuals follows. Materials lost by the weathering of acid igneous rocks.—The weathering" of acid igneous rocks results in the loss of lime, mag- nesia, soda, potassa, and silica. The percentage loss of the various constituents approximately follows the descending order in which they are named. For purposes of comparison alumina may be regarded as constant. The ratio of calcium to magnesium lost in the weathering of an acid igneous rock can only be given in terms of tendencies. In Table V the figures for the ratio of cal- t Edward Steidtmann, ‘‘A Graphic Comparison of the Alteration of Rocks by Weathering with Their Alteration by Hot Solutions,”’ Economic Geology, III, 381-409. 392 EVOLUTION OF LIMESTONE AND DOLOMITE 393 cium to magnesium lost are based on the assumption that alumina has remained constant. The percentage loss of calcium averages higher than that of magnesium, a tendency generally character- istic of metamorphic processes. TABLE V WEATHERING OF AciID IGNEOUS ROCKS Ca/Mg Ca/Mg PERCENTAGE Loss Rock FRESH IY ND) BONG S| | eral | SOURCE Rock Lost Ca Mg meGramites: 6. sa. 2.8 Bisse 5 at 31 Watson, Granites of Geor- gia, 318 Pw Granitessesy ee. 4.9 7.6 98 64 Tbid., 315 BeGranites.. we Bas ae 88 Al Ibid., 321 Am GLaAMiterne er 6.9 5 59 82 Ibid., 320 Granites ila 3.4 4.6 43 32 Ibid., 309 GuGranites: 2.4: 5 238 2 50 Ibid., 312 7Granitesw on. 5 oO) 84 83 Ibid., 312 8 Granite ...... 3.8 3 74 06 Ibid., 327 OmGramitenss oa. 3.9 Ae 78 70 Ibid., 327 ToOMGranites. .. 1 Deal: 2.58 77 65 Ibid., 325 Tr Granitessiiis: a3 4.8 80 54 Average Georgia nam Granitessc....c. a3 no mag- | 2 fo) Merrill, Rocks and Rock nesium Weathering (Dist. of lost Columbia), 207 Tom GramMitere ace: 13 no mag- | 21 fe) Ibid. nesium lost 14 Phonolite..... BER 12 17 5 Ibid. (Bohemia), 108 15 Andesite...... B ol 9.5 77 28 Ibid. (Grenada), 208 TOL OVeNMtes cs. -0.: 16500) Te 67 78 Clarke, U.S.G.S. Bull. 330, 412 E77 GMEISSs s/o, 4.2 Se LOO 76 Ibid. IAVETAC CH ea 3.45 +4.12 59-7 50 Materials lost by the weathering of basic igneous rocks—In the weathering of basic igneous rocks, the percentage losses of cal- clum and magnesium average about even, in Table VI. The ratio of calclum to magnesium lost varies between wide margins, from one to infinity. The materials selected for analyses may have unequaled value in the problem. Materials lost by the weathering of an average igneous rock.—The ratio of calcium to magnesium in an average igneous rock (Clarke) is about 1.37. The ratio of calcium to magnesium lost by its weathering, calculated on the basis of average calcium and mag- nesium decrements, is about 1.85. 394 EDWARD STEIDTMANN TABLE VI WEATHERING OF Basic IGNEous Rocks Ca/Mg Ca/Mg PERCENTAGE Loss Rock FRESH MATERIALS SOURCE Rock Lost Ca Mg Te Diabasesn cic BaD 4.1 25 9.5 | Merrill, Rocks and Rock Weathering (Medford, Mass.), 218 2. Dialbasea meee ie) TS 83 61 Ibid. (Spanish Guiana), 222 BeiBasaltzeee sae 1.0 a iit 47 96 Ibid. (Crouzet, France), 223 A BAGEMWS s Ss oe oe 1.2 TA! 84 7h Ibid. (Kammer Bull.), 222 Si Wiabaseere sme 6 6 98.7 98.2 | Merrill, Am. Geol., XXII, 93 6) Diabases... 2) 04 6 98.6 | 98 Ibid., 95 iy ADSKOIAMESS oo uae 1.84 1.85 07-3 | 97.17| Merrill, Rocks and Rock Weathering, 225 8 Augite diorite. TO, Baws 59 76 Morozewicz, Zeit. Kryst. Min., CXXXIX, 612 9g Diabase..... : 1.08 BW) 84 42 P. Holland and Dickin- son, Proc. Liverpool Geol. Soc., VII, 108 ToMBOulder mak to | infinity W7/s || | CS) Helen Mine, Michipicoten (unpub. monograph) Tr Gabbronse es 2.92 Tey 8.9 6 Gabbro, Allen Junction, Minn. (unpub. mono- graph) 12 Diabase..... 2 AG Tats 92 82 Dike, Penokee-Gogebic (unpub. monograph) (AVCE AS Crete tere: Te 27 1.62 GOe3\ 2 Oms5, ie TABLE VII WEATHERING OF LIMESTONE Ca/M. Ca/M, Rock Frock Rack Meee Rack Source Carboniferous lime- SOME ewer eee 148 14.9 Ann. Geol. Rept. Arkansas (1890), 179 Average of three an- DLV SESH een ee ene I.4 2 Bull. 7, Va. Geol. Surv., 97 Kn OK eae es eee 1.6 Or Russell, Bull. 52, U.S.G.S., 25 Galena, vee axe aoe TS Tey Bull. 14, Wis. Geol. Surv., 15-16 Galena. seen eeery TS 66 Ibid. Galenaye ia eerie 5 64 Ibid. aArashrances eae U9 2.38 Hilterman, Die Verwitierungs- Produkte von Gesteinen der Trias Formation, Frankens. Inaugural Dissertation, Er- langen, 1889 Plattentkallkcete esi 66.0 1.82 F. W. Pfaff, “‘Ueber Dolomit und seine Erstellung,’’ Newes Jahrb., XXIII, 538 Krebscheeren Kalk.. . 61.0 THOT Ibid., 539 IAWeLa Seam aia 20.5 2.61 EVOLUTION OF LIMESTONE AND DOLOMITE 395 Materials lost by the weathering of limestones.—In the present stage of earth evolution, the principal contribution of calcium and magnesium from the weathering of the sediments probably comes from limestones. A compilation of the calcium magnesium ratio of several fresh and weathered limestones seems to indicate that the percentage loss of calcium is somewhat higher than that of magnesium and that the amount of calcium given off by the weathering of limestone greatly exceeds magnesium (Table VII). Résumé of weathering.—It follows from the facts stated that the weathering of igneous rocks and sediments results in the loss of more calcium than magnesium and that in general the percent- age loss of calcium is greater than that of magnesium. Materials lost by dynamic. metamorphism.—The dynamic meta- morphism of sediments as well as igneous rocks seems to bring about certain definite chemical changes. It is difficult, however, to measure these changes since it is uncertain whether any of the elements are stable in any given case. All that can be done in the problem of determining the relative stability of lime and mag- nesium under dynamic condition is to compare the magnesium and calcium ratios in the unaltered materials with the calcium magnesium ratios in their metamorphosed equivalents, as shown by Table IX. Secondly, to compare the importance of calcium and magnesium in the minerals of the unaltered and metamor- phosed rocks. Table IX appears to indicate that the ratios of calcium to magnesium are lower in the metamorphosed phases than in the unaltered; that is, the percentage of lime lost by dynamic metamorphism appears to be higher than that of mag- nesium. Concordant with this apparent chemical change is the fact that the minerals which are developed under conditions of dynamic metamorphism are predominantly magnesium and potassium bearing, rather than calcium bearing, such as the micas, chlorites, and amphiboles of Table VIII. It appears that mag- nesium is better adapted to dynamic condition than calcium. A selective removal of calcium from the zone of anamorphism to the zone of katamorphism and ultimately to the ocean seems to be a logical sequence. 396 EDWARD STEIDTMANN TABLE VIII Tur PERCENTAGES OF CALCIUM AND MAGNESIUM IN CALCIUM- AND MAGNESIUM- BEARING MINERALS, CHARACTERISTIC OF DyNAmMIC METAMORPHISM Mineral Ca Mg ANGLINOLItES favs scat ee nee 0.5 Teo Variable Anthophyllites ere aie eee 0.20 28.69 Variable VN ea oo vemai ae etal vanl element cious Area II.4 8.6 Variable BIO ELE enc yeene area e recut eau eC [EN Uae etree Sara 14.3 Variable Gh Ori tee Pei Us ere eae eyes 20.0 Variable Gordieritems ) ecu os oe eeu reeegert: 6.1 EHonnblende te crrrs ete rere 8.7 8.64 Serpentine ss ec Sunpeerycece nl. aden ale 25.8 Spleen eee ec, aee ckcueeeraene aera yom olen TT. 15 Scapoliteevene ce secs cotusmuacien: Qe nim es [enn lay eel sroummallane= Wee eae eteet et eee Teed! 8.9 Variable Tremolite see ccs a ee On 4 14.9 Variable Vesuviamites santa sre m-mec sed Dek 1.57 Variable Wollastonite as meee see SAC Ok als denis, deniers ZiOISICS oye telah oe cena etn ee 71 lint [yan an ee ic alee TABLE IX DyNnAMIC METAMORPHISM Ca/Mg Ca/Mg Rock Original Altered Rock Rock Ih. Average of 12 clays angisoilsecne ace 15 Ae Clarke, Bull. 168, U.S.G.S., 1900, p. 296 Average of 78 shales... er aL Clarke, Bull. 330, U.S.G.S., 1908, p. 27 Average of 9 slates of Vermontyas cena .25 | Van Hise, Mon. 47, U.S.G.S., 1904, p. 895 Dolomite nines see ite HOR Mon. 46, U.S.G.S. Altered to talc schist, PDe2tG= 222 BURIED c Gon ob Ts5 .66 Gab bromusey: eon 95 .44 | Bull. 62, U.S.G.S., p. 76 Gab bromerw ie eicr eine 95 TO) 1i|) Bull 62) UES:G:S:. Up. 470.0 sAltered! more than preceding case GreenStoner ieee ie, 180) 1.05 | Bull. 62, U.S.G.S., p. 91 Gabbro diorite...... 1.45 .89 | Bull. 62, U.S:G.S., p. 80 MORE; occ do Tyee .62 Materials lost by contact metamorphism.—Contact metamor- phism of the sediments tends to develop minerals of complex constitution and high specific gravity. The materials in excess of the requirements for the development of adapted minerals tend to be removed, and in part may reach the sea. The relative EVOLUTION OF LIMESTONE AND DOLOMITE 307 instability of calcium at contacts as compared with magnesium is suggested by the tendency toward increase in the magnesium content of altered sediments, as compared with their unaltered equivalents. cee, Lables XC XI, XIN The averages of the tables are misleading since they do not represent the tendency of the majority of cases. TABLE X Contact METAMORPHISM OF SLATES BY AcrID INTRUSIVES - Ca/Mg Ca/Mg Rock Fresh Altered Rock Rock Chloritic phyllite... . Our? ©), 22 Altered by granite (Neues Jahrb., 1897, p. 156) Slatempiscins sire oe 0.18 0.05 Altered by hornblende granite 50’ from contact (Hawes, Am. Jour. Sc., Mt. Willard, N.H.) Slaten aaa ete: 0.18 0.14 | Ditto, 15’ from the contact SlateAe eons es ees 0.18 By Ditto, 1’ from the contact SIAtesr cs see cme °.18 pit Ditto, at contact WaguiiGulchy 39.4... 0.39 0.87 Contact with quartz diorite (Bull. r50, U.S.G.S.) Morenci shales...... 0.82 0.55 Contact with porphyry (Bull. 229, U.S.G.S., p. 348) Composite. 342... 0.62 0.38 | Of 6 slates by acid intrusives AMVEINGS 5 wig dos oo - 0.33 O93 7, TABLE XI Contact METAMORPHISM OF SLATES BY BAsic INTRUSIVES Ca/Mg Ca/Mg Rock Fresh Altered Rock Rock Composites. 225)... 88 .36 | 8 adinoles altered by diabase (Roth, ; Geol., III) Icenmeschiefer:....... 49 .56 | Composite of 3 slates altered by dia- base (ibid.) Wenneschictersm. a. 55 .30 | Ibid., 147 Slattesip geet erence 09 .47 | Composite of 3 slates altered by dolerite (Crystal Falls, Mon. 36, U.S.G.S.) Slatesm tem e neers 19 .32 | Clausthal altered by diabase dikes (Groddeck, Jahrb. der kéniglich- preuss. Landesanstalt, 1855, pp. 1-53) Witgestioviey 5 4°55 ab soso oe 04 .34 | By gabbro intrusive (Mon. 43, U.S.G.S., 170) Carboniferous....... Rou IO. Shale by peridotite dike (Bull. 348, OeSHGES=9 5343) Averare arene 1.04 Te 5 3098 EDWARD STEIDTMANN TABLE XII LimESTONE Contact METAMORPHISM Ca/Mg Ca/Mg Rock . Fresh Altered Rock Rock HMomestake:. 0... 27 I.O1 By andesite (Leith and Harder, ‘‘ Utah: The Iron Ores of the Iron Springs,” Dist. Bull. 338, U.S.G.S.) Wihite: Kmnobsn.. eee 18.4 74.0 By acid intrusive (Kemp, Ueto Eco. Geol., II, 1907) IMIOrENCi aan nea: Bates) || to} Altered by porphyry (Eco. Gale IDLO, 1907) Bimehamtee ae Gee 54 34 Altered by porphyry (Prof. Paper 28, U.S.G.S.) IBN. Goa a obec 13.2 39 Altered by porphyry (zbid.) Chanarchilloy: 25: 25 57 Chile, altered by greenstone (Woesta, Ueber das Vorkommen der Chlor. Jod. Brom. Verbindungen in der Natur, Marburg, 1870) Slightly altered lime- SLONGm nie re 7.4 D8 F. D. Adams, Jour. Geol., XVII (1909) Slightly altered lime- Stone: eae ce bhai TIS Ibid. Materials lost from rocks by hot solutions.—The alteration of rocks by hot solutions along fissures also shows a marked tendency toward rapid removal of lime, and a much slower rate for the removal of magnesium. In some cases calcite, epidote, and other lime-bearing minerals develop in basic igneous rocks, but often calcium is practically absent from the secondary minerals. The compilation on page 399 has been made showing the calcium and magnesium ratio of fresh and altered rocks adjacent to ore-bearing fissures. Résumé of calcium and magnesium in the products of metamor- phism.—The data on metamorphism which have been presented indicate that the percentage loss of calcium which rocks sustain through metamorphic processes tends to be higher than that of magnesium. In fact, it could be shown that the percentage of calcium lost tends to be higher than that of any other element. Exceptions are noted, of course. In view of this tendency, Salis- bury’s' estimate that the disintegration of 55,000,000 cubic miles of average igneous rock would yield the common salt of the sea, while the disruption of three or more times as much rock would be required to yield the limestones, is suggestive. tR. D. Salisbury, ‘‘The Mineral Matter of the Sea,” Jour. Geol., XIII, 476-77. EVOLUTION OF LIMESTONE AND DOLOMITE 399 TABLE XIII Tue Catcium MAGNESIUM RATIO OF FRESH ROCKS AND THEIR EQUIVALENTS ALTERED BY Hor SoLtutTions ADJACENT TO ORE-BEARING FISSURES Ca/Mg Ca/Mg Rock Fresh Altered Rock Rock Gramitezrns ses soto 1.00 18 Amphibole schist... . TeI5O TS, Granites rseie.s ne 1.50 67. Rhonolites ss. sees. 5-40 2.4 Granodiorite........ 1.04 5.0 Amphibole schist... . 34 Tee Ibimestonen 5. en 4s LO) 1p AMG 5 on 0066 oo ae 1.64 585 .67 35 54 57 .28 a 00 Monzonite porphyry ar 17 2 09 ae Goat Monzonite porphyry Te) Bs2 Diorteres anes ae anor 1.4 1.03 Hornblende andesite . 2.50 oh. Ge Hornblende dacite. . . 2.67 —2.86 Butte granite........ 1.95 .62 Butteyoramites. =... 1.95 53 Butte! granite... .. 1.95 SoG) | West Australia, Pilbara, Goldfield. Quoted by Lindgren, Eco. Geol., I, 540 Kalgoorlie, West Australia. Quoted by Lindgren, zb7d., 530-44 Lindgren and Ransome, ‘Cripple Creeki ay Cole Ph 7raUESiGasentod. Ditto Lindgren, “‘Placer Co. Cal.,” A.I.M.E. (1901), 586-87 | Ditto Limestone altered by hot springs. Spurry) Aspen, Colo. Mons 3i, WESLGES 2 TO J. E. Spurr, “No. 2 Tonopah,” Nev. PAP 42 seo . E. Spurr, ‘‘No. 3 Tonopah,” zbzd. . E. Spurr, ‘No. 4 Tonopah,” ibid. . E. Spurr, ‘‘No. 5 Tonopah,” zbzd. . E. Spurr, ‘‘No. 6 Tonopah,” bid. . E. Spurr, ‘No. 7 Tonopah,” zdid. . E. Spurr, ‘‘No. 8 Tonopah,” zbzd. Lindgren, ‘‘Clifton-Morenci,” P.P. 43, pp. 168-69 Ditto, No. 3 Ditto, No. 4 Ditto, No. 5 Boutwell, ‘Bingham District,” P.P. 38, p- 178 “Willow Creek District, Idaho,” 20 Anne Rept. U-S.G:s., bt. Ui 21132 “No. 3, Hauraki Gold Fields,” Eco. Geol., IV (1909), 637 “No. 2, Hauraki Gold Fields,” zbid., 638 “No. 2, Sericitized Granite,’ unpub- lished investigation, University of Wisconsin | “No. 8, Granite Mineralized,’ unpub- lished investigation, University of Wisconsin “No. 9, Hard Silicified Granite,’’ unpub- lished investigation, University of Wisconsin It is obvious that if metamorphism continued until all rocks were separated into end products, the residuals remaining in place and the materials lost transported to the sea, it would result in a running down of the calcium content of the lands, and a relative increase in magnesium. 400 EDWARD STEIDTMANN In the deep zones of high pressure and temperature, where there is only slight mobility of the residual materials, this result may be reached and perpetuated for a long time until, in the course of geologic ages, they finally become the shallow zones of low temperature and pressure. Here the residuals of metamor- phic processes as well as the materials lost are forever in a state of motion in response to the movements of the atmosphere and the hydrosphere, controlled by gravity and the sun. Thus the prod- ucts of metamorphism are redistributed into the sedimentary rocks, and these in turn are reworked and redistributed. In this redistribution lies the potentiality of an increase of the ratio of calcium to magnesium of the lands with geologic time. Has sedimentation increased the ratio of calcium to magnesium of the lands during geologic time? The sedimentary rocks are derived from other sediments and from igneous rocks, ultimately they are derived from igneous rocks. Clarke’s average igneous rock is generally accepted’ as representing the approximate composition of the primitive litho- sphere. The criticism may be offered that this average is neces- sarily not based on a study of the volumetric importance of the various igneous rock types in the primitive lithosphere. It has also been maintained that the igneous rocks themselves are very largely derived from the fusion of sediments, which may be so but has not been proven. The recurrence of certain predominant igneous rock types at various times and places suggests that the composition of magmas has not been materially influenced in the way which one would expect from regional subfusion of sedi- ments. It seems probable that the primitive lithosphere had a composition between rhyolite and basalt, which is expressed in Clarke’s average. Approximations, not finalities, seem all that can be hoped for in this problem. By an ingenious graphic method, W. J. Mead' estimates that the average igneous rock is equivalent to shales, sandstones, and limestones of Clarke’s average compositions in the ratio of 80: 11:9. F. W. Clarke? has made a similar estimate, based on average 1 W. J. Mead, ‘‘ Redistribution of the Elements in the Formation of Sedimentary Rocks,” Jour. Geol., XV (1906), 238. 2 FW. Clarke, “‘Data of Geochemistry,” Bull. 330, U.S.G.S. (1908). EVOLUTION OF LIMESTONE AND DOLOMITE 401 chemical compositions, in which he distributes the average igneous rocks into shales, sandstones, and limestones in the ratio of 80:15: 5. An earlier estimate by Van Hise’ divides the sedimentary rocks into 65 per cent shales, 30 per cent sandstones, and 5 per cent limestones. A computation made by myself, using Mead’s method, shows that a composite Georgia? granite made from Watson’s analyses is nearly equivalent to a mixture of composite Georgia clay (Wat- son’s) and average sandstone (Clarke’s) in the ratio of 55:45, not enough lime and magnesia being present to be available for lime- stone. Another computation by myself shows that a composite basic rock made up from composites of diabase, gabbro, basalt, and peridotite in the ratio of 6:6; 6:1 is equivalent to average shale and limestone (Clarke’s) in the ratio of 88:12. The upshot of ail these computations and estimates seems to be that the predominant igneous rock types are equivalent to a large per- centage of clastics, predominantly mud or shale, and a relatively small percentage of limestone, hence the same would be true of the primitive lithosphere, regardless of whether it was entirely rhyolite or entirely basalt. Under the theory of the stability of oceanic and continental segments, the redistribution of the primitive lithosphere into sediments may have taken place along one or the other of two uniformitarian directions. The redistribution materials may have been deposited upon the continents and in the oceans in such proportions as to leave the composition of the lands unchanged. This might be termed “‘integral”’ redistribution, because it leaves the composition of the lands as a whole as it was before. Obviously “integral”? redistribution of the redistribution materials to the nth power would not change the composition of the lands. But redistribution has certainly changed the composition of the lands with respect to one element at least—sodium. That the lands contain less sodium now than in the past, in consequence of leach- ing and the accumulation of non-sodiferous sediments on the tC. R. Van Hise, ‘‘Treatise on Metamorphism,” Mon. U.S.G.S., XLVII (1904), 940. 2 Watson, Bull. No. g-A, Geol. Survey of Georgia. 402 EDWARD STEIDTMANN lands, is clearly shown by Becker’ in his recent contribution, ‘“‘The Age of the Earth.” Instead of leaving the composition of the lands as before, redistribution might result in a selective withdrawal of certain elements from the lands and possibly the retention of others. This may be termed “‘selective”’ redistribution. Redistribution has been selective with respect to sodium, resulting in a progressive decline in the contribution of sodium from the lands to the sea. It probably has been selective with respect to potassium, causing only a slight accumulation of potassium in the sea as compared with sodium. The question is raised here whether selective redis- tribution may not have caused an actual progressive increase in the calcium content of the lands and a correlative progressive decrease in magnesium, which in turn may have been ‘connected with a similar progressive change in the ratio of calcium and mag- nesium contributed to the sea, and of the calclum and magnesium carbonates deposited in the sea. It has been pointed out that regardless of whether the primary lithosphere was rhyolite or basalt, redistribution would result in a large proportion of clas- tics, predominantly mud, and a small proportion of limestone. If redistribution has been integral with respect to clastics and lime- stones, it would follow that the sediments exposed on the conti- nents are predominantly clastics and subordinately limestones. This test will be applied here to the continental interiors, the continental margins, the epicontinental seas and the deep seas, so far as the progress of my studies permits. The geologic record of the continental interiors —The greater portion of the surface of the lands consists of sediments. Major Tillo? estimates that the Archaean and younger eruptives constitute only 24.3 per cent of the known area of the continents. It follows from obvious reasons that the greater part of the calcium and magnesium now being delivered to the sea by the rivers comes from the sediments exposed on the lands, and the proportions of calcium and magnesium in the rivers will be roughly proportional UGE Becker, “abhe Nee of the Earth,” Smithsonian Inst. Miscellaneous Collec- tions, LVI (1910), No. 6. 2 Quoted from Berghaus’ Atlas der Geologie (1892). EVOLUTION OF LIMESTONE AND DOLOMITE 403 to the amount of calcium and magnesium in the sediments and to the relative solubility of calcium- and magnesium-bearing minerals in the sediments. In his discussion of ‘‘The Metamorphic Cycle,” C. K. Leith? says: Averages of sections made from field observations give uniformly a lower? percentage of shales and higher of limestones. An average of twenty-one sections from different parts of the United States shows thirty per cent of limestone. If the difference of proportion determined by the chemicals and field methods is a real one, as inspection of the data seems to indicate, the significant questions are raised, (1) whether there may not be a concen- tration of limestones on the continental areas, their complimentary shales and muds being in the deep sea, (2) whether limestone may not be concen- trated in the upper, observed part of the lithosphere, because of its known inability to remain in the deep seated zones of high pressure and temperature. That the Paleozoic sediments of the Mississippi Valley show a surprising concentration of limestones amounting to from 23.6 to 66.6 per cent of the sections averaged and a marked deficiency of shales and sandstones is brought out in an admirable study made by Miss F. W. Carter.4 The results of this study are com- piled in Table XIV. TABLE XIV TABLE SHOWING THE RELATIVE PROPORTIONS OF LIMESTONES, SHALES, AND SAND- STONE IN THE PALEOZOIC OF THE MISSISSIPPI State Limestone Shale Sandstone Pennsylvania....... 23.6 47 25 Vall aiid ays eee iene Hae 52 23 Bias Onion ae race, ele Bono 4l.4 23.6 SE Mnchicambenns eee MD 38.3 TOL lmeianareccced) nese 30 30.3 30.6 WWisconsimgane eee a. 52 38.3 9.5 IM DIOUMNESONES Go teo on ae 50.9 40.4 TS Moat tes cite, oh Ses Omer 2000 9.7 IVINSSO UTI ey are 66.6 22.8 10.4 @OKlahomayiys. 4.5. Bie BO}. 2) TO. 7, Colorado—eastern. . . Di ats 4.8 57.6 Colorado—central.. .. er, 20.9. 21.9 ~C. K. Leith, “The Metamorphic Cycle,” Jour. Geol., XV (1907), 304. 2 Lower than the percentage gotten by distributing an average igneous rock into average sediments. 3 The chemical method of W. J. Mead (of. cit.). 4 Unpublished thesis (1910), University of Wisconsin. 404 EDWARD STEIDTMANN The composition of the lmestones of the Paleozoic of the Mississippi Valley averages that of magnesian limestones, with a lime percentage higher than that of a normal dolomite. The unconformities in the Paleozoic of the Mississippi Valley represent the removal or lack of deposition of both limestones and clastics, mostly clastics, as follows from the compilation below, also made by Miss Carter (Table XV). Both sedimentation and erosion seem to have worked hand in hand toward the con- centration of limestones. ; TABLE XV TABLE OF UNCONFORMITIES IN THE PALEOZOIC OF THE MISSISSIPPI VALLEY Kind of Rock Eroded Location Extent Amount Eroded Summary First base of St. | Widespread | Limestone Less than one- Peter sandstone L. magnesian half thickness Second base Ma- | Widespread | Galena lime- Negligible quoketa shale stone Took of Third top of Si- | Widespread | Salina and Ni-|} Slight EN oe lurian agara lime- umiesroue {=} stone Fourth top of | Widespread | Limestones and | Much Mississippian shales Fifth ever since | Widespread | Shale, sand- More than pre-| Took off more Carboniferous stone, and ceding com- clastics than some lime- bined limestones in stone preceding It seems to follow that the sediments of the Mississippi Valley were either derived from sediments already high in limestones, or else the complementary muds have been carried elsewhere, to the margins of the continents perhaps. But even granted that they were derived from sediments high in limestones, it is diffi- cult to escape from the conclusion that ultimately redistribution was selective. Concentration of limestones on the lands began somewhere at some time. A selective withdrawal of muds from the continents began.somehow, for the Paleozoic sediments of the Mississippi Valley show a proportion of limestones far in excess of the proportions gotten by redistributing either rhyolite or basalt, the two dominant magmatic differentiates. A remarkable preponderance of limestone is also evident from EVOLUTION OF LIMESTONE AND DOLOMITE 405 the following averages made from sections in the interior of China, described by Blackwelder in Researches in China. ; | Section Shales Sandstone | Limestone | Unknown see Per cent Per cent Per cent Per cent Sinian system, Shantung, Northeastern China (Cambro-ordovician, unconform- tay AO TUlIIM CSEOIME) lysine ies ie le inieieos cues 14 nas 86 Shantung (Carboniferous) Hi eeu 20 Shantung, Permo-Mesozoic............ 36 40 ue 24 Shansi-Wu-Tai District. Average of Paleo- zoic (unconformity on limestone)..... Clastics cae 80.5 19-5 Eastern Ssi Chuan and Lower Yang Tzi Gorges. Paleozoic section (unconformity on top of upper Carboniferous limestone) 17.6 6. 75.6 On The continent of Europe shows a similar dominance of lime- stones over clastics. In the southern province of sedimentation, the record is nearly continuous from the Cambrian to the Pliocene, and presents a proportion of limestones far in excess of the ratio gotten by distributing an average igneous rock into the sediments. Another peculiarity of the sediments on the continental interiors is that they are generally less disturbed and less anamorphosed than the sediments on the margins of the continents. The mar- - ginal distribution of mountain ranges and volcanoes harmonizes with this generalization. The fact that the sediments of the continental interiors are generally less anamorphosed than those of the margins is significant in regard to their chemical denudation. Anamorphism tends to cause the decomposition of carbonates and the development of complex silicates, but the silicates are less easily dissolved, hence the relatively small amount of ana- morphism of these sediments increases their importance as sources of calcium and magnesium in river waters. Taking the lime and magnesia contents of Clarke’s average sediments merely as objects of illustration, the following table suggests how important the concentration of limestone on the continents may be in changing the ratio of calcium and magne- sium in the river waters from what it would be if the lands had the composition either of an average igneous rock, rhyolite or basalt. 409 EDWARD STEIDTMANN TABLE XVI Percentage of Percentages of Rock Percentage | Percentage Ratio Average Paleozoic Sediments CaO MgO CaO:MgO Sediment of Missouri : (Mead) (Carter) Shales (Clarke).... Beye 2.44 Iyer 80 22.8 Sandstone (Clarke).| 5.50 1.16 reat TI 10.4 Limestone (Clarke) | 42.57 7.80 ip eyaial ) 66.6 Average igneous rock (Clarke).... 4.79 2.30 Te Alcer Rhyolite (Osann). . 1.43 38 Bear Basalt (Osann).... 8.91 6.03 TAT The ratio of lime to magnesia in the average sediment is about the same as in the average igneous rock, 1.4:1. The ratio of lime to magnesia in the Paleozoic sediments of Missouri would be about 4:1 if their composition is like that of Clarke’s average sediments. The numerical values are not positive, but they point to the probability that the concentration of limestones on the continental interiors may have had a surprising effect on the lime and magnesia content of river waters, and ultimately on the chemical deposits of the sea. The record of continental margins.—Chamberlin’ has pointed out that the sediments which fringe the margins of the continents are characterized by a greater number of unconformities and more intense metamorphism than the sediments of the continental interiors. The imperfections of the marginal record therefore make it impossible to make a fair comparison between the lime- stone content of the marginal sedimentary column and that of the continental interiors. It is perhaps significant that the mar- ginal sediments of late Tertiary and more recent times are pre- dominantly clastic, which suggests a synchronous relation between continental expansion and the deposition of clastics. Deposition within the too-fathom line during continental expan- ston.—It is significant that in the present geologic epoch of con- tinental expansion, the area of the epicontinental sea is limited to about 10,000,000 square miles, perhaps less than a third of what it has been during periods of great marine expansion. It is also significant that the present period of continental expansion «T. C. Chamberlin, Geology, III, 526. EVOLUTION OF LIMESTONE AND DOLOMITE 407 is not favorable to limestone building in epicontinental seas. The preponderance of clastics now forming on the shallows sur- rounding the lands is such that the sediments within the troo- fathom line are generally spoken of as consisting entirely of muds and sands, although important limestone-building areas are found around Florida, Yucatan, and on the Australian Great Barrier reef. The dominance of clastics seems to be related to climatic conditions and the rejuvenation of streams which has accom- panied the rejuvenation of the lands. But shallow, epicontinental seas in times past have been important areas of limestone deposi- tion when their expanse was greater than now. Deposition beyond the too-fathom line —The area of the ocean is estimated by Murray as 143,259,300 square miles. The littoral and shallow-water zones comprise about 10,062,500 square miles, consequently the deep-sea area covers about 133,186,800 square miles. The calcareous deep-sea deposits of terrigenous origin, coral muds, and coral sands have an area of about 2,556,800 square miles or about 1.9 per cent of the deep-sea area. Of the pelagic deep-sea calcareous deposits, the globigerina ooze com- prises 49,520 square miles, pteropod ooze 400,000 square miles, or a total of 49,920,000 square miles, 37 per cent of the deep-sea area. The total area of deep-sea calcareous deposits thus constitutes about 39 per cent of the deep-sea area. The terrigenous non-calcareous muds have a total area of about 16,050,000 square miles, or about 11 per cent of the deep sea. The pelagic non-calcareous deposits have an area of about 64,670,- ooo square miles, approximately 48 per cent of the area of the deep sea, of which red clay represents 51,500,000 square miles and diatom 00ze 10,880,000 square miles. In total, the non-calcareous deep-sea deposits cover about 59 per cent of the deep-sea area. The content of calcium and magnesium in samples of deep-sea deposits collected by the Challenger expedition has been compiled in Table XVII. It shows that calcium is more abundant than magnesium, the ratio of calcium to magnesium being about 13:1. The report on deep-sea deposits by the Challenger expedition concludes that the average calcium carbonate content of the deep-sea bottom is about 408 EDWARD STEIDTMANN TABLE XVII i Mean Denth Ratio of Cal- Macnee ‘ Approximate Per- Deposit athe! Maen Bereeneave eae centages ol (free of Coraltsandian ers: 176 TQ).22 1.8 34.6 coral ss and muds 1.40 Green sand........ 440 30: .64 20.0 ) Green mud........ 513 10.20 “59 Red smiuditer seer 623 43:1 04 TO}55 .07 Coralomudiasae AOE Tiers, Wee see B Area) Seale eee Vocanic muds..... Loci eA Ie Ne pte Salo) Volcanic sands..... 16:1 .99 TOMO pel ag Pteropod ooze..... 1,044 76:1 42 RD .28 Bluetmuds.7: 3.4. 1,411 Apia 56 2.71 10.00 Diatom ooze...... 1,477 28:1 a2 9.17 7.60 Globigerina ooze .. 1,996 Ig: 1.38 26.3 34.50 Radiolarian ooze.. . 2,004 Nes 1.84 4.19 1.60 Rediclay a0 es si2: 3,730 Sa .70 3.48 36.00 92.66 37 per cent, of which fully 90 per cent is derived from the remains of calcareous organisms living near the surface of the sea. ever, the 37 per cent calcium carbonate at the sea bottom merely represents the difference between solution and deposition. tion of the calcareous remains according to the Challenger report is a very important process, resulting principally from the genera- tion of carbonic acid by the decay of the dead organisms. It is to this fact that the decrease in calcium carbonate with depth is supposed to be due. TABLE XVIII TABLE SHOWING RELATION OF CACO; TO DEPTH OF WATER, TAKEN FROM CHALLENGER REPORT 14 cases under 500 (73 7 24 42 68 ce 65 ce 8 co 2 I from 500 to 1,000 + | T,0001tO1 500 1,500 tO 2,000 2,000 to 2,500 2,500 to 3,000 3,000 to 3,500 3,500 to 4,000 Over 4,000 fathoms a9 Average Percentage CaCO; 86. CaCO, 66. CaCO; 70. CaCO, 69.55 CaCO, 46.73 CaCO; 17. CaCO, CaCO, CaCO; trace How- Solu- See Tables XVII and XVIII. EVOLUTION OF LIMESTONE AND DOLOMITE 409 The calcium content of the ocean is a variable controlled by its solution and deposition in the ocean and its introduction from the lands. It is therefore barely possible that calcium is now accumulating in the sea, partly from direct chemical reasons, the calcium content of the ocean being below the saturation point, and partly because the shallow-water area, most conducive to the biochemical deposition of calcium carbonate, is relatively limited during the present epoch of continental expansion; and partly because the shallow waters bordering the continents are now con- siderably polluted by mud and other land débris which depreciates the shore zone as a habitat for lime-secreting organisms. Present climatic conditions also restrict the life zones favorable to shallow- water limestone deposition. Murray and Irvine’ have concluded from experimental evidence that the calcium carbonate of the sea is probably nearly constant in quantity, since the precipitating agents of the sea probably maintain a balance between the intro- duction and deposition of calcium carbonate, despite the fact that the calcium carbonate content of the sea is below the satura- tion point. Whether or not calcium carbonate is actually accumu- lating in the sea seems uncertain, when the wide range of pre- cipitating conditions controlled by the temperature, pressure, and the relative abundance of living and decaying organisms is con- sidered. Nor would annual analyses of sea water give any clue, since it has been estimated that it would require about 680,000 years to accumulate the calcium carbonate now in the sea at the present rate of contribution from the land, a fact which in itself may be significant of the possibilities in this problem. Judging from the selective solubility of calcium carbonate with respect to depth, it seems that the widening of the epicontinental sea to approximately 30,000,000 square miles during the Carboniferous, as estimated by Chamberlin, must have given a tremendous impetus to the deposition of calcium carbonates on these extensive shallows. Here wave agitation might cause mechanical precipi- tation, and would minimize the carbonic acid content of the waters which might otherwise be influenced by organic decay. Evapora- tion would tend to cause concentration. Thus in shallow waters, t Proc. Roy. Soc. Edinburgh, XVII (1890). 81. 410 EDWARD STEIDTMANN calcium carbonate may become so unstable through these and other causes as to result in direct chemical precipitation as shown by the well-known case described by Willis’ in the Everglades off the Coast of Florida and that of Lyell? in the mouth of the Rhone. In shallow, warm waters near the lands, the deposition of calcium carbonate through lime-secreting organisms is very much more rapid than in deep waters. While 90 per cent of the accumulations of calcium carbonate on the floor of the present deep sea come from the skeleta of free- swimming organisms which thrived within the photobathic zone in shallow waters, the remains of both the free-swimming and benthos organisms augment the rate of accumulation. Further- more, the chances for the preservation of skeleta are many times better in shallow water than in the deeps, as shown by the decrease in the calcium carbonate content of marine deposits with depth. In sinking through miles of water, the remains of pelagic organisms often dissolve before reaching the bottom. Not only is there a very clear dependence of abundant limestone deposition on shallows in the present seas, but from the physical evidence of ripple marks, etc., Schuchert concludes that North American Paleozoic lme- stones were probably all deposited in less than 300 feet of water. The food supply is another very important factor which attracts lime-secreting organisms to warm, clear, shallow seas near the continents. The activity of the lime-secreting organisms would be further stimulated by the climatic moderation and uniformity which seem to accompany periods of oceanic expansion. The warming of the seas, consonant with oceanic expansion according to Cham- berlin’s hypothesis, would diminish its capacity for carbonic acid and decrease the solubility of calcium carbonate. But the shallow epicontinental sea, it seems, would be most susceptible to solar heating, hence from a combination of causes, mechanical, physical, chemical, and organic, limestone building in the shallow seas would probably be intensified in more than arithmetical ratio to the increase in the area of shallow water. The total contribution of calcium from the land would be lessened because of the decreased 1 Jour. Geol, Me 8og)ns ts 2 Principles of Geology (12 ed.), I, 426. EVOLUTION OF LIMESTONE AND DOLOMITE All area of the lands. Could not a depletion of the calcium carbonate content of the sea result from intensified deposition on the sub- merged continents? What then? The solubility of the calcium carbonate toward the shallows from the deeps, a process now in operation as shown by the results of the Challenger expedition, would be accelerated. The selective accumulation of limestones on the continents as shown by geologic sections would be con- summated. Significance of the deposition of muds in the ocean basins.—The composition of river muds is variable, depending upon the com- position of the lands over which the rivers flow. The longer the delta region of a river, in general, the smaller probably will be the amount of soluble materials in the muds. The Nile and Mississippi muds may be regarded as typical of the larger streams of the world. In an analysis of Mississippi mud'* the ratio of lime to magnesia is 1.11:1 (Table XIX). The ratio of lime to magnesia in an analysis of Nile? mud is 1.82:1. TABLE XIX TABLE SHOWING LIME AND MAGNESIA Ratios or NILE AND Mrssissitppr Mups CaO MeO : Percentage Percentage Ratio of CaO to MgO INilesmud ee aes 4.85 2.64 82 Mississippi mud... .. 1.83 1.04 Sr The analyses of Nile and Mississippi muds show a relatively high content of magnesia, as compared with other sediments. It follows that if any large proportion of muds is lost from the lands through deposition in the ocean basins, it would mean a selective abstraction of magnesia from the lands, considering the quanti- tative importance of the muds. The ratio of suspended material to total dissolved solids in the Mississippi at Memphis, according to Dole’s yearly average, is 2.3:1. Mellard Reed has estimated that the proportion of suspended to dissolved materials in the river waters of the world is 66:33, or 2:1. Approximately one- half of the dissolved material is calcium carbonate. The ratio ™ By C. H. Stone, Science, XXIII (1906), 634. 2 Analysis D., Bull. 330, U.S.G.S., 420. 412 EDWARD STEIDTMANN of suspended materials or muds to calcium carbonate is about 4.4:1, which argues for a great deficiency in muds in the sedi- ments now forming as compared with the proportion got by distribut- ing average igneous rock under conditions most favorable to the deposition of clastics, namely, the condition of continental expan- sion. This is in line with the fact that the present lands represent an accumulation of limestones. Of the muds now carried to the sea, the major portion are deposited on the continental shelves, and therefore have the potentiality of again becoming a part of the land surface. J. W. Barrell* estimates that from 50 to 70 per cent of the solids brought down to the sea by rivers is deposited within the too-fathom line. But many of the large world streams, the Amazon, the Congo, Indus, Ganges, and others, have their terminations near the too-fathom line. Amazon muds have been traced to a distance of 300 miles from the mouth. Barrell esti- mates that from 20 to 50 per cent of the muds from the rivers are deposited beyond the too-fathom line, and are thus permanently withdrawn from the lands. ‘This estimate is entirely in harmony with the deficiency of clastics in geologic sections, and with the probable withdrawal of magnesium from the lands which is regis- tered in the decreasing magnesium content of limestones in going up the geologic time scale. But the loss of muds from the lands may have been even greater in the past, since many large streams in various latitudes have submerged channels which in some cases extend to the edge of the 1too-fathom line. On the other hand, the percentage loss of muds undoubtedly was relatively much less during periods of widespread continental submergence. But during such periods, the total mud transported was very much less, owing to the smaller relief of the lands and their floral blanket resulting from the moderate, equitable climatic conditions which appear to have accompanied the expansion of the seas. During such periods, muds were accumulating on the lands, until periods of continental uplift, like the present, accelerate their transpor- tation toward the continental margins and the deep sea. Does the ratio of calcium to magnesium in the sea show a selective loss of magnesium from the lands with geologic time?—The ratio of calcium to magnesium in the river waters of the world, taking tJ. W. Barrell, Jour. Geol., XIV, 346. EVOLUTION OF LIMESTONE AND DOLOMITE 413 Clarke’s data, is approximately 6:1. In the ocean, the ratio of calcium to magnesium is 0.35:1. The relative amount of calcium abstracted from sea wateris evidently many times greater than that of magnesium. At the present time, a large proportion of the calcium is being deposited in the deep sea, and is thus either temporarily or permanently withdrawn from the land. A very large proportion of the calcium delivered to the sea has, however, been returned to the lands in the form of limestone deposits; in fact there seems to have been a relatively greater return of calcium carbonate to the lands than of the complementary clastics.: In view of the excess of limestone on the lands, it seems highly probable that the high magnesium content of the ocean represents a selective withdrawal of magnesium from the lands during geo- logic time. This may be one factor which could have caused a decline in the proportion of magnesium contributed to the sea, in the same way as the sodium contribution has declined with geologic time, because of its accumulation in the sea. Résumé of results of sedimentation and their effect on the ratio of calcium to magnesium of the lands.—1. The marine sediments which are revealed to the geologist on the continental interiors were deposited during periods of widespread continental sub- mergence. It is a significant fact that the sediments on the continental interiors probably represent several times as much limestone as could be gotten by redistributing an average igneous rock, or average rhyolite or basalt. The gain in limestone seems to be due mainly to a loss of the complementary shales and select- ive deposition of limestones on the continental interiors. The sedimentary mantle covers about three-fourths of the known area of the continents. The ratio of calcium to magnesium in the average igneous rock is about 1.37 to 1 (Clarke). The lowest ratio of calclum to magnesium in any group of limestones in Daly’s* compilation is 2.93:1, and the maximum 56.32:1. The ratio of calcium to magnesium in Clarke’s average limestone is about 5 to 1. The ratio of calcium to magnesium in the average shale (Clarke) is about 1.48 to 1; that of the average sandstone (Clarke), 5.5 tor. The dominance of sedimentary over Archean and erup- tive terranes and the high percentage of limestones over muds TR, A. Daly, Bull. Geol. Soc. of America, XX, 153-70. 4t4 EDWARD STEIDTMANN indicate a higher ratio of calcium to magnesium of the present lands than in the primitive lithosphere. 2. The marginal sediments show a tendency toward more in- tense anamorphism, a greater number of and more profound uncon- formities than those of the interior, and a dominance of clastics in those sediments which were deposited during continental expan- sion. More intense anamorphism! of the border sediments involves a selective retention of magnesium in them. 3. Marine sediments on the present continental shelves within the 1too-fathom lineconsist predominantly of clastics, from 50 per cent to 70 per cent of the river-borne sediments being deposited here. This seems to afford a fair perspective of the nature of sedimentation during continental expansion. 4. Areally the calcareous deposits constitute a minority of the deep-sea deposits. The rate of accumulation of the terrigenous deep-sea muds is probably vastly greater than that of the cal- -careous deposits. Furthermore, a considerable portion of the calcareous deposits goes back into solution, and has therefore the potentiality of returning to the lands. The permanent with- drawal of terrigenous muds from the land areas unquestionably exceeds that of calcareous deposits. This selective withdrawal suggests one factor in the causation of the loss of muds from the continental interiors. Since muds not only tend to absorb more magnesium than calcium, but actually show a high magnesium content when compared to other sediments, it is not improbable that the permanent loss of muds from the continents also involves a selective and permanent loss of magnesium. 5. The selective retention of land-derived magnesium in sea water may have been an important factor in causing an increase in the ratio of calctum to magnesium of the lands. HAS THE RATIO OF CALCIUM TO MAGNESIUM IN THE RIVER WATERS INCREASED WITH GEOLOGIC TIME? The lands have been shown to represent a higher content of limestone than could have been gotten from the redistribution of the principal igneous rock types which are generally accepted t [bid., 37, “‘ Dynamic Metamorphism.” EVOLUTION OF LIMESTONE AND. DOLOMITE 415 as approximating the composition of the primitive lithosphere. In tracing the evolution of the limestones and dolomites to chemi- cal changes in the sea, it is found highly probable that the ratio of calcium to magnesium in the streams is higher at the present time than in the streams of the primitive lands. This will develop from the following considerations. The influence of the terranes on the calcium magnesium ratio of underground water and streams.—Unfortunately the data on the calclum magnesium ratio of streams and underground water cannot be regarded as a satisfactory basis for correlation with the calcium magnesium ratio of the terranes over which they flow. The chances of error in water analysis and the variability of the composition of streams make a single analysis or even a group of analyses a questionable basis of correlation. The yearly average stream compositions based on daily samples gotten by the United States Geological Survey’ for the streams of the United States east of the rooth meridian constitute a creditable exception. Judging by the data available, it seems that underground waters and streams have a higher ratio of calcium to magnesium than the terranes through which they flow. This agrees with the fact that the metamorphism of rocks results generally in a higher percentage loss of calcium than of magnesium. From Orton’s’ figures, the average ratio of calcium to magnesium in the Niagara limestone of Ohio is 1.72. The rock waters as reported by Orton in the Niagara limestone have the following ratios of calcium to magnesium: LOCALITY Ca/Mg Sidineven@hiOwsne Asie eee as Sp. se Gea ee 4.0 Cehimara@ iO Reh wa ep ere eer cine oe 1.92 Moumitaimelarkey ODIO Ween len fale ee Beans Houmtaimebarky@OMiOn jase). nee ee 2052 leita @iitayateete We ne emer a ekalcriay weaeee 22s lard bun Gaetan. cy. ucvsan dacs cetera erate 2150, An average of 66 analyses of well waters from sandstones given in Bull. 4, of the University of Illinois, yields a ratio of calcium «R. B. Dole, ‘‘Water Supply,” Paper 236, U.S.G.S. 2 Edw. Orton, Nineteenth Ann. Rept. U.S.G.S., Part to. 416 EDWARD STEIDTMANN to magnesium of about 2.6. Twenty-four analyses of well water from dolomites gave an average calcium magnesium ratio of 2.3. None of the crystalline terranes from which stream analyses are reported can be regarded as equivalent in composition to an average igneous rock, in which the ratio of calcium to magnesium IS about: 1.37 tOnr. The analyses are of unequal value. Only those of the Chippewa and Wisconsin are based on yearly averages. The calcium ratio of the others may be too high or too low. In the following table, the ratio of calcium to magnesium varies from 2.03 to 4.91. TABLE XX TABLE SHOWING THE CAtctuM MacGNestum Ratio IN STREAMS FLOWING OVER CRYSTALLINE ROCKS River Mg Ca Source Arkansas River, Canyon, BGs Clarke, Bull. 330, U.S.G.S., 59 Pigeon |Re) Manne csa: I BA al Or deom Ottawa R. Low water (a)...| ~ 3.50 | Daly, Bull. Geol. Soc. Am., XX, 159 Ottawa R. High Water (d).. I 3.82 . | Ibid. Ottawa R. mean (a) and (0) I 3.69 | Lbid. Ottawa R. -(St/Anne)..... I 4.91 Ibid. From granite terrane, aver- age Ovamalyses ao. ge. ie I 3.20 | J. Hanamann, “Bohemia,” Ar- chiv. Natur-Landesforschung Bohmen, IX, No. 4 From mica schist, av. 6 an- BLY SESieee sens toca ah cee I 2.48 Tbid., X, No. 5 Wisconsin River........... I 2.03 | Average of one year. Dole, W.S. Paper, U:S.G.S., 236 Average of one year. Dole, W.S. Paper, U.S.G.S., 236 Chippewa River........... I is) ~I e) The following calcium magnesium ratios of streams on shales and dolomites are based on yearly averages of daily samples reported by Dole. TABLE XXI Tue Catctum MAGNESstumM RATIO IN STREAMS FLOWING OVER DOLOMITE AND SHALES River Sample Mg Ca Fox River, Elgin, Ill............| Niagara dolomite I 1.70 | Average of I year Fox River, Niagara dolomite and Ottawa espe Cincinnati shale I 1.87 | ce oe ethic EVOLUTION OF LIMESTONE AND DOLOMITE TABLE XXI—Continued Al] River Sample Mg Kankakee, Kamniakees yi tiatn. » Niagara dolomite I Rock, Rocktord, Ml sea: 32: Dolomite and shale I Rock, Sircrabbares 10M oar oleae it anes I White, Nepales Whovelss aoe a0 be “ Sele I White, Indianapolis, Ind... .. oy Sea pe I Cedar, Cedar Rapids, Ia..... Silence I Hudson, ei somspNe Vine cece = it cnet I Wabash, Logansport, Ind...... Se pet gts I Wabash, Vincennes, Ind....... “s Ue a I Tllinois, ean Sallesilll eee ces. ng is I Illinois, aoe), Wc osne oops He pany I Illinois, Kampsville, Ill... .. ie ie sh: I Towa, lower Cityaila. =. a. a Soe ees I Maumee, Moledon Owe ssi. a eae I Little Vermilion, Streator lee testa = iY er I Little Wabash. Garris; Ws ois oo 6 hele is see I Miami, DENN, Ons aacaneses is ites I Cache, Mounds, Dl. 2.2... ue ES aia I Big Vermilion, IDeinyallls. Jo was ooo ts Saat I Big Muddy, Murphysboro, Ill..... BY hs er I Embarass, Charleston, Ill........ ss Semen I Embarass, Lawrenceville, Ill..... ES Sas I Grand, Grand Rapids, Mich. . : apie I Muskingum, Zanesville, O......... 2 De aS I Sangamon, IDyeernnwte, INN ois ds lo 8 bes y eaten I Sangamon, Spungiields lees: i. iY (oh ie I Sangamon, Chandlerville, Ill..... Sofa I Average of I year “cc co be (73 (73 co Ce 73 (73 cc 66 6c (75 coe (a9 73 co 66 73 6c co Oe (73 (73 coe (a3 a3 (cana) cc cc co ee (a3 (73 coe (a3 (73 cc 66 73 cc coe ce (75 cc 66 (73 (75 cc 66 a3 “ce co 66 (73 (73 co ae “ (73 (7nd (a3 cc cc 66 (73 (73 cc 66 cc (73 cc Ge “ce (73 (7a ce ce cc be ce 73 G15 oc (73 ce 6 ce (73 cee cc (a3 ce 6 ce 418 EDWARD STEIDTMANN The average calcium to magnesium ratio of these streams is between 2.5 and three. The ratio of calcium to magnesium of a normal dolomite is 1.61; that of an average shale 1.47 (Clarke). The calcium to magnesium ratio of the streams is probably higher than that of the terranes through which they flow. An average of five water analyses on phyllite reported by Hanamann shows a ratio of calcium to magnesium equal to 2.37. The stream waters from limestone areas show a high calcium ratio. See Table XXII. TABLE XXII TABLE SHOWING THE CALCIUM MAGNESIUM RATIO IN STREAMS FLOWING OVER LIMESTONE River Mg Ca Source (Rhamesena cee reer I Tr s6): |) Bulls 330) 0, SiGiSe 75 Meuse, Liége, Belgium..... I TORO i oid 75 SelmerateB CrCyaeneie eee I 46.0 | Ibid., 76 Isoiresats@nléansie. a= ea2 I 10.7 Ibid., 76 Rhone at Geneva......../: I 16.8 | Ibid., 76 Kentucky River, Frankfort .| > 1 Ro Ibid., 66 Cumberland at Nashville. . . I 10.6 Tbid., 66 The following table of averages suggests the influence of the terrane on the run-off. TABLE XXIII Terrane No. of Analyses | Ca/Mg eiMeStONeSieeee ween eee 7 15 Phy llitesis ates setters ters 5 2.37 Hanamann Crystallineia ee eee II 3.36 Dolomites and shales ..... 20 2.91 Sandstone muses amet 66 26 ID olomitemees tea see 24 Om The influence of climate on the calcium magnesium ratio of streams.—Clarke! has pointed out that the streams of humid, more or less forest-covered portions of North America are normally carbonate waters in which calcium is the principal base, while rivers in arid climates tend to be high in sulphates and chlorides in which calcium may or may not be the principal base. The tF, W. Clarke, Bull. 330, U.S.G.S., 72. EVOLUTION OF LIMESTONE AND DOLOMITE 419 magnesium content of streams in arid climates tends to be high, as shown by the following table. TABLE XXIV Catcium MaGNEstumM RATIO IN STREAMS OF ARID COUNTRIES Streams Mg Care| Source SacramentopRes Caley ry eter. I anton Clarkes Bull9330, Ues:Ges-.0/0 Sanulvonenzopen Cally te oa. age I 2.01 | Ibid., 70 SantayClanas Rew Calle ac nema a I Dex || Lhitelee, 7) IMassiont Greeks Cale eo ecient I 2.1830) Lids, 70 ColdiSprmer@reekes Cala a. I 2.00 || Lbid., 70 WonouGreekel Calla aiae) eaten a © I 2 ATA Ot 70 Santa Ynez R., Gibraltar, Cal..... I 2.93 | Lbid., 70 (@hrehhipRe wl ceriawin arsine ott. I 1.8r | Lbid., 70 Chelios cra acta rege sense ar. I Taso) |e lbid.. 70 @lreliigRewAlc eriate tress fcaceroaysies I 2.88 | Ibid.,-70 TAZ OSWRGHMA NG Kena wr sctey tales ecstasy Si I 4.67 | Ibid., 69 IRGOR GRAIG Cs eXt othe. cag ches ster Tee 20 llbid=. 66) REcosm Rue Nees s asin hotness Tey eae aie LOtde 00 Colorado River, Yuma, Ariz....... I 3.30 | Ibid., 69 (GilaehRunver-wATizg-c cision a esate: Ty ese 7a bra. 206 Salle River, NAVAw eu cube concen cde I DF Oise |yelbtd 800 ColoradouwRe, Austin; Tex... 55... I 3 | Dole, W.S. Paper 236, U.S.G.S., 56 Rio Grande; Waredo, Vex... 2.0. ..- r | 4.5 | Ibid., 96 IBIAOS, WWENGO), IMGs os coe ogame nos Te Omceun|ellO7deeeG@ Where the terrane is exceptionally calcareous, however, cal- cium may predominate considerably. The insolubility of lime under arid conditions is illustrated by Hilgard’st composite soils, Table XXV. TABLE XXV Catcium MaGNEstumM Ratio IN Sorts FROM ARID AND Humip REGIONS Soil ; Mg Ca Average of 466 soils from humid regions of southern part of OTe CS Eat esiimare meme ct mek oe tts c Wu wmanha a malt ndie a Manoten Als late I sts) Average of 313 soils from arid portions of United States....... I Ts Apparently soils in arid climates contain about twice as much calcium in proportion to magnesium as those of humid climates. Influence of the belt of cementation on the calcium magnesium ratio of underground waters——The materials carried in solution by underground waters in the belt of cementation undergo various abstractions and additions on their way to the sea. The cements 1E. W. Hilgard, Bull. No. 3, U.S. Weather Bureau (1892), 30. 420 EDWARD STEIDTMANN of the limestones and sandstones undoubtedly contain much more calcium than magnesium, although no estimate of their rela- tive proportions can be given. Since the ratio of calcium to mag- nesium in the average sandstone (Clarke’s) is 5.50 to 1, it is prob- able that the ratio of calcium to magnesium of the cementing materials in the sandstones is even higher. The solutions which percolate through shales, clays, and other silicates are known to suffer an exchange of bases and other changes through the interaction of water solutions and silicates. This interaction is dependent upon the condition of chemical equilibrium between the solutions and the silicates. Kiilenberg’ and other experimenters have shown that soils absorb more potassa and magnesia than lime and soda. The great absorption of potassa by soils and the very slight absorption of soda has been interpreted as the reason why land plants utilize potassa more largely than soda. TABLE XXVI RATIO OF CALCIUM TO MAGNESIUM IN THE CHLORIDE WATERS OF THE DEEP COPPER MINES OF MICHIGAN, AS COMPARED WITH THE SURFACE WATER Deep Mine Waters Ratio of Ca: Mg CrandgEleventicalshant meccw cua ets caer 62 Tamarack Junior (very strong)........... 310 (Cems tlalis Boles ape wos ca Gee aowud es 412 CYandeHesr7thileviele eS aeene re et: 475 Ramiarack qs OOsec tame rere semen TE (trace of magnesium present) Mrimountain. othwlevelk | 7145.2. oe ace (trace of magnesium present) Quincy minel(very Strong) sees ea. | 4,300 (Gyiunbavey/iwabhoVes ye Akins) Crees Abie Se AH yoo 3,078 SUPA ceswa tenet mn art iui eaten ARS The influence of silicates in the belt of cementation on the calcium magnesium ratio of underground waters is suggested by a comparison of the calcium magnesium ratio of the surface and deep waters of the copper mines of Lake Superior. The deep waters are probably the modified residuum left from the cycle of deposition which developed the ores and gangue minerals. The result of cementing processes has been a concentration of calcium in the solutions. Magnesium has evidently been forced out of solution by the conditions of chemical equilibrium. See Table XXVI. * Mittcil. d. Landw. Centralvereins fiir Schlesien, Heft 15, p. 83, quoted by E. C. Sullivan, Bull. 312, U.S.G.S., 16-10. EVOLUTION OF LIMESTONE AND DOLOMITE 421 The relative absorptive power of the crustal materials for calcium and magnesium has not been adequately determined. Certain it is that many shales, slates, muds, and soils have a higher content of magnesium than of calcium. The probability that the selective withdrawal of muds from the lands to the deep sea has involved a selective loss of magnesium from the lands has been pointed out on p. 414. The average calcium magnesium ratio of the solutions contributed to the sea.—The Mississippi‘ at New Orleans, which may be regarded as a mixture of the waters from the average Paleozoic terrane, shows a ratio of calcium to magnesium of 3.81 to 1. The average calcium to magnesium ratio of 73 streams east of rooth meridian of the United States observed daily at 94 stations for a period of one year is about 4 to 1. In Sir John Murray’s well-known com- position of 19 streams of the world, the calcium to magnesium ratio is 4.4 to 1. From Mellard Reade’s? data, the ratio of cal- clum to magnesium in the materials of chemical denudation is 8.25 to 1. A better based figure for the average composition of the streams of the earth is that recently made by Clarke. The ratio of calclum to magnesium in Clarke’s average as previously cited is about 6 to I. While the ratio of calcium to magnesium of the solutions con- tributed to the sea is higher than it would be if the lands had the composition of an average igneous rock, the largest streams do not seem to show the high calcium to magnesium ratio that one would expect from the amount of limestone on the continents. However, the Mississippi is about the only large stream whose composition is accurately determined. It may be that the large amount of suspended material in rivers tends to lower the calcium ratio, since the muds, particularly of humid climates, tend to be high inmagnesium. The arid nature of about one-fifth of the land area is another factor which may cause a retention of calcium by the land. If this hypothesis is correct, a compensating increase mn the calcium ratio of rivers will be contemporaneous with oceanic expansion. RB. Doles Wes. Paper 230, US.GiS., £7. 2 Mellard Reade, Chemical Denudation in Relation to Geologic Time (1879), 1-61. 3F. W. Clarke, Study of Chemical Denudation, Smithsonian Institution, Vol. LVI (2910), No. 5, p. 8. 422 EDWARD STEIDTMANN Conclusion: Increase of the ratio of calcium to magnesium in rivers with geologic tume.—Evidence has been presented to show that the ratio of calcium to magnesium of stream water is influenced primarily by the ratio of calcium to magnesium of the terranes which they drain, being generally higher than that of the terranes. Climate exerts a modifying influence. Aridity lowers the ratio of calcium to magnesium of the stream waters, causing a concentration of calcium in the soils, while humidity has the opposite effect. The interaction of the salts carried in solution by streams and ground waters with the land materials, particu- larly those high in clay, results in a greater loss of magnesium from the waters than of calcium, thus tending to increase the ratio of calcium to magnesium of the streams and ground waters. Regardless of any theory of the origin of the earth, geologic evidence points to the igneous rocks as the primitive source of the sedimentary rocks. The streams of the primitive lithosphere, therefore, probably approached in their chemical character the present streams, flowing over crystalline rocks, subject to climatic and other modifications. Such streams have a ratio of calcium to magnesium approximating 3 to 1. The best figure given for the average calcium to magnesium ratio of the streams of the world is approximately 6 to 1. The latter figure probably should be greater, considering the abundance of limestone on the conti- nents. Little doubt therefore remains that the proportion of calcium to magnesium in the streams is now higher than in earlier stages of the earth’s history. It is also highly probable that the increase in the ratio of calcium to magnesium in the rivers has been continuous with geologic time, because of the progressive increase in this ratio in the limestones deposited during geologic time, and because of the selective deposition of limestones on the submerged continents. STATEMENT OF HYPOTHESIS The conclusion has been reached that dolomites develop pre- dominantly in the sea rather than by the metamorphism of lime- stones after their emergence from the sea. Hence the decline in the percentage of dolomite in going up the geologic column seems EVOLUTION OF LIMESTONE AND DOLOMITE 423 to indicate that less and less dolomite was deposited in successive periods of geologic history, thus pointing to a progressive change in the conditions of deposition. Of the four factors controlling the deposition of carbonates in the sea, viz., temperature, pressure, life processes, and chemical composition, only the last two show any probability of progressive change with time. There is no evidence for a change in the nature of life processes. There is evidence for a change in the chemical composition of the sea, specifically for an increase in the ratio of calcium to magnesium contributed to the sea from the lands, which will appear from the following considerations. The present lands contain a much larger proportion of limestones than could be gotten by redis- ' tributing a granite or basalt, generally accepted as being equiva- lent to the materials of the primitive lands. It is inferred from a consideration of the relation between the composition of river waters and the terranes which they drain that the present rivers have a higher ratio of calclum to magnesium than those of the primitive lands. The accumulation of limestones on the lands of increasing calcium content, with time, seems to be related to a reworking of the land over and over again along certain selective lines. A higher percentage of calcium than of magnesium tends to be lost from all kinds of rocks when subjected to all kinds of metamorphic processes. Hence there is a continuous selective removal of cal- cium from the lands of the sea, as is evidenced by the fact that the ratio of calcium to magnesium of rivers tends to be higher than the ratio of calcium to magnesium of the lands which they drain. This involves a selective retention of magnesium in the clastics. ‘The transportation and deposition of clastics is at a maximum during periods of continental expansion and at a mini- mum during periods of continental submergence. The clastics are therefore deposited mainly on the margins of the continents and in the deep sea. Those deposited in the deep sea are per- manently lost to the lands and with them goes a selective loss of magnesium from the lands. The carbonates, calcium and mag- nesium, are deposited mainly in shallow epicontinental seas during periods of continental submergence, in consequence of organic 424 EDWARD STEIDTMANN and inorganic agencies. The percentage of calcium carbonate which is deposited from the sea is higher than that of magnesium carbonate, and from field and laboratory evidence it is inferred that the proportions of the two carbonates deposited are in some direct relation to their proportions in the rivers which bring them to the sea. Hence, there is a selective return of calcium to the lands. It is therefore inferred that the evolution of the limestones and dolomites has been in response to the gradual increment of calcium over magnesium in the solutions contributed to the sea, a tendency arising primarily from physical-chemical causes, aided or accelerated by organic processes working harmoniously with the inorganic environment. For illustration, it may be assumed that sedimentation began during continental expansion when the lands had the composition of an average igneous rock. From the known results of meta- morphic processes, it would follow that the solutions contributed to the sea had a higher calcium to magnesium ratio than the lands from which they were derived, and that the residuals had a higher proportion of magnesium to calcium than the original rocks. A part of the residuals, particularly the muds, are subject to selective transportation to the continental margin and the deep sea. Accept- ing the hypothesis of the permanence of oceans and the continents, those deposited in the deep sea are permanently removed from the continents. The calcium and magnesium salts interact with the materials of the ocean bottom and enter into the constitution of silicates, carbonates, and other compounds, or they may inter- act with other constituents in the water, and be precipitated prin- cipally as the carbonates. Magnesium would tend to interact more actively with the muds of the bottom than calcium. Calcium would tend to be more insoluble in shallow water than in the deeps, while the opposite tendency probably characterizes mag- nesium. Magnesium salts in general are more soluble. in sea water than calcium salts. Organic precipitation, apparently only an adaptation to conditions of chemical equilibrium already existing, would be particularly effective in abstracting calcium from warm, shallow seas. The relative solubility of the materials EVOLUTION OF LIMESTONE AND DOLOMITE 425 precipitated would depend upon conditions of equilibrium con- trolled mainly by temperature, concentration, the amount of carbonic acid in the air and ocean, and organic processes. As a result of the preceding selective influences, the calcium to magne- sium ratio of the limestones would tend to be higher than that of the solutions contributed to the sea. However, contemporaneous with continental expansion, as at the present time, limestone deposi- tion would be at a minimum, which might involve a concentration of calcium in the sea until more favorable conditions of precipi- tation arise. Limited areas of shallow water, vigorous erosion, continental climates, and other supplementary conditions make periods of continental expansion more favorable to the deposition of clastics than limestones. Gradually the lands waste away, the ocean advances over the continents, partly in consequence of fill from the land, in part, perhaps, -as a result of secular earth movements which cause a shallowing of the ocean basins. The rivers carry less and less débris to the sea, and deposit it farther and farther inland from the margin. On the submerged continental areas, covered by shallow seas, which now may be three or more times as extensive as they were during the preceding period of continental expansion, chemical and biochemical processes combine in making this an era of limestone building. From experimental and field evidence, the inference is drawn that the ratio of calcium to magnesium in the deposited limestones is influenced primarily by their respective rate of contribution from the land, and modified by selective organic and inorganic agencies working to a common end. As postulated by Chamberlin, with the expansion of the seas, the zonal, diversified continental climates tending toward aridity and refrigeration yield to more uniform, mild atmospheric con- ditions. A widening of the life zones favorable to limestone deposition follows. Thus in the Devonian, corals thrived in the now ungenial climate of Hudson Bay. With world-wide climatic moderation, a new condition of equilibrium is established between the carbon dioxide of the sea and air. Warm water absorbs less carbon dioxide than cold. The sea begins to contribute its excess of carbon dioxide to the air, in consequence of which the calcium 426 EDWARD STEIDTMANN of the sea becomes still more insoluble. The atmospheric condi- tions and the land relief favor the floral blanketing of the earth, thus stimulating chemical denudation and creating an effective screen for the retention of clastic materials on the land. As the waters again withdraw from the lands into the hollows of the sea in response to secular earth movements, they leave composite sediments behind them whose ratio of calcium to mag- nesium is higher than that of the average igneous rock. With topographic rejuvenation, the muds are again carried toward the margin and to the deep sea, or possibly at times directly to the deep sea as suggested by submerged stream channels, the con- tinuation of existing streams which in some cases extend to the margin of the deep sea. Various parts of the earth are subjected to regional metamorphism and secular uplift, particularly the continental margins, causing selective removal of calcium and the concentration of magnesium in the residuals. The marginal sediments, dominantly clastics, by virtue of position, relief, and a combination of other factors, are in a most favorable position to be removed from the land and swept into the deep sea, where they would be permanently withdrawn from the land. The depo- sition of clastics is again at a maximum, that of limestones at a minimum. Asthe pendulum swings from one extreme to another, it marks a curve of progressive change in the composition of the lands and in the ratio of calcium to magnesium in the salts con- tributed to the sea, consummated by an accumulation of limestones on the continents of progressively higher calcium content, both a reflex and a cause of changes in the land composition, and by the withdrawal of the complementary muds toward the margins and the deep sea, slow during periods of oceanic expansion, but tre- mendously accelerated during periods of oceanic retreat, selective concentration of magnesium in the deep zones of high temperature and pressure, in the clastics, and in the sea—a never-ending cycle of selective causes and cumulative effects, recalling the words of Faust: Wie alles sich zum Ganzen webt, Eins in dem anderen wirkt und lebt. EVOLUTION OF LIMESTONE AND DOLOMITE 427 SUMMARY The problem under discussion is, Why does the dolomite con- tent of the geologic column decrease with time? Is it due to a secondary alteration of limestone after emergence from the sea, roughly proportional to time, or is it due to a gradual decline in the primary development of dolomite in the sea? If the latter, what factors controlling the deposition of dolomite have changed during geologic time, temperature, pressure, life processes, or the chemical composition of the sea ? . The conclusions reached are: dolomite develops predominantly in the sea, therefore the decrease in the dolomite content of the sediments in going up the geologic column is mainly due to a decrease in the proportion of dolomite developed in the sea with time. The factors of deposition whose progressive change has probably controlled the decline of dolomite development in the sea are life processes and the chemical composition of the sea. There is no definite evidence for a change in the nature of the life processes in their relation to dolomite deposition. There is evidence for a change in the chemical composition of the sea; namely, the fact that the present ratio of calcium to magnesium of the streams is probably more than twice that of streams draining crystalline terranes, comparable in composition to the primitive lands. Accept- ing uniformitarianism, it follows that the present streams have a much higher ratio of calclum to magnesium than the primitive streams. It has been indicated that solutions high in magnesium and low in calcium are more favorable to the development of dolomite than those which are low in magnesium and high in calcium. It is therefore highly probable that the chemistry of the primitive sea was more favorable to the deposition of dolomite than the present ocean. The increase in the proportion of calcium to magnesium in the streams is believed to be due to selective processes whose effects have been cumulative with time. Rock alterations tend to result in a higher percentage loss of calcium than of magnesium, the materials lost being largely transported in solution to the sea. 428 EDWARD STEIDTMANN A higher percentage of magnesium is retained in the residuals of rock decay than calcium, but erosive processes are constantly removing the residuals toward the margins of the lands, and during periods of continental expansion a considerable proportion is swept into the deep sea and permanently lost from the lands. In consequence of a combination of organic and inorganic agencies, the maximum deposition of limestones and dolomites is on the submerged lands during periods of oceanic expansion. ‘The per- centage of calcium precipitated is higher than that of magnesium, but the proportions of calcium to magnesium which are precipi- tated bear some direct relation to their ratio in the rivers which bring them to the sea. With the progressive elimination of clastics and magnesium from the lands with geologic time, and in their place the gradual accumulation of calcium in the form of limestone, the proportion of calcium to magnesium contributed by rivers to sea has increased with time. The writer is indebted to C. K. Leith for suggestions and criticisms. DIFFERENTIATION OF KEWEENAWAN DIABASES IN THE VICINITY OF LAKE NIPIGON E. S. MOORE The Pennsylvania State College In recent numbers of Economic Geology two papers have been published describing the differentiation products of the quartz- diabases of the Nipissing District, Ontario. Since these diabases have generally been regarded as of Keweenawan age, certain differentiation products of the Keweenawan diabases in the vicin- ity of Lake Nipigon are also of interest. On the north shore of Lake Superior and extending northward beyond Lake Nipigon there are masses of diabase and gabbro which intrude the older crystalline rocks in the form of batholiths, dikes, and bosses and the sediments in the form of the “Logan sills.” Although there are differences of opinion regarding the geological age of these rocks, the writer concurs with those who regard, as closely related in origin, the great amygdaloidal basalt flows of Keweenaw Point, the Duluth gabbro, the ‘Logan sills,” the Sudbury Nickel eruptive, and the Cobalt diabases, as well as many other masses of diabase in intervening areas. The great igneous activity of this region seems to have been the result of extensive crustal adjustment centered around Lake Superior and diminishing in intensity as a greater distance from the center was reached. It is probable that on the northern side of the Lake Superior basin the intrusive masses were being injected into sedi- ments which had already been formed while the alternate deposits of sediments and lava flows were being deposited on the south side and that a close relationship exists between all portions ot this great series of sediments and extrusive and intrusive igneous rocks. While a general description of the petrography of these rocks is given here, the object of this paper is to call attention to certain tW. H. Collins, Econ. Geology, V, No. 6, p. 538; R. E. Hore, ibid., VI, No. 1, JO. Sits 429 430 E. S. MOORE evidences of differentiation which have already been mentioned by Dr. A. P. Coleman and the writer and to add additional notes to the descriptions of this phenomenon. PETROGRAPHY OF THE DIABASES The Keweenawan rocks around Lake Superior have been described petrographically in detail, by Irving, Bayley, Van Hise, and many others. In the vicinity of Lake Nipigon the rocks are in many respects similar to those around Lake Superior and they have been described with less detail by Coleman, Wilson, and other geologists. The greater portion of the shores and the islands of this lake consist of basic rock, either diabase or gabbro. Thin sections almost invariably show the ophitic texture more or less well developed, and, although in many places the diabase grades toward gabbro, the greater portion of the rock is diabase. In the sills, diabase always seems to be found, and the same state- ment may be made of the smaller bodies of the rock, while some of the larger batholithic masses, which have suffered some differ- entiation, more strongly resemble gabbro. Structurally the rocks form bosses, large and small, batholiths, or very large irregular masses, dikes, and sills. The dikes are often large, as some were seen in the Onaman Iron Range area 150 ft. in width, and these seem to represent offshoots from the main diabase mass in the vicinity of the lake. The sills, known as the “Logan sills,” form beds from two to several hundred feet in thickness. These masses lie between beds of sandstone, shale, or dolomitic limestone, or between these sediments and the under- lying Archean rocks, and in all cases studied they present evidence of their intrusive character. Columnar structure is a character- istic of nearly all of the larger masses, especially of the larger sills. In macroscopical characters these basic rocks generally present a monotonous appearance. They vary in grain from coarse to medium fine and in color from brownish to nearly black. Some of them weather rapidly to granular incoherent masses, and, in the early stages of this weathering, they exhibit in many places cleav- age surfaces with a bronze tint. In many cases the ophitic texture is readily recognized in the hand specimen, but in the masses DIFFERENTIATION OF KEWEENAWAN DIABASES A431 which tend to become coarse grained and to separate into little aggregations of feldspar and magnetite this texture is lost to a large extent and the rock becomes more like a coarse gabbro. In one place on the shore of Lake Nipigon some sand was collected which showed poikilitic texture where feldspars were inclosed in augite. In microscopical observations these rocks usually show labra- dorite, augite, or diopside, and ilmenite or magnetite. Olivine is widespread but is not always present and in specimens without olivine quartz has been found, but it is lacking in many specimens. Biotite appears in small quantities and titanite was found in one section. Since the latter mineral occurs near a dike of acid rock and is not commonly developed in diabases or gabbro, it is believed to be due to the influence of this dike, as some of these acid dikes carry titanite. Although these rocks are on the whole comparatively fresh, certain alteration products occur. The olivine frequently shows serpentine and iron oxide as alteration products, and the augite and diopside, although usually quite fresh, often contain second- ary amphiboles and actinolite. In a specimen from ‘ Haystack Mountain,” north of Lake Nipigon, a crystal of magnetite occurs partially surrounded by a mass of actinolite needles which, on revolving the stage of the microscope, show rotary extinction (Fig. 1). These needles seem to be the product of alteration of an augite crystal whose growth began around the magnetite and they resemble similar fibrous growths which W. S. Bayley de- scribes as occurring around magnetite in Fic. 1.—Magnetite par- - ‘ i tially surrounded by augite the basic rocks of the Lake Superior region, which has altered to actin- although he does not ascribe a secondary lite needles (greatly en- origin to them.t In a specimen from the jarged): shore of Lake Nipigon, opposite “‘Two Mountain” Island, the diopside and magnetite are intergrown to some extent and the latter sometimes occurs as a fringe along the border of crystals of the former. Although much of the magnetite associated with the t Journal of Geology, I, 702-10. 432 E. S. MOORE diopside is primary, some crystals of the diopside which are par- tially altered to secondary amphiboles contain also undoubted evidence of alteration to magnetite and hematite. While these are unusual alteration products for diopside, analyses of this mineral from gabbro sometimes show as much as 15 per cent of Fic. 2.—Photomicrograph of diabase showing ophitic texture (crossed nicols; X40). iron oxide. The pyroxene is readily recognized as diopside by its characteristic color and extinction angles. In texture the ophitic character is usually well developed, as the labradorite generally occurs as lath-shaped, nearly euhedral crys- tals, which penetrate the augite and diopside (Fig. 2) and in some cases are surrounded by them, giving also a poikilitic texture. The rock might therefore be called, to apply the term suggested DIFFERENTIATION OF KEWEENAWAN DIABASES 433 by A. N. Winchell for such textures, a poikilophitic rock.t In sections of diabase from “Haystack Mountain” the augite is frequently twinned with two members, and instead of the usual stout crystal it occurs in long, narrow forms, somewhat lath shaped, and in this respect resembling the feldspars. DIFFERENTIATION PRODUCTS; PEGMATITE DIKES The differentiation products of the Keweenawan rocks of the Lake Superior region have been frequently mentioned. Clements states that the gabbro in Minnesota shows undoubted evidence of differentiation in the large masses of anorthosite and the patches of magnetite and titaniferous iron ore.2, W. S. Bayley describes peridotites and pyroxenites as very basic phases of the gabbro in his description of the Lake Superior region. From Lake Nipigon A. P. Coleman describes picrite and other very basic phases of the diabase and also certain acid dikes which are described as post-Keweenawan but closely related to the Kewee- nawan basic rocks and perhaps differentiation products of them.‘ These rocks are described as having a pegmatitic or micropeg- matitic texture and as having the composition of granite or grano- diorite. In ‘Haystack Mountain” north of Lake Nipigon the writer found similar dikes and from their relationships suggested that they represented an acid phase of the diabase magma.’ The later observation of similar dikes in the Duluth gabbro near Duluth, Minnesota, confirmed the belief that these rocks are differentiation products of the diabases and gabbros. The rock at “‘Haystack Mountain” is a coarse diabase, rather gabbro-like, and shows small, dark patches of titaniferous magnet- ite and in places lighter blotches consisting largely of feldspar. The magnetite is sufficiently abundant in part of the hill to influence the compass so that prospectors were led to record mining claims t “Use of ‘Ophitic’ and Related Terms in Petrography,” Bull. Geol. Soc. Am., XX (1010), 661-67. 2 U.S. Geol. Survey, Monograph XLV, 397-424. 3 Jour. of Geology, II (1894), 814-25. 4 Bureau of Mines of Ontario, XVII (1908), 163-64. 5 Ibid., XVIII (1907), 162. 434 E. S. MOORE upon it. Besides these small segregations of feldspar there are irregular dike-like masses of similar, light-colored rock and a few fairly distinct dikes, all of small size and varying from one-half inch to a foot in width. These dikes are rather fine grained and in thin section show the following characters. The texture is usually micropegmatitic and one section is composed of about 60 per cent feldspar, 30 per cent quartz, 10 per cent hornblende, and a little magnetite and hematite. The feldspar is chiefly orthoclase with a little albite and the rock is a granite. Another section contains very little hornblende and a little epidote, the rock being composed almost entirely of feldspar in the proportions of 65 parts orthoclase to 35 parts plagioclase. This rock is a syenite grading toward a monzonite. Still another section is from a dike which might be regarded as a monzonite. It contains a little enstatite, epidote, and titanite, while the greater portion of the rock is feldspar and in the proportions of about 66 per cent albite and oligoclase and 34 per cent orthoclase. A fourth section is from an augite-syenite dike in which the orthoclase makes up 75 per cent and the sodic variety 25 per cent of the feldspar. There are a good many small augite crystals and the micropegmatitic texture is well developed. A section of a dike from near ‘Two Mountain” Island is also an augite-syenite. : The most interesting dikes in the region are those occurring on Flat Rock Portage near the south end of Lake Nipigon. At this point a large mass of rock, described by Coleman as a sill of epi-basalt or fine-grained diabase-porphyrite, is cut by a pegmatite dike varying in width from three inches to one foot. Where the diabase has suffered columnar jointing the fissures have filled with acid rock similar to the flesh-colored or pink dikes described above (Fig. 4). The surface of this sill is flat and fine grained, and when the glacier passed over it interesting chatter-marks were left. The pegmatite dike is medium coarse grained, flesh colored, and appears to be composed very largely of feldspar. Under the microscope the pegmatitic intergrowth of various minerals and the graphic intergrowth of quartz and feldspar are well developed. The rock consists of quartz in proportion of ro per cent, epidote to per cent, and sodium-calcium feldspar 80 per cent. The DIFFERENTIATION OF KEWEENAWAN DIABASES 435 feldspars show a wonderful development of the zonal structure (Fig. 3). In the rapid growth of the crystals the zones of calcium- and sodium-rich material have developed mostly at their ends, and they have thus been drawn out to excessive linear proportions. The indices of refraction indicate that the most calcic feldspars form the central zone and the more sodic follow outward. The usual order of rate of weathering of the sodic and calcic feld- Fic. 3.—Photomicrograph of a section from a pegmatite dike at Flat Rock Portage showing unusual development of zonal structure in sodium-calcium feldspar (crossed nicols; X4o). spars does not hold in many of these crystals, as the second zone even with lower index of refraction often shows much more exten- sive alteration than the central zone. The alteration products are epidote and a mineral or mixture of minerals which has yellow- ish polarization colors and is believed to be epidote, kaolin, and zeolites. A peculiar influence of the pegmatitic intergrowth of the minerals is the crystallization of epidote in some of the zones 436 E. S. MOORE replacing the feldspar. This arrangement was seen where the path of an epidote crystal was cut across by that of the growing, zonally built feldspar. In a couple of cases a group of epidote crystals is crossed by feldspar, but the simultaneous extinction of all parts of the epidotes shows them to be parts of the same crystal and in one case particularly a portion of the epidote crystal has passed through the feldspar, forming one of the zones of the crystal. In these cases the epidote is undoubtedly primary, although considerable second- ary epidote occurs from alteration of the feldspars. An interesting example of a crushed feldspar is seen in this section where the crystal has been broken into slivers and the fragments surrounded by quartz. This must have been due to pressure, although the rock as a whole does not show evidence of excessive pressure beyond the undulatory extinction of some of the quartz grains. In his description of the copper-bearing rocks of Lake Superior, Irving describes some sections from dikes of red rock in the Duluth gabbro which would indicate that they are probably similar to those dikes described above.* These acid dikes appear to be differentiation phases of the Keweenawan diabases and gabbros because they occur, with one exception, in these rocks only, and, so far as observed, only in the larger masses and not in the thin sills which are too small to pro- duce them by differentiation. This one exception is a dike 30 inches wide cutting quartzite and the overlying diabase near Ombabika Narrows.? This dike might be due to the rising of the liquid from some large diabase mass below through a fissure extend- ing into the overlying rocks. There are no other bodies of acid rock in the region later in age than the Keweenawan diabases, and the pegmatitic and micropegmatitic textures suggest end phases of a magma. These dikes probably fill crevices in the diabase formed in the solidified exterior of a large mass, due to adjustment of pressures during processes of cooling, and the acid material rose from the still hot and more acid lower portions of the mass. The tact that these dikes are so much more basic on the whole than U.S. Geol Survey, Monograph V, 119-20. 2 Coleman, Bureau of Mines of Ontario, XVII, 164. DIFFERENTIATION OF KEWEENAWAN DIABASES 437 the aplites of the Cobalt area may be due to the tact that the magma from which they separated was on the whole more basic. The ophitic texture of the diabases indicates 2 magma early saturated with calcium. The alkalies, being in small quantity, were mostly left over until the end of the crystallization period and then united with the remaining aluminium and silica to form potassium or sodium feldspars, while the small excess of silica occurring in a few places Fic. 4.—Acid dikes filling columnar joint fractures in diabase. went to form quartz. It is thus assumed that the magma became saturated with the more basic materials first, and the remaining acid materials, still liquid, were in some cases crowded toward the lower and central portion of the mass to escape into the fissures when opened, and form dikes. ; In the case of the pegmatite dike in which the feldspar is largely calcic and occurs with the silica, it is probable that the excess of magnesium and iron caused the rocks to become saturated with 438 E. S. MOORE these elements, and the augite and olivine, separating out earlier, caused some of the calcium and aluminium to be left over to enter the dike. It is interesting to note that in the Sudbury and Cobalt areas, where the Keweenawan rocks have suffered very great differen- tiation compared with that in the Lake Nipigon region, there are extensive ore bodies connected with them, while there is nothing but a little iron ore in the Nipigon region, and this occurs at the contact with sediments and may be leached from them. While differentiation of magmas and thus the separation of the metals, as well as other elements, from the magmas, may be only one factor in the development of mineral veins as well as mag- matic segregations, it seems probable that all data collected on this subject will show this is one of the important factors. Other things being equal, if the igneous rocks are the source of the metals, those magmas which show the greatest differentiation should be the most favorable for the production of ores, whether they supply metal-bearing solutions directly to the veins—a process quite conceivable in some cases—or whether they cause segregation of the metals so that they can readily be dissolved by meteoric waters in sufficient quantities to form ore deposits in veins. GENERA OF MISSISSIPPIAN LOOP-BEARING BRACHIOPODA STUART WELLER INTRODUCTION The correct specific determination of the loop-bearing brachio- pod shells of the Mississippian faunas has always been attended with difficulty. This condition led the writer to undertake a critical study of a large amount of material, in order to determine, if possible, the true criteria for the determination of species. In this study the internal characters of the shell, as well as the exter- nal configuration, have been taken into consideration. In the earlier literature all the shells of this type were included in the genus Terebratula, but since the appearance of Hall and Clarke’s great work on the Genera of Paleozoic Brachiopoda,’ it has been the usual custom to refer all of them to the genus Dielasma, although Girty has described one form as a member of the genus Harttina” As a result of the present investigation, however, it has been found that forms which have been commonly included in a single genus, and in some instances, even, forms which have been referred to a single species, in reality represent several perfectly distinct generic types. The method used in the investigation is that which has been formerly used in the study of the Rhyncho- nelloid shells. Specimens have been ground from the posterior extremity and at short intervals the ground surface has been pol- ished and a careful drawing made of the cross-section of the shell parts shown. From such a series of cross-sections of any shell it is easy to recognize the character and position of the internal lamellae, such as the median septum, the hinge-plate, the socket- plates, the bases of the crura, etc. In the course of the study it has been shown that the characters which can be considered as of t Paleontology of New York, Vol. VIII, Parts x and 2. 2 “Farttina indianensis Girty,’’ Proc. Nat. Mus., Vol. XXXIV, p. 203. 439 440 STUART WELLER generic value in these shells are to be found in the rostral portion of the brachial valve. Six and possibly seven good generic groups have been recognized, and will be defined here. DIELASMA King Description.—Shell terebratuliform. Pedicle valve with or without a median sinus, the beak strongly incurved, the foramen large and encroaching upon the umbonal portion of the valve; internally with well-developed dental lamellae. Brachial valve usually without mesial fold; internally the crural plates are sepa- rate from the dental socket-plates, they diverge from the apex of the valve with an elongate attachment to the inner surface of the valve, the free portion of the brachidium is short, with diverging MW. (UI WH, Nat, YD NS Fic. 1.—A series of fourteen cross-sections (23) of the rostral portion of the brachial valve of Dielasma formosa Hall. descending lamellae; between the crural plates for the full length of their attachment to the inner surface of the valve, is a concave, transverse plate for muscular attachment, which joins the inner surface of the crural plates a little above their base; this plate rests against the inner surface of the valve along the median line for a portion or the whole of its length, or it may be free throughout; when attached along the median line a pair of slender cavities, triangular in cross-section, converge from the general cavity of the shell toward the beak, but when the transverse plate is not attached along its median line there is a single, broad and low cavity beneath the plate, extending toward the apex; anteriorly this plate extends to a greater or less distance beyond the attach- ment of the crural plates and is pointed, rounded, or emarginate in front; its surface is marked by concentric wrinkles parallel with MISSISSIPPIAN LOOP-BEARING BRACHIOPODA 441 its anterior margin which are usually discontinuous along the median line. Remarks.—The genus Dielasma was established by King with Terebratula elongatus Schl. as genotype, and although he defined the genus primarily upon the presence of prominent dental lamellae in the pedicle valve, and on the form of the loop, his illustrations of the internal casts of the species under the name Epithyris elongata’ show that the crural plates are separate from the socket walls, one of the most essential features of Dielasma as here defined. David- son’? gives illustrations of the same species which exhibit all the essential generic characters of Dielasma most perfectly. The interpretation of the genus by Hall and Clarke is identical with that here given, but those authors included certain species in the genus without sufficient investigation of their internal characters, which are fundamentally different; it has in fact been the usual custom among American workers, since the publication of Hall and Clarke’s work, to refer all Mississippian terebratuloid shells to the genus Dielasma. In specimens preserved in the condition of internal casts the generic characters of Dielasma are always very obvious, the posi- tion of the crural lamellae, separate from the socket-plates, being indicated by a pair of slits diverging from the beak of the brachial valve; when the transverse muscle-bearing plate is attached along its mesial line a second pair of diverging slits are present between those formed by the crural lamellae, and the finger-like casts of the slender cavities beneath the transverse plate are clearly shown, whether they are actually present or broken off. In speci- mens having the shell preserved the shell substance is frequently translucent enough to show the position of the internal lamellae as dark lines, in which case the genus can be recognized at once, and when the shell is opaque it is usually easy to determine the generic characters by the judicious use of a needle, without injur- ing the specimen as to its external form and characters upon which the various species are differentiated. * Monog. Perm. Foss. England, Pl. 6, Figs. 37, 41 (1850). 2 Brit. Foss Brach., 11, Permian, Pl. 1, Figs. 18, 20. 3 Pal ae ven Ve th app. 203-04: 442 STUART WELLER Genotype.—D. elongata (Schl.). Other species, D. formosa (Hall), D. shumardianum (Miller), D. fernglenensis Weller, D. burlingtonensis White. GIRTYELLA Nn. gen. Description.—Shell terebratuliform. The pedicle valve sinuate, with a large, subcircular or subovate, oblique foramen which encroaches upon the umbo; the brachial valve frequently sinuate and often with a slight median fold in the bottom of the sinus. Internally the dental lamellae are well developed in the pedicle valve. In the brachial valve the socket-plates are joined by a concave hinge-plate which is imperforate at the apex and is sup- ported by a median septum; the inner sides of the dental sockets OD a a Neo Nt NL Fic. 2.—A series of nine cross-sections (23) of the rostral portion of the shell of Girtyella indianensis (Girty), three of which show both the pedicle and brachial valves, the others showing only the structure of the brachial valve. retreat from the margins of the valve anteriorly beyond the point of articulation, and become the bases of the crura which are still joined by the concave hinge-plate and are also supported by lamellae resting against the inner surface of the lateral slopes of the valve. The brachidium short, its free portion apparently being like that of Dielasma and not reaching to the middle of the shell. Remarks.—Members of this genus have commonly been included in the genus Dielasma, but they differ fundamentally from that genus in the presence of a median septum supporting the hinge- plate of the brachial valve, and in the origin of the bases of the crura from the socket-plates. In his description of the species which is selected as the genotype, Girty referred the form to the - genus Harttina on account of the presence of a median septum in MISSISSIPPIAN LOOP-BEARING BRACHIOPODA 443 the brachial valve, but the brachidium of Harttina is elongate, like that of Cryptonella, reaching nearly to the front of the shell, while that of Gzrtyella is short, like the brachidium of Dielasma. Genotype.—G. indianensis (Girty). Other species, G. turgida (Hall), G. brevilobata (Swall.). DIELASMOIDES n. gen. Description.—Shell terebratuliform. Pedicle valve bisinuate toward the front in the genotype, the two depressions separated by a low, broadly rounded mesial elevation; the foramen large, oblique, encroaching wholly upon the umbonal region. Brachial valve with a slight mesial flattening or depression anteriorly in the genotype. Internally the dental lamellae are well developed in the pedicle valve; in the brachial valve the socket-plates are sup- ported at their inner margins by a pair of lamellae which pass obliquely toward the floor of the valve to which they are joined OCOWDQQyuvst Fic. 3.—A series of seven cross-sections (X23) of the rostral portion of the shell of Dielasmoides bisinuata n. sp., in the last three of which only the brachial valve is shown. near the median line; between these lamellae, the outer walls of the valve and the socket-plates, are a pair of cavities narrowly triangular in cross-section which expand anteriorly and open out into the general cavity of the valve; the crura originate from the anterior extensions of the inner walls of the socket-plates. Form of the brachidium not known. Remarks.—The characters of the rostral cavity of the brachial valve in this genus differ from those of Dzelasma in the absence of any special crural lamellae distinct from the socket-plates. The two rostral cavities, narrowly triangular in cross-section, have a superficial resemblance in the two genera, but the narrow base of the triangle in this genus is formed by the socket-plate, while in Dielasma it is formed by the basal portion of the crural lamellae, and the special muscle-bearing plate between the bases of the crural lamellae of Dielasma is absent in this genus. This form is 444 | STUART WELLER perhaps to be compared with Girivella as the genus most closely allied to it. If the concave hinge-plate of Girtyella, which is sup- ported by a median septum, be depressed along its median line to such an extent that the concave plate itself rests directly upon the floor of the valve, the median septum being eliminated, then we would have essentially the characters shown in this genus. The bisinuate folding of the anterior portion of the pedicle valve may be a generic character, but this cannot be determined from the single species so far recognized. Genotype.-—D. bisinuata n. sp., St. Louis (?) oolite, Lewis County, Mo. CRANAENA Hall and Clarke Description.—Shell terebratuliform. Pedicle valve with or without a median sinus and with well-developed dental lamellae ©) a, 7, J wy ray We ef Fic. 4.—A series of ten cross-sections (X23) of the rostral portion of the shell of Cranaena iowensis (Calvin), from all but the Ae of which the pedicle valve has been omitted. Gt eS NS NS Fic. 5.—A series of six cross-sections of the rostral portion of the brachial valve of Cranaena sp. undesc., residual chest, Springfield, Mo. of moderate length internally, the foramen large, oblique, and encroaching upon the umbonal portion of the valve, the beak incurved. Brachial valve without median fold, even in those species with a well-defined sinus in the opposite valve, but some- times with a slight mesial depression near the front margin; inter- nally the well-developed socket-plates are connected transversely MISSTSSIPPIAN LOOP-BEARING BRACHIOPODA 445 by a concave hinge-plate which is perforated at the apex of the valve posteriorly; upon the inner or concave surface of the hinge- plate a pair of ridges originate at or near the anterior margin of the perforation and continue anteriorly across the plate, from the front of which they are produced into the crura. These crural ridges divide the hinge-plate into three equal divisions, or into two equal lateral divisions and a broader central one, and in some species the crural ridges are accompanied by similar ridgelike thickenings upon the opposite side of the hinge-plate. The brachidium short and Dielasma-like, not reaching to the middle of the valve. Remarks.—This genus differs fundamentally from Dvzelasma in the origin of the crura from the thickened crural ridges of the hinge-plate, rather than from crural lamellae resting upon the inner surface of the valve, and in the absence of a special plate in the brachial valve for muscular attachment, the muscles being — attached directly to the inner surface of the valve. From Girty- ella the genus differs in the perforation of the hinge-plate, in the absence of a median septum in the brachial valve, and in the origin of the crura from crural ridges of the hinge-plate rather than from the anterior extension of the inner extremities of the socket- plates. Genotype—-C. romingerit (Hall). Other species, C. iowensis (Calvin), C. n. sp., residual chert, Springfield, Mo. HAMBURGIA N. gen. Description.—Shell terebratuliform. Pedicle valve not sinuate in the genotype and only known species, the foramen large, oblique, encroaching upon the umbonal region; brachial valve without fold or sinus. Internally the pedicle valve is supplied with well- developed dental lamellae; the brachial valve with well-developed socket-plates which retreat from the lateral margins of the valve anteriorly beyond the articulation of the valves; they are con- nected transversely by a deeply concave hinge-plate which is separated from the inner surface of the valve by an exceedingly low and broad cavity; upon the inner or concave side of the hinge-plate a pair of ridges originate toward the apex and diverge 446 STUART WELLER slightly while becoming stronger anteriorly, finally passing into the bases of the crura; shortly in front of the point of origin of the crural ridges on the hinge-plate the socket-plates are rapidly reduced in height and soon become obsolete, beyond which point the hinge-plate is not connected with the inner surface of the valve, but becomes a concave plate joining the bases of the crura and terminating anteriorly in a short distance. The complete form of the brachidium is not known, but it is probably short, not reaching the mid-length of the valve. Remarks.—This genus is perhaps most closely allied to Cranaena, from which it differs in the extreme concavity of the hinge-plate, the cavity between it and the inner surface of the valve being reduced in height, and in the absence of the perforation of the VW NY. Fic. 6.—A series of nine cross-sections (X 23) of the rostral portion of the brachial valve of Hamburgia typa, n. sp. hinge-plate at the apex, which is, perhaps, the most diagnostic character. The genus is totally distinct from Dielasma, in which the crural plates originate as ridges upon the inner surface of the valve instead of upon the concave surface of the hinge-plate. The concave, transverse plate between the bases of the crura is somewhat similar in the two genera except that it is not connected along its median line to the inner surface of the valve in Hamburgia, but in Dielasma the inner surface of this plate furnishes attachment for the adductor muscles, which is apparently not true in Hamburgia. Genotype—H. typa, n. sp., Hamburg oolitic limestone of Kinderhood age, Hamburg, III. DIELASMELLA N. gen. Description.—Shell terebratuliform, compressed. Pedicle valve with well-developed dental lamellae of moderate length. Brachial valve without median septum or true hinge-plate, the socket- MISSISSIPPIAN LOOP-BEARING BRACHIOPODA 447 plates well developed, retreating from the lateral margins of the valve anteriorly and becoming differentiated into two portions, a basal portion which joins the inner surface of the valve and is directed obliquely inward, and a distal portion which is abruptly bent in a subgeniculate angle so as to be directed obliquely out- ward; the portion included in the angular bend of the two plates is produced anteriorly into the bases of the crura, and just before the crura become free a narrow transverse band joins their bases. The characters of the brachidium not completely determined, but it is believed to be of the short, Dzelasma-like type. Shell structure finely punctate. Remarks.—In the arrangement of the internal features of the apical portion of the brachial valve this genus is perhaps more closely allied to Cranaena than to any other of the generic types here recognized. It differs from Cranaena chiefly in the reduction a w we SE es male ba rs i z = “SS” Fic. 7.—A series of six cross-sections (24) of the rostral portion of the brachial valve of D. compressa (Weller). of the hinge-plate to a narrow, transverse band joining the crural bases, while in Cranaena it is elongate, with a comparatively small apical perforation, and with the crura originating as a pair ot ribs diverging anteriorly from near the apex. The difference in shape, viz., the compressed shell and the erect beak of the pedicle valve, are other features which easily separate the members of this genus from all species of Cranaena which have been recognized. Genotype.—D. compressa (Weller). Glen Park limestone. ROWLEYELLA n. gen. Description.—Shell terebratuliform, with the valves subequally convex. The beak of the pedicle valve perforated by a subcir- cular foramen which encroaches wholly upon the umbonal region, the delthyrium broadly triangular and wholly closed. Internally each valve is supplied with a strong median septum which, in the pedicle valve, reaches nearly to the center of the valve, that of the brachial valve being somewhat shorter. 448 - STUART WELLER Remarks.—The relationships of this genus cannot be certainly determined from the material available for study. The general contour of the shell at once suggests its affinity with the tere- bratuloid, loop-bearing shells, as also does the character of the foramen of the pedicle valve, and so far as it can be observed the delthyrium and its covering. The shell structure has not been certainly determined. Upon one example a punctate structure is slightly suggested, but that characteristic structure of the tere- bratuloids cannot be said to be demonstrated. The characteristic of these shells which is most foreign to the terebratuloids is the strongly developed mesial septum of the pedicle valve, which evidently supports a well-developed spondylium, and the presence of this character in association with a strong median septum in the brachial valve suggests the family Pentameridae, but it has not been determined that a cruralium accompanies this median septum of the brachial valve. This feature of a median septum in the pedicle valve has not been recognized heretofore among the loop-bearing terebratuloids, although there is perhaps no reason why it should not exist, but a median septum in the brachial valve occurs in at least one genus of these shells. In any event the characters exhibited by these shells exclude them from any described genus among either the terebratuloids or the pentameroids. The presence of a pedicle median septum is sufficient to differentiate the genus from any other of terebratu- loids, if indeed it be one of these shells, and the characters of the foramen and delthyrium differentiate it from any pentameroid. The shell perhaps agrees most closely in the sum of all the char- acters present with the spire-bearing Camarophorella, but there is no evidence, in the specimens observed, of the presence of a brachial platform associated with the median septum of that valve, as is true in Camaro phorella. Genotype.—R. fabulites (Rowley), Burlington white chert, Louisiana, Mo. PHYSIOGRAPHIC STUDIES IN THE SAN JUAN DISTRICT OF COLORADO" WALLACE W. ATWOOD The University of Chicago The studies during the past field season were carried on near the southern and southwestern margin of the San Juan Mountains and over the adjoining plateau district. Investigations were planned for the purpose of working out the complete physiographic history of the district. The courses of the Pleistocene glaciers were indicated on the maps and the deposits left by those glaciers differentiated. In connection with these studies it was possible to differentiate the moraines of two distinct glacial epochs in each of the large canyons examined. Beyond the terminal moraines of each epoch and extending for many miles down stream, terrace remnants of valley trains were recognized. It was evident from the position of the younger glacial moraines and younger outwash valley trains that there had been a notable amount of valley deepening in hard rocks during the interglacial epoch. This suggests that the mountain area had been elevated by at least several hundred feet relative to sea-level during the Pleistocene period. The glacial features on the south slope of the range did not differ from glacial features which have been fully described by various writers who have become familiar with glacial phe- nomena in the high mountains of the West. In examining the areas which rose above the upper limit of ice action on the south slopes of the mountains, certain gravel-strewn surfaces were found. The gravels were beautifully polished and of very resistant material. They were composed chiefly of quartzite, quartz, red jasper, flint, cherts, and greenstones. Much of the material was less than half an inch in diameter, but some of the pebbles ranged between one and two inches in their longer t Published with the permission of the Director of the U.S. Geological Survey. 449 450 WALLACE W. ATWOOD axes. The surfaces on which these gravels were found were along the crests between the great mountain canyons and on the tops of mesa-like hills near the base of the range. If the gravel-strewn surfaces were extended they would unite and form a plain of gently rolling topography. That plain would slope away from the core of the range, show a distinct warping at the base of the range, and pass off over the upland surfaces of neighboring plateaus. The nature and distribution of the gravels suggested that they were remnants of stream deposits in channels which formerly crossed the present inter-canyon ridges. They appeared to be the deposits of streams which flowed over low gradients and sug- gested further, by their distribution and the distribution and relations of the surfaces on which they were found, a deformed peneplain. In following this ancient erosion surface southward _ and southwestward over the plateau district it seemed that certain of the outlying mesa surfaces would correspond in age to this peneplain surface, and it was anticipated that on such out- lying surfaces a mantle or scattering of still finer gravels might be found. But these mesa surfaces were found to carry a heavy mantle of bowlder-gravels, in which the larger masses ranged up to three and four feet in diameter. In these bowlder-gravel de- posits certain special bowlders could be recognized which came from outcrops in the mountain areas, and it appeared that they must have been washed out and deposited as a portion of a great alluvial fan about the margin of the mountains. These bowlder- strewn surfaces were followed southward nearly fifty miles from the base of the range, and at that distance the larger bowlders seen ranged to at least three feet in diameter. The interpretation of this bowlder-gravel mantle and its relation to the erosion surfaces upon which it rests and the erosion surfaces in the mountains which seem to correspond in age to those underneath the bowlder-gravels is that at the close of the cycle of erosion during which the pene- plain as described was developed, there was a general uplift in the district, which was emphasized in the San Juan dome. The head- waters of the streams in the uplifted dome were so rejuvenated that they carried together with sands and gravels many large bowlders to the base of the range and spread that material out as PHVSIOGRAPHIC STUDIES IN SAN JUAN DISTRICT 451 alluvial fans over the neighboring plateaus. The streams which crossed the plateaus were first rejuvenated in their lower courses, and as rejuvenation worked upstream across the broad plateaus the growth of the great alluvial fans ceased and their dissection began. The high-level bowlder-gravels are, therefore,.a deposit which marks the beginning of a new period of erosion in the moun- tains and a temporary period of alluviation about the base of the mountains. The bowlder-gravels are, therefore, not of exactly the same age, but a little younger than the small peneplain gravels of the mountain area. Below the summit elevations in the mountains and in the neighboring plateaus there are other broad bowlder-capped mesa- like forms which appear to represent the base to which the streams worked when the peneplain was first deformed. The bowlder capping on these mesas is at places as much as thirty feet in thick- ness. The Florida and Fort Lewis mesas just south of the San. Juan Mountains are typical of this bowlder-mesa stage in the dissection of the area. Another uplift associated with the more or less continuous growth of the mountains deformed the graded surfaces of the bowlder-mesa stage, again rejuvenated the streams, and opened another cycle of erosion. The surfaces to which the streams then worked are represented by broad, open valleys of late maturity or early old age in the softer rocks, and by canyons in the harder rocks. This cycle of erosion has been for convenience referred to as the Oxford stage, for there is an excellent development of the typical lowlands of this stage near the village of Oxford, a few miles southeast of Durango. It immediately preceded the first epoch of glaciation as recorded by the moraines and outwash deposits found on the south slope of the range. In the mountain-canyons the bowlder-mesa and Oxford stages are both represented by rock benches which in some instances carry stream alluvium. These studies have opened certain large problems in the rela- tionship of the mountains to the plateaus, and suggested a close correlation in the physiographic histories of the two provinces. Are the high-level bowlder-gravels resting on a true peneplain ? Of what age is this peneplain? In the area examined during the 452 WALLACE W. ATWOOD past season it is known to truncate Wasatch beds. Is it not, however, as late as late Miocene? Does it correspond in age to the great peneplains of the Grand Canyon district?" Are the high-level bowlder-gravels bordering the Front Range of the Rocky Mountains, the Big Horn Mountains,? the Livingston Range? of the same age as these about the San Juan Mountains ? Do they rest on peneplain surfaces? Is the Blackfoot Peneplain of Montana described by Bailey Willis‘ of the same age as the one observed in this region? Are the Wyoming conglomerates about the base of the Uinta Mountains, placed in the Pliocene by the King Survey,’ of the same origin and age as the high-level bowlder-gravels south of the San Juans? What is the relation- ship of certain bowlder deposits found near the summit and near the core of the San Juans, and certain ancient stream gravels which have been found by Stone, near the summit of the Front Range, to the bowlder-gravels about the bases of these ranges? Numer- ous other correlations in the Rocky Mountain areas and in the Pacific Coast mountains are suggested. Has there been with each period of mountain growth in the Cordilleran region of North America a rejuvenation of the streams which affected the headwaters long before it affected the middle courses of the rivers, and did these rejuvenated headwaters distribute the bowlder-gravels in each case on the neighboring pla- teaus? How far were such deposits carried and how were the larger bowlders transported? Numerous cases may be cited where bowlders ten to twelve feet in diameter have traveled at least twenty-five miles from their sources over surfaces of very low gradient. If such bowlders can travel twenty-five miles on gently sloping surfaces, is it not possible for them to travel much farther than that over low gradients? How are huge bowlders transported over nearly horizontal surfaces? How far have climatic changes affected the work of streams in the Rocky Moun- tain region? Are the reported glacial deposits in southeastern *H. H. Robinson, Am. Jour. Sci., 4th Series (1907), XXIV, 109-29. 2 Salisbury and Blackwelder, Jour. Geol., II (1903), 220-23. 3 Bailey Willis, Bull. Geol. Soc. Am., XIII (1902), 329-30. 4 Op. cit., 310. 5 Hague and Emmons, Rep. of goth Parallel Survey (1877), II, 64-65, 188-89 ff. PHYVSIOGRAPHIC STUDIES IN SAN JUAN DISTRICT 453 Utah," in which there are granitic and gneissic bowlders one to five feet in diameter, the origin of which is at present unknown unless it be the San Juan Mountains, true glacial deposits or stream deposits? Could mountain glaciers from the San Juan Range have reached southwestward one hundred miles from the base of the range? Is there not some other explanation for the coarse bowlder deposits reported in that portion of the plateau district ?? Has there been a continuous or periodic growth of the San Juan dome during late Tertiary and Quaternary times ?3 How are the great systems of fissures which cut the late Tertiary volcanics . related in age to the recent deformative movements? Co-operative work by all who are engaged in field studies in the Rocky Mountains and plateau provinces should prove of great value in promoting the solution of these problems. ™D. D. Sterrett, U.S. Geol. Survey, Mineral Resources (1908), Pt. II, 825 (1909). 2.W. M. Davis, Proc. Am. Acad. Aris and Sci., XX XV (1900), 345-73. 3 Cross and Spencer, U.S. Geol. Survey 21st Ann. Rep. (1900), Pt. II, 100. THE VARIATIONS OF GLACIERS. XVE HARRY FIELDING REID Johns Hopkins University The following is a summary of the Fifteenth Annual Report of the International Committee on Glaciers.’ REPORT OF GLACIERS FOR 1909 Swiss Alps.—Of the ninety glaciers measured in 1909, only two . have been advancing for three successive years, the Scex-rouge and the lower Grindelwald Glacier; the latter has advanced 59 meters in two years. Nine other glaciers have advanced slightly during the last year but it is not certain that they are in a stage of advance. A general retreat is dominant in the Swiss Alps. Eastern Alps.—Thirty-nine glaciers were measured, and the retreat is general, although in many cases it is slow. The Lang- taler Glacier and the Grosselendkees seem to be stationary, and the Mitterkarferner has made a small advance.‘ Italian Alps.—The retreat, which has been general for some years, seems to be continuing without change.® French Alps.—Observations on the snow-fall and the variation in the length of glaciers have continued, and maps of some glaciers are being made on a scale of 1 : 10,000. In the Mont Blanc range the retreat is nearly general, though slight; the Glacier des Bossons has advanced a little more than one meter. The ends of the glaciers have in general diminished in thickness with a corre- sponding diminution in the velocity of flow. In the Tarentaise and Maurienne the retreat is also general, but feeble. In the Dauphiné we find that the snow-fall has been distinctly heavier since 1906 with the result that the glaciers on the northern side of the Pelvoux massif have grown thicker and are beginning to advance; and the 1 The earlier reports appeared in the Journal of Geology, Vols. III-XIX. 2 Zeitschrift fiir Gletscherkunde (1911), V, 177-202. 3 Report of Professor Forel and M. Muret. 4 Report of Professor Briickner. 5 Report of Professor Marinelli. 454 THE VARIATIONS OF GLACIERS 455 whole appearance of the reservoirs indicates that the advance in this region will become general. In the southern part of the same massif, although the increase in the snow-fall is also marked, the retreat of the glaciers continues. In the Pyrenees the snow-fall has been heavy since 1906 and the small glaciers show a marked tendency to increase in size.’ Swedish Alps.—A number of glaciers have been measured and the changes resulting during a variable number of years up to 1909 indicate a slight increase in general, though a few of the glaciers are apparently stationary and one or two slightly retreating? _ Norwegian Alps.—The difference in the behavior of the glaciers of Jotunheim, in the interior of Norway, and those of Folgefon and Jostedalsbraé, near the coast, which was commented on in the last report, continues. The glaciers near the coast are generally advancing, whereas those in the interior are generally retreating. Russia.—The greater part of the glaciers observed in the Cau- casus between 1899 and 1907 have retreated. A few have remained stationary and only two have made an advance. A number of glaciers in the Altai and Muss-tau mountains, in Siberia, have been visited but no careful measurements made.* REPORT ON THE GLACIERS IN THE UNITED STATES FOR IgIo§ There are a number of small glaciers in Colorado which, on the whole, show a tendency to become smaller, but their variations from year to year are extremely slight.° The Hallett Glacier shows no measurable change since 1909 (Mills). The Carbon Glacier on the northern side of Mt. Rainier is in marked recession (Matthes). The United States Geological Survey has been continuing its * Report of M. Rabot. 3 Report of M. Oyen. 2 Report of Professor Hamburg. 4 Report of Colonel Schokalsky. 5 A synopsis of this report will appear in the Sixteenth Annual Report of the Inter- national Committee. The report on the glaciers of the United States for the year 1909 was given in this Journal (XIX, 83-809). 6 All available information regarding these glaciers has been collected by Judge Junius Henderson, ‘‘Extinct and Existing Glaciers of Colorado,” University of Colo- rado Studies (1910), VIII, 33-76. 456 HARRY FIELDING REID explorations and surveys of Alaska, and several glacier regions have been mapped. A number of glaciers north of Juneau are in rapid recession. The glaciers in the neighborhood and north of the head- waters of the Copper River (in the neighborhood of latitude 633° N and longitude 145° W) seem to be retreating slowly (Brooks). Fresh moraine, extending for nearly two miles at the end of Nabesna Glacier, shows that it is retreating rather rapidly. The Chisana Glacier, 15 miles to the east, has a very clean end; a com- parison of photographs taken in 1899 and 1908 shows surprisingly little change in the aspect of the glacier, though at one place a slight recession has taken place. Frederika Glacier, entering White River valley from the north, when seen by Dr. C. W. Hayes in 1891 ended ‘‘in a nearly vertical ice cliff... . about 250 feet high. At the foot of the cliff there is a small accumulation of gravel and ice fragments apparently being pushed along by the advancing mass.”* In 1909 Mr. Stephen R. Capps says “its surface is remarkably smooth and slopes down evenly to a thin edge in front.” The Frederika Glacier has evidently changed from an advancing to a retreating glacier in the interval. Exactly opposite Frederika Glacier another glacier, in retreat in 1891, is now advancing.” Such spasmodic cases are probably produced by sudden acces- sion of material due to avalanches or land slides, rather than to simple variations in snow-fall or temperature. The Ruth Glacier, rising on Mt. McKinley and extending many miles to the east, is slowly diminishing in size (Rusk). Professor U. S. Grant and Mr. D. F. Higgins have published a general map of Prince William Sound, showing the location of all the glaciers, on a scale of 4 inches to the miles They have also published descriptions, pictures, and detailed maps of the ends of several of these glaciers.4 The last observations were made in « “Expedition through the Yukon District,” Nat. Geog. Mag. (1892), IV, 153. 2 Glaciation on the North Side of the Wrangell Mountains, Alaska,” Jour. Geol. (1910), XVIII, 56. 3 “Reconnaissance of the Geology and Mineral Resources of Prince William Sound, Alaska,” U.S. Geological Survey, Bulletin No. 443, Washington, 1910. 4 “Glaciers of Prince William Sound and the Southern Part of the Kenai Peninsula, Alaska,”’ Bull. Amer. Geog. Soc. (1910), XLII, 721-33. THE VARIATIONS OF GLACIERS 457 the summer of 1909 and we note that the Shoup Glacier was practically stationary and was fully as large then as it had been for several decades. The Columbia Glacier was found, at its eastern edge, about 500 feet in advance of its position of 1899, and this advance seems to have taken place principally since the summer of 1908. Professor Grant found indications that the glacier was well in advance some fifty years ago and before that date was considerably smaller. The Meares Glacier seems to be a little in advance of its position of 1905, and the general condition of the vegetation in the immediate neighborhood indicated that the glacier in 1909 was probably as far forward as it has been during the last one hundred years or more. Professor Lawrence Martin conducted another expedition for the National Geographic Society in 1910 to study the Alaskan glaciers. He sends me the following notes: Fairweather Range.—La Perouse Glacier advanced approximately a quarter mile between September 4, 1909, and June 10, 1910, and was destroying forest on the latter date, as it had previously done in September, 1895. Yakutat Bay.—Nunatak Glacier advanced 700 to 1,000 feet between July 6, 1909, and June 17, roro0, after retreating steadily at least 24 miles from 1890 to 1909. Hubbard Glacier did not continue to advance as rapidly as seemed possibly would be the case in 1909, parts of the front advancing 600 feet between 1909 and rgto while other parts retreated 500 to 1,000 feet. In 1910 Lucia Glacier had probably nearly ceased the great advance which was in progress in July, 1909. Nunatak Glacier is the ninth ice tongue in the Yakutat Bay region to advance since 1899, following a long period of contin- uous retreat or stagnation. In each case listed below the advance is thought to be the result of great accessions of snow and ice by avalanches during the earthquakes of September, 1899. Glacier Date of Advance Length of Glacier Galiamomnncncynaren cnn toca, After 1895 and before 1905 2o0r 3 miles Unnamed Glacier...:...... IQOI 3 or 4 miles Elaenke rrr ae 1905-6 6 or 7 miles INET OVA Ayatsen easy fone ue aie fy 1905-6 8 miles Nariegatedia waeyaccaccens). 1905-6 Io miles IMR VAM ER Meee asides eee le es 1905-6 to miles} lid demvgatse eee eee eas nt 1g06 or 1907 16 or 17 miles NSW Ciaanes Fc eee tami nes oie. cie 1909 17 or 18 miles INjumattalke ee acwemutusas: arorduevacacte IgIo 20 miles * Between Haenke and Hubbard glaciers. { Excluding expanded lobe in Malaspina. 458 HARRY FIELDING REID Prince William Sound.—Columbia Glacier advanced 600 feet between August 24, 1909, and July 4, 1910, and at least 132 feet more between the latter date and September 5, 1910. In College Fiord the Harvard, Yale, Radcliffe, Smith, Bryn Mawr, Vassar, Wellesley, and Barnard glaciers were advancing much more actively in toro than in 1909, and were destroying forest at their borders, as were the Meares Glacier in Unakwik Inlet, the Harriman, Baker, Roaring, and Cataract glaciers in Harriman Fiord, and the Blackstone Gla- cier in Blackstone Bay. Harvard glacier had advanced 100 to 150 feet, Yale 750 feet, and Harriman 300 feet between 1899 and ro10. Barry and Surprise glaciers in Harriman Fiord retreated 23 and 14 miles respectively from 1899 to 1910, different parts of the Barry retreating 500 to 1,600 feet of this distance between 1909 and ro10. Valdez and Shoup glaciers in eastern Prince William Sound and Nellie Juan Glacier in Port Nellie Juan remained unchanged from 1908 to 1910, as did Chenega, Princeton, and Tiger glaciers in Icy Bay, where there was a six or seven mile retreat between 1787 and 1908, most of it later than 1898. Portage Glacier in Passage Canal had a great advance between 1794 and 1880, filling a pass from Prince William Sound to Cook Inlet to a height of over 1,000 feet, where there was previously a low canoe portage and no glacier. Copper River.—Miles Glacier retreated about 1,700 feet from 1900 to 1906 and readvanced 1,800 feet from 1906 to 1910. Grinnell Glacier advanced slightly between 1909 and 1or1o. Different parts of the front of Childs Gla- cier advanced 920 to 1,225 feet between 1909 and June, roro, in midglacier, where the front is undercut by Copper River. On the north bank of the river where the margin of the glacier ends on the land and was stagnant in 1909, it advanced 1,500 to 1,600 feet up to June 10, 1910, and 204 feet more up to October s, 1910. The glacier front developed lobes so that some parts advanced faster than others. The rates per day through the summer of 1910 were as follows: ; ADVANCE IN METERS | RATE PER Day IN FEET DaTES Days Fastest Average Fastest Average une! to tom ius 2 Oesaeaceer ect 49 124 116 Ds Does) allio ORE OPAIOR Orme ae eaiater 8 26 23 3.25 2.87 INGLES AMO VaGOKeas AIC g's g cae S 5 41 8 See 1.60 Age Dt COPA RET einen cane 6 27 4 4.5 0.66 Aug 17, tov AUS 20m cries 12 42 ae) Boi 1.58 INiNer BAOV AHO) SONS Cla sono casas 21 37 27 1.76 1.28 Sept. 19 to Oct: 55... .: rata Hee 17 118} rE} 0.7 0.44 In midglacier there was a relative retreat of the advancing ice front from June to September, while the north border continued to advance strongly, as shown above. ‘This retreat was due to undercuttings during the summer THE VARIATIONS OF GLACIERS 459 rise of the river and was followed by a strong advance in late September and early October when the level of the river fell and its undercutting power was weakened. This oscillation is shown in the accompanying table. Dates Variation of Glacier Stage of River 1909 to June Io, t9to.......| Advance, 920—-1,225 ft. Fall, 14 ft., May 6 to June to iunesToktoy Auge iE). sass: Retreat, 450 ft. Rise, 6 ft. PAN OA mToTe COMA Gee Titres oars aay! Retreat, 65 ft. Level, about stationary ANUS: 057 (WO) === 5 Hla doe aac Retreat continued....... Rises tag ite LOO CE Se une et pies Advance, 390 ft., plus unknown retreat above | Fall, 9 ft. As this advance of Childs Glacier seriously threatened a $1,400,000 steel railway bridge which in October, to10, was only 1,575 feet from the north margin of the glacier, the behavior of Childs Glacier during the winter of 1910- r1, when Copper River is low and weak, will be of much interest. The diminu- tion of movement on the north bank suggests, however, that the advance is practically over. The advance of Grinnell Glacier is also of interest, for this ice tongue occupies a strategic position with reference to the railway, which traverses its stagnant outer portion. In this portion, however, there was no disturbance in 1gio, the advance affecting another part of the glacier. The Allen Glacier, whose stagnant outer portion is traversed for 54 miles by this railway, remained unchanged from 1909 to 1910. It appears, therefore, that the glaciers about Prince William Sound give some indication of a general, but not very large, advance. CoRRECTION.—In the Report on the Glaciers of the United States for 1908 (Journal of Geology, XVII, 671), the name ‘‘Matamaka Glacier”? should have been “ Matanuska Glacier.” The regular meeting of the International Committee on Glaciers took place in Stockholm on the 2oth of August, tgto, in connection with the Eleventh International Geological Congress.’ The retir- ing president, Professor Edouard Briickner, presented the report of the Committee to the Congress. He called attention to the origin of the Committee, which was first appointed by the Sixth International Geological Congress in 1894, and has been collecting information regarding the variations of glaciers ever since; and emphasized the importance of the work « “Ta Commission internationale des Glaciers au Congrés géologique international, Stockholm, aott, 1910.” Zeitschrift fiir Gletscherkunde (1911), V, pp. 161-76. 460 HARRY FIELDING REID of Professor Finsterwalder, who, as retiring president in 1903, laid down the fundamentals of a mathematical theory of glacier varia- tions. He then reviewed the information collected regarding the variations of glaciers. In the Alps the retreat of the glaciers con- tinues steadily, although a few glaciers of the Oetztal and some others have made small temporary advances. The retreat has lasted for several decades. A graphic representation of the varia- tions of 26 glaciers in the Swiss Alps, including the Mont Blanc group, shows that, for the greater number of them, the retreat has lasted since the beginning of the nineteenth century, and that the advance which occurred about 1850 was but an episode in the general retreat. Since that time the retreat has been still more marked. In the Scandinavian Alps the variations have been some- what different; in this region an advance occurred in the beginning of the twentieth century. It began in the Jostedalsbra and the Folgefon and progressed toward the north; but the advance was confined to the coast region and the glaciers in the mountains of central Scandinavia did not participate in it. This advance must also be looked upon as an unimportant event in the general retreat. The glaciers of the Caucasus, of the Tyan-Shan, the Altai, the Highlands of Pamir, and the Himalaya are in retreat, though here also special cases of advance have been noted. Among the glaciers of the United States and Canada the retreat is general and this is true to a still more marked extent in Alaska. Between 1892 and 1907, the retreat of the glaciers has increased the area of Glacier Bay by 19 square miles. Of great interest are the sudden remarkable advances of the glaciers of Yakutat Bay, which Pro- fessor R. S. Tarr has described and imputed to the great increase in snow-supply due to avalanches incited by the earthquakes of 1899. Professor Hauthal has described the rapid advance of the Bismarck Glacier, in South America, since the end of the last century. Professor Kilian described the variations of glaciers in France. _ The importance of the water from glacial streams has led the Min- ister of Agriculture to give material aid to the observations of glaciers. Professor Dr. F. W. Svenonius spoke of the difficulty of making THE VARIATIONS OF GLACIERS 461 observations among the Swedish glaciers on account of their distance and inaccessibility, but nevertheless the Swedish Geological Sur- vey has published a collection of six essays by the leading Swedish glacialists which give an excellent account of what is known of these glaciers." t Dr. Axel Hamberg has been elected an ordinary member of the Committee to represent Sweden, succeeding Dr. F. W. Svenonius, who has retired and been elected a corresponding member. Professor R. S. Tarr has also been elected a corresponding member. Two corresponding members, Mr. W.S. Vaux, Jr., and Professor E. Hagen- bach-Bischoff, have died since the 1906 meeting of the Committee. The following officers were elected to serve until the next meeting of the International Congress of Geologists: Honorary President, Prince Roland Bonaparte of Paris; Active Presi- dent, M. Charles Rabot of Paris; Secretary, M. Ernest Muret of Lausanne. PETROLOGICAL ABSTERACES! AND TR EVILYES EDITED BY ALBERT JOHANNSEN ARSCHINOW, WLADiImIR. ‘‘Ueber die Verwendung einer Glashalb- kugel zu quantitativen optischen Untersuchungen am Polari- sationsmikroskope,”’ Zeitschr. Kryst., XLVIII (1910), 225-29. Fig. i. A simple apparatus for making quantitative measurements by tilting a thin section under the microscope. The author claims to be able to make measurements with as great a degree of accuracy as may be made with Fedorow’s or Klein’s Universaltisch. The instrument consists of a glass hemisphere, 50-60 mm. in diameter, which is centered upon the stage of the microscope and rotated by hand. The section is fastened to the flat side of the hemisphere with cedar oil or glycerin, and with the cover-glass down. The determination of planes of extinction and so on are made as with the Fedorow Univer- saltisch, and the angle of rotation is measured by means of two grad- uated metal strips, attached 90° apart, to a movable ring around the equator of the glass hemisphere, and themselves capable of being moved on pivots. By raising the tube of the microscope above these rings, the angles at which they cross may be read, and this determines the amount of rotation of the glass hemisphere. For certain measurements a small glass hemisphere, 8-15 mm. in diameter, is attached to the upper sur- face of the slide. ALBERT JOHANNSEN Bastin, Epson S. ‘‘Geology of the Pegmatites and Associated Rocks of Maine,” Bull. U.S. Geol. Survey No. 445, Washington, TOM. | Ro s52, platoiss. co eimape Te In this bulletin on the pegmatites of Maine, Doctor Bastin has given not only local descriptions but has made an important contribu- tion to the general literature of the pegmatites as well. The work is divided, practically, into three parts: a general discussion of pegmatites 462 PETROLOGICAL ABSTRACTS AND REVIEWS 463 and in particular those of Maine, local descriptions by counties, and descriptions of the economically important minerals. Granite-pegmatites are defined here as differing but little from the granites of the state in mineral composition, but are characterized, not necessarily by coarse, but by extreme irregularity of grain. They occur in dikes or sill-like masses, generally of sheet-like form and some- times of considerable size. The contact with the country rock is gen- erally sharp, indicating very little assimilation by the pegmatite even where it is of batholithic dimensions. Contact metamorphism around the pegmatites is no greater than that near granite contacts, and indi- cates, according to the writer, that the amount of mineralizers present was but little greater than in the latter rocks; less than ten times as great, probably. Genetically the pegmatites are related to the asso- ciated granites and are probably contemporaneous with them. Where particularly abundant, they form, apparently, the roofs above granite batholiths. An examination of the quartz grains indicates, in the coarser varieties, that the crystallization began slightly above 575° C. and ended at a lower pom EUL Gs The finer-grained varieties may have crystallized entirely above 575°. Among the minerals of economic importance found in the Maine pegmatites are the feldspars, orthoclase and microcline rose and smoky quartz, amethyst, muscovite, tourmaline, beryl of various colors, and topaz. The occurrences, compositions, properties, and uses of these minerals are discussed. ALBERT JOHANNSEN CovuyatT, J. ‘‘Les roches sodiques du désert arabique,” Compies Rendus de V Académie des Sciences, CLI (1910), 1138-41. In a region east of the Nile, near longitude 34° 18’ E., latitude 24° 40’ N., there are dikes and stocks of nepheline syenite with much varia- tion in texture, also tinguaite and sélvsbergite. Four analyses of the syenite show SiO, 60.1 to 56.5 per cent; Na.O g.0 to 10.6 per cent; KOA. se toes-2per cent: Fe,O,;, FeO 3.2 to 6.1 per cent; very low MgO and CaO. In the quantitative classification the rocks are mias- koses and laurdaloses. The syenites are related to volcanic eruptives of Cretaceous age and posterior to a series of trachytes, andesites, and basalts. F.-C. CALKINS 464 PETROLOGICAL ABSTRACTS AND REVIEWS Duparc, L., AND Pampuit, G. “Sur Vissite, une nouvelle roche filonienne dans la dunite,’’ Comptes Rendus de l Académie des Sciences, CLI (1910), 1136-38. The rocks described form dikes in massive dunites of the platinum deposits in the basin of the river Iss. They consist mainly of horn- blende, with subordinate pyroxene, labradorite in some cases, magnetite, and apatite. Five. analyses are given. SiO, ranges approximately from 33 to 47 per cent; Fe.O, from 3 to 9 per cent; FeO from 14 to 9 per cent; CaO from 16 to 11 per cent; MgO from to to 7 per cent; total alkalies 2 to 3 per cent. In the quantitative classification, the rocks fall in auvergnose and three unnamed subrangs. F. C. CALKINS Duparc, L., AND WuNDER, M. “Sur les Serpentines du Krebet- Salatim (Oural du Nord),” Comptes Rendus de l’ Académie des Sciences, CLII (1911), 883-85. Describes dunites and harzburgites more or less completely altered to antigorite and bastite. Five analyses. of these rocks and one of the inclosed calcareous hornfels are given. F. C. CALKINS GRANDJEAN, F. ‘‘Sur un mesure du laminage des sédiments (calcaires et schistes) par celui de leurs cristaux clastiques de tourmaline,’ Comptes Rendus de l’ Académie des Sciences, CLI (1910), 907-9. The author finds tourmaline an unfailing constituent of shales and limestones. The crystals of this mineral in deformed rocks show a middle portion normal in color and apparently undeformed, and ragged terminal portions of paler hue. The terminal zones are considered due to elongation of the tourmaline, and their average length (about 30 measurements is ordinarily sufficient) gives a co-efficient of deforma- tion for the rock. F. C. CALKINS GRoTH-JACKSON. The Optical Properties of Crystals. New York: John Wiley & Sons, 1910. Pp. xiv+309, figs. 121, colored plates 2. In spite of a number of good books on optical crystallography which have appeared within the past few years, no work has quite taken the PETROLOGICAL ABSTRACTS AND REVIEWS 465 place of Groth’s classical Physikalische Krystallographie, and it is with great pleasure that this translation of certain parts is welcomed. The only criticism that can be made is that Professor Jackson did not trans- late the entire work. The translation, in general, follows the form of the original and includes all of the ‘Optical Properties” in Part I, with additions, here and there, from the parts in the original devoted to systematic descrip- tions of crystals and methods of crystal investigation. The translation seems to be good, although, in places, the sentences, closely following the German, are rather long. A slight error is introduced, on p. 15, where the number of vibrations per second of red and violet light are spoken of, by the translation of the German billion (10%) as billion (10°). The book is well gotten up, and the line drawings, apparently from wax plates, are sharp and clear. ALBERT JOHANNSEN Howe, J. ALLEN. The Geology of Building Stones. New York: Longmans, Green & Co.; London: Edward Arnold, 1g9ro. 12mo, pp. vilit+-455, pl. 8, maps 7, figs. 31. This work, the fourth of Arnold’s Geological Series, under the general editorship of Dr. J. E. Marr, apparently is intended primarily for architects. It treats of the rock-forming minerals and the rocks in non-technical language and gives the principal properties of each. The decay of building stone is discussed, and methods of testing are described. The author says, ‘There is no help: sooner or later, in the course of practice, the architect or engineer will have the need of some geological knowledge forced upon him.” If the little knowledge is not a dangerous thing, this book may serve a useful purpose. ALBERT JOHANNSEN Lacroix, A. “Le cortége filonien des péridotites de la Nouvelle Calédonie,’ Comptes Rendus de VAcadémie des Sciences, CLIT (1911), 816-22. The peridotites of Nouvelle Calédonie are cut by narrow dikes form- ing a gabbroic and a dioritic series. In both, gradations can be traced between a leucocratic extreme (anorthosite) and a melanocratic extreme, (pyroxenite or hornblendite). Nine analyses are given which prove that six of the rocks fall into previously unnamed subdivisions of the 466 PETROLOGICAL ABSTRACTS AND REVIEWS quantitative classification. These are: Ouenose (III. 5. 5. 4-5); Caledonose (I. 5. 5. 4=5);) Dhiose” GV. a[2)5 tl2]4* 2) Naketose (QV. 2l3ky aleier2)5 shoghosea(hiny a aie4— 5) In the gabbroic series, Si0,, Al,O;, and CaO increase together, while FeO and MgO decrease. Because of the basicity of the feldspars the most feldspathic phase is the poorest in SiO,._ In the dioritic series, the proportion of lime is nearly constant, silica varies irregularly, Al,O; and alkalies increase with the feldspar content. There are also dikes composed almost wholly of magnesiochromite; these locally contain chrome-bearing diopside and bronzite, and are associated with anorthosites. F. C. CALKINS Leiss, C. ‘Neues Mikroskop Modell VIb fir krystallogra- phische und petrographische Studien,’ Zeztschr. Kryst., XE VAN (toro) 24 42s higeate A large microscope similar in construction to the Hirschwald micro- scope (Fuess VIa). It differs, however, in having an Abbe condenser and Ahrens polarizer, and a large, flat micrometer stage. Like the VIa microscope, the upper and lower nicols can be rotated simultaneously. This does away with the cap nicol and permits the use of a large tube, iving an extra large field. § 5 g ALBERT JOHANNSEN SKEATS, ERNEST W. ‘‘The Volcanic Rocks of Victoria,’ Austra- lian Association for the Advancement of Science, 1909, 173-235. Pl. 4, numerous analyses. This paper was read as the Presidential Address, Section C, of the Australian Association for the Advancement of Science. It contains a summary of the present knowledge of the Victorian volcanic rocks and has appended a bibliography of 268 items, dealing wholly, or in part, with these rocks. The geographical distribution is shown on a map and the geological range was determined to be from Basal Ordovician ( ?) to recent. Petrographically the rocks are rhyolites, dacites, basalts, quartz porphyries, granite porphyries, diabases, serpentines, quartz keratophyres, melaphyres, sdlvsbergites, limburgites, and the new rocks anorthoclase trachyte, anorthoclase-olivine trachyte, olivine- anorthoclase basalt, olivine-anorthoclase andesite, and macedonite. Petrographical descriptions, not 'as complete as might be desired, espe- PETROLOGICAL ABSTRACTS AND REVIEWS 467 cially those dealing with new types, are given, and the geographical and geological relations are shown. The new rock terms proposed are: Anorthoclase trachyte—This type was previously described by Professor Gregory as trachyphonolite. As described by Skeats, in a corrected copy of his paper, it is ‘“‘a dark-greenish rock. Large phenocrysts of anorthoclase are numerous. The ground mass has sometimes a fluidal arrangement of laths of anorthoclase, in other cases the crystals are stouter and the structure orthophyric. Small crystals of aegirine are scattered through the rock, a little green glass, a few sec- tions of nosean, ilmenite, and occasionally apatite are also present.’’ From the description it does not appear that any other feldspar occurs, although the statement, in another place, that “anorthoclase is the dominant felspar,”’ suggests that another is present. Anorthoclase-olivine trachyte-—Spoken of as “more basic than the rock just described.”’ It resembles the former rock but contains, in addition, more or less olivine. Olivine-anorthoclase basalt—‘‘A still less acid type... . . It differs mainly from the last type in the greater abundance of olivine and less frequent anorthoclase.”” In the opinion of the reviewer this description would hardly justify the use of the term basalt. _ Macedonite is a non-porphyritic, basaltic-looking rock and in the annotated copy is said to “consist largely of minute felspars, a colour- less to green interstitial mineral, either glass or chlorite, serpentine or chlorite pseudomorphs after olivine, some light-brown biotite and purplish, fibrous apatite prisms. Octahedra of perofskite occur, some of which are opaque, others of a dark grayish-green colour. The exact relations of this rock are difficult to determine. Chemically it is in some respects intermediate between the tephrites and the orthoclase basalts, but mineralogically it is quite distinct. Its nearest relations are with the mugearites, from which it differs in the ratio of soda to potash and in the small amount of olivine present.” The writer does not say what kind of feldspar is present, but if the analysis is computed in the Quantitative System of C.I.P.W., the norm shows orthoclase, 20.02 per cent, albite, 29.87 per cent, and anorthite, 18.63 per cent. As computed by the reviewer the rock is a Shoshonose. Olivine-anorthoclase andesite—This is a porphyritic, subsiliceous andesite. It contains lath-shaped plagioclase and granular or ophitic augite, magnetite, and olivine as its normal constituents. Corroded phenocrysts of anorthoclase occur and connect this type with the alkali rocks. 468 PETROLOGICAL ABSTRACTS AND REVIEWS While exact and detailed descriptions may seem tedious in an address, it would be desirable in printed descriptions of new types of rocks that they be made as complete as possible and that the relative amounts of the different constituents be stated. For such rocks, clear-cut definitions should be given. The paper is a well-written summary of what is known of the vol- canic rocks of Victoria, and one is always thankful for contributions containing careful analyses and complete bibliographies. ALBERT JOHANNSEN Watson, THomas L. ‘Intermediate (Quartz Monzonitic) Char- acter of the Central and Southern Appalachian Granites,”’ Bull. Phil. Soc., Univ. Va., I (1910), 1-39. By comparing the analyses of granites from different parts of the Appalachian region, the author finds that they are, in general, of mon- zonitic character, the soda being molecularly equal to or greater than potash. Comparing the western granites with the Appalachian rocks, he finds that “the eastern type shows stronger granite affinities and the western type stronger quartz diorite affinities.”’ In general the granites of the eastern region are of similar composition, containing acid oligo- clase and some albite in addition to potash feldspar; the ratio aver- aging 1.88 to 1. All of the granites, from Alabama to New England, as well as the subsilicic gabbros, diabases, pyroxenites, and peridotites, “have been derived from a common parent body of magma intruded, in most cases, at different times,” says the writer. The age of the massive granite is stated to be early or later Paleozoic, while the gran- ite-gneisses (gneissoid-granites) are pre-Cambrian. Numerous analyses, all of them partial, are given. ALBERT JOHANNSEN REVIEWS The Coming of Evolution: the Story of a Great Revolution. By JoHn W. Jupp. London and Edinburgh: The Cambridge University Press, 1910. Pp. 171. The numerous addresses which were delivered in various parts of the world in connection with the recent Darwin Centenary seem to have had for their common burden the revivification of all science by the revolution which Darwin introduced in the biological field. Seldom has it been pointed out, and never before in so convincing a manner, that the acceptance of evolution for the organic world was a direct out- growth of its demonstration in the field of geological science. It was the publication by Sir Charles Lyell in 1830-33 of his Principles of Geology, giving currency to continuity or uniformitarianism in the realm of inorganic nature, that laid the foundations of modern geology and paved the way for modern biology as well. Darwin was first a geologist, and his great debt to Lyell he was ever ready to acknowledge. Says Professor Judd: “Were I to assert that if the Principles of Geology had not been written, we should never have had the Origin of Species, I think I should not be going too far; at all events, I can safely assert, from several conversations I had with Darwin, that he would have most unhesitatingly agreed in that opinion.” Huxley has given his verdict that ‘consistent uniformitarianism postulates evolution as much in the organic as in the inorganic world.” In dedicating the second edition of his favorite work, the Narrative of the Voyage of the Beagle, Darwin wrote: ‘‘To Charles Lyell, Esq., F.R.S., this second edition is dedicated with grateful pleasure, as an acknowl- edgment that the chief part of whatever scientific merit this journal and the other works of the author may possess, has been derived from studying the well-known admirable Principles of Geology.’ ‘To Leonard Horner he wrote: “I always feel as if my books came half out of Lyell’s brain.”’ In the Origin of Species Darwin refers to “Lyell’s grand work on the Principles of Geology, which the future historian will recognize as having produced a revolution in Natural Science.” The Coming of Evolution, first in the geological and later in the biological field, has fortunately now been told by a veteran geologist 469 470 REVIEWS and one who enjoyed the friendship of all the great leaders in the move- ment—Huxley, Hooker, Scrope, Wallace, Lyell, and Darwin. Of those who were on terms of affectionate intimacy with both Charles Lyell and Charles Darwin, Professor Judd is perhaps the unique survivor. It is this intimate personal relationship to the chief actors in the great drama, combined with a peculiarly simple and graceful style of writing, which makes the fascination of this little book. At every turn of the page the reader is surprised by the reference to some remark of Lyell, Darwin, or Huxley, which sheds a flood of light upon the psychology of the whole movement. The great success of the Principles of Geology seems in some measure to have been due to Lyell’s study of the causes of failure of the Theory of the Earth by the illustrious Hutton, whose death occurred the year Lyell was born. On the basis of his extended observations, Hutton as early as 1785 wrote the oft-quoted, “I can see no evidence of a beginning, and no prospect of an end,” a blunt statement which antagonized the church, then especially active in hunting heresy. Furthermore, his work was written in a heavy and cumbrous style. Profiting by this example, Lyell schooled himself in graceful, accurate, and forceful expression, and at some pains and with favoring fortune was able to avoid a clash with the established church. In no small measure this was due to an extremely favorable notice of his Prin- ciples in the Quarterly Review, then the champion of orthodoxy. With the geologists of the official Geological Survey, Lyell was less fortunate, and in spite of the general popularity of his epoch-making ideas, they were bitterly fought by the official class of geologists and only slowly won support in this field. Professor Judd’s fascinating story of the coming of evolution should find a wide circle of readers, especially among students of natural science. Wr. North American Index Fossils: Invertebrates. By AMADEUS W. GRABAU AND HERVEY WOODBURN SHIMER. Vols. I and II. New York: A. G. Seiler & Co., 1909 and 1910. With the rapid accumulation of special literature in the field of systematic paleontology, and the growing inaccessibility of many of the older works except to those having access to large libraries, it is ever becoming more and more difficult for the non-specialist to identify his species of fossils. At the same time, with the growing refinements in stratigraphy, it is ever becoming more important to the stratigraphic REVIEWS 471 geologist to give close attention to the fossil faunas present in his rock formations, and to have accurate identifications of his fossils. It is, therefore, a pleasure to notice the appearance of such a work as North American Index Fossils by Grabau and Shimer. In the two volumes of 853 and gog pages respectively which: com- prise this work, approximately 1,500 genera and 4,000 species are defined, a large portion of the species being accompanied by illustrations incorporated in the text, the figures being copied from various sources for which credit is always given. The species selected for definition have been chosen to include, first, those most characteristic of impor- tant stratigraphic divisions, i.e., those of wide geographic and limited stratigraphic range; secondly, those having a wide geographic distri- bution even though their stratigraphic range is also great, i.e., the very common American species; and thirdly, forms which it is important that students of structural and anatomical paleontology should under- stand. The species are arranged chronologically under their respective genera, the genera being arranged systematically under their proper families, orders, classes, and phyla. Brief discussions of the structural features of each phylum and class are included, but except in the case of the Arthropoda, no definitions of subclasses or orders are given. Under each class is given a brief bibliography of the more important literature, which will be of use to such as wish to carry their studies beyond the limits of the work. A decided innovation is the inclusion of extensive analytical keys to the genera under each of the classes. These keys are probably the most elaborate ever attempted for fossil invertebrates, and will doubtless be of much value to those using the books, although it must be kept in mind always that such keys can never be of so great utility in the classification of fossils, which are fre- quently if not usually represented by more or less incomplete speci- mens, as in the classification of living organisms. The closing pages of the second volume are given up to a series of appendices, as follows: A, Summary of North American Stratigraphy, Tables of Geological Formations (50 pages); B, Faunal Summary, Tables Showing Distribution of Species Described (50 pages); C, General Bibliography of North American Invertebrate Index Fossils and Fossil Faunas (1832-1909) (89 pages). In this bibliography the titles are arranged in accordance with the geological systems, those for each system being grouped geographically; D, Hints for Collecting and Preparing Fossil Invertebrates (16 pages); E, Glossary (36 pages) and General Index. 472 REVIEWS These volumes have been prepared primarily for the non-specialist, more especially for workers in stratigraphic geology who have not received special training in paleontology. For such workers, as well as for geological students in colleges and universities, and for amateur paleontologists and collectors of fossils, the volumes will prove to be of great value. S. W. Olenellus and Other Genera of the Mesonacidae. By CHARLES D. WALcoTT, Smithsonian Miscellaneous Collections, LV, No. 6. In his memoir on the Olenellus fauna, published in the Tenth Annual Report of the United States Geological Survey, in 1891, Walcott recog- nized seven American and three foreign species of Olenellus, included in three subgeneric groups, Olenellus proper, Mesonacis, and Holmia. ‘The present contribution represents the advance of knowledge concerning this highly interesting group of Cambrian trilobites since the appearance of the earlier memoir. Thirty-four species, including two varieties, are now recognized, besides two undetermined ones, thirty-six in all, arranged in ten groups which are given full generic rank, the entire group of forms being elevated to a family under the name Mesonacidae. ‘Twenty-four of these forms are American and twelve foreign, the foreign representa- tives being known only from northwestern Europe. With the restriction of the genus Olenellus to include only one group of these species, it comes about that this genus is no longer characteristic of the entire Lower Cambrian, as has commonly been assumed since the publication of the earlier memoir, but occurs only in the uppermost division of the series. In the present paper the Lower Cambrian is divided into four faunal zones, designated, beginning with the oldest, (1) Nevadia zone, (2) Elliptocephala zone, (3) Callavia zone, (4) Olenellus zone, each named from the leading Mesonacid genus present in the fauna. Aside from these four index genera the following are recognized: Meson- acis Walcott, Holmia Matthew, Wanneria n. gen., Paedeumias n. gen., Peachella n. gen., and Olenelloides Peach. In their genetic relations the genera discussed are assumed to diverge along two lines from the primitive Nevadia. The one line includes Callavia, Holmia, and Wanneria in order, the last of which is supposed to give origin to Paradoxides of the Middle Cambrian. The second line of descent springing from Nevadia includes Mesonacis, Elliptocephala, Paedeumias, and Olenellus in serial order, the last of these genera giving origin on the one hand to Peachella and on the other hand to Olenelloides. REVIEWS A473, Some question may be raised, perhaps, as to the legitimacy of the assumption of such a phylogenetic origin of Paradoxides. The most diagnostic character of the entire family Mesonacidae is the absence of a facial suture, although well-developed compound eyes are present. Elsewhere among the trilobites, where the free and fixed cheeks have become anchylosed, with the consequent disappearance of the facial suture, as, for instance, in the Devonian genus Phacops, this character has appeared at the termination of a long phylogenetic line in which all the earlier members possess functional facial sutures. The facial suture is so characteristic of every order and every family of trilobites, save the Mesonacidae, that one is forced to the assumption that it was a character of the primitive stock from which all have sprung. It there- fore seems necessary to assume that the ancestors of the Mesonacidae possessed a functional facial suture, and that the absence of this character in this group of genera is indicative of its terminal position in a long phylogenetic line whose pre-Cambrian history is unknown to us. Since such a character when once lost cannot be restored again, it would follow that Paradoxides with its functional facial suture could not have origi- nated from any member of the Mesonacidae. Might it not be assumed that Paradoxides arose from a totally distinct phylogenetic line in a different early Cambrian biologic province, perhaps southern Europe, and later migrated into the North Atlantic province where it occurs in strata generally younger than those bearing the Mesonacid faunas ? Under such an interpretation it would be necessary to grant that some- where Paradoxides may have been contemporaneous with at least a portion of the Mesonacid faunas in North America, and this contem- poraneity may even have extended to the North American shore of the North Atlantic basin. The paper adds much to our knowledge of these very ancient faunas of the earth, and the author is to be congratulated upon the success of his most persistent search for these rare fossil forms. Not the least attractive portion of the paper are the twenty-two beautifully executed half-tone plates. 5. W. Elements of Geology. By Exiot BLACKWELDER AND Haran H. Barrows. Pp. 475; figs. 485; pls.16. New York: American Book Co., 1911. This is not a manual or reference book, but an elementary textbook intended primarily for use by young students in the high schools, acade- 474 REVIEWS mies, and institutions of similar grade. The book was written in the belief that it is the function of a text as well as the duty of a teacher to develop in the student the power to reason. This spirit pervades the work throughout. The method is essentially analytical and the text explanatory rather than descriptive. Abundant use is made of questions which are ingen- iously devised to guide the student’s mental operations and to lead him unconsciously through certain desired chains of reasoning. Many of the questions are inserted in the text—a practice which makes the stu- dent stop and think and, by causing him to tie his ideas together, inciden- tally and unconsciously brings him to see the interrelation of the dif- ferent geologic agents and processes. The treatment throughout indicates a continuous desire to prevent the student from forming hard-and-fast conceptions of processes and geologic features that are necessarily often variable. There is a steady determination to compel the student to maintain a critical open mind and at the same time to draw close distinctions in the use of variable terms, as in the relative heights of hills and mountains and of plains and plateaus. Sometimes, however, this most laudable endeavor threatens to overstep itself and lumber up the text with hypercritical qualifications. In an elementary textbook where space is severely limited unessential discriminations crowd out more weighty matter, while the student on his part may come to give too much thought to precision in little things at the expense of a grasp of great things. But this is only another item in the ever-present question of where to draw the line. The text is clear, direct, and well written. In some cases, as in chap. i, the opening paragraph is a bit wobbly, but when the initial groping for just the right line is past and the topic is well under way, the chain of ideas, like the language, flows evenly and gracefully along without effort. Poise and balance characterize the treatment of facts and principles. The essential features are treated clearly though concisely, and the minor features are subordinated or left out where their omission does not weaken the presentation of the main topics. Unessential facts have been carefully pruned. Keen discrimination is apparent here. The departure of the authors from current practice in the arrange- ment of material will be most conspicuously seen in the omission of separate chapters on vulcanism and earthquakes. This was done in the belief that volcanoes and especially earthquakes are exceptional REVIEWS 475 and local phenomena and that although spectacular and ever interesting to the popular mind, they are not entitled to the same space in such a work as are the more general geologic processes. The main features of vulcanism and volcanic rocks are, however, quite adequately treated in the chapter on the composition of the earth, while volcanic moun- tains as surface features appear in the very excellent chapter entitled “The Great Relief Features of the Land.” The proper handling of historical geology in brief space is a difficult task. There is a great deal of ground to be covered and a great mass of material to be judiciously picked over. Unless the work is well done, the residue left is apt to be a dry bone skeleton with the flesh and blood largely gone. In the historical portion of this work the salient and vital points are made to stand out clearly. This is particu- larly true of the life history. In part this is secured by a sprightly use of paragraph headings to feature the various vicissitudes through which life forms have passed in their long history. With these in mind, the significance of the discussion is more readily grasped and the details are more easily retained. The authors have treated the Tertiary as a “Period,” giving it the same rank in the geologic time scale as they do the Comanche or the Cretaceous. After stating that it is divided into the Eocene, Mio- cene, and Pliocene epochs, the Tertiary is discussed largely as a unit. The Tertiary presents many rich problems for advanced students, especially its mammalian evolution and its diastrophism, but these are perhaps beyond the reach of a beginning class. The authors, believ- ing that the points of newness or striking facts are largely over by the time the Tertiary is reached, have apparently thought it best to curtail the treatment and advance rapidly to the close of the history. A feature which cannot be too highly commended is the extensive use of three-dimension diagrams to portray the operation of geologic processes. This, in the reviewer’s opinion, is much more expressive than the ordinary style. The set of three block diagrams on p. 146 which picture the successive development of youthful, mature, and old topography, illustrating not only the surface development of the streams but the simultaneous lowering of the land toward peneplanation, shows the possibilities of the method. By reducing the size of the illustrations, a very large number have been successfully introduced and add very greatly to the effectiveness and attractiveness of the book. It is a veritable picture book with most of the pictures new to geologic readers. 476 REVIEWS Finally it may be said that the general scheme and mode of treatment of the book follow the lead of the comprehensive treatise of Chamberlin and Salisbury, and the fundamental views which give distinctive charac- ter to that work find reflection in this. R LSC, Geology of the Kiruna District (2). Igneous Rocks and Iron Ore of Kiirunavaara Luossovaara and Tualluvaara. Academical Dis- sertation by Per A. Geter, for the degree of Doctor of Philosophy. By the permission of the philosophical faculty of the University of Upsala. Stockholm, 1910. Pp. 278; 2 geologic maps. The district is in northern Lapland. The rocks, which are generally regarded as pre-Cambrian, include greenstones, conglomerates, syenite porphyries, magnetite ores, quartz porphyry, phyllites, sandstones, etc. They are strongly folded and in general stand nearly vertical but other- wise do not show pronounced metamorphism. The textures are well preserved. A typical ore body is the one of Kiirunavaara which forms the backbone of a mountain about 748 meters high. This ore body is over 5 kilometers long and some 96 meters wide. Other ore bodies are somewhat smaller. The ore zone is included between quartz por- phyry and syenite porphyry. The minerals of the ore are magnetite, hematite (subordinate), fluor-apatite, augite, amphibole, biotite, titanite, tourmaline, zircon, etc. Generally there is enough apatite to place the ore above the Bessemer limit. The ore minerals are intergrown like those of an igneous rock and contacts between ore and country rock are in places gradational. All of the minerals of the ore except tourmaline are primary constituents of igneous rocks near by. Rock textures indicate that the ore mass has crystallized quite in the same way as an igneous rock—these include trachytoidal flow structure, skeleton forms of magnetite, and the ophitic distribution of augite. The ores are believed to be of magmatic origin and the writer is inclined to the view that the associated syenites are effusive in character. He does not agree with De Launey, who held that the ores were deposited at the surface from gases and hot solutions by pneumatolytic-sedimentary processes. The writer does not feel sure as to the nature of the differentiation processes which have resulted in the product, but does believe that such an origin is proven. REVIEWS 477 The Edmonton Coal Field, Alberta. By D. B. Dowttnc. Canada Department of Mines, Geological Survey Branch, 1g1o. 59 pages, 2 maps. _ The area primarily considered is on the Saskatchewan River, in and near Edmonton, but a short discussion of the surrounding coal fields is included. The coal is lignitic or semi-bituminous, and occurs near the middle and at the top of 700 feet of brackish water deposits, the Edmonton formation, at the top of the Cretaceous, and in Tertiary sandstone above. The lower horizon, the Clover Bar seam, is worked at Edmonton, and 80,000,000 tons are estimated to be available in an area of 14 square miles. We Ae. Preliminary Memoir on the Lewes and Nordenskiold Rivers Coal District, Yukon Territory. By D. D. Cartrnes. Canada Department of Mines, Geological Survey Branch, 1g10. 70 pages, 2 maps. The development of the Whitehorse copper deposits was the incentive for the investigation of the available coal resources in the district described in this report. The important formations of the district are the Braeburn limestone (carboniferous ?), the Laberge series, con- glomerates, shales, sandstones, etc., and Tantalus conglomerates, Jurasso- Cretaceous. Tertiary volcanics have broken through these formations and overflowed them in many places. Important coking coal seams occur in the Tantalus conglomerates and near the top of the Laberge series, but they are available only near the navigable water, such as the Lewes River and Lake Laberge. Wiel: Geology of the Nipigon Basin, Ontario. By A. W. G. WILson. Canada Department of Mines, Geological Survey Branch, IQIO. 152 pages, I map. The region covered by this excellent report is underlain mainly by Laurentian gneisses and granites, but scattered over it are areas of greenstones and green schists, called Keewatin. A few bands of Lower Huronian rocks are known. Lying on the eroded surface of these formations is a series of conglomerates, sandstones, shales, and dolo- mitic limestones classed as Keweenawan, although the author believes 478 REVIEWS they might be younger than pre-Cambrian. The youngest rock is a diabase, which occurs as intrusive sheets and flows. The evidence for and against the diabase occurring as a volcanic flow is fully discussed, the conclusion being that as now known they are basal residuals of former extensive flows. The glacial geology is briefly discussed, the author concluding that ice erosion was very limited, except locally. The physiographic features are considered, also the economic geology, but no deposits of any value are known. Wi Aes The Geology and Ore Deposits of the West Pilbara Goldfield. By H. P. Woopwarp. Bull. No. 41, Western Australia Geologi- cal Survey. Pp. 142; 5 geological maps; 1 mining plan; 25 figs. The first part of the bulletin is devoted to a general discussion of the physiography, geology, and petrography of the district, which occupies the triangular portion of the northwest division of the state included between the Fortescue and Yule rivers. The southern part of the area is a high tableland which drops abruptly to the wide, low coastal plain forming the northern part. The oldest rocks in the region are metamorphosed sedimentaries— clay slates and shales—that have been intruded successively by dolerite, gabbro, and granite. The last is thought to have altered some of the clay slates and dolerites to crystalline schists. A period of sub- sidence was accompanied by an outburst of volcanic activity in the form of fissure eruptions of very fluid basic lava. Subsidence continued, and marine beds are found above the last lava flow. Re-elevation and denudation have given rise to the present topography. The various formations are described in some detail, and petrological notes on seventy specimens are appended. The second part of the bulletin is devoted to a more detailed descrip- tion of the country and the mining centers visited. The lodes are most frequently found in the altered sedimentaries. They carry, in addition © to gold, varying amounts of pyrite, chalcopyrite, and galena. Little evidence regarding the genesis of the lodes is presented. Much of the material is of greater interest to the cneine et and the investor than to the geologist. A DPB: REVIEWS 479 A Review of Mining Operations in the State of South Australia during the Half-Year Ended December 31, 1910. No. 13. Issued by T. DUFFIELD, Secretary for Mines. Adelaide, 1911. J210)5 Bylo, Mp This paper gives statistics on leases, claims, subsidies, men employed, prices, and various industrial and technical features of the mining districts of South Australia. Notes on recent development work, including assays of samples and amount of boring, tunneling, etc., done on various properties make up a large part of the review. An interesting method of draining the southeastern district has been approved by the government geologist. The plan is to sink borings or shafts into a porous stratum underlying the swamp areas and allow the water to escape through underground channels, saving the expense of extensive ditches necessary for surface drainage. Small areas have been drained into natural sink holes with very encouraging results. eran IB)e aye Report on the Iron Ore Deposits along the Ottawa (Quebec Side) and Gatineau Rivers. By FRitz CIRKEL. Canada Department of Mines,-Mines Branch. No. 23, 1909. Pp. 147; plates 5; maps 2. The area covered by this report is about goo square miles, extending from Ottawa 100 miles up the Ottawa River and 83 miles up the Gatineau. Deposits of magnetite and hematite ore have been known for over sixty years and attempts have been made at various times to develop them, but without success. The present report is the result of a com- prehensive examination of the region to determine the possibilities of development of the deposits. One important factor is the available water power which is described in detail in the appendix. E. R. L. Maryland Geological Survey, Vol. VIII, t909. WILLIAM BULLOCK CLARK, State Geologist. This volume, which is entirely economic in its nature, contains the following reports: Part I, “Second Report on State Highway Con- struction,” by Walter Wilson Crosby, pp. 29-95; Part I, ‘“‘ Maryland Mineral Industries, 1896-1907,” by Wm. Bullock Clark and Edward B. Mathews, pp. 99-223; Part III, “Report on the Limestones of Mary- land with Special Reference to their Use in the Manufacture of Lime 480 REVIEWS and Cement,” by Edward Bennett Mathews and John Sharshall Grasty, PP. 2257477: E. ROE. Missourt Bureau of Geology and Mines. Biennial Report of the State Geologist for the Years 1909 and 1910. By H. A. BUEHLER AND OTHERS. The report contains a summary of the present and proposed work of the bureau and the following chapters descriptive of work now in progress: “The Principal Coal Fields of Northern Missouri,’ by Henry Hinds, pp. 26-35; “Reconnaissance Work,” by V. H. Hughes, pp. 36-54; and “The Geology of the Newburg Area,” by Wallace Lee, pp. 55-63. E.R. L. Mississippi State Geological Survey, 1907. ALBERT F. CRIDER, Director. The volume contains the following reports: Bulletin No. I, ‘Cement and Portland Cement Materials of Mississippi,” by Albert F. Crider, pp- 73; Bulletin No. II, “Clays of Mississippi, Part 1, Brick Clays and Clay Industry of Northern Mississippi,’ by William N. Logan, pp. 255; Bulletin No. III, “The Lignite of Mississippi,’ by Calvin S. Brown, pp. 71. BoE: The Geology of the Whatatutu Subdivision, Raukumara Division, Poverty Bay. By JAMES HENRY ADAMS. New Zealand Geo- logical Survey, Bulletin No. 9 (New Series). Wellington, LO10: 7p. 48; maps 5;, plates. The Raukumara division lies on the eastern side of the North Island of New Zealand and consists of a series of rolling ridges of moderate height separated by deeply cut river valleys. The rocks belong chiefly to the Whatatutu series which are upper Miocene in age and which are folded into irregular anticlines and synclines. Indications of oil have been found at various points within the region and the object of the survey was to obtain information as to the possibilities of development. With this end in view the anticlines and synclines were mapped and described with considerable care. Fossils are abundant in some locali- ties but have received little attention in this report. E. RoE: ae i There are quantities of dust floating in the air of the-ordi- Al, nary, schoolroom, brought in _~ from’ the streets and raised ’ from the floor by the constant movement of the children’s feet. Science has proved that dust is a favorite nesting place for disease germs. | It follows that at every breath the children are in danger of being infected by the germs contained in the floating dust they inhale. The best known prevent- ive of disease-carrying dust is Standard Floor Dressing. Standard Floor Dressing catches all dust the instant it settles on the floor and holds it there, together with the germs the dust contains. At the end of the day dust and germs are easily swept away without again ris- ing into the air. The air is thus kept untainted; the spread of disease is checked at the outset. Our free illustrated booklet on dust. dangers and how to avoid them con- tains information of special value to principals of schools and all others in a position to promote hygienic conditions among children. Post- - paid on request. Write for it today. Not intended for household use. - Standard Oil Company (Incorporated) Purify your Refrigerator! GET THE GENUINE | Hi OA Baker’s Chocolate] | igeaces Px locaton foul odors. vand sickness by Keeping in your refrigerafora sponge sprinkled occasionally with Flatts Chlorides. Wash the sponge ice a week Blue Wrapper — Yellow Label Trade Mark on the Back FINEST IN THE WORLD PAIS SSCS NAO FA a al OT 8 NW BEE A MEL OE For Cooking and Drinking Ths e oO dor less | WALTER BAKER & CO.Ltd.|| || Dasadmfectanrte || Established 1780 DORCHESTER, MASS. Sold by Druggists Everywhere ‘ cae ne 3 Write to the manufacturer, Henry B. Platt, 42 Cliff Street, New York. for free hook and sample bottle. Fe ; y have been established over 60 YEARS. By ‘ow P ‘ system of payments every family in moderate cir : cumstances can own a VOS€ piano. We take ola; instruments in exchange and deliver the new pianw in your home free of expense, Write for Catalogue D and explanations, vose & ONS PIANO CO... Boston r \ VOLUME XIX : ; NUMBER 6 ‘EEE JOURNAL or GEOLOGY A_SEMI- QUARTERLY EDITED BY ) THOMAS C. CHAMBERLIN AND ROLLIN D. SALISBURY anes : With the Active Collaboration of SAMUEL W. WILLISTON ALBERT JOHANNSEN WILLIAM H. EMMONS Vertebrate Paleontology Petrology Economic Geology STUART WELLER WALLACE W. ATWOOD ROLLIN T. CHAMBERLIN Invertebrate Paleontology Physiography Dynamic Geology ASSOCIATE EDITORS SIR ARCHIBALD GEIKIE, Great Britain; GROVE K.GILBERT, National Survey, Washington, D.C. HEINRICH ROSENBUSCH, Germany CHARLES D. WALCOTT, Smithsonian Institution THEODOR N. TSCHERNYSCHEW, Russia HENRY S. WILLIAMS, Cornell University CHARLES BARROIS, France JOSEPH P.IDDINGS, Washington, D.C. ALBRECHT PENCK, Germany _ JOHN C. BRANNER, Stanford University HANS REUSCH, Norway igs RICHARD A. F. PENROSE, Jr., Philadelphia, Pa, GERARD DEGEER, Sweden WILLIAM B. CLARK, Johns Hopkins University “ORVILLE A. DERBY, Brazil WILLIAM H. HOBBS, University of Michigan T, W. EDGEWORTH DAVID, Australia FRANK D. ADAMS, McGill University BAILEY WILLIS, Argentine Republic CHARLES K. LEITH, University of Wisconsin SEPTEMBER-OCTOBER, to11 CONTENTS _ PRELIMINARY STATEMENT. CONCERNING A NEW SYSTEM OF QUATERNARY LAKES IN THE MISSISSIPPI BASIN - - - - - - - ~ Evucene Westey SHaw 481 GRAVEL AS A pore Ne ROCK = ye Joy Lyon Rica | 402 THE CRETACEOUS AND TERTIARY FORMATION OF WESTERN NORTH DAKOTA. BUTS Bz SPTVD RONG MUON TSN eG, LEONARD 507 ON THE GENUS SYRINGOPLEURA SCHUCHERT - - - - - - Grorce H.Gimty 548 PRELIMINARY NOTES ON SOME. IGNEOUS ROCKS OF JAPAN. I - = -S. Kézu s55 “PRELIMINARY NOTES ON SOME IGNEOUS ROCKS OF JAPAN. II-°- -S.Kézu- 56x PRELIMINARY NOTES. ON SOME-IGNEOUS ROCKS OF JAPAN. Ill - -S. Késu 566 “TRAE NV ATTEN ASSIS Sa Silence ei gr en are eevee ca pi, Peer cee Seat Che Aniversity of Chicago press CHICAGO, ILLINOIS AGENTS: '- THE CAMBRIDGE UNIVERSITY PRESS, Lonpon anp EpinsurcH WILLIAM WESLEY & SON, Lonpon TH. STAUFFER, Le1rzic \ ' THE MARUZEN-KABUSHIKI-KAISHA, Toxvo, Osaka, Kyoro Che Journal of Geology Published on or about the following dates: February 1, March 15, May 1, June 15, August 1, September 15, November 1, December 15. Vol. XIX CONTENTS FOR SEPTEMBER-OCTOBER, ft No. 6 PRELIMINARY STATEMENT CONCERNING A NEW SYSTEM OF QUATERNARY LAKES IN THE MISSISSIPPI BASIN” - - - - - - - - - - - - - EuGENE WESLEY SHAW 481 GRAVEL AS A RESISTANT ROCK - - - - - - - - - SEAS - - JoHn Lyon RicH 492 THE CRETACEOUS AND TERTIARY FORMATIONS OF WESTERN NORTH DAKOTA AND EASTERN MONTANA - - - - - - - - - - - - - - A. G. LEONARD 507 ON THE GENUS SYRINGOPLEURA SCHUCHERT - eee - - - - = GroRGE H. Girty 548 PRELIMINARY NOTES ON SOME IGNEOUS ROCKS OF JAPAN. I - - - - - - - 8. Kézu 555 PRELIMINARY NOTES ON SOME IGNEOUS ROCKS OF JAPAN. II - - - an eS - 8. Kézu 561 PRELIMINARY NOTES ON SOME IGNEOUS ROCKS OF JAPAN. II - - - - - -S. Kézu 566 RE VDE W'S ah aie einen as lar at ar (Mie ei oie UN hcesicsobeon | (Vce (gas mi etna Stee Tae ED ONS Hes i se LIA 6) The Journal of Geology is published semi-quarterly. |The subscription price is $3.00 ‘per year; the price of single copies is 50 cents. 4] Postage is prepaid bythe publishers on all orders from the United States, Mexico, Cuba, Porto Rico, Panama Canal Zone, Republic of Panama, Hawaiian Islands, Philippine Islands, Guam, Tutuila(Samoa),Shanghai. 4] Postage is charged extra as follows: For Ganada: 30 cents on annual subscriptions (total $3.30), on single copies, 4 cents (total 54 cents); for all other countries in the Postal Union, 53 cents on annual subscriptions (total $3.53), on single copies, I1 certs (total 61 cents). {[Remit- tances should be made payable to The University of Chicago Press and should be in Chicago or New York exchange, postal or express muney order. If local check is used, 10 cents should be added for collection. The following agents have been appointed and are authorized to quote the prices indicated: For the British Empire: The Cambridge University Press, Fetter Lane, London, E.C., and William _ Wesley & Son, 28 Essex Street, Strand, London. Yearly subscriptions, including postage, 145. 6d. each; single copies, including postage, 2s. 6¢. each. For the continent of Europe: Th. Stauffer, Universitatsstrasse 26, Leipzig, Germany. Yearly sub- scriptions, including postage, M. 14.80 each; single copies including postage, M. 2.60 each. For Japan and Korea: The Maruzen-Kabushiki-Kaisha, 11 to 16 Nihonbashi Tori Sanchome, Tokyo, Japan. Yearly subscriptions, including postage, Yen 7.00 each ; single copies, including postage, Yen 1.25 each. Claims for missing numbers should be made within the conten following the regular month of publication. The publishers expect to supply missing numbers free only when they have been lost in transit. Business correspondence should be addressed to The University of Chicago Press, Chicago, Il. Communications for the editors should be pddreseed to them at The University of Chicago, Chicago, Ill. Contributors are requested to write scientific and proper names with particular care and in citations to follow the form shown in the pages of the Journal. — Separates, if desired, must be ordered in advance of publication. Separates of leading articles will be printed as ordered, of which 50 (without covers) will be furnished gratis, the actual cost of the remainder (and covers, if desired) to be paid for by the author. The table below shows the approximate cost of separates consisting of plain text or text with line engravings. The actual cost may vary from the figures given, and will depend upon the amount of work in re-making the pages into forms, press work, paper, binding, etc. Separates containing half-tones may be expected to cost somewhat more than the rates given, the increase depending upon the number of cuts and the amount of work required upon them. Number of copies 50 100 150 ; 200 etterspress, for. uAvpages OmlesS@ 2) s. ace > $1.40 $1.60 $1.80 $2.00 Wetterapress ton sOpagesomlessy se]. ume chi I.90 2.20 2.50 2.80 Letter-press, for 16 pages orless.. ; .. . 3.20 4.00 4.60 5.30 Single plates (1 double=2 single). . . . . .90 1.10 1.30 1.50 , Covers, with title (paper like Journal cover) . [.20 1.69 2.00 2.40 Entered March 20, 1893, at the Post-Office at Chicago, IIl., as second-class matter, under Act of Congress, March 3, 1879. THE IOURNAL OF GEOLOGY SLIEVMIBER OCLOBE FR, Tor PRELIMINARY STATEMENT CONCERNING A NEW SYSTEM .OF QUATERNARY LAKES IN THE MISSISSIPPI BASIN* EUGENE WESLEY SHAW U.S. Geological Survey It is a significant fact that in but few places do the Mississippi and Ohio rivers flow on consolidated rock. Throughout most of their courses they flow over bodies of silt, sand, and gravel 50-100 feet in thickness. The lower half or third of each tributary also flows over a thick unconsolidated mass, which is similar to those on the larger streams, except that in general it is less coarse. For examples, the Wisconsin River in southwestern Wisconsin is working 50 feet or more above a hard rock channel; Big Muddy River in southern Illinois flows between mud banks in a broad, shallow valley with a buried channel 4o feet below; and away east in Pennsylvania the Monongahela does not flow over bed-rock at any point within the limits of the state. Thus, not only the valleys of the Mississippi and Ohio, but the lower part of almost every tributary valley in the northeast central states, and probably in a considerably larger territory, is partly filled with loose sedi- ment, and in Illinois, Indiana, and Kentucky the filling on the tributary streams consists largely of clay, a brief description and interpretation of which are the objects of the present paper. t Published by permission of the Director of the U.S. Geological Survey, Washing- ton, D.C. A more complete description is to be published by the Ill. Geological Survey. 481 482 EUGENE WESLEY SHAW The upper surface of the clay forms a terrace which is generally so broad and so low that it is scarcely perceptible, though it is commonly separated from the flood-plain by a low scarp. This terrace is almost perfectly horizontal, and since the flood-plain rises up stream the terrace and flood-plain finally merge. However, since the flood-plain itself on the tributaries is nearly horizontal (for the streams have but little fall) the flood-plain and terrace on some rivers are distinct for 40 miles or more, although vertically they are almost nowhere more than 4o feet apart. Another characteristic of these valleys is that in places they anastomose. Many valley floors connect through divides with neighboring valley floors. Some of the connecting parts are broad and resemble bays in the sea; others are narrow and strait-like; and the severed parts of the divide are massive. In many places the flat valley floor surrounds hills that stand up sharply like islands. These features of the lower parts of valleys tributary to the Mississippi and Ohio—the broad bottoms in hilly country, and the irregularly branching valleys—point toward valley filling. And well-sections and exposures support this indication, showing that bed-rock is far below the present streams. Detailed description of the clay.—The clay varies from greenish- gray to purplish-gray in color and from medium plasticity to “oumbo.” The lower part is evenly stratified.and in places finely laminated. The upper part has less distinct stratification and is characterized by irregular concretionary masses of lime. Around the border and in the up-stream parts of the deposit there are lenses of fine sand, but considering the formation as a whole, sand forms a remarkably small part. With the exception of the concretionary lime, some particles of which are as small as sand grains, most of the deposit is without perceptible grit. In ground plan the bodies of clay are very irregular and even anas- tomosing—shapes that would be expected of valley fills in a country of medium to low relief (see Fig. 1). The surface of the clay in each valley is horizontal and lies from 5 to 75 feet above low water. But the altitude varies from valley to valley. Near Cairo the surface of the clay is 345 feet above sea; at Galena, Illinois, 400- miles up the Mississippi, it is 650 feet; and there is a corresponding QUATERNARY LAKES IN THE MISSISSIPPI BASIN 483 increase in altitude up the Ohio. Thus, although the deposit along each tributary and its branches is usually isolated and lies at a different altitude from that on every other stream, the different bodies have such a regular arrangement and have so many char- acters in common that there can be little question but that they are closely related, and they appear to be in large part lake deposits, but in smaller part stream deposits, so that they may be referred to as fluvio-lacustrine. IS MILES. Fic. 1.—Lake Muddy, in southern Illinois. One of a series of lakes, now extinct, caused by a rapidly growing valley filling on the Mississippi and certain other streams, the filling forming a dam across the mouths of tributaries. The lakes stood at dif- ferent altitudes, being controlled by the altitude of the Mississippi at their various outlets; each was in a continual state of fluctuation, the position of its surface at any moment being controlled by the stage of the Mississippi, and for a part of the time each was intermittent. The narrow part of Lake Muddy near the outlet was in a narrow, high-walled part of the valley, due to uplifted hard rocks. With the approach of every flood on the Mississippi water gushed up through the narrow part of the lake to the broader inland, a part carrying with it fine sand which, with interbedded lake silt, formed a delta at the lower end of the lake, fronting toward the head of the lake. 484 EUGENE WESLEY SHAW Shore features were generally poorly developed, though 12-15 miles northeast of Madisonville, Kentucky, 60 miles by water from the Ohio River, there are beautifully developed and well-preserved each-ridges. These ridges are very symmetrical, being 20-50 feet wide, and 8 to 10 feet high (see Fig. 2). They are composed of sand and fine gravel and are situated across the mouths of small tributary valleys. The reason for the excellent development of gravel ridges at this place is the generous available supply of loosely cemented conglomerate, probably Late Tertiary in age, com- posed largely of well-rounded quartz and flint pebbles. Elsewhere there was not a large amount of well-rounded pebbles within reach of the lake and so far no other well-developed ridges have been found. At numerous places where the bank of the lake was easily eroded there is some suggestion of wave cutting, but the evidence has been almost obliterated by recent erosion. One reason for the general poor development of shore features is that owing to the rise and fall of the rivers the lakes were continually fluctuating and were intermittent for a part of the period of their existence. Thus, particularly in districts of low relief, the shores of the lakes did not stand in one position long enough to develop shore features. Good collections of fossils were obtained, the fauna consisting of nearly a score of species of gastropods and lamellibranches, and undoubtedly many more species, including perhaps vertebrate and plant remains, might be found. Most of the forms collected inhabit lagoons and the quiet parts of streams. One of them (Campe- loma) is a scavenger living in decaying animal matter. Others frequent lily ponds. Some, such as Vertigo, are northern forms, being found at present from Wisconsin northward. The lime masses are probably secretions of blue-green algae, though at present they show little organic structure. They are more abundant in the thinner parts of the formation, and this may be correlated with the fact that lime-secreting algae flourish in very shallow or intermittent waters. Previous work.—Bodies of this clay have been regarded as glacial drift; a lowland phase of the loess; an old normal flood-plain deposit; a back-water deposit from glacial floods on the larger QUATERNARY LAKES IN THE MISSISSIPPI BASIN 485 DR fem ny = H 4 S ts." Hn - a SWaew . \ re : 3 me Ca 4 Fic. 2.—Part of Madisonville, Ky., topog. sheet, U.S. Geol. Survey, showing beach of extinct Green Lake (named from Green River which now drains the lake bed) and an ‘‘island hill.” The thickness of the lake deposit here is about 30 feet and the surface 381 to 385 feet above sea. The “‘island hill” is one of many peculiar partly-buried hills which rise sharply from the flat surface of the material deposited around them so that they bear a strong resemblance to islands rising above a water surface. The beach ridge shows well in the topography between Philips’ store and the McLean County line. 486 EUGENE WESLEY SHAW streams; a deposit due to subsidence; a deposit due to climatic change; and in southwestern Wisconsin a closely related but pre- dominantly stream-laid deposit has been attributed to glacial floods and deposits in the Mississippi Valley. The clay is not glacial drift, for it contains no stones and little sand; and much of it lies outside the glacial boundary. More- over, it is found only in the lowest places and its upper surface is horizontal without regard to the underlying surface of hard rock. It is not loess, for it fills all depressions up to certain altitudes and is not found at higher positions. Its thickness and others of the characters already described show that it is not a normal flood- plain deposit. It could scarcely be a simple back-water deposit from glacial floods without the help of a valley train, because that would require that the rivers have a sustained depth of about two hundred feet for the thousands of years it must have taken the clay to accumulate. A subsidence of the surface might lead to the development of a few bodies of clay having the shape and arrange- . ment of those under discussion, but warping so complex as to cause the regular arrangement and shape of so many bodies of clay would be inconceivable. Nor could the deposits have been produced by climatic change, for such deposits slope down stream and these are horizontal. Finally, the limited up-stream extent of the clay, the fineness of the material, the horizontality of the surface, and the fact that the clay abuts against thick bodies of coarser material on the large rivers, indicate that most of the clay accumulated in lakes produced by valley fillings, the master drainage lines of the region. In order to understand the cause and history of the lakes it is therefore necessary to look into the history of the large rivers. Valley filling on the Mississippi and Ohio.—The deposits on the Mississippi and Ohio consist principally of sand, but there is con- siderable gravel and silt, the gravel being more abundant at the base and the silt at the top. Most of the material lies below extreme high-water stage, and hence the surface forms a flood-plain, but here and there bodies ot sand and gravel stand about 30 feet above the reach of high water, the upper surface in such places forming a terrace at the altitude of the valley filling on near-by tributaries. QUATERNARY LAKES IN THE MISSISSIPPI BASIN 487 Apparently the river valleys were once filled to a position as high as the surface of the filling on the tributaries, but have now been partly cleared out, the surface of the fill being lowered about 30 feet. The part remaining is about 150 feet thick and extends about 120 feet below low water, the range between high- and low- water stages being about 30 feet (see Fig. 3). In this connection it seems worth while to note that when the discharge of a stream is increased, the vertical distance between the bottom of the channel and the flood-plain is also increased, and this comes about not alone by scouring out the channel, but also by building up the alluvium. Thus, without any change in size of load, it is possible to produce thick alluvium by simply “increasing the volume of water. To return to the lakes themselves: they differed from most bodies of quiet water in that the position of the surface varied greatly every year, for it was controlled by the various stages of the rivers. If the range between high and low water had been the same that it is now the surfaces of the lakes would have fluctuated between limits about ro to 4o feet apart. But the lakes formed a huge reservoir so that with the same discharge as at present the rivers would not have risen nearly so much in times of flood. Indeed, to raise the surface of the lakes and rivers one foot, it took over one hundred billion cubic feet or nearly a cubic mile of water; moreover, every rise of 5 or to feet would double the dis- charge of the rivers, so that tremendous floods could be taken care of without great increase in depth of water. Terminology.—It seems probable that the rather extensive development of deposits and resulting topographic features such as are described in this paper will lead to the introduction of some new descriptive terms. Perhaps it will be found convenient to use “‘contragradation”’ or “‘dam gradation” for that kind of stream aggradation which is caused by an obstruction, or, more broadly, decrease in velocity, and perhaps to invent still other terms for the aggradation due to increase in load and decrease in volume. In case the obstruction develops so rapidly as to produce ponded water, such as is described in the present paper, the deposit is on the whole very fine-grained and the top nearly horizontal though “‘YINOUL dy} OF JINOS dy} WOI JUL] JOOIIP B UL JOU pUY JOA APpN]T JO asin0d sfoyM oY} UO 9pRUT SAL UOTIIOS ay} ApjoeI109 sapyord oy} MoYs 0} Jopio uy ‘“AoT[eA Iddississtpy oy} Jo (¢ 9us0jslo]q) Surusdosep pideva v Suryvorpur ‘AoyeA iddisstsstpy ay JO 7eyI 0} poysnf{pe you st AoeA Appnyy jo opyoud yoor-poq oy, “MOT ATawiatyx9 C07 CQ pur ‘MoT Dd 07 g ‘9}¥IJOpOU BuTOq gq 01 VW wlory [Vy oy. ‘urvosys oy} JO ALOYSIY oY} SJOopor opyord yuaseid oy T, ‘Soyovel WivII}s JY} WOT] Jo}VM Youd sv Ivy sv iddississtpy dy} UO 19}VM MOT PUP YSTY Udaajoq asurr oy} Aq po][oryUOD Suteq souvysIp oy} ‘WI0}}0q JoUURYS JUsSeId 9yy aaoqe Ysty pur [e}yuoziu0y st ure[d-pooy qyuosaid oy} Jo qed JaMOT oY, “[eyUOZIIOY st YyOva jo doy ayy pur ‘ssUTT]Y JUaJayIp OM aq 0} Woes dI9YT, ‘sULIISNOV] = qaed ur pure [erang wed ul ‘Avp pue js poureis-oug AroA yavd jsou oY}. 10F st ‘stay}O AuvUT UT Jey} dyI] ‘AoT[VA Jaa Appnyy ur ty [euoyem Sumy oy, “ssodap sary iddisstsstp_ JO worses-ssoio pue sysodep Jeary Appny, JO woroes [vulpnysuo[—’ ‘org a) a SATIN OS ov 2 ( 2 \ = , oN S SS aaa —<——<— Al \ a Sa SO Niwid Goold iNsSaud ars iS \ SS RESIN Se Tis aiv7T sO dol HLS NNN Ws ATeva 130 dOL os ie : ouogsAydnyy S N § 488 QUATERNARY LAKES IN THE MISSISSIPPI BASIN 489 more or less concave. For the resulting topographic feature, the bottom of Muddy River Valley may be taken as a type and Muddy may be an acceptable name for it, referring, as the name does, both to a particular type and to a principal character of the deposit, and the streams which flow over it, and also to the general character of the country where the feature is developed. On the other hand, in case aggradation keeps pace with the growth of the dam the material is in general coarser and the upper surface rises up stream, though at a less rate than the original stream channel. For this topographic feature the surface of the deposit forming a low terrace along Big Sandy River in eastern Kentucky may be taken as a type and called a Sandy. Perhaps also it will be found desirable to speak of the island-like hills sur- rounded by the deposit as Island Hills, and the hill bearing the town of Island, in Kentucky, may be taken as a type. Summarizing.—Vhe inferred history of the lake deposit reads about as follows: In middle or late glacial time the rivers were flowing on beds about too feet below their present ones. Whether this great depth was attained in an interglacial epoch by a regional uplift or was reached through the deep scouring of glacial floods has not yet been determined. The tributaries entered the flood- plains of the Mississippi and Ohio on channel bottoms only about 40 feet lower than those in use today and their flood-plains were near the position of their present channel bottoms, these positions being controlled by low- and high-water stages on the master streams. As at present, at low-water stage there was no standing water in the tributaries, but at high water the deep channels were filled by back water from the rivers, thus forming long, narrow winding lakes. When aggradation began on the Mississippi and Ohio, both low- and high-water marks on them and on the tribu- taries rose. At low water there were embryo, perennial lakes in the channels of the tributaries at their mouths and at high water the flood-plains were covered more deeply than before. The area covered both at low- and high-water stages gradually extended until low-water stage reached the altitude of the former flood-plain. From this time on there were perennial bodies of quiet water of considerable size on each tributary, and wedge-shaped masses of 4QO EUGENE WESLEY SHAW lake deposit about 80 feet thick at the lower ends and thinning out to a feather edge up stream, accumulated on the old flood-plains. Nearly all the material deposited in the lakes was fine sediment such as would be carried in suspension, and the lakes seem to have been filled with this material up to certain concordant positions, probably to the natural position of a flood-plain or just below the high-water mark of the time. - 2s A A Re 0) QGri37FLOop-PLAIN SILT SSS LAKE DEPOSIT O | 2 3 4 5 MILES Fic. 4.—Diagram showing arrangement of principal deposits and surface features along Beaucoup Creek, Perry and Jackson counties, Illinois. The filling thickens and the flood-plain becomes narrower down stream. When the lake became extinct the bed became a great swamp. The stream first cut into the lower end of the fill, draining that part of the swamp and developing a narrow flood-plain below the sur- face of the lake silt. With further downward cutting the new flood-plain was lowered and extended up stream and the swamp area reduced. Meanwhile, stream deposits continued to accumulate at the upper end of the lake bed. Many other valley bot- toms are similar, having a peculiar swampy central portion. QUATERNARY LAKES IN THE MISSISSIPPI BASIN . 491 When the Mississippi and Ohio finally became not only able to carry all the load delivered to them but a little more, they began to cut down again. Perhaps even before this time the lakes had become intermittent, being drained except at times of high water, for they were almost filled with sediment. The great flat lake bottoms became swamps, and channels began to deepen again at the former outlets. At the same time the swamps themselves began to be drained at the lower ends. The process of swamp draining has continued to the present time, and on medium-sized streams there now remain only 10-20 miles of swamp, the lower 20-50 miles having been drained (see Fig 4). GRAVEL As: A RESISTANT ROCK JOHN LYON RICH Cornell University, Ithaca, N.Y. INTRODUCTION The thesis which this paper will endeavor to establish may be stated as follows: Gravel, in its relation to the agencies of denudation, is, under certain geological conditions, a highly resistant rock. To these agencies it will, in general, offer greater resistance than ordinary igneous or sedimentary rocks, with a few possible exceptions. On the validity of this thesis hinge important deductions as to the normal course of topographic development in cases where such gravel plays a prominent part in the geological structure of a region. It is my purpose (1) to point out the theoretical reasons for the resistant nature of gravel deposits; (2) to show from an actual occurrence in nature that the gravels do behave as the theoretical considerations would lead us to expect; and (3) to sketch, by way of suggestion, the normal course of development of topography in a region where alluvial fans of coarse material are accumulating at the base of mountains. By way of suggestion there will be further a brief application of the principles brought out to certain well- known topographic features. Except for the descriptive portion of the paper, which will be clearly distinguished, the article is an analytical study made mainly for the purpose of determining the influence of certain types of rocks upon the processes and rate of denudation, and of calling attention to what appears to be a normal cycle of denudation and the topo- graphic development of mountains in an arid region, and to a lesser extent in a humid region as well. At the present time no attempt will be made to review exhaustively the literature of the subject. 1 Published by permission of the Director of the U.S. Geological Survey. 402 GRAVEL AS A RESISTANT ROCK 493 RESISTANT QUALITIES OF GRAVEL It is natural to look upon gravels as weak rocks which may easily be removed by the agencies of denudation. While this is doubt- less true for sand or possibly even for fine gravel, it is a view which does not hold true of gravel of a coarser nature such as accumulates at the base of mountain ranges either in arid or humid climates, or of river gravels of the coarser type such as the Lafayette gravels of the Mississippi Valley. There are two essential reasons for the resistant quality of gravel as regards denudation. These are: (1) the selected nature of the material; (2) its porosity. As regards the first of these, gravel is a composite rock made up of units, each of which is selected on the basis of its ability to withstand the action of the agencies of de- struction to which all rocks are subjected. These agencies of dis- integration are both mechanical and chemical. With respect to the mechanical, gravel may be looked upon as a residue which has survived the rolling, pounding, and abrasion incident to its trans- portation along the stream course, an experience which, if the jour- ney be a long one, effectively grinds down and destroys all but the most resistant of the materials subjected to it. From the standpoint, also, of rock decomposition gravel is par- ticularly resistant, for it is a rock whose component materials are severally selected on the basis of their ability to withstand such decomposition. In a region of normal development where there has been no interference with normal conditions by such accidents as glaciation, the stream gravels represent, in the main, only rocks or fragments of rocks which, by virtue of their resistant qualities, have been able to survive unchanged the decomposition and mechanical disintegration which has effectively destroyed the rock surrounding them. They have undergone successfully the ordeal which has destroyed the neighboring rock. They are therefore still able to resist further subjection to the action of the same agen- cies of change. Physically also, gravel is especially fitted to resist disintegration because in it the component fragments are reduced to compact units unbroken, as a rule, by fractures or other lines of weakness. The surfaces are generally smoothed and give little opportunity for 404 JOHN LYON RICH the attack of frost or for the entrance of percolating water, while the comparatively small size of the units diminishes the activity of insolation as a disintegrating agent. For these reasons, while a massive quartzite, for instance, may be as resistant as gravel to disintegration due to either mechanical abrasion or chemical decom- position, it will be more likely, especially if in larger masses, to suffer more from the effects of insolation and frost. From the foregoing it is clear that stream gravels, particularly the coarser ones, may properly be looked upon as concentrates of the most resistant elements of the rocks from which they are derived. It follows that a gravel of such a nature will be more resistant to the agencies of disintegration than the original rocks. From its very nature and origin a gravel deposit should be expected to offer great resistance to the normal agencies of sub-aerial denudation. This resistant quality is particularly significant in the development of the topography of gravel deposits, since, disintegration being at a minimum, bodily removal of the component units of the gravel is necessary for their reduction; and, as we shall point out later, bodily removal, too, is at a minimum except along the immediate courses of good-sized streams. In the latter situations this may be readily accomplished, but away from the actual stream course the removal of the material must necessarily be very slow. The importance of this point in relation to the dissection of the alluvial fans along the base of a mountain range will be more fully elaborated on a subsequent page. The second characteristic of gravels which makes them resistant to the disintegrating and erosive forces which would wear them down is their porosity and the consequent comparatively slight development of surface drainage on the gravel areas. A gravel deposit of moderate coarseness offers the maximum of favorable conditions for the absorption and storage of the rain which falls upon its surface. This hinders the formation of small surface streams, and since, as we have seen, disintegration is at a minimum, and the removal of the gravel is almost entirely dependent upon the transporting action of such streams, the gravels are doubly pro- tected from removal. % From the foregoing theoretical considerations we should expect GRAVEL AS A RESISTANT ROCK 495 that gravel would be one of the most resistant of rocks as far as its relation to the processes of disintegration and removal is concerned. Compare, for instance, the relative ease with which weathering and erosion break up and remove the rocks from an area of granite and from one of moderately coarse gravels in a similar situation. In the case of the granite there is always a greater or less number of joints or fissures through which water may enter and perform its work of disintegration either by direct chemical decomposition or by the subsidiary agency of frost. In contrast to this there are the smooth, usually fissureless surfaces of the gravel units. The granite is made up of a variety of minerals, some of which are easily attacked by the weathering agents. Some, it is true, are as resist- ant as the most resistant components of the gravel but in every case these are small, being limited in size by the texture of the granite. There is too, in a rock of complex mineral composition, the factor of pulling apart of the mineral grains by differential expansion and contraction. The less resistant minerals, by weathering away and breaking down, leave the harder and more resistant ones free to be removed by the surface waters. Since, in general, the size of the grains is comparatively small, in a granite scarcely exceeding one centimeter in diameter, the resistant materials are readily removed by the s'reams in the form of sand, while the products of the more thor- ough disintegration of the less resistant minerals are easily carried away in suspension or solution or may even be, in considerable measure, picked up and carried off by the winds. Thus we see that a granite is much more vulnerable to the attacks of the weathering agencies than a coarse gravel. What is true of granite is also true in varying measure of any of the less resistant sedimentary or igneous rocks, such as shale, soft sand- stone or limestone, diorite, etc. In the case of quartzite and cer- tain of the lavas it is a question which would disintegrate more rapidly, these or the gravel. The latter has in its favor, as a resist- ant rock, the factor of porosity and the slight effect of insolation. All the above considerations apply particularly when the slope is low. On very steep slopes the lack of coherence of the gravel, JOHN LYON RICH i Cy a us aaers aH Be: Bees of ¢ ae ¢ the features m show Wee N.M The gravel area is shown by the dotted pattern. ’ gle ilver City quadran 1.—A portion of the S the text. Fic. described in GRAVEL AS A RESISTANT ROCK 497 combined with the effect of gravity and the rapid mechanical erosion, would doubtless cause more rapid removal of gravel than of granite on account of the dominance of the factor of bodily transportation. PIEDMONT GRAVELS NEAR SILVER CITY, NEW MEXICO Near the town of Silver City, N.M., best shown between there and the smaller town of Central, seven miles directly to the east (see U.S.G.S., Silver City quadrangle), there lies a gravel deposit of Piedmont nature which presents points of particular interest in connection with the thesis just presented. The road from Silver City to Central follows closely along the inner or mountainward margin of this deposit (Fig. 1) which extends from here southward for 40 or 50 miles as a part of the gravel fan of a great interior basin which, with its tributary basins, covers a large area in the south- western part of the state. The gravel plateau, as it will be called, in the portion under consideration between Silver City and Central, has a slope to the south of about too ft. per mile and strikes approximately east and west. It is characterized by great uniformity and evenness as seen from any point on the plateau surface. One sees merely a monotonous plain of gravel, horizontal as one looks to the east or west, but sloping always toward the south. This appearance of great evenness applies, however, only to the remnants as seen from a point on the plateau surface, for, particularly near the northern or mountainward margin, it is considerably dissected by streams which, flowing outward from their sources in the mountains, have carved long, usually nearly straight and parallel valleys down the dip of the plateau (Fig. 2). These valleys are cut to a depth of about 150 ft. at the maximum. As one follows them out toward the desert plain they gradually become shallower and finally dis- appear altogether. Degradation there gives place to aggradation. The gravel of the plateau is composed of rocks found in place in the mountains and the whole character and relationship of the deposit points clearly to its origin as a Piedmont accumulation of gravel spread out from the adjacent mountains to the north at some earlier time before dissection set in. 4098 JOHN LYON RICH While from its nature as a Piedmont accumulation, the gravel of the plateau has not suffered complete elimination of the less resistant elements, it is, nevertheless, an assorted mass in which rocks of the more resistant kinds strongly predominate. A list of a few of the more common of these will give a fair idea of the nature of the gravel and of the extent to which the more resistant rocks dominate. ‘The list follows: green quartzite, white quartzite, light-colored rhyolite, basalt, diorite, epidotized granodiorite, garnet rock, and magnetite from the Hanover ore deposits. Fic. 2.—Looking down one of the valleys which crosses the gravel plateau from the lowland to the desert beyond. Note particularly the character of the valley; its narrowness and lack of tributaries. Compare this with the broad valleys developed on the bed-rock of the lowland as shown in Fig. 3. Lone Mountain in the background. The coarseness of the material varies. Individual bowlders of large size are buried in a matrix of smaller bowlders, pebbles, and sand. This combination gives a rock of very porous nature, capable of absorbing quickly the water which falls uponit. At the same time the removal of the finer material of the matrix leaves the coarser bowlders and pebbles concentrated at the surface where they form a very effective protective covering—effective against either rain erosion, wind, or decomposition. The most conspicuous feature in connection with this plateau is the fact that it is now separated by a lowland from the moun- GRAVEL AS A RESISTANT ROCK 499 tains which supplied the material for its construction. Nor is this lowland the site of a stream valley. It runs, on the contrary, parallel to the strike of the beds and is crossed directly by the course of all the streams which flow from the mountains out through the dissected plateau to the desert beyond. A good idea of the nature of the lowland may be gained from the photograph, Fig. 3 (see also the map, Fig. 1). This shows the low- land in the foreground and to the left; the even-topped gravel plateau on the skyline; and, sloping down toward the observer, Frc. 3.—Looking southeast from the Central road three miles east of Silver City. This view shows clearly the even-topped gravel plateau and its inward slope toward the lowland in the foreground. the inner scarp of the plateau facing the lowland at the divides between streams. The view is looking southeast from the Central road three miles east of Silver City. On.the interstream ridges the difference in elevation between the inner lowland and the tops of the plateau surface varies between 50 and roo ft. In going northward along the tops of the divides, toward the mountains, one must travel from 3 to 2 miles before he again encounters ground as high as the tops of the gravel plateau. If the dip slope of the plateau surface is projected across the low- land toward the mountains the present land surface is not inter- sected within a distance of about 4 miles, on the average, from the 500 JOHN LYON RICH general line of the gravel plateau. Such a projection may be assumed to be a minimum original slope, for it makes no allowance for an increased slope of the plateau surface nearer the mountains, as must have been the case if the gravels once covered the lowland. This feature is illustrated in the two profiles shown in Fig. 4, drawn to scale from two different points well out on the plateau northward across the lowland to the base of the mountains. Amile E of Cross Mt /Jountains. Lowland Gravel limit Plateau Surface Lrmiles N of Apache Jee $000 Gravel Limit Lowland aoe = yy, 1 A 1 Horizontal Scale in Niles o ie Verétcal Scale 17 Feet. Fic. 4.—Profiles across the gravel plateau and the lowland from points on the desert to the foot of the mountains, showing the projected gravel surface and the relations of the gravels to the mountains. With topographic relations as they are at present the nearest possible source of the gravels of the plateau is separated from it by a lowland averaging 4 miles in width. It will be at once evident that, at the time of the formation of the plateau, the lowland could not have existed in its present relation. Any one of three things may have happened to bring about present conditions: (1) The gravels may have been removed by erosion from the area between their present limit and the mountains; (2) there may have been faulting by which the lowland was relatively lowered; or (3) the mountains GRAVEL AS A RESISTANT ROCK 501 may have worn back and the lowland developed by differential erosion since the deposition of the gravels. Opposed to the first of these alternatives is the fact that the gravel plateau ends abruptly along a relatively straight line. There are no outliers of gravel between this general line and the mountains. It is highly improbable that streams flowing nearly parallel and not more than a mile apart should strip all signs of the gravels from the upper four miles of their course, while in their lower course, where they flow across the gravel plateau, they should be in relatively narrow valleys with almost no tributaries and should have done little more than to cut their way through the plateau without having. been able to widen their valleys to any great extent (Fig. 2). A second objection is the fact that the line of contact between the gravels and the underyling rock slopes upward toward the moun- Plate = — sf Profile of Stream. Fic. 5.—Sketch showing the relation between the gravels of the plateau and the underlying rock which indicates that the gravels néver reached much nearer the moun- tains than now. tains at such an angle that it would intersect the projected line of the plateau surface at a point not far within the present limit of the gravels (see Fig. 5). In other words the gravels thin toward the mountains at such a rate that they would wedge out within a short distance from their present limit, and the lowland is accord- ingly developed in the bed-rock. A third objection is raised by the fact that in the gravels of the plateau there are aggregations or nests of huge lava bowlders, some of them 15 feet in diameter, which indicates that the moun- tians must at one time have been closer, for bowlders of such size are too large to be carried far by water, particularly by water flowing on a slope of too ft. per mile, which is approximately that of the plateau surface. The second alternative, faulting, seems highly improbable, for there is no evidence whatever of the presence of faults along the 502 JOHN LYON RICH line between the gravels and the lowland. At Silver City a tongue of the gravel plateau extends with accordant grade directly across the line of any fault which might have uplifted the gravels farther east. Further conclusive evidence of the lack of fault relation- ships of the lowland is furnished by the fact that the contact between the gravels and the underlying rock runs down the valleys and up across the interstream ridges in a perfectly normal manner, and with such wide divergence from a straight line as to preclude the possibility of an explanation by faulting. The failure of the first two hypotheses to account for the inner lowland leaves only the third, that of differential erosion. This calls for, first, the formation of the plateau as a Piedmont plain of accumulation at the base of the mountains; later, the cessation of active aggradation, possibly because of a lowering of the moun- tains through erosion; the initiation of a degradational phase of activity; and finally, the gradual erosional retreat of the mountain front and the reduction of the intermediate land at a rate faster than that of the gravels, leaving them standing in their present relations. Important in this connection is the nature of the rock composing the lowland. It is, in the main, a series of soft Cretaceous shales cut by dikes of a moderately resistant igneous rock: In parts of the lowland the shales are absent and the bed-rock is igneous. This, however, makes little difference in the nature of the resulting topography. Everything is worn down to a nearly uniform level lower than that of the gravels. A discussion of all the possible causes of the change in the phase of activity from one of aggradation to one of degradation would be out of place here. Two such may, however, be mentioned. The first is change in climate, the second, a lowering of the mountains by erosion with consequent relative increase in the factor of decom- position, over that of disintegration and transportation, brought about by the lessened slope. If the same process of differential erosion continues, the moun- tains will eventually become much reduced in height while the gravels, suffering less by erosion, will stand relatively higher and may finally come to dominate the topography of the surrounding GRAVEL AS A RESISTANT ROCK 503 country. With respect to the drainage to the south, the extent to which this process can be carried is limited by the base level of the interior basin to which the streams are tributary. A factor which must profoundly affect the topographic develop- ment of the whole region is the Gila River with its tributaries which, passing within 20 miles of Silver City, to the northwest, drains a large proportion of the mountain area. The Gila drains directly to the sea, and being a good-sized permanent stream, whose vailey is some 1,600 ft. lower than the gravel plateau, it is actively pushing its headwaters southeastward into the drainage area of the interior basin in which the plateau is situated. The divide, in one place, now lies only 6 miles from the gravel plateau and is only 4oo ft. higher. The Gila affords opportunity for the free removal of the waste from the mountains. Short and steep slopes combine to increase its effectiveness. Eventually the normal outcome of processes now in operation should be that the mountains would become lowered; the interior lowland between the plateau and the mountains would, by capture, become tributary to the Gila; and the plateau itself, remaining higher on account of its superior resistance to erosion, would ter- minate in a scarp overlooking the lower lands to the north. SIMILAR FEATURES IN OTHER REGIONS Other areas are known where gravel deposits of a nature similar to those on the plateau east of Silver City occupy a similar topo- graphic position and seem to show much the same history. A good example, with which the writer is familiar, is the Bishop conglomerate of southwestern Wyoming and northeastern Utah. This represents a Piedmont gravel accumulation derived from the Uinta Mountains, and at one time skirting entirely round their base. Subsequent erosion has so lowered the mountains that over considerable areas, particularly at the eastern end, they are actually lower than the tops of the gravel-capped plateaus which represent the eroded remnants of the Piedmont gravel deposits. This condi- tion has been described by the writer in an earlier paper.t_ Between «“The Physiography of the Bishop Conglomerate, Southwestern Wyoming,’ Jour. Geol., XVIII, No. 7 (1910), 601-32. 504 JOHN LYON RICH the mountains and the plateau are valleys sometimes Io to 15 miles wide, and as much as 2,500 ft. deep (zbid., p. 622). Other observers who have worked on the soul side of the range report similar conditions there. The resistant qualities of the gravel are particularly well illus- trated by the Bishop conglomerate. The plateaus have remained with little change while general erosion has lowered the surround- ing country nearly 1,000 ft. on the average. In point of origin and later development, the Bishop conglomer- ate is thought to represent exactly the same type of phenomenon as we have described from the Silver City region; the only difference being that, in the former case, the process has been carried farther and the results are just so much the more striking. CYCLE OF MOUNTAIN DEVELOPMENT If the above analysis is correct, as it Seems to be, both from the theoretical side and from field observation, the influence of gravel deposits is an important factor to be considered in the cycle of development of mountain topography. This-cycle is admittedly complex, involving many factors, but for the purpose of clearly presenting the point especially in mind at the present time, it is not necessary to follow each of the factors involved. On the con- trary, the consideration of the subject will be confined, as far as practicable, to a brief outline of the manner in which gravels, by reason of their selected nature, suffer less than other rocks. At the initiation of the cycle of mountain development let us postulate the following ideal conditions: A mountain range, or simple fault block of moderately resistant and varied rocks sharply uplifted above the surrounding country. Free drainage from the foot of the mountains to some base level, either of interior or of exterior drainage, lying at a considerably lower elevation. In order to give the maximum of favorable conditions, we will postu- late further that the climate is semi-arid so that vegetation plays a subordinate role. Granting these initial conditions, and assuming that there are no further crustal movements, let us trace the development of the mountain range. GRAVEL AS A RESISTANT ROCK 505 At first, with steep, exposed slopes, disintegration, through frost and insolation, and erosion will be rapid. The streams, while powerful enough to carry the loosened material down the steep slopes, will be unable to transport it across the lowland below. Piedmont fans of coarse gravel will accumulate along the mountain base. As time goes on the fans will continue to grow at the expense of the mountains. During this stage the fans are the seat of con- tinual deposition, the mountains of continual waste and removal. Finally there must come a time when the mountains have become so lowered that the streams are no longer flowing over steep slopes. As this stage is approached, disintegration and decomposition within the mountain area will become relatively more important and the rocks will be reduced to a finer condition before being car- ried off. The streams will no longer be overburdened with sediment too coarse to be carried beyond the base of the mountain. At ‘this point the upbuilding of the fans at the immediate mountain base must cease while the locus of deposition is shifted farther out because the stream load, being of a finer nature, may be carried to a greater distance before deposition occurs. This is the turning-point in the history of the mountain range. From now on, both mountains and fans will be subject to denuda- tion or degradation. If both the fans and the mountains were worn down at an equal rate, the whole area would merely lose in elevation without any marked change in the relations of mountains and gravels. Since, however, according to our thesis, the gravels will suffer from erosion less than the rocks of the mountains, differential erosion becomes an important factor. As the slopes decrease and decomposition plays an increasingly important réle while the material furnished to the streams becomes finer and less in amount, the burden of the streams becomes less and they are able to cut where deposition was in progress before, and will sink their channels into the Piedmont fans. Since the mountains are lowered faster than the gravels, a low- land will gradually develop, beginning first near the position of the inner margin of the gravel at the time of the change from aggrada- tion to degradation. If the base level of the streams is sufficiently low, this lowland may eventually come to include the whole of the 506 JOHN LYON RICH mountain area. If the streams crossing the Piedmont gravel fans sink deeply enough they may finally cut entirely through the gravel into the underlying rock. In that case we will have a plateau between the streams, capped by gravels of a composition corre- sponding to that of the rocks of the lowland which occupies the site of the original mountains, but lying at a level higher than the summits of these mountains as they now exist. Various combinations of factors will modify in different ways the course of development as sketched above, but the general principle involved should hold true, and the results should be in harmony with this principle as modified by the particular factors dominating in any one case. EXAMPLES ILLUSTRATING THE TYPE OF DEVELOPMENT ABOVE OUTLINED As examples of the influence of the slower differential erosion of gravel deposits the following may be mentioned: The region east of Silver City; the Uinta Mountains and the associated Bishop conglomerate, both described in the preceding pages. The Cats- kill Mountains of New York, in their relation to the old lowland to the east, are a possible illustration of the principle. THE CRETACEOUS AND TERTIARY FORMATIONS OF WESTERN NORTH DAKOTA AND EASTERN MONTANA A. G. LEONARD State University of North Dakota CONTENTS INTRODUCTION PIERRE SHALE Distribution Characteristics Fossils Yellowstone—Bowman County area Dawson County area Fox HILis SANDSTONE Distribution Cannon Ball River area Contact with Lance formation Little Beaver Creek area Dawson County area Variability of Fox Hills LANCE FORMATION Distribution South-central North Dakota area Section on Cannon Ball River Section on Heart River Bismarck section Divisions of the Lance formation Little Missouri area Contact of Lance and Fox Hills Plants Oyster bed Dinosaurs Yellowstone Valley area Glendive section Fossils Missouri Valley area Section on Hell Creek Fossils Age and relationship of Lance formation 5°7 508 A. G. LEONARD Fort UNION FORMATION Character of the beds Coal Plants Invertebrates Vertebrates WHITE RIVER BEDS White Butte area Little Bad Lands area Sentinel Butte area Long Pine Hills area INTRODUCTION The Cretaceous formations represented in the area under dis- cussion are the Pierre shale and Fox Hills sandstone. Overlying the latter is a non-marine formation, which has variously been called ‘‘the Ceratops beds,” “‘Lower Fort Union,” ‘‘Somber beds,”’ ‘““Laramie,”’ “Hell Creek beds,’ and °‘Lance formation.’’ (The United States Geological Survey has recently adopted the name “Lance formation,” derived from the term ‘Lance Creek beds,” which was applied to the deposits by J. B. Hatcher, and this name is employed in the following pages. The age of the Lance formation is still unsettled, some geologists regarding it as part of the Fort Union and thus early Eocene in age, while others believe that it included, or is part of the Laramie and is, therefore, Cretaceous. The Tertiary formations are represented by the Fort Union and White River. Western North Dakota is particularly favorable for the study of these formations, since they are excellently exposed in the Little Missouri badlands and along the valley of the Missouri and its tributaries. Bowman and Billings counties afford a con- tinuous section extending from the Pierre shale up through the Fox Hills, Lance formation, and Fort Union to the White River beds of the Oligocene, involving a thickness of some 2,150 feet of strata. The data here presented were gathered during seven seasons of field work in North Dakota and Montana, a portion of the time as assistant on the United States Geological Survey, and a portion CRETACEOUS AND TERTIARY FORMATIONS 509 under the auspices of the North Dakota Geological Survey. The work in Montana was confined mostly to Dawson and Custer counties; the Yellowstone River having been followed from its mouth to Miles City, and a trip being taken north from Miles City to the Hell Creek region and across the Missouri River to Glasgow. PIERRE SHALE The Pierre shale is exposed along the Missouri River for a dis- tance of about twenty miles north of the South Dakota line, or as far as the mouth of Big Beaver Creek in Emmons County; in eastern Montana it appears along the Missouri Valley from a point probably as far west as the Musselshell River to the station of Brockton, on the Great Northern railroad, or a distance of nearly 180 miles; it also occupies a small area on Little Beaver Creek in northwestern Bowman County, North Dakota, which is probably continuous with the Pierre outcrop on the Yellowstone River, twelve miles above Glendive. The Pierre formation is a bluish gray to dark gray, sometimes almost black, jointed shale, which often weathers into small, flaky fragments. The rock commonly shows yellow spots or stains of iron oxide. The topmost beds of the Pierre contain numerous calcareous concretions varying in size from a few inches to six and eight feet in diameter. Some of these concretions are rich in invertebrates, which are characteristic of the upper forty or fifty feet of the Pierre, while others are barren of fossils. Many are cut by a network of calcite veins which are commonly lighter colored than the matrix. The following species, identified by Dr. T. W. Stanton, were collected in the Little Beaver Creek locality, Bowman County, North Dakota: Ostrea pellucida M. and H. Lunatia. Avicula lingueformis E. and S. Anisomyon patelliformis M. and H. Inoceramus cripsi var. barabini Mor- Margarita nebrascensis M. and H. ton. Fasciolaria? (Cryptorhytis) flexi- Chlamys nebrascensis M. and H. costata M. and H. Yoldia evansi M. and H. Pyrifusus. Nucula cancellata M. and H. Haminea ? occidentalis M. and H. Lucina occidentalis Morton. - Scaphites nodosus Owen vars. brevis Protocardia subquadrata E. and S. and plenus. Callista deweyi M. and H. Nautilus dekayi Morton. 510 A. G. LEONARD From the locality on the Yellowstone, at the mouth of Cedar Creek, the following marine shells were secured, from the upper fifty feet of the Pierre: Avicula nebrascana M. and H. Scaphites nodosus Owen vars. brevis Avicula linguaeformis E. and S. and plenus. Inoceramus sagensis Owen. Limopsis parvula M. and H. Inoceramus cripsi var. barabini Mor- Yoldia evansi M. and H. ton. Lucina subundata M. and H. Modiola meeki E. and S. Protocardia subquadrata E. and S. Veniella subtumida M. and H. Dentalium gracile M. and H. Callista deweyi M. and H. Vanikoro ambigua M. and H. Anchura americana E. and S. Margarita nebrascensis M. and H. Haminea occidentalis M. and H. Fasciolaria (Piestocheilus) culbert- Pyrifusus newberryi M. and H. soni M. and H. Lunatia concinna M. and H. Baculites ovatus Say. Scaphites nodosus var. quadrangu- Nautilus dekayi Morton. laris M. and H. Chlamys nebrascensis M. and H. The beds which outcrop at the latter locality on the Yellow- stone, twelve miles above Glendive, Montana, are brought above river level by an anticlinal fold, the dip of the strata here being 20. S. 52° W. The Bowman County outcrop is probably caused by the same anticline, since the strike of S. 38° E. shows that the fold so well exposed on the Yellowstone, if continued in that direc- tion, would include the Little Beaver Creek locality. That the two areas of outcrop are continuous seems probable from the fact that ammonites and other marine shells are reported to have been found at several intervening points on Cabin and Cedar creeks. There are extensive outcrops of Pierre shale along the Missouri River and its tributaries in the northeastern corner of Montana, in Dawson and Valley counties. At the mouth of Big Dry Creek, fifteen miles south of Glasgow, the shale rises 200 feet above the river, and it is also well shown on most of the creeks entering the Missouri from the south for a distance of eighty or one hundred miles west of the Big Dry. Among these is Hell Creek, on which 150 feet of Pierre are exposed above creek level. Among the most common fossils occurring in the calcareous concretions of this locality are ammonites and baculites. CRETACEOUS AND TERTIARY FORMATIONS 511 In the southeastern corner of Custer County, Montana, as a result of the Black Hills uplift, the Pierre shale outcrops over an area of considerable extent, overlying the Benton and Niobrara formations, which also appear at the surface. FOX HILLS SANDSTONE The Fox Hills sandstone is the most recent of the marine for- mations of the Great Plains region. It is very variable in character and undergoes considerable change in composition and appearance from one locality to another. It is exposed on the Missouri River as far north as old Fort Rice, about eight miles above the mouth of the Cannon Ball River; it appears on Little Beaver Creek, a tributary of the Little Missouri in southwestern North Dakota; on the Yellowstone a few miles above Glendive, Montana; it occurs in the Hell Creek region, and also on the Missouri River, near the town of Brockton, Montana. On the lower Cannon Ball River, for a distance of ten or twelve miles above its mouth, the Fox Hills formation is exceptionally well shown. In many places it forms cliffs rising abruptly from — the water’s edge, and the cuts made for the new branch line of the Northern Pacific afford many excellent exposures. It rises eighty to ninety feet above the Cannon Ball River, or approximately 1,080 feet above sea-level. The Fox Hills sandstone when unweathered is gray with yellow patches, but in weathered outcrops it is yellow or brown in color. The rock is rather fine-grained and, for the most part, so soft and friable that it can be.crumbled in the hand. Cross-bedding is very common and the rock contains great numbers of large and small ferruginous sandstone concretions or nodules, many of these likewise exhibiting cross-bedding. The nodules are apparently due to the segregation of the iron into irregular patches cementing the sand into firm, hard masses, considerably harder than the sandstone in which they are imbedded. In many places the iron has impregnated certain layers and formed indurated ledges, which resist weathering and project beyond the softer portions (Fig. 1). The nodules vary in size from an inch and less to six and eight feet. Small, irregular, twisted or stem-like forms are 512 A. G. LEONARD abundant at certain points. Some portions of the rock are so completely filled with these brown concretions that they constitute the main bulk of the formation, and the gray, loosely cemented sandstone forms a kind of matrix in which the hard nodules are imbedded. In the process of weathering these more resistant nodules project far beyond the softer rock, and at the base of slopes and scattered over the surface they are exceedingly abundant. Fic. 1.—The Fox Hills sandstone on Cannon Ball River, North Dakota, showing hard ledges and concretions on a weathered surface. Where the rock has only a few concretions, and therefore, where the iron has not been segregated to as large an extent at certain points, the sandstone is of a yellow color, due to the disseminated iron oxide. On the other hand, where the brown ferruginous nodules are thickly scattered through the beds, the rest of the rock is gray, the iron having been largely leached from it and concen- trated in the nodules. Many of the latter are of good size and spherical in shape, and it is these which have given its name to the CRETACEOUS AND TERTIARY FORMATIONS 513 Cannon Ball River, since they occur abundantly along that stream. The following Fox Hills fossils were collected on the Cannon Ball River about ten miles above its mouth:' Tancredia americana M. and H. Avicula nebrascana E. and S. Callista deweyi M. and H. Protocardia subquadrata E. and S. Tellina scitula M. and H. Mactra warrenana M. and H. Ostrea pellucida M. and H. Mactra ? sp. Avicula linguiformis E. and S. Scaphites cheyennensis (Owen). The first three in the above list occurred in a bed of sandstone forty feet below the top of the formation, while the others were from a higher horizon, ten feet below the top of the Fox Hills. About three miles below this locality specimens of Mactra warrenana M. and H., Dentalium gracile M. and H.? and Cinulia cincinna (M. and H.) ? were .collected. On Long Lake Creek, a tributary of the Missouri River, from the east the sandstone yielded the following: Avicula linguiformis E. and S., Tellina scitula M. and H., and Chemnitzia cerithiformis M. and H.? The contact of the Fox Hills sandstone with the overlying Lance formation is well shown in the bluffs on the north side of the Cannon Ball River, about ten miles above its mouth. Here the two formations are seen to be conformable, the top of the Fux Hills being marked by a light gray, almost white sandstene, which exhibits cross-bedding (Fig. 2). This bed is one foot to eighteen inches thick. Sedimentation was apparently continuous from Fox Hills time on into the period when the Lance beds were being formed. East of the Missouri River, in Emmons County, the sandstone is present on Beaver Creek, extending up the valley of that stream almost to Linton, and having an elevation of nearly 150 feet above the creek, near its mouth. About 160 miles west of the Missouri River, the Fox Hills sandstone is exposed in a small area on Little Beaver Creek, in the northwest corner of Bowman County, North Dakota. The section here is as follows: t Identified by Dr. T. W. Stanton. 514 A. G. LEONARD Feet Sandstone, massive, light greenish gray, weathers to yellow color...... 50 Sandstone ledge, yellow:..252 220 sss ees ane haus chaos Ue ates che 8-10 Clay, sandy, finely laminated and formed of alternating light and dark laminae. Contains nodules of iron pyrites. Exposed above creek Cig: RR Rae he ee RE Un enn ye RIM 5 linn agi am It Nenad a oe 25 In this upper sandstone, Dr. T. W. Stanton collected several marine fossils characteristic of the Fox Hills, including Leda Frc. 2.—The Fox Hills and Lance formations on the Cannon Ball River. The contact is at the hard ledge on which the man is standing. (Yoldia) evansi, Tellina scitula, Entalis? paupercula, and Haly- menites major. Where exposed in bluffs along Little Beaver Creek, at several points the gray sandstone shows an uneven, eroded surface, which the writer has described as an unconformity.*. It may, however, be due to the action of currents in the shallow sea of Fox Hills time, as suggested by Dr. Stanton, in which case no long t Fifth Biennial Report, N.D. Geol. Surv., 44. CRETACEOUS AND TERTIARY FORMATIONS 515 time interval between the deposition of the sandstone and the overlying Lance beds would be indicated by the eroded surface of the Fox Hills. . In the vicinity of Iron Bluffs, on the Yellowstone twelve miles southwest of Glendive, Montana, the Pierre is overlain by 150 feet of sandstones and shales, the age of which is in doubt, though the beds have the stratigraphic position of the Fox Hills. The lower seventy-five feet is composed of shales and sandstones while the upper half is formed of a brownish sandstone. The only fossils found in these beds at this locality are some plants, which are too fragmentary to be identified. The Fox Hills sandstone is well exposed on Hell Creek, a trib- utary of the Missouri River in northwestern Dawson County, Montana. Lying above the dark gray Pierre shale, with its fossiliferous concretions, are 1oo feet of shales and sandstone belonging to the Fox Hills. The formation is here composed of light gray to yellow, more or less sandy shale, with some layers of nearly pure sandstone. About eight feet below the top, there is quite a persistent bed of fine-grained yellow sandstone with a thickness of eleven feet (Fig. 3). The beds are lighter in color and, for the most part, more sandy than the Pierre shale. From concretions near the summit of the Fox Hills on Hell Creek, Mr. Barnum Brown collected the following shells :* Cardium (Protocardium) subquad- lLunatia concinna M. and H. ratum E. and S. Cylichna scitula? H. and M. Nucula cancellata M. and H. Baculites ovatus Say. Tellina scitula M. and H. Scaphites conradi Morton. Yoldia evansi M. and H. Chemnitzia cerithiformis M. and H. Crenella elegantula M. and H. Mactra? nitidula M. and H. Piestochilus culbertsoni M. and H. Actaeon (Oligoptycha) concinnus M. Anchura (Drepanochilus) americana and H. 513; BNC! Along the Missouri River valley, over too miles northeast of Hell Creek, and near the station of Brockton, on the Great Northern Railroad yellow sandstones interstratified with gray clay are found overlying the Pierre.?, These beds are probably to be referred to t Bull. Am. Mus. Nat. Hist., XXIII, 827. 2 Carl D. Smith, Bull. U.S. Geol. Surv., No. 381, 38. 516 A. G. LEONARD the Fox Hills formation. Their thickness is about 200 feet and they are well exposed in the river bluff south of Brockton. Asa rule the sandstone is soft, but in places there are hard concretion- like masses, which after weathering stand out as ledges or as cannon-ball shaped masses imbedded in a matrix of softer rock. The material shows much irregularity of bedding, is in places cross-bedded, and is extremely variable in character horizontally. Fic. 3.—The Fox Hills formatioa on Hell Creek, Moatana, showing sandstone ledge (A) near the top. The Fox Hills sandstone probably occurs also about the Pierre shale area in southeastern Custer County, Montana. The variability of the Fox Hills formation is well illustrated by the foregoing description of its outcrops. In some places, it is composed wholly of sandstone, in others it is mostly a sandy shale, while in still others it is partly sandstone and partly shale. When shales are present they are generally arenaceous and are commonly CRETACEOUS AND TERTIARY FORMATIONS ley most abundant toward the base of the formation, where, in some places, they pass gradually into the Pierre shale. It will be seen from the above lists that some of the fossils occurring in the upper part of the Pierre range up into the Fox Hills. The top of the latter is better defined than its base, the change from it to the overlying Lance beds in some places being abrupt, but generally the two are conformable. The Fox Hills beds vary in thickness from seventy- five to two hundred feet. LANCE FORMATION The Lance beds have a wide distribution in North Dakota and eastern Montana, as well as in northwestern South Dakota and northeastern Wyoming. The largest area in North Dakota is in the south-central part of the state, where this formation occupies a large part of Morton county and all of the Standing Rock Indian Reservation, outside the Fox Hills and Pierre outcrops; east of the Missouri River, it covers southern Burleigh and the greater part of Emmons County, together with adjoining portions of Kidder, Logan, and McIntosh counties. In the southwestern corner of North Dakota is a second smaller area stretching along the Little Missouri River for a distance of over fifty miles in western Bowman and southern Billings counties. In eastern Montana the Lance beds are found along the Yellowstone River from the vicinity of Forsyth to a point about fifteen miles below Glendive. South of the Yellowstone, these beds are exposed along the valleys of the Powder and Tongue rivers and their tributaries. The badlands occupying a wide strip of country on the south side of the Missouri River in northern Dawson County are for the most part formed of Lance beds, and they extend as far east as Brockton. According to C. D. Smith™ the formation is found on the Fort Peck Indian Reservation, and the beds also occur west and north of the reservation in Valley County, Montana. South-central North Dakota area.—In Morton County, North Dakota, numerous good outcrops of the Lance formation appear along the Missouri, Cannon Ball, and Heart rivers and many of the smaller streams (Fig. 4). The beds are found along the Mis- t Bull. U.S. Geol. Surv., No. 381, 39. 518 A. G. LEONARD souri River to within eight or ten miles of Washburn, where they disappear below river level and are replaced by the Fort Union. On the North Fork of the Cannon Ball they extend almost as far west as the Hettinger County line, and on the Heart River they reach to within a few miles of the Stark County line. Along the boundary between North and South Dakota the western border Fic. 4.—Bluff of Missouri River near old Fort Rice, showing the lower Lance formation. of the formation is not far from Haynes, on the Chicago, Milwaukee and Puget Sound Railroad. In passing down the North Fork of the Cannon Ball River from the western edge of Morton County to the junction with the South Fork, and thence down the Cannon Ball River to its mouth, one traverses the entire thickness of the Lance formation from the Fort Union above to the Fox Hills below. About ten miles below the Hettinger County line; inisec. 5, (eiasey Nee Ren com\Venmtne CRETACEOUS AND TERTIARY FORMATIONS 519 contact of the Fort Union and Lance beds is well shown in the following section, exposed in a high bluff of the river: Feet Inches 15. Shale, light gray and yellow, to top of bluff......... beens ta 16 AROMA CHOCOlatE DOW an choi hers We cessive ais aie spe be ele I 6 RE SUAS, INVENT Facey se aOR ee Re 6 12. Shale, light yellow, soft, and readily crumbled............... 6 TER 5. UME, WEA aE rg UR a CE 15 HO, (COM. s 5, grasa ese Ais eck oe ecu ea Er eae 4 2 Oo SMG, CAPS Seg coe cic aaa eC EOL Re a a eae a 23 Bo (COfilly 5 6 bed Bie Sie SBN Abts See tae ee Sire Oa 2 FEMS Welle ate args ee as Tice TRIG, 6 caus arasei ns wn ¥ aches aieue, coe I 6 Ga C Oa nnn pr crm Mir Men NU reine SMe tl Do, Mia edie nid’ lulls Bb arail Bi FeAl sanyo Mtoe raya was. Aas sce cece Wed Slehaey aang wee Slew ike) Amma (© ©) ci see oso ER Sas sevens Gran. Salles or ocen 0 Hie de I 6 Beroanestonesuehtvorayicross-bedded...2..:....2.65s6 540200: 25 Paonalewsandy sorOwi. with MUCAMTOM, 2.05 ..0. 00.500. 6. 000 er 4-6 1. Sandstone, soft, yellow, with concretions and some thin limoni- HICEStheakcuexpOsedsaMOve MIVEES 2am 4 dasa. sce agen. 50 163 9 Nos. 1, 2, and 3 of the above section belong to the Lance for- mation, while the other members are Fort Union. As is the case at a number of points, a coal bed (No. 4) occurs at the contact, and there are also two workable beds above this. The upper sandstone of the Lance formation extends down the river seven or eight miles below this section, forming in many places vertical cliffs rising from the water’s edge. Then a dark shale appears beneath the sandstone as shown in the following section, which is seen about ten miles below the previous one, in secs. 29 and Bon 23Ne Re 83) W-: Feet Koa ndstonesottsvellowsitontopror blithe is. tee. aS nae hk eee 20 4. Shale, dark gray to black, alternating with thin-bedded, shaly sand- SEO TV Cree eric See Mee a rear ta ned as ia AI Socal cy Rial a al shasta vant ins rants yelianegele I5 Bennalewdarksoray toublack. when Mmoist,.:.4. 92-206. 6222. ss ee els 70 PeSandstones yellows wathehard ledge neat top... .W.c-css. cs 4500s ee oe 20 1. Shale, dark gray to black, sandy, exposed above river................ 25 150 Only twenty feet of the upper sandstone of the Lance formation appear at this point, and the bluffs are here formed largely of the underlying black shale. 520 A. G. LEONARD The beds near the middle portion of the Lance formation are well exposed in the bluffs on the south side of the Cannon Ball River near Shields where the following section occurs: Feet Inches Soil and subsoil: 23. sen ees iis ee oe one ee rs 4-5 Sandstone, yellow to gray. soft and friable... ..:.. 25... 31 Shaleveray amd) yell owe en iene ececrenpe oe) ie tem ee yee eect? ie) Sandstone, gray and yellow, containing thin shale layers and brown, carbonaceous: streaks 0%).3) otc ss Meee 38 Shale, gray, contaminguiron COncretions. 8 9... Gece ciat ce pels 6 Shale, black and brown, carbonaceous; containing dark brown fEFRUMINOUS*COMCKELIONS: 20) le sess ss Geese ee 6-10 Shales gray = cuts Me ram acter eter ee eer eee 15 6 Shale-brown, carbonaceous tinct at se tate ta canes east I Ghale,jeraly sarily sect. ei na etc Greta ae al eae ye 3 Shale: brown, carbonaceous.) viet vec a. ein gilts eign ane ge te 18 Cae ee ee eats etre fe aeoalee catoetul St ar elimi Tn UCR iy La cc eae ae 6 Shalewblackscoallivic en sac teen ete sO Uae mes et eens I 4 Shalesrorarya Samclyse, tease tte tas Sc tele ton anaes ce geen ee a Bees II Shales brow Cat bOnaceOUsi.a Miva ec) scks cr eae ae I 6 Shale every, SAMGy cass es che vie tcc Gegupntere) acta aha eon cement I 6 Shale “browimncarbonace Osis!) ss s.cec het cene ere weenie eee 2 6 Ghia Fesigrreh yoeeee icant Cue ae tea et notes cutie Inked get ern a a le metre eae etc ae 7 6 Sandstone, gray, soft, with shale layer near middle, 2-4 feet (oN (el ey ORM Ea ROME RAD Siena MLL one UME cAtn on UI earn aitabte Ale ato 44 Glyale sera: cue eas cetceteres ay ae alc ie ate di ae ge eee ees 8 We xpOSCCs LO MIVEL, tetas ore we ee uthrenen eect ee eee eee 20 Bo) 22° Up vere vene ee LOM et ca eR oria eens OE, AI nV ORNA aida oMint ce 210 One of the characteristics of the Lance beds, exhibited in many widely scattered localities, is well shown in this section; namely, the many brown, carbonaceous layers which are present, often forming a conspicuous feature of the formation, as along the Little Missouri River. It will be noted that the beds are here composed about evenly of shales and sandstones, though the latter are con- fined to three thick members. . About thirty miles below Shields, and ten or twelve miles above the mouth of the Cannon Ball, the lower Lance beds are exposed, together with the underlying Fox Hills sandstone, as shown in the following section: CRETACEOUS AND TERTIARY FORMATIONS 521 eet AD atisCpoTaye aT ySam asic pees cea ceayaiene ee aasuck geile i TANI Ree Bin eee 2 Sinaleteclania colored ere treeg ccs eter aie SN dns. OM el na moe ewe Melee ae 27 Sandstone, soft, with many thin, brown, carbonaceous laminae......... 11 SAMAS LOM CMe IO WARSONE us Seat che ane Tw inaee at Sy Sh eee 16 Shale, brown, carbonaceous, with two coal seams, one 3 inches and the CRM che mil CMes mtn Chon Peet ya nM sm ht gr. cia tate ath Gace seh gee may 8 SIMBUG, CARBINYc: : 3.88 dso fate tne OY ae RLY Te ae 3 SAINGISTHOMNG STEN Le 5 Ooi lees eR eg Se RRR eee a ea eG 8 SIMGIKS, PAREN 6a 5 eR coc aN sea ese er Soc are ie a ree gO 4 Shale wkOwlrCa TD OMACeOUSsai ieee icseeye wuss ae sig Hees in eel eke: 3 Sandstone and shale in alternating layers, the former predominating; colorsadarkyerayeeprown, andeyellow? 2.5. .<..2 +4 lotsa oe dee Sa Shalesidarkerave witha few, browm bands... ..0:.0......¢c.1.s. ss eu: 22 Sana scome re NO xg elt ll Semon erence Ate ae tl iat ox. ule sctcte Manele thins 80 TROLL ga ec seth ea eae Nr Oe eg ER gt LA 241 From the lower sandstone of this section, Fox Hills shells were collected. The Lance beds here rest conformably on this sandstone, and there appears to have been a gradual change from the marine conditions of Fox Hills time to the fresh-water conditions under which the Lance beds accumulated, with continuous deposition throughout. The strata forming the upper 350 feet of the Lance formation, comprising the upper, massive sandstone, and the underlying dark shales, are very well exposed in the valley of the Heart River, in Morton County. For a distance of five or six miles below the bridge on the Glen Ullin—Leipzig road, this valley is a narrow gorge walled in by sandstone cliffs. This rock, which forms the upper member of the Lance formation, is-a massive, gray, brown, and yellow sandstone, having a thickness of approximately one hundred feet. The underlying shales are dark gray to black, when moist, and weather to a yellow color. They are cut by several sets of joint cracks and along these cracks the change from gray to yellow first takes place, the gray, unweathered material being left in the areas inclosed by the joints. Near the surface the shales are weathered and oxidized throughout, but at some depth the yellow color is confined to narrow bands on either side 522 A.G. LEONARD of the joint cracks. In places, these beds are composed of thin layers of black shale and gray, very sandy shale, or sand. On the Heart River, south of Almont, the following section appears, embracing portions of both the Lance and Fort Union formations: Feet 5) ‘Sandstone; yellow, Soft, massive a9. ter nen or eee errs ete eee 50 a. Shales yellowandélight gray <4 sso we es ea ete a ee 61 gt Sandstone x white ales © sts, sista ease oe ceo tne cere ae a hs ee ee 30 2. Sandstone, yellow and brown below, gray toward top. The upper sandstone.ot the Wancetormatione: .1c,s5 ener) eee eee ee 95 tShales. darkccolored es eerie ata etic ane ot tern ate ayant eee 180 416 Nos. 1 and 2 belong to the Lance formation, while the three upper numbers are Fort Union. On the Heart River in the vicinity of Mandan and in the bluffs of the Missouri near Bismarck, the Lance beds are made up chiefly of the dark shales, as is evident from the two sections which follow. The first is exposed at the east end of the Northern Pacific railroad bridge over the Missouri River. Feet Inches Drift, resting on the eroded surface of the Lance formation. .... 15-20 Shale, dark gray to black, with thin, light gray streaks; cut by many joint cracks several inches apart. Faces of the joints StaimedsbyiPONk o0s.uc cecrti taeie aoe ease eee ne ie 42 Shales sandy blackcams tits atone: che vena tc aes beeen iio ie Cree I Slialles Polack oes 08 Ws SAE anda eee ayens ite tare adam in tog ua ers 3 6 Sandstone darkieray toiblack ye. anne valence ei ce seeks stellen i Shales Polack & aioe seh 0) ue haart Sy ine are fare eotea eae neaine tat on ost eS 2 6 Sandstone vello we): y5 cece sheers Sev cso mcrae ee eee eee eae 4 Shale, dark gray to black, alternating with yellow, fine-grained sandstone-and! sandy shaleacia 672, aeoem te ces ee tine 22 Slate: Polack es es Le ae NY rec Mee tack Uda nada stay 30 Umexposeditomiviersle vel seis ae ete ie se ae ease ei Seca ae 15 MO CAN Sh sche kG Pe Sel Ne Sine Mea: ane PAU i ng ON ee I4I The second section appears on the south side of the Heart River two miles above Mandan, and is as follows: CRETACEOUS AND TERTIARY FORMATIONS 523 SOilPS an Cliygmarmery rere ts Cr fe ae al Rei el aa Me ee ee 3 Scams CISTOCE Me mpstenetn tise irate Rey tata Gt Mat) ah adele a eetet spent 20 Shale, gray and black, mottled; arenaceous in part, the sand being very fine; sandy layers have yellow color. Some por- tions contain considerable carbonaceous material, which gives the rock its black color. Shale cut by several sets of joints running irregularly in many directions, but all making large angle with the horizontal. These joint cracks are filled with gypsum and the sides stained with iron. The mottled character shows on weathered face of the bluff, where there are large blotches of black on the gray surface............ 28 Shale, dark gray and yellow, some layers sandy; more thinly peddedkthanioverlyingimemlber. fyi. pels. ee eal. ee oe 7 6 Sandstone, soft, fine-grained, gray, and yellow................ 7 6 Sandstone, argillaceous, forming hard projecting ledge........ 2 Shale, dark gray to black, alternating with bands of laminated, AIM e=- ST AINE HClO wWeSAMGe cay, eset chet le rie es aba hese nals as 3 Shalerdarkveray to blacks whenmnoist <).%. 206). 000s ons sews ) 6 Sandstone’ soft and incoherent, yellow... .29.505.......5..04 I Winexposeditopriviermevielim artic etn Ne aac cis mane stalls an 3 27 eS ° = i) He 4 e) oO oO In the vicinity of Long Lake, in southeastern Burleigh County, and in the railroad cuts along the Linton Branch of the Northern Pacific, in northern Emmons County, the sandstone and shales of the Lance formation are well exposed, and they outcrop at a ‘number of points about Linton. The eastern boundary can be determined only approximately on account of the heavy mantle of drift, which covers the bed-rock. The Lance formation of south-central North Dakota, as shown on the foregoing pages, consists of three members: an upper sand- stone about one hundred feet thick, a middle member composed of dark shales with a few sandstone layers and having a thickness of 200 to 250 feet, and a lower member made up of shales and sand- stone in alternating layers. This latter member has a thickness of 350 feet or over, and the maximum thickness of the entire Lance formation is probably not far from 700 feet in this region. Fossils occur sparingly in this area. =1.526—1.531. The extinction angle on (o10) is +5 to +9, and on (cor) o to +1: The characteristic undulatory extinction is well marked. The plane of the optic axes is approximately perpendicular to (oro), and the negative acute bisectrix is nearly normal to (201). The apparent optical angle measured on the section nearly parallel to (201), by Mallard-Becke’s method, is 84° 44’ and 2V is 52° 20’, the mean index of refraction being assumed as m»=1.528. The dispersion is p>v. Fic. 1 Fic. 2 The groundmass feldspars are also alkaline feldspar and occur in elongated prisms, simple Carlsbad twinning being commonly present. They are arranged as in trachytic fabric and the fluxion is especially marked around the phenocrysts. Aegirine-augite, as phenocrysts, is almost absent, and the largest crystal, which was observed in 5 thin sections, measures 1.5 mm. in length, but the average length of the prisms is o.2 mm. The mineral is of bluish-green color, and is somewhat pleochroic from bluish-green to the same color with yellowish tint. The greatest extinction angle measured with respect to the ¢ axis, gave 45°. As inclusions magnetic grains are common; and brown glass, appatite needles, and feldspar laths can be detatched, the last being very scarce. 558 S. KOZU Olivine occurs in some specimens, as a very rare accessory in an anhedral form. Magnetite is very scarce as phenocrysts. Minute euhedral to anhedral crystals are disseminated in the groundmass, and form about 4 per cent of the whole. It is also associated with the aegirine-augite. Apatite is conspicuous as very minute needle-shaped crystals. Chemical characters —Separate analyses were made of the rock and of the porphyritic anorthoclase. For the purpose of analysis of the mineral, the phenocrysts were picked out of the weathered rock, in which the minerals remain on the surface in a favorable state to be taken off from the matrix. The feldspar material is quite fresh, but the surface and the inner portions along the parting and cracks are stained by decomposed products from the matrix and inclusions. To purify it as much as possible, it was crushed into 1-2 mm. grains and was digested in dilute hydrochloric acid at 80°C. for 24 hours, until it turned white in appearance. But an intimate association with impurities rendered it impossible to prepare a thoroughly clean sample, so that the results of analysis are somewhat unsatisfactory. The chemical analysis, made by S. Kawamura in the laboratory of the Imperial Geological Survey of Japan, is as follows: ok ON a a Ran Ee elie Aen dec net ate 64.98 Ya © eee eaten Seu rede, Geri ner Emenee: 7 19.62 Pes @ eres bortiheee, aia Re Sauer se reer anes ue cee eae 0.98 NT Oe ee es MORN, Aa 1 kare ae en dean ieee 0122 CAO rasp) clots Maaco ede ck kee eae oot Sane 3.48 INE ag os ina cation the aay Ce a ne 4.86 TOE Paps Reece cana gigs Mel cara ance te aera s, Ral eee 5.83 99.93 By the withdrawal of the excess of silica, lime, magnesia, and iron as impurities mainly due to inclusions, the chemical composi- tion of the mineral is shown approximately by the following ratios: Si Qs eee cerca cep ce re tered Nea tage pres ee 63.08 Ta a Par rae ictal iy ease Dae rd ire Red PS ce 21.80 EY © LPS Ma RR rss tu, EME ei Na hn aR aN ee 8 3.24 [iE © FANE Reem entity hme a UE ain cece ARAL 5.39 UO Bien ecoregion 3 fea ye 6.49 NOTES ON SOME IGNEOUS ROCKS OF JAPAN 559 From these figures the formula of the anorthoclase is found to be Or,.,3Ab;Any. The analysis of the rock, by K. Takayanagi, and that of the pantelleritic trachyte, by Fornstner, are given in the following table: * Loss on ignition. A=Soda-trachyte, Matsu-shima, Kytsha. B=Augite-andesite (pantelleritic trachyte of Rosenbusch), Porto Scauri, Pantelleria. The norms, calculated from these analyses, are as follows: Ratios from the norms OuartzZaw.... Diopside. . . Hypersthene Magnetite. . liimvenitereens patiter usc. €20/ oa . OVS 09 OV ST Bet: (Or et 99-3 B 43 5 ie ed: 30 54 45 BPAPS -95 OV HK J (oy oy (OP Say sy 99.51 mn i) CF fp IS AS (6) 4.90 560 SKOZE! In the quantitative system, the rock from Matsu-shima would be classified under the name of laurvikose, near pulaskose. There is a close resemblance in chemical characters between this rock and the pantelleritic trachyte which was described by Fornstner as augite-andesite, and is also laurvikose. The relationship between them is shown in the above tables. PRELIMINARY NOTES ON SOME IGNEOUS ROCKS OF JAPAN. IP S. KOZU Imperial Geological Survey of Japan Il. QUARTZ-BASALT Locality.—Kasa-yama, near Hagi city, prov. Nagato. Occurrence.—The rock occurs as a lava flow erupted at the vol- cano Kasa-yama, which consists merely of an isolated cone of small size, 112.5 meters above the sea-level and about 1,300 meters in diameter across its base. In the summit, there is a perfectly preserved crater, 25 meters in diameter and 13 meters in depth. This small and regular cone stands in strong contrast to the topog- raphy of the environs, where the geology is mainly composed of granites and mesozoic sedimentaries, and especially to that of table-lands or flat islands formed by basalt flows which poured out here and there through the ground. Age.—The eruption of the rock appears to be Diluvium and the latest of the basalt in this region, which seem to have been erupted at the period from the close of Tertiary to Diluvium. Megascopic characters—The specimen collected from the lava dam near the Shinto shrine at the eastern foot of Kasa-yama is noteworthy for containing abundant quartz as porphyritic grains in a hypocrystalline groundmass. It is black in color and vesicular with small and irregular cavities, but has a high specific gravity. The quartz, varying in size from 1 mm. to 5 mm., shows an irregular outline, but sometimes almost hexagonal. Though the percentage of quartz grains varies in different portions of the lava, generally they are distributed uniformly and are clearly dis- tinguishable from the groundmass by their color, as seen in the photograph. Besides these there are only a few crystals of yellow ‘Published by permission of the Director of the Imperial Geological Survey of | Japan. 501 562 S. KOZU olivine as megascopic phenocrysts. This mineral is 2 mm. in diameter, and is also fresh in aspect. Microscopical characters —The mineral components are olivine, augite, plagioclase, magnetite, and apatite, with phenocrystic quartz. The microscopic phenocrysts are not abundant; among them the olivine is most common, then follows the augite in nearly equal amounts; the plagioclase occurs subordinately. The ground- Frc. 1.—Quartz-basalt. 3. The white grains are quartz. mass is hypocrystalline in texture and consists of lath-shaped plagioclase, prismatic or granular augite, and magnetite crystals, with abundant interstitial glass of light-brown color, clouded by numerous globules. Olivine belongs to the earlier crystallization among the mineral ingredients of the rock, and is almost free from inclusions with the exception of a few crystals of magnetite and glass, which are very rare. It forms anhedral to subhedral shapes with finely ragged outline, and about it minute granules of pyroxene may be observed. NOTES ON SOME IGNEOUS ROCKS OF JAPAN 563 The olivine is entirely fresh and remarkably but irregularly cracked. Augite is very faint yellowish or nearly colorless, and is more abundant than olivine, though it is rarely present as microscopic phenocrysts. As phenocrysts, it is anhedral, but in the groundmass it is well shaped; elongated prisms are common. In rare instances, twinning parallel to the orthopinacoid may be seen in the larger crystals. In the reaction-border about the phenocrystic quartz, augite is the only mineral constituent and is imbedded in brown glass. Inclusions of mag- netite, apatite, and glass are sparingly present. Plagioclase is basic lab- radorite and appears in well-formed, long _ pris- moids with polysynthetic twinning according to the Carlsbad and albite law. Zoning is almost absent. Minute grains of pyroxene and magnetite are present as inclusions in small quantity, with also a few of glass. Quartz occurs as a por- phyritic constituent, and the average diameter is about 2mm. The outline of the mineral in thin section is usually irregular, but sometimes shows the bipy- ramidal form referable to crystallographic faces, as seen in the microphotograph (Fig. 2). Each grain of quartz is fringed with a reaction-border, consisting of elongated prism and grains of augite imbedded in brown glass. The minute prismoids are arranged quite regularly. They are grouped radially, each group _ containing a few crystals that converge toward the outer side of the border, as seen in Fig. 2. In triangular, interstitial spaces between each radial group granular augites are scattered irregu- larly. In some instances, the deep invasion of the brown glass, Fic. 2.—Bipyramidal quartz with reaction border. X23. 564 S. KOZU with very fine crystals of augite, is observed along the cracks in the quartz. Glass inclusions with gas bubbles are present in bipyramidal shapes, and ruptures starting from the four corners of the rhombic sections are well marked in thin sections of the mineral nearly parallel to the optic axis, as shown in Figs. 3 and 4. FIG. 3 Fic. 4 Chemical characters —The analysis of the rock, shown in column A in the following table, was made by T. Ono in the laboratory of the Imperial Geological Survey of Japan. The analysis of the quartz-basalt from the north base of Lassen peak, described by Diller, is given in column B. A B SIO eA ay Cas ee, ene oa eee ee 56.08 SOn Sr We G) ar encarta et eeaes n R ey nl ns ehe Tike} 16 18.10 He Ohad uterine. eee nen ae 2.46 4.26 KeOr - 6.97 2.68 Mg @ x Sie esa tees be eh cee Bare Avis? OF Ieee ambit Nene meta Pais oyster ps th Ones INA ON Sais ore ht tins lc ca ben Megereee: 2.02 Bn KOs eae eka crise aerate kamen a 1.50 se 30 1c DO We eran PIN Nec aren er hese Onis 0.69 BETO er eus ede ci ate rele ae. aan Toit 0.48 PAO) ea, RR LW cee hk eal a EE tr O.14 IMI OM eee Seay nee ena 0.34 O.11 Ba QE ces 2a Sen ear Mine terete 0.04 Cr Ors eee ie er a ey ea arse tr SHOE ee spe ene ieee Wa di aecie aa te 0.04 1 DSU G Teme Sg nie ale ree es ca hema ca tr QQ. 22 100.10 A. Quartz-basalt (lava). Kasa-yama, prov. Nagato. B. Quartz-basalt (lava). North base of Lassen Peak, California. * Loss on ignition. NOTES ON SOME IGNEOUS ROCKS OF JAPAN 565 On comparing the analyses, it is obvious that the two rocks are closely similar in chemical characters, but the rock of Kasa- yama differs slightly in containing lower magnesia, lime, and alka- lies, and higher iron and titanium; the low value of the first com- ponent especially does not satisfy Harker’s hypothesis with regard to the plotting of his diagram. Norms, calculated from the analyses, are as follows: A B OE CUA SENG Scie Gs Ae ee 14.8 10.9 Orthoclasevee sentir foe fleck eee 8.9 he D AM OTHE Seo LAN i ie cca eo ea ae 16.8 26.7 ANOLON SOUNDS Bnd Bre cin Se ae Ram 35°83 31.4 GCoruidl wine Gece 55 Se 0.2 an IDO) ORTGIS 2G 5 aoe cate ee are ese ate 7.1 ela DETSEMECIMC Miia tee asain oc 16.9 8.6 IMACS os Geir fa ane Baul 6.3 TMS MTS eee setae s nee ne Dai 0.9 99.0 99.1 A B al ite” pHOCFooGd HO GMO GOdooU dG On 3.30 aS %S) 2 = Ss ca R EER CARMI Neat EAD ectct EE ER 0. 24 Only K.0’+Na,0’ Cad’ areata ict a aCe ae 0.38 Ons K,0’ NaO’ cc 0. 50 O25 By the Quantitative System these rocks are classified as bandose. PRELIMINARY NOTES ON SOME IGNEOUS ROCKS OF JAPAN. IE S. KOZU Imperial Geological Survey of Japan III. ALKALI-FELDSPAR-BEARING BASALTIC ROCK (FUKAE-GAN) AND ALKALI-FELDSPAR-BEARING BASALT Localities —Alkali-feldspar-bearing basaltic rocks were collected from Fukae-shima in the Goto Islands; alkali-feldspar-bearing basalts from Madara-shima, an islet northwest of the Yobuko port, prov. Hizen; from O-shima, near the Island of Iki; from Uramino-taki, near Omura city, prov. Hizen. These localities are in the northern part of Kydshd or its outlying islands. Occurrence.—The rock type, associated with olivine-basalt on the one hand and with soda-trachyte? on the other, appears to have an extended distribution over the northern part of Kytsht. At Fukae-shima, this rock group forms the plateau and some striking dome-shaped hills standing on it, as seen in the photo- graphs (Figs. 1 and 2). There are well preserved or strongly breached craters in the summit of each dome. The hills, in great part, consist of ashes, lapilli, and slaggy lava, in which finely shaped bombs may be found abundantly. The plateau is of hard lava. Age.—Near the close of Tertiary to Diluvium. The specimens for the following descriptions were collected by the writer from Fukae-shima; by Y. Otsuki from Madara-shima and O-shima; and by D. Sat6 from Uramino-taki. They may be classified in two groups by the mineralogical and chemical char- acters. I. Alkali-felds par-bearing basaltic rock (Fukae-gan in Japanese) .—— The rocks of this group collected from Fukae-shima are transi- t Published by permission of the Director of the Imperial Geological Survey of Japan. 2 Jour. Geol., XTX (1011). 566 NOTES ON SOME IGNEOUS ROCKS OF JAPAN 567 tional forms in both texture and mineralogical characters, owing to their crystallinity, but are closely alike in chemical properties. They are represented by three types as described below. As the first type, a bomb ejected from the volcano Ondake: was selected for the following description and chemical analysis: View from the northwest Fic. 1 Megascopic characters.—This rock type is aphanitic and black in color. Scattered magnophyric feldspars are the only con- stituents visible to the naked eye. The groundmass is vesicular, and the vesicles are small and round. The phenocrysts are of fresh aspect and show rather euhedral forms, short prismatic or tabular, and in some instances are of considerable size, reaching 20 mm. in length (Fig. 3). Olivine, which is present as abun- dant microscopical phenocrysts, is scarcely recognizable even by the aid of a lens. 568 SOLU Microscopical characters —The microscopic phenocrysts in this type of rock are of olivine in great part, with subordinate andesine. The groundmass is hypocrystalline and is filled with small and round vesicles (Fig. 4), looking like the outline of leucite. Feldspar.—This mineral shows distinctly two different habits. The phenocrysts are stout prismoid, sometimes tabular, and some- ha: Oshima ALS Z Koitabeshima Oilabeshema ——_———___. SSS = —————. Midake Usudahe pa ee ee View from the north west Fic. 2 what rounded. They are andesine (Ab,An,), with a mean index of refraction slightly higher than ny=1.554. They contain abun- dant small particles forming an outer zone, and round it usually, thin and clean layers with different composition, but the difference between each layer is not pronounced. Twinning is scarce, and in rare instances undulatory extinction can be observed. Feldspars forming the groundmass are slightly more sodic than the pheno- crysts, and occur in elongated or rather slender shapes. They are NOTES ON SOME IGNEOUS ROCKS OF JAPAN 569 marked by irregular cracks, filled with isotropic, low refractive, and colorless, substance. Twinning according to the albite law is common. Olivine occurs in two-sized crystals. The larger ones are abun- dant and play an important réle as microscopic phenocrysts. Their shapes are equant or prismoid. In many instances, the outlines of crystals are irregular by invasion of groundmass, sometimes Fic. 3.—Andesine phenocryst, natural size extremely narrow and deep, parallel to crystallographic faces, showing the successive growth of the mineral, but the general out- lines are referable to crystal forms (Figs. 5 and 6). Though distinctly cracked they are entirely fresh and inclose clouded glass, but are free from other inclusions. Augite forms magnophyric crystals which are very rarely seen in hand specimens. The minute grains in the base, showing high refraction, appear to be augite. 570 S. KOZU Magnetite clouds the base as minute grains or dusty particles, and their abundant presence affects the color of the rock. Apatite Fic. 4.—Microphotograph of the first type of the first group, magnified 30 times. The minerals seen in the figure are olivine micro- phenocrysts and andesine prisms. appears mostly as inclusions in feldspar. The second type collected from Ohama in Fukae-shima is more crystalline. Alkali- feldspar appears locally in the crystalline part as the border of the plagioclase of the groundmass. The augite crystals are comparatively few and occur in small anhedral forms. The magnetite crys- tals are more numerous in this type of rock than in the first one, are somewhat larger, but are also anhedral in shape (Hig. 7): The third type collected from Masuda in Fukae-shima is holo- crystalline, and in some parts has typical ophitic texture. The IESTGsn5 Fic. 6 Fics. 5 AND 6.—Showing irregularly outlined olivines. X55 mineral components are andesine, alkali-feldspar, augite, olivine, magnetite, and apatite. The andesine is distinctly cracked, with NOTES ON SOME IGNEOUS ROCKS OF JAPAN 571 invasion of colorless, low refractive and isotropic substance as in the above types. The alkali-feldspar occurs as the border of almost all crystals of andesine in the groundmass. The augite is light purple in color, and is xenomorphic toward plagioclase. The mag- netite frequently occurs in crystal form. Il. Alkali-feldspar-bearing basalt.— This group differs from the above in the presence of labradorite in the place of andesine, as the essential component. The specimen from Madara-shima is dark reddish gray in color with semiwaxy luster. It is holocrystalline, fine granular, and Fic. 7.—Microphotograph of the Fic. 8.—Microphotograph of the second type of the first group. X3o. third type of the first group. X3o. inconspicuously porphyritic, with not abundant magniphyric feldspar and less pyroxene. Under the microscope the rock consists of labradorite, alkali- feldspar, augite, olivine, titaniferous iron ores, and apatite. The labradorite is subhedral to euhedral, twinned according to the Carlsbad and albite laws, and commonly prismatic in shape. Zonal structure is rarely seen. Each of the feldspar crystals com- posing the groundmass is enveloped by a shell of alkali-feldspar. The augite is light greenish yellow with purple tinge, and is sub- hedral to anhedral, stout prismatic to equant. The larger ones are indistinct phenocrysts; the minute grains are interstitially dis- tributed in the groundmass with magnetite crystals. The olivine 572 SUKOZE as microscopic phenocrysts is subhedral to anhedral, and altera- tion into iddingsite is commonly visible along cracks and in marginal portions. The texture of the groundmass is somewhat inter- sertal, and is characterized by divergent arrangement of prisms of plagioclase enveloped by alkali-feldspar, with interstitial gran- ules of augite, olivine, and magnetite (Fig. 9). A more distinctly crystalline and coarser type is a specimen from O-shima, an islet, near the Island of Iki. Megascopically the rock, more or less decomposed, is evidently holocrystalline, but the indi- vidual crystals are scarcely recognizable, though the di- verse arrangement of prismoid feldspars, 1.5 to 2 mm. long, is well marked in the hand specimen. The color is light gray, On account of the abundant feldspars, and is dotted by dull reddish brown spots produced by decompo- sition of the olivine. Rare, Fic. poe Micronnotormaph of the alkali- inconspicuous phenocrysts are feldspar-bearing basalt from Madara-shima. tabular, white feldspar; irreg- A38; ularly shaped black augite; and equant, dark reddish olivine. All of them are less than 3 mm. in diameter. Under the microscope (Fig. to), the texture is transitional from doleritic to intersertal, as the augite is xenomorphic toward feld- spar in one case and automorphic in the other. The mineralogical constituents are as before, but the presence of broad bands of alkali- feldspar enveloping labradorite is especially noticeable (Fig. 11). In some crystals, the alkali-feldspar has more than three times the volume of the labradorite, that is,o.75 mm. in length and 0.09 mm. in width (Fig. 11). In general, the labradorite is in extremely elongated prisms, twinned according to the Carlsbad and albite laws. The augite is anhedral to subhedral, prismatic to equant. In color it is light purple. The magnetite frequently occurs in NOTES ON SOME IGNEOUS ROCKS OF JAPAN 5S anhedrons 0.35 mm. in diameter. The apatite is noticeable in elongated prisms. The most finely grained variety is from Uramino-taki. It is light gray, compact with a few vesicles, and nonporphyritic, some- times with nodular olivine. There are groups of scaly, blackish brown mica in the vesicles. Under the microscope it is almost holocrystalline and granular. The mineralogical components are the same as in the previous variety, with a small quantity of biotite, which usually occurs in cavities. The biotite is reddish brown and strongly pleochroic. Its apparent optical angle (2£) varies between 37.5° and 29.5°. Fic. to Fic. 11 Chemical characters —Of the first group of rocks a complete analysis was made of the first type (Bomb) and two partial analyses of the second (B) and third (C) types, by K. Yokoyama. Of the rocks of the second group a complete analysis of a specimen (D) collected from Marada-shima was made by T. Ono. The three analyses A, B, and C of the first group show a close relationship in chemical characters, notwithstanding they have different mineralogical components due to their crystallinity. For- eign rocks that have a close resemblance in chemical characters with the rocks of this group are olivine basalt (E) and orthoclase- bearing doleritic basalt (F) of New South Wales, described by 574 S. KOZU G. W. Card. For the sake of comparison, these analyses are given in the following table with those of the rocks under consideration. A B c D E FE G SiO sn eee | 48 . 33 49.15 48.70 | 52.109 48.98 53-21 49.24 ALOT see | 16.29 eer ore 19.74 16.88 17.84 15.84 HO ec Bed: 4.72 3.30 3.80 6.09 FeOne ea eee 8.73 6.28 7.29 522 7.18 Me Qin ances ease. ©. 2.24 5. 27 2.96 2.02 CaO si es ek. gare: er S250 6.99 8.86 6.48 5.26 INE WON: Garant 5 Peo 50) 3.64 885 3.48 3.30 3.36 Gai COE See tale eet er I.49 1.61 TA 2.04 2.11 3.03 Pe io) HO ares. et Raateaped / 0.52 0.65 1.08 HOS fy oeee S| a8 | noe | aes LO ae ine) real n.d. n.d. 0.06 0.02 n.d. LO Hu a iouia ea nane | 2.40 n.d. 1.28 I.O1 1.84 | AO VAG Wee ti se le MOR7O) n.d. ©.30 0.44 TeAU7, Ro) el guar te Mian. (iemesayelt | n.d. none 0.09 n.d. Clana a creer fee angele Ve n.d. 0.02 o.1I n.d. Vin Oe eae Ori n.d. 0.31 0.32 0.29 ClO aaa Ming 0.06 0.06 Ba On2r Potalge warn 99.99 99.99 | 100.41 99.88 | 100.46 SPiGas oie: 2.562 | on 2.869 2.768 2.79 Bomb ejected from the volcano Ondake, Fukae, the Gotd Islands, Ana- lyst, K. Yokoyama. Lava erupted from the volcano Ondake, Ohama, Fukae, the Goto Islands, Analyst, K. Yokoyama. Lava erupted from the volcano Ondake, Masuda, Fukae, the Goté Islands, Analyst, K. Yokoyama. A B C D. Lava, Madara-shima, prov. Hizen, Analyst, T. Ono. E. Olivine-basalt, one and a half miles north of St. George’s head, N.S.W. F. Orthoclase-bearing doleritic basalt, south side of Croobyar Creek, N.S.W. G. Mugearite, Druim na Criche, 5 miles S.S.W. of Portree, Skye. The norms calculated from these analyses are as follows: A D E F G GIO Hae Anew vee st WA mle Wiig aa ep oe ee 3.2 eat 4.9 see Orthoclase Wer Seatac ohlnap alsa ah aleher is suceaes 8.9 11.7 12.2 17.8 12.2 Albite._ KATO R OE re are melo eee 30.4 20.3 28.8 20.7 44.0 RUNTIME Grado cnguavecdvabose ds 23.9 32.3 24.7 24.5 13.6 Sodium Chlorid Gees sis mca mone SS ane aan 0.4 pep Diopside science eerie mo 2 14.6 3.0 3.4 Hypersthene Ran te einen a Tented Wk 2.7 12 Ras 10.8 Teal Olivine. SPOON AE Rc nace ears ee II.5 9.7 te 7.6 Magne titer tetra tr eer heals 6.7 4.9 5.6 8.8 Imenite EMS ey arr ote et a rip ee aie hints 4.6 2.4 2.0 Bx IADALIEO Saree ceca ee ae ieee 1.9 0.7 1.3 Boi 99.5 97-5 98.0 97-9 97-3 NOTES ON SOME IGNEOUS ROCKS OF JAPAN 575 Ratios calculated from the norms are as follows: | A D E F G Sal | | ae Wie ae Fem tne 1.74 3.64 2.03 | 3.16 2.51 Q4+L | | ESS eee ence 0.04 2.009 x0 Reo Popa soAsAbe Soomecn meme cones 0.86 | 0.66 | 0.87 0.99 2.16 = SSR GOS COEUR ARNO aoe Ree ee mie On2 Seo) Ons 0.40 0.59 0.26 By the quantitative system, A, D, E, and F would be classified under the name andose and G under akerose. From the tables given above, it is clear that the rocks from Fukae differ from normal basalt, in containing a high percentage of alkalies in proportion to the silica contents, especially of soda, forming normative andesine Ab,;sAn,,, which is slightly more calcic than the modal plagioclase. Though the alkali-feldspar is not present as a recognizable mineral in the first type (Bomb) of the first rock group, its molecule is to be looked for in the glass-base. In the second and third type, the alakli-feldspar is seen in the modal state. The chemical resemblance between the rock of Fukae, A, and the olivine-basalt from St. George’s Head, E, is very close. The differences between them are lower potash for orthoclase, slightly higher soda for plagioclase and higher normative ilmenite in the Fukae rocks, compared with the olivine-basalt, of St. George’s Head. Generally the rock is characterized by properties mineral- ogically and chemically intermediate between the mugearite, G, described by Harker, and the olivine-basalt described by Card, though it is very near to the last rock, and it differs from shosho- nite described by Iddings in being dosodic. The rock from Madara-shima, D, differs slightly from the Fukae rock in the lower value of magnesia and in higher percentages of silica and potash, and of alumina which increases the normative anorthite. It has a close resemblance in chemical characters to the orthoclase-bearing doleritic basalt, F, from Croobyar Creek, New South Wales, described by Card. . REVIEWS Gypsum Deposits of New York. By D. H. NEWLAND AND HENRY LEIGHTON. New York State Museum Bulletin 143, Albany, TOUOs = Da OAe The bulletin presents a concise but complete description of the gypsum deposits and the gypsum industry of the state of New York. The workable deposits are restricted to the Salina state of the upper Silurian and are pretty generally confined to a single formation of this series, the Camillus shale. The geology of the Salina series is carefully and clearly set forth. Considerable attention is given to general questions relating to the origin of gypsum, its properties, and the theory of its transformation into plasters. The reviewer is pleased to note that the section devoted to the description of mines and quarries is much shorter than is usually found in a report of this character. 1d IR Jog Report on a Part of the Northwest Territories Drained by the Winisk and Attawapiskat Rivers. By Witi1amM McINnnEs. Geol. Survey of Canada, No. 1008. Pp. 54; Figs. 5; Map 1. In this report the author gives the results of a reconnaissance survey of the country to the southwest of Hudson Bay. Adjacent to the bay there are gently folded Silurian limestones and dolomites, probably of Niagaran age. Outside this belt comes a belt of bowlder clay 160 miles in width, overlain by post-glacial marine clays, which, below the Boskineig fall in the Winisk River, have an altitude of 350 feet above sea-level. Beyond this again is the Laurentian peneplain, of Archean granites and schists. This is the customary rocky-lake country, heavily drift covered in places. Glacial striae on exposed rock surfaces indicate a glacial movement toward the S.S.W. The writer also gives a general description of the canoe routes, flora and fauna of the country, climate and possibilities of agriculture. HeCuc 576 i i nT ee Keep The Dust Down In schoolrooms and gymnasiums it is of the greatest importance that everything be done to benefit the health even in the smallest degree. The harmful effects of too much dust in the air are well known. The constant shuffling of feet in the schoolroom, and the more violent exercise in the gymnasium, stir up dust and circulate it in dangerous quantities. It is of the greatest importance to general health that the amount of q this floating dust should be reduced. IDAIR \\\ \« x NK Ww ane : i N WH es IN N holds down all dust that settles, and prevents its circulation in the air. Standard Floor Dressing is a special preparation, and vegetable and animal germs cannot find subsistence in it. They are held = down and swept away at the end of the day. Many schools, symnasiums and stores, where the importance of reducing dust was | “| recognized, have been quick to avail themselves of the properties of Standard Floor Dressing. | Illustrated booklet sent free—A booklet on “Dust Danger and How to Avoid It” will be mailed to you free immediately upon receipt of your request. It contains much valuable information and is a book you should have. Not intended for household use. Standard Oil Company (Incorporated ) o>] STANDSRD DL COMIN Sy GET THE GENUINE Baker’s Chocolate Sy 3S ic ae ha La nro 8 SS WAY : coy vere BAN | a wil LIKE F230: Blue Wrapper — Yellow Label Trade Mark on the Back FINEST IN THE WORLD For Cooking and Drinking WALTER BAKER & CO. Ltd. Established 1780 DORCHESTER, MASS. PIANO in your home free ef expense. Write for Catalogue D and explanations, 7 3 live on filth and leave Flies it everywhere. They spread disease and their presence is an indication of unsanitary condi- tions. They can not exist in the house that is properly screened and disinfected. The Odorless Disinfectant A colorless liquid, powerful, safe, and economical. Sold in quart bottles only, by druggists, high-class grocers and house-furnishing dealers. Manufac- tured by Henry B. Platt, New York and Montreal. have been established over 60 YEARS. By our system of payments every family in moderate om cumstances can own a VOS€@ piano. We take old instruments in exchange and deliver the new piano iy tS 2 VOSeC & SONS PIANO CO., Boston, Mass. i iiss VOLUME XIX NUMBER 7 THE JOURNAL or GEOLOGY A SEMI-QUARTERLY EDITED By THOMAS C. CHAMBERLIN AND ROLLIN D. SALISBURY With the Active Collaboration of SAMUEL W. WILLISTON ALBERT JOHANNSEN WILLIAM H. EMMONS Vertebrate Paleontology Petrology Economic Geology STUART WELLER WALLACE W. ATWOOD ROLLIN T. CHAMBERLIN Invertebrate Paleontology Physiography Dynamic Geology ASSOCIATE EDITORS SIR ARCHIBALD GEIKIE, Great Britain’ GROVE K. GILBERT, National Survey, Washington, D.C. HEINRICH ROSENBUSCH, Germany CHARLES D. WALCOTT, Smithsonian Institution THEODOR N. TSCHERNYSCHEW, Russia HENRY S. WILLIAMS, Cornell University CHARLES BARROIS, France JOSEPH P.IDDINGS, Washington, D.C, ALBRECHT PENCK, Germany JOHN C, BRANNER, Stanford University HANS REUSCH, Norway RICHARD A. F. PENROSE, Jr., Philadelphia, Pa. GERARD DEGEER, Sweden WILLIAM B. CLARK, Johns Hopkins University ORVILLE A. DERBY, Brazil WILLIAM H. HOBBS, University of Michigan T. W. EDGEWORTH DAVID, Australia FRANK D. ADAMS, McGill University BAILEY WILLIS, Argentine Republic CHARLES K, LEITH, University of Wisconsin OCTOBER-NOVEMBER, 1911 CONTENTS HEI TOUMNN ORE T 40 Sons 8 ee ie Oe Gaieoer (CALVIN PEE WENORO OW ASOSTASY «202, 2 \ 2 2e4 6s, 72 = ee SAR MON| Lewis SPECULATIONS REGARDING THE GENESIS OF THE DIAMOND Orvittr A. DERBY PRELIMINARY NOTES ON SOME IGNEOUS ROCKS OF JAPAN. IV - - -S. Kézu FACTORS INFLUENCING THE ROUNDING OF SAND GRAINS - - Victor ZrectER THE UNCONFORMITY BETWEEN THE BEDFORD AND BEREA FORMATIONS OF | NORTHERN: OHEO <5 200-0 sea eZ - - - - WILBUR GREELEY BURROUGHS TTC IRTUNG sy a EUR ane ge Nee eames mt sia WO on SIMCDINES TRUTSILICC 00 (On Rea 2 elas ee na EY CS Ae The Untversity of Chicago press CHICAGO, ILLINOIS AGENTS: THE CAMBRIDGE UNIVERSITY PRESS, HoNDoN AND EDINBURGH WILLIAM WESLEY & SON, Lonpon TH. STAUFFER, LeErpzic THE MARUZEN-KABUSHIKI-KAISHA, Toxyo, Osaka, Kyoto 577 603 627 632 645 655 660 661 669 The Fournal of Geology Published on or about the following dates: February 1, March 15, May 1, June 15, August 1, September 15, November 1, December 15. : Vol. XIX CONTENTS FOR OCTOBER-NOVEMBER, 1911 No. 7 THE IOWAN DRIFT 2) 32) St ie ee Oe eee ee Oo SAMUEED CALVING ta77 THE THEORY OF ISOSTASY - - - - - .- 9- = = = “=. =. = =. Harmon Lewis 603 SPECULATIONS REGARDING THE GENESIS OF THE DIAMOND - - = = Orvitte A. DerBy 627 PRELIMINARY NOTES ON SOME IGNEOUS ROCKS OF JAPAN. IV - - - - -S. Kézu 632 FACTORS INFLUENCING THE ROUNDING OF SAND GRAINS - - - - - - Victor ZIEGLER 645 THE UNCONFORMITY BETWEEN THE BEDFORD AND BEREA FORMATIONS OF NORTHERN OHIO- -- - - ~- ee =) WILBUR GREELEY BURROUGHS 655 EDITORUAL§ (oe eo Se age AE Re GORE ea mcd SPN eke ar Ato ee be ee eee ee (Coe eG) REVIEWS? 2 200 Si oss ieeeeaion navi cola. aeea tng taal ew inh ep Naat <2 een eee We) ea OIE RECENT PUBLICATIONS (5 uctoe ict ap Ea eee ON Oe lg 2 SVE ee aie ate eto og The Journal of Geology is published semi-quarterly. §| The subscription price is $3.00 per year; the price of single copies is 50 cents. | Postage is prepaid by the publishers on all orders from the United States, Mexico, Cuba, Porto Rico, Panama Canal Zone, Republic of Panama, Hawaiian Islands, Philippine Islands, Guam, Tutuila(Samoa),Shanghai. ‘j Postage is charged extra as follows: For Canada, 30 cents on annual subscriptions (total $3.30), on single copies, 4 cents (total 54 cents); for all other countries in the Postal Union, 53 cents on annual subscriptions (total $3.53), on single copies, 11 cents (total 61 cents). | Remit- tances should be made payable to The University of Chicago Press and should be in Chicago or New York exchange, postal or express muney order. If local check is used, 10 cents should be added for collection. The following agents have been appointed and are authorized to quote the prices indicated: _ For the British Empire: The Cambridge University Press, Fetter Lane, London, E.C., and William Wesley & Son, 28 Essex Street, Strand, London. Yearly subscriptions, including postage, 14s. 62. each; single copies, including postage, 2s. 6d. each. For the continent of Europe: Th. Stauffer, Universitatsstrasse 26, Leipzig, Germany. Yearly sub- scriptions, including postage, M. 14.80 each; single copies including postage, M. 2.60 each. For Japan and Korea: The Maruzen-Kabushiki-Kaisha, 11 to 16 Nihonbashi Tori Sanchome, Tokyo, Japan. Yearly subscriptions, including postage, Yen 7.00 each; single copies, including postage, Yen 1.25 each., Claims for missing numbers should be made within the month following the regular month of publication. The publishers expect to supply missing numbers free only when they have been lost in transit. Business correspondence should be addressed to The University of Chicago Press, Chiesa. Ill. Communications for the editors should be addressed to them at The University of eiieaee Chicago, Ll. Contributors are, requested to write scientific and proper names with Berio care and in citations to follow the form shown in the pages of the Journal. Separates, if desired, must be ordered in advance of publication. Separates of leading articles will be ‘printed as ordered, of which 50 (without covers) will be furnished gra/is, the actual cost of the remainder (and covers, if desired) to be paid for by the author. The table below shows the approximate cost of separates consisting of plain text or text with line engravings. The actual cost may vary from the figures given, and will depend upon the amount of work in re-making the pages into forms, press work, paper, binding, etc. Separates containing half-tones may be expected to cost somewhat more than the rates given, the increase depending upon the number of cuts and the amount of work required upon them. Number of copies 50 IO0O 150 200 letter=press. for 4 pages onlessy Va0%. 7 vss tes $1.40 $1.60: $1.80 $2.00 Wetter-presss or S,, pases On lessa seer. esse 1.90 220) 2.50 2.80 Wetter-press, for 16\pagesior less7) = ee lh 3.20 4.00 4.60 5.30 Single plates (1 double=2 single). . . . . .90 1.10 1.30 1.50 Covers, with title (paper like Journal cover) . T. 20 1.69 seen e200; 2.40 Entered March 20, 1893, at the Post- Office at Chicago, Ill., as second-class matter, under Act of Congress, March 3, 1879. Ce ine ae Es ROUWRNAL OF GEOLOGY OCGROBER-=NOVEMVMBER, ror THE IOWAN DRIFT SAMUEL CALVIN CONTENTS INTRODUCTION Is there an Iowan Drift ? Has the Term Iowan Been Correctly Applied ? CAUSE OF CONFUSION IN USE OF THE TERMS KANSAN AND IOWAN EFFECT ON TEXTS AND Maps oF MAKING CERTAIN PROPOSED CHANGES IN USE OF THE TERMS EVIDENCE CONCERNING THE IOWAN DRIFT AND ITs GEOLOGICAL RELATIONS 1. The Iowan Drift IS 2. The Iowan Drift is Young as Compared with the Kansan 3. The Iowan Drift is not a Phase of the Kansan 4. The Iowan Drift has very Intimate Relations to Certain Bodies of Loess 5. The Iowan Drift is not Related to the [linoian INTRODUCTION Is there an Iowan drift? Whatever the reader may think about it, the question seems to be in order. Three papers have recently appeared in which the Iowan drift receives more or less attention." Two of the papers go so far as to question the very existence of such a sheet of drift as that which geologists have for Frank Leverett, ‘‘Weathering and Erosion as Time Measures,” American Journal of Science, XX VII, May, 1909. , “Comparison of North American and European Glacial Deposits,”’ Zeitschrift fiir Gletscherkunde, IV, Berlin, 1910. T. C. Chamberlin, ‘‘‘Comparison of North American and European Glacial Deposits,’ by Frank Leverett;”? a review of the second paper, Journal of Geology, XVIII, No. 5, July-August, roto. Vol. XIX, No. 7 SET 578 SAMUEL CALVIN some time been calling Iowan. The third paper raises the question whether, even if there is such a drift, the name it has been wearing should not be applied to something else. It is possible that the questions raised by these papers may never be settled to the satis- faction of everyone, because men do not always see alike; but a few facts bearing on the subject may be worthy of consideration. CAUSE OF CONFUSION IN USE OF THE TERMS KANSAN AND IOWAN The doubt as to the correct use of the terms Kansan and Iowan is the one that deserves first and most serious attention. This doubt has arisen naturally and for admittedly good reasons, but it is all due to the fact that in the earlier discussions of the Pleisto- cene deposits of northeastern Iowa it was assumed that there were but two drift sheets east of the Wisconsin lobe which occupies the north-central part of the state. The two supposed drifts were named by McGee? the Upper and the Lower till. The view that there are but two tills in this area was adopted by Chamberlin in his classic contribution to the third edition of The Great Ice Age, by James Geikie,? and the name East-Iowan was given to what was assumed to be McGee’s Upper till, while what was taken to be the Lower till was called Kansan. There are, however, three drift sheets in the region, and the attempts to describe three formations in terms of two led to confusion. In some cases the upper and middle sheets were described as a unit; in others the lower and middle were treated as one; much more frequently the lowest was ignored, and the descriptions of the “lower” and “upper” tills were drawn from the other two. The presence in certain localities of a forest bed or interglacial gravels, which it was assumed always lay between what the authors described as upper and lower tills, as East-lowan and Kansan, complicated matters still further. There are, indeed, many positive references in the original texts to this forest and gravel horizon—since called Aftonian—as the plane of separation between the two drift sheets : Paper on “The Pleistocene History of Northeastern Iowa,’ by W J McGee, Eleventh Annual Report of the United States Geological Survey, Part I, 1891, pp. 189-577; and other papers by the same author. 2 The Great Ice Age, 3d ed., chaps. xli and xlii, ‘Glacial Phenomena of North America,” by Professor T. C. Chamberlin, pp. 724-74, 1895. THE IOWAN DRIFT 579 at that time credited to the region; but if the texts relating to the . subject are carefully read and the maps published in connection with them are examined, it will be seen that the view that the lower till, the Kansan, lies below the Aftonian is untenable. For example, the description of the materials and prevailing color of the upper till on p. 476 of the Eleventh Annual Report is true for only the third of the drift sheets and is at variance with the facts if intended to include the middle till. The sdme is true of the reference to the large granite bowlders as ‘“‘the most conspicuous element of the upper till,’ on p. 481. On the other hand, the characteristics assigned to the lower till in the comparisons made between it and the upper on p. 479, are all features that belong to the middle drift sheet; in no true sense are they descriptive of the sub-Aftonian. It is true that at the end of the paragraph there is a reference to the ‘‘forest bed” as a plane of separation between the upper and lower tills, but the characters which the author saw and so cor- rectly and graphically described belong to a super-Aftonian till and to nothing else. If now we turn to the chapters on “‘ Glacial Phenomena of North America,” contributed by Professor Chamberlin to Geikie’s Great Ice Age, we shall see again how the preconception that there were but two drifts where three actually exist, led unavoidably to con- fusion. It is no reflection on anyone that such confusion crept in in the earlier discussions. Some things are unavoidably overlooked ° by the pioneer who opens up for us new fields of science, and we can but admire the genius and the insight of the masters who taught us how to read the complicated history of the Pleistocene deposits of the Mississippi Valley. As in the Eleventh Annual Report, so in the Great Ice Age, it is two super-Aftonian tills that are most frequently referred to in the text, and most accurately represented on the map opposite p. 727 as, East-Iowan and Kansan. The distinguishing characteristics of the upper and middle drift sheets could not be more clearly or more succinctly stated than is done on p. 760 of the work cited where, speaking of the East-lowan, it is said: In Iowa the granitic types predominate. Immense bowlders are freely scattered over a portion of the surface. As greenstones prevail in the lower 580 SAMUEL CALVIN till, there is a petrological as well as a stratigraphical basis for separating the two formations. .... The most notable feature of this drift sheet is its connection with the main deposits of loess. .... The East-Iowan till sheet is, however, associated with loess of such exceptional extent and nature as to make this epoch especially notable on account of this relationship. As already stated, the till graduates at its edge into loess that spreads away from its border. The statements quoted can be interpreted in but one way. They are true in case the term East-Iowan was applied to the uppermost of the three till sheets in Iowa east of the Wisconsin moraine. ‘There is just one till in which “granitic types predomi- nate.”’ There is just one of which it can be said that ‘immense bowlders are freely scattered over a portion of the surface.’”’ There is just one that bears the described relation to the loess. While greenstones occur in all three of the till sheets of the area under consideration, it is in the middle sheet that they are most con- spicuously prevalent. The statement on p. 756 of The Great Ice Age, clearly implied if indirectly made, that “the Kansan formation emerges from beneath the overlapping East-Iowan formation to the extent of 200 miles at the west”’ can apply only to a sub-Iowan, but super-Aftonian drift. It cannot possibly apply to the sub-Aftonian for the reason that at the time it was written the known natural exposures of the sub-Aftonian were confined to a very limited area. A number * of outcrops of this formation have since been recognized and recorded, but it may be questioned whether the aggregate of all the now known exposures of sub-Aftonian would be equal to more than one or two square miles. Certainly there are no known areas of anything approaching 200 miles in extent in which the lowest of our drift sheets emerges from beneath anything so as to justify its representation ona map. Other statements that can apply only to what geologists have recently and consistently been calling Kan- san occur on p. 757. The Kansan, we are told, “is greatly worn in regions where the denuding agents have worked under favorable Conditions seas 4.4. In other regions of flat surface and low declivity the degradation is less marked, and extensive remnants of the original surface-plane have been preserved.” The first super- Aftonian drift fulfils the conditions of the parts of the text quoted and of many others; the sub-Aftonian does not. THE IOWAN DRIFT 581 While admitting, then, all that may be claimed for the frequent references to the Aftonian soils, forests, peats, and gravels, it must also be admitted that the descriptions in the early texts, which treat of the texture, color, and petrological contents of the Kansan and the Iowan, are based on observations made on two super-Aftonian drifts. If there could remain a particle of doubt on this point after reading the texts, the doubt would be dispelled by an examination of the map opposite p. 727 of The Great Ice Age. The drifts of the two areas represented as Kansan and Iowan respectively are both super-Aftonian, and, considering the state of knowledge at the time, the map is remarkably correct. The eastern edge of the Iowan could scarcely be better drawn today. With the exception of a few points which would be mere microscopic dots on a map of this scale, the whole area mapped as Kansan is covered with super- Aftonian till. There is not a single known natural outcrop of sub-Aftonian in the Kansan area east of the Iowan margin. There are no known outcrops of sub-Aftonian in Illinois, Missouri, or northeastern Kansas where the map shows extensive areas of Kansan. It is only very recently that the presence of sub-Aftonian has been demonstrated in Nebraska; but even here it occurs in vertical sections at the base of bluffs, in such position that it could not well be represented on maps of moderate size. In Nebraska, as in practically all the rest of the area mapped as Kansan, it is a super-Aftonian drift that occupies the Kansan area on the map. In all the earlier texts and maps it is a super-Aftonian drift to which the name Kansan was most persistently and most consistently applied. EFFECTS ON TEXTS AND MAPS OF MAKING CERTAIN PROPOSED CHANGES IN THE USE OF THE TERMS KANSAN AND IOWAN As has been said, the imperfection of knowledge at the time the Iowan and Kansan drifts were named led to confusion and incon- sistencies of statements, and these are of such character and extent as to make it now utterly impossible to apply the proposed names in any conceivable way that will be in full accord with all the state- ments of the texts. The frequent and positive references to the horizon of the gravels and forest beds must be admitted and must be given full weight in determining the particular drift sheets to 582 SAMUEL CALVIN which the names Kansan and Iowan should be applied. On the other hand, the original descriptions of the lower and upper till— of the Kansan and the Iowan—must have careful consideration, and the evidence of the map in The Great Ice Age, above cited, must be taken into account. ~The descriptions would have to be rewritten and the map redrawn to make them consistent with the view that the Kansan is sub-Aftonian. If the term Kansan is transferred to the sub-Aftonian, and the term Iowan to the drift next above,’ practically the whole area represented on the map as Kansan would have to be changed to Iowan. The Iowan would then extend into southern Illinois, would cover southern and western Iowa, northern Missouri, eastern Nebraska, and northeastern Kansas. With the transfer of the term to the sub-Aftonian the Kansan would be represented on the map by a few dots and thin lines that could be seen only with the magnifier, the whole area comprising an aggregate of only a few sections; and in the present state of knowledge we could not be certain that Kansas has a cubic foot of Kansan (sub-Aftonian) drift. We are face to face with the fact that any application of the terms Kansan and Iowan involves some inconsistencies, is at variance with some of the statements in the original publications; and so long as we seem to need the terms and have to use them, it is only a question of how to use and apply them so as to do least violence to the original maps and descriptions. If the map and descriptive texts referred to may be taken as representing the intent of the authors, the practice of applying the terms which has been followed, and which seems now to come in for a certain amount of mild condemnation, is the only one that is reasonably consistent or possible. For it must be admitted that if the sub-Aftonian is to be called Kansan, and the first super-Aftonian drift is to be the Iowan, more than nine-tenths of the original descriptions are wholly erroneous and misleading, and the map in The Great Ice Age showing the distribution of these drifts is altogether meaningless and at variance with the facts. Recent usage in the application of the terms Kansan and Iowan is based on what seemed to be, and still seems to be, the only reason- * Some such shift as this seems to be favored by what is said in the Journal of Geology, July-August, 1910, pp. 473-74. THE IOWAN DRIFT 583 able interpretation of what the authors had in mind when describ- ing the physical characteristics of the two drift sheets and mapping their areal distribution. A departure from this usage, which would make the sub-Aftonian till Kansan and would apply the term Iowan to the old, weathered till above the Aftonian, with its blue color, its strikingly conspicuous array of greenstones, and with relations to the loess so entirely different from the relations correctly described in the text as pertaining to the Iowan, would necessitate the making of radical and revolutionary changes in the map and descriptive texts above noted. It surely accords better with what was pub- lished at the time the names were applied to let recent usage remain unchallenged and unchanged. EVIDENCE CONCERNING THE IOWAN DRIFT AND ITS GEOLOGICAL RELATIONS The surprising attitude toward the Iowan drift, expressed in the papers by Leverett, is something difficult to understand. A very little field study in the right places will demonstrate: 1. The Iowan drift is. 2. The Iowan drift is young as compared with the Kansan. 3. The Iowan drift is not a phase of the Kansan. 4. The Iowan drift has very intimate relations to certain bodies of loess.: 5. The Iowan drift is not related to the Illinoian. Tue Iowan Drtrt Is To discuss the question of whether there is an Iowan drift dis- tinct from the super-Aftonian till that has been called Kansan is like undertaking a task that one knows is absolutely useless and unnecessary. For, while the Iowan is thin, and in places is absent, there is here a very substantial drift sheet overlying the Kansan and possessing distinctive characters of its own. The Iowan is separated from the Kansan by a ferretto zone in some places and by weathered gravels in others, while its characteristic topography and remarkable bowlders proclaim its presence throughout exten- sive areas where no sections are available. Buchanan gravels, distributed by volumes of water from the melting Kansan glaciers, 584 SAMUEL CALVIN and disposed in numerous sheets and ridges, were laid down throughout northeastern Iowa, on top of the Kansan till, over the area which, but a short time previously, had been abandoned by the retreating ice. They are true outwash gravels. Exposure during the long interval between the Kansan and the Iowan has wrought profound changes in the decomposable granites and other Ramee Fic. 1.—View in the old gravel pit near Doris, Buchanan County, Iowa, showing Buchanan gravels, a deposit contemporaneous with the closing phase of the Kansan, overlain by the younger Iowan drift. constituents of the Buchanan deposits; and now these gravel deposits become unimpeachable witnesses to the fact that glaciers belonging to a stage long subsequent to the Kansan distributed a new sheet of till differing from the Kansan in composition, color, and petrological contents. The Iowan is yellow; the Kansan, while normally blue, sometimes weathers yellow; where yellow Iowan rests directly on yellow weathered Kansan, the line of contact may not be as distinct and satisfactory as some observers THE IOWAN DRIFT 585 might wish; but where the Buchanan gravels intervene, the fact that there is a drift younger than the Kansan and perfectly distinct from it is as clearly indicated as that there are two drifts separated by the Aftonian horizon. Leaving out, therefore, all the evidence from the multitudes of well sections and all other natural or artificial exposures that do not show the intervening aqueous deposits, a few of the scores of points where a young till overlies super-Kansan gravels may be Fic. 2.—View in the same pit a few rods west of point shown in Fig. 1, taken while work of excavation was in progress, showing the uneven line of contact between the old gravels and the younger, overlying Iowan. The irregularity in the contact line may be due to plowing or gouging by the Iowan ice. cited. The best-known of these is the old Illinois Central gravel pit near Doris in Buchanan County, a point that has been fre- quently mentioned. Here are gravel beds with a maximum thick- ness of fifteen feet or more. The deposit furnished many hundreds of carloads of railway ballast annually for a number of years. In the central part of the pit the gravel was taken out down to the blue Kansan till, and balls of the same blue till are included in the deposit. At the east end of the excavation there are at least ten feet of yellow till above the gravel, recording a later, newer stage of glaciation. The thickness of the later deposit is variable, for it 586 SAMUEL CALVIN was laid down on an uneven surface; but at the point illustrated in Fig. 1, the section of the till, including the black loam at the top, is about eight feet. That the later and newer till is uncon- formable on the gravel is shown in Fig. 2. The typical bowlders of the Iowan drift belong to this overlying till; there is nothing corresponding to them in the blue Kansan. In the process of excavation a number of the Iowan bowlders were undermined and allowed to fall into the pit. One such, perched on the brink of the excavation, is shown in Fig. 3, and a larger-sized companion, completely undermined, has fallen in. The typical, young, un- eroded, bowlder-dotted surface of the younger drift, which stretches away from the margin of the old working, is illustrated in Fig. 4. A quotation or two from the report on Buchanan County, Iowa Geological Survey, VIII, may be pertinent. On pp. 239-40 we read: A very common relation of Pleistocene deposits is illustrated in the well section on the land of J. W. Welch in the southwest quarter of Section 28, Bufialo Township. The record shows: Feet 25 Wark soilandiyellow tlle caso ce ee 4 2. Reddish, ferruginous sand and gravel.............. 23 iebluerclayapenctrated anc sci ieee eit aera e I No. 3 of this section is Iowan drift, No. 2 is Buchanan gravel, and No. 1 is Kansan till.1 In the same quarter section another well shows, Feet Ze Soilandayellows tills eesess ere eran oe orerecreeeren ce ane 22 2aueddish pravelicco. ss wisace ween sete a ene ee ae ee II t. Blue‘clay, with pockets of sande joss seen aes 19 Although the thickness varies considerably, the members of this last section are severally the same as the corresponding numbers of the one above. In another part of the same report, p. 209, it is recorded that “the eastern part of Fairbank Township is a very level, dry plateau in which a sheet of Iowan drift varying from two or three to thirty feet in thickness overlies an extensive bed of Buchanan gravels. The plateau is a unique piece of prairie land, without the usual undulations, and without any indications of imperfect drainage. The underlying gravel seems to afford an easy means of escape for the surplus surface waters.” * Owing to an error in proofreading the terms Iowan and Kansan are transposed on p. 240 of the volume cited. THE IOWAN DRIFT 587 Many similar cases could be cited, but it surely is not necessary to multiply arguments in support of a fact that is so perfectly obvious as the existence of the Iowan drift (Figs. 1 and 2). There is no sheet of till that has more distinctive characters, more definite stratigraphic relations. A glacial deposit showing thicknesses of 4 feet, ro feet, 22 feet, 30 feet, a deposit with distinctive bowlders of enormous size, a deposit that is young, fresh, uneroded, and separated from the Kansan by a weathered ferretto zone and pro- foundly altered gravels, is certainly a very real and substantial Fic. 3.—View in the Doris gravel pit, showing undermined Iowan bowlders; one is still perched on the brink of the excavation; the larger companion has fallen into the pit. _ thing that may not be disposed of by referring to it as ‘‘only the weathered surface of a drift,” or by the use of such a qualifying phrase as “so-called Iowan.” That there are two gravel horizons in this region—one Aftonian, the other Buchanan, one below the blue Kansan drift, the other above it—is indicated by two wells near the northwest corner of Section 22, Buffalo Township, Buchanan County. One of the wells, 152 feet deep and ending in gravel (Aftonian) which lies beneath the Kansan, furnishes a constant stream of water an inch in diameter; the other, which is not flowing, is 25 feet deep and ends in a bed of Buchanan gravel which overlies Kansan. 588 SAMUEL CALVIN From the other counties included in the Iowan area comes evidence of the distinctive character of the Iowan drift similar to the evidence from Buchanan. Probably the banner county in Iowa for inter-Kansan-Iowan gravels is Floyd. Here are scores of exposures, occupying every conceivable position from the flood plains of the streams, like the Little Cedar, the Cedar, and Flood creeks, which carried off the waters from the melting Kansan ice, to the highest points on the broad, monotonously level, uneroded divides; and in every case within the Iowan area where these old, weathered gravels are known to be present; they are overlain by deposits indicating a much later and newer glacial episode. It will be sufficient to note only a few of the numerous cases which have been observed. At the old brickyard west of the fair-ground in Charles City the material used was a five-foot bed of loam and yellow clay carrying Iowan bowlders four to five feet in diameter, and overlying the valley phase of the Buchanan gravels, which, in the neighborhood of Charles City, attain an enormous develop- ment. For some miles above Charles City, on the west side of the Cedar River, the old valley gravels may be seen passing under a thin bed of young, bowlder-bearing loam and clay which covers the gentle slopes and passes up over the flat divides. A boring with a post auger within the bowlder-strewn area went through the thin edge of the Iowan loam into the underlying gravels. The holes dug for some recently set telephone poles along the road and a small stream trench some distance up the slope in the field show the same relation of bowlder-bearing loam to the Buchanan beds. A young glacial deposit overlies super-Kansan gravels; at one point in the trench the gravels rest on the blue Kansan till. The same story is told, though in a slightly different way, by the majority of the many ‘‘mound springs” of this part of Iowa. Mound springs are peaty, boggy places on the hill slopes, due in most cases to the presence of upland gravels lying on impervious blue Kansan, and all covered by the younger sheet of Iowan. The gravels in such cases are reservoirs holding large quantities of water, and this escapes on the slopes near the plane of contact between the reservoir and the underlying clay at the point where the con- ditions are most favorable, presumably where the Iowan cover is THE IOWAN DRIFT 589 thinnest. The dry upland slopes above the level of the peat, as well as the dry slopes below that level, are liberally sprinkled with Iowan bowlders imbedded in loam and clay and ranging up to more than 12 feet in diameter. These bowlders are in themselves ade- quate evidence of a glacial invasion at a time subsequent to the gravel-forming phase, for they were not transported and deposited by either wind or water. Fic. 4.—View looking north from the margin of the Doris pit, showing the young, uneroded, bowlder-strewn Iowan drift plain; a very typical view in the area occupied by this young drift. : Typical examples of mound springs, easily accessible from Charles City, and showing the stratigraphic relations of Kansan, Buchanan, and Iowan deposits, occur on both sides of the railway in the north half of Section 2, Township 95, Range 15. Preliminary to laying a water pipe from the springy belt to the barn on the land of Mr. W. E. Waller, south of the railway, a shallow well was dug on the dry ground just above the peat; and this passed through the cover of Iowan bowlder-bearing loam and clay, and through the thinned edge of the rusty gravels, down far enough to make a water-tight basin in the blue Kansan. A deeper well near the barn, with 12 feet of gravel and 60 feet of blue clay, may be cited 590 SAMUEL CALVIN to show the constant relation of the prevailing gravels of the region to the typical Kansan drift. The same relation is shown in the fine Pleistocene section which occurs a few rods north of the Mitchell County line, not far from the southwest corner of Section 14, Township 97, Range 17. Here, in the south bank of Rock Creek, is an exposure of typical Kansan, blue in color and breaking into the characteristic polyhedral blocks, with an exposed thickness of 50 feet; at the top is a discolored, weathered zone three to four feet thick; next in order is a gravel bed, rusted and rotted, thick- ness about two feet; and all is covered by a thin loamy deposit carrying many fresh bowlders of varying size, belonging to a post- Buchanan stage of glaciation—the Iowan. But it is certainly unnecessary to offer additional evidence along this line. Cases of the kind already cited may be multiplied indefinitely. Fortunately the Buchanan gravels are especially well developed in northeastern Iowa, and in the Iowan area they uniformly afford indubitable evidence of a younger, newer, later stage of ice invasion. Outside the Iowan area, as at Colesburg, Delaware County, on the east, and at Iowa City on the south, the Buchanan gravels are covered with heavy deposits of loess, and without the least suggestion of later glaciation. ‘Some very impor- tant event, later than the deposition of the gravels, an event which caused the deposition of a body of till ranging up to 20 or 30 feet in thickness and carrying bowlders more than 12 feet in diameter, occurred within the Iowan area and did not occur outside of it. What was that event? Observations in and around the area lead unavoidably to but one conclusion, a conclusion that admits of no question: The Iowan drift 1s. Tue Iowan Drirt Is YouNG AS COMPARED WITH THE KANSAN The superposition of the Iowan till and the great Iowan bowlders on the weathered Buchanan is all the evidence needed to demon- strate that the Iowan is younger than the Kansan. The freshness of thé granites in and on the Jowan—many with the sharp angles caused by fracture unaffected by weathering (Fig. 10), while the granites of the Buchanan are very largely in an advanced stage of THE IOWAN DRIFT sol decomposition—lends strong support to the view that the Iowan is separated from the Kansan by a very long interval of time. The relative youth of the lowan may, or may not, be indicated by the fact that in places the formation is still very calcareous up to the grass roots. A concrete illustration of calcareous Iowan is seen in a shallow well near the northwest corner of the southeast quarter of Section 21, Township 95, Range 17. It should be stated Fic. 5.—View at the Dykeman quarry in Section 26, Township 97, Range 17, showing part of an area of thin drift, in which neither Iowan, Kansan, nor Nebraskan can be recognized. that the later investigations show that this young drift is variable as to the amount of the lime content; for in such localities as that just cited it seems to be as rich in calcium carbonate as the Wis- consin, while in other places it gives no reaction with acid. The original statement concerning this constituent of the Iowan drift was based on facts which remain true for the localities which had then been tested; but the writer has long since ceased to attach much importance to the acid test as a basis for determining the 592 SAMUEL CALVIN relative age of drift. The splendid piece of work by Dr. R. T. Chamberlin in the St. Croix region’ shows how a very young drift may exhibit no trace of lime, while a much older one may give vigorous reactions. In each and every case the amount of calcium carbonate present at or near the surface of a deposit of drift will depend on the original composition of the till and the movements of the ground waters. The same drift sheet gives very different reactions in different localities. The work in Taylor County in t1g10 showed very large quantities of lime carbonate in the form of segregated sheets and nodules, distributed along the joints in the highly weathered zone of the old Kansan. The lime came practically to the surface and was turned up among the grass roots by the plow. Along the roadsides, where the highway had been recently worked, it was breaking down to powder and mixing with the crumbling clay; and a sample of the old drift taken at the very surface might have given such energetic reactions to the man with the acid bottle as to lead him to think that he was dealing with the youngest glacial deposit in Iowa. Just why it is that both the old Kansan and the young Iowan should be so very calcareous up to the grass roots in some localities, while showing no traces of lime in others, could be explained, in some cases at least, on the basis of physical characteristics and relation to surface and sub-surface drainage; but it will be sufficient here simply to record the fact and repeat the obvious inference that acid tests applied to drift sheets are of exceedingly small importance in the determination of relative age. The acid bottle, intelligently used, has its place; but the user must be careful to recognize its limitations. : Among the evidences of youth in the Iowan drift is the fact that, in its typical areas, it is uneroded and imperfectly drained. The area selected for illustration in Fig. 7 of the article from the A meri- can Journal of Science, and again in Fig. 5 of the paper reprinted from the Zeitschrift fiir Gletscherkunde, and offered as “the type of erosion”’ in the Iowan drift or as something ‘‘showing topography of a so-called Iowan drift plain,” is a somewhat unfortunate and misleading choice for the reason that it is representative of but a ‘ Rollin T. Chamberlin, “‘Older Drifts in the St. Croix Region,” Journal of Geology, XVIII, No. 6, September-October, 1910. THE IOWAN DRIFT 593 small fraction of the real Iowan drift plain. It embraces the headwaters of Otter Creek, the valley of which belongs to one of the numerous small, exceptional areas in which there is no drift of any kind, neither Nebraskan, Kansan, nor Iowan. Within the southern edge of the map, and at Hazelton, less than a mile farther down the valley, the Niagara limestone is exposed in natural cliffs or quarry faces, forming continuous exposures along the streams in Sections Fic. 6.—View taken east of the diagonal road in Section 28, Township 95, Range 15, showing typical driftless hills in the belt of country west of the Cedar River. 2 and ro of Hazelton Township. A part of Section 2 is included in the map. There are outcrops below the village of Hazelton; and in the roads on the sloping sides of the valley, up to the summit of the slopes, rain, wash, and wear of traffic have exposed the fos- siliferous Niagaran dolomite, so thin and meager is the Pleistocene in this anomalous area. A full discussion of the Niagaran outcrops in this practically driftless valley will be found on pp. 217-20, Iowa Geological Survey, VIII, published 1898. Describing one of the quarries in Section 10, the report says: ‘‘The height of the 5904 SAMUEL CALVIN vertical quarry face is about fourteen feet. The upper two or three feet is made up of soil, reddish-brown residual clays, and decayed fragmentary limestone.” It will be noted that there is no recognizable drift; and if the area mapped is to be used to prove that there is no Iowan, with equal force, fairness, and cogency certain parts along its southern margin, together with the whole Fic. 7.—View on north side of the road passing through the middle of Section 21, Township 84, Range 18, showing the marly, fossiliferous phase of the Lime Creek shales at the surface, with no overlying drift of any age. valley of Otter Creek southward, might be used to prove that there are no glacial deposits of any sort within the whole state of Iowa. It would be possible to select a great many points within the Iowan area, that are driftless or practically so, where Pleistocene deposits are wholly absent or are represented by thin beds of sandy loam or a few stray bowlders. One such begins in the eastern edge of Independence and extends eastward over the stony hills for more than a mile. This is part of a belt some miles in length bordering THE IOWAN DRIFT 595 the Wapsipinicon River. The valley of the Cedar River, the anomalous characteristics of which are recognized and noted, if not explained, in Jowa Geological Survey, XIII, 298, 306, affords numerous examples (Figs. 5, 6). For some unaccountable reason the parts of the state occupied by the Lime Creek shales have an unusual number of driftless patches, some of which have dimensions of several miles. A rather small, but typical area of the kind occurs Fic. 8.—View on ridge in Section 3, Washington Township, Chickasaw County, Towa, showing the largest bowlder in the county rising out of a heavy growth of small grain. in Section 21, Township 84, Range 18 (Fig. 7). Large areas, almost continuous, occur over the ten-mile stretch between Mason City and Rockwell; and on the south side of Lime Creek there is a belt, practically driftless, two or three miles wide, all the way to Rockford. There may be bowlders in these areas, even where the other constituents of the drift are absent; and in no small propor- tion of the territory under consideration, ‘‘the soil through which the farmer drives his plow is made up of decomposed shales of 596 SAMUEL CALVIN Devonian age.” The collector may gather Lime Creek fossils in the pastures and cultivated fields. The peculiarities of these areas, so far as they are seen in Cerro Gordo County, are noted in Iowa Geological Survey, VIII, 175, where, years ago, the statement quoted was published. With the exception of the sub-Aftonian, or Nebraskan, which does not give character anywhere to parts of the glaciated territory large enough for mapping, each drift sheet has its characteristic topography which prevails over the major part of its area, and each has its exceptional phases which affect but a small percentage of its surface. There is a broad belt of typical Iowan between the Wapsipinicon and the Cedar, north of Walker. With the exception of a narrow strip west of the larger river, the broad area between the anomalous Cedar and Flood creeks in Mitchell, Floyd, and Franklin counties is as strikingly level, uneroded, and free from drainage courses as much of the typical Wisconsin, and in some places it is also quite as calcareous. Flood Creek is simply a prairie stream that scarcely breaks the monotony of the plain that extends from the Cedar to the Shell Rock; through its entire course north of Nora Springs, even the Shell Rock flows in a young, shallow trench cut in the otherwise unbroken Iowan plain. Areas such as these—scores of miles in length and width, with scarcely a drainage trench outside the channels of the larger streams—illustrate the real type of erosion in the Iowan; these show the topography of a real Iowan drift plain, and it is scarcely necessary to add that that topography is characteristic of youth. The typical bowlders of the Iowan are coarse feldspathic granites in no way remarkable for their power to resist the destruc- tive agencies of weathering, and yet very little decomposition has taken place amongst them since they were left exposed at the time of the retreating Iowan ice. In some way, either before or during transportation, many of the bowlders were fractured, and in such cases the angles are still comparatively sharp (Fig..10), while bowl- ders of corresponding texture in the Buchanan gravels or in the weathered zone of the Kansan drift are completely decayed. Topography, bowlders, and stratigraphic position all unite in support of the theses: THE IOWAN DRIFT 507 The Iowan drift is young as compared with the Kansan. The Iowan drift is not a phase (weathered or unweathered) of the Kansan. The statement on p. 282 of the paper on “‘Comparison of North American and European Glacial Deposits,” to the effect that the Iowan bowlders ‘‘are found chiefly in shallow draws, called sloughs, Fic. 9.—View in Section 14, Township 95, Range 16, showing a fresh, planed bowlder of the Iowan type in a dry, cultivated field. at the heads of the valleys or drainage lines,” has no special signifi- cance even if it were fully justified; but the fact is that the Iowan bowlders occur in various relations to the rather featureless topog- raphy of the region to which they are confined. The largest mass of granite in Chickasaw County is on the highest ridge of the whole region, midway between Devon and Alta Vista. There are three Iowan bowlders in Floyd County especially noted for their com- manding size, and each is located on dry upland. One of these (Fig. to) occurs less than two miles southwest of Charles City, in the northwest quarter of Section 14, Township 95, Range 16. It 598 SAMUEL CALVIN lies in a cultivated field on the long gentle slope above the road which follows the north line of the section. The dimensions above ground are 27X21X1r feet. Some of the faces are surfaces of fracture, and the angles remain sharp and unaffected by weather. A fragment of smaller size, evidently split off from the larger mass during the time of transportation, equally fresh as to angles and Fic. 10.—View in Section 14, Township 95, Range 16, showing one of the largest and most typical of the Iowan bowlders on dry ground, with sharp angles unaffected by weathering. general surface, lies a few rods to the northeast. A very fine bowlder (Fig. 11), more massive than the Charles City specimen, unbroken in transportation, lies near the southeast corner of the city park in Nora Springs, and there are neither “‘draws”’ nor “sloughs” anywhere near it. Probably the largest bowlder in the state, the largest so far recorded, occurs in a dry pasture near the southwest corner of Section 22, Township 94, Range 15 (Fig. 12). It is more than forty feet in length, a block of characteristic Iowan THE IOWAN DRIFT 599 granite of royal proportions. In some parts of the Iowan area, notably in the region between the Cedar and Little Cedar east of Charles City, the bowlders are distributed in trains which stretch across the country from northwest to southeast without respect to sloughs, while intervening spaces of essentially the same topography are practically free. The fact is, however, that the bowlders may be anywhere; upland or lowland seems to make no special differ- Fic. 11.—View in Nora Springs, Iowa, showing a large and very typical Iowan bowlder on dry upland, near the southeast corner of the city park. ence; their distribution follows no constant rule, except one: typical Iowan bowlders are strictly limited to the area of the Iowan drift. THe Iowan Drirt HAs CERTAIN VERY INTIMATE RELATIONS TO CERTAIN Bopies oF LOESS The discussion of the loess and of Calvin’s attitude toward it, on pp. 298-99 of the Berlin paper, is based on so many misappre- hensions that the task of straightening out the tangle is one too hopeless to be undertaken. There are bodies of loess belonging to different ages, but there is one loess that stands in intimate and close relation to the Iowan drift. The view that the loess is chiefly 600 SAMUEL CALVIN an interglacial deposit is in no way inconsistent with the earlier view—and the view still entertained so far as the source of the deposit is concerned—‘‘that the loess is a. silt derived from the finer materials of the Iowan drift.” That a certain deposit of loess was derived from the Iowan is a conviction that grows stronger and stronger as the work is prosecuted farther and farther in the Fic. 12.—View near the southeast corner of Section 22, Township 94, Range 15, showing what is probably the largest Iowan bowlder in the state; and this lies in a dry pasture. field; and outside the paper under consideration there has never been any ‘“‘abandonment of the view that there is an Iowan drift correlating with the loess.”’ The Buchanan gravels are an interglacial deposit. They are not of glacial origin, and they lie between two sheets of drift. The fact that they are interglacial, however, gives no adequate ground to infer that they were not derived from the Kansan, or that there has been an abandonment of the view that there is a Kansan THE IOWAN DRIFT 601 drift correlating with the Buchanan gravels. The Iowan loess is related to the Iowan drift in much the same way that the gravels are related to the Kansan. The earlier view was that the loess was deposited at the time of. maximum development of the Iowan glaciation, when the Iowan area was still covered with ice. The only modification of that view at the present time is that loess deposition took place after the Iowan ice had retreated to a greater or less extent, after an interglacial interval had actually begun. By such retreat extensive mud flats were left, and as these dried before becoming covered with vegetation, strong winds coming, probably, from the ice fields to the north, carried fine sand and dust from the bare surfaces and deposited them beyond the edge of the Iowan area, out upon the old, eroded Kansan. For the development of loess three things seem to be necessary: (1) a gathering-ground of extensive bare and dry surfaces, such as would be furnished by the part of the Iowan area from which the ice had retreated; (2) winds to transport the materials from the dried mud flats; (3) anchorage such as would be furnished most extensively by the vegetation of the extra-marginal Kansan surface. The bare Iowan area afforded no anchorage, but it was an excellent source of supply. Waters carried and sorted materials from the Kansan till and deposited the interglacial formation called Buchanan gravels; winds picked up from the Iowan till such materials as they could trans- port, and deposited amidst the vegetation of the extra-marginal territory the interglacial formation known as the Iowan loess. The genetic relation of the loess to the Iowan drift is not so very unlike the corresponding relation of the Buchanan gravels to the Kansan; and so far as genetic relationships are concerned, there has been no abandonment of the view originally proposed. The color, composition, and calcareous content of the Iowan loess are in perfect accord with the hypothesis just expressed; its geographic distribution around the lobed margin of the Iowan area agrees also with the view; the great thickness of this loess at and near its inner margin, and its thinning out with increasing distance from the source of supply, corroborate all the other lines of evidence; while the great amount of eolian sand associated with it in a narrow belt surrounding the lobes of Iowan drift lends additional support. 602 SAMUEL CALVIN The Missouri River loess and all other loess deposits which have evidently been derived from the broad flood plains of near-by rivers, have a similar distribution relative to their source; they are thickest and coarsest near the gathering-ground and become thinner as the distance from the base of supply increases. All the facts connected with the origin, composition, and distribution of the loess are perfectly explicable without resorting to the hypothesis that “‘a considerable part’ was derived from the great plains east of the Rocky Mountains.”’ Studies in the field afford overwhelming evidence that, genetically and geographically, the Iowan drift has very intimate relations to certain bodies of loess. Tue Iowan Drirt Is Not RELATED TO THE ILLINOIAN It is scarcely necessary to discuss the suggestion that the lowan may be correlated with the Illinoian. Parenthetically it may be said that if the Iowan and the Illinoian represent the same stage of glaciation, the name IIlinoian becomes a synonym for lowan, and we shall be reduced to the painful necessity of referring to one of our most beloved drift sheets as the “‘so-called Hlinoian.” But no such calamity awaits the Illinoian. The Iowan is much the younger of the two. As indicated by the structural and genetic relations above noted, the Iowan —a little later probably than its maximum stage—is practically contemporaneous with the loess; and as the Berlin paper, with noteworthy lucidity, correctly states on p. 299; “the Sangamon interval separates the loess from the Illinoian stage of glaciation so widely that there would seem to be no relation between loess deposition and Ilinoian outwash.” The same long interval, the same wide separation, exists between the Iowan and the Illinoian stages of glaciation. The two drifts are not related in time or in any other way. All the facts which may be gathered from the most thorough investigations in the field, support this last proposition: The Iowan drift 1s not related to the Illinotan. THE THEORY OF ISOSTASY HARMON LEWIS University of Wisconsin TABLE OF CONTENTS SECTION I. INTRODUCTION General Definitions Section IJ. THE Greopetic Work oF JOHN F. HAyForp Brief Description of His Work and Methods Criticism of Hayford’s Work Criticism of method of finding degree of completeness of compensation Criticism of C-solution Further Considerations Possibilities of an incomplete compensation Changes in formulae required by shallow depth of compensation Summary SecTIon III. Tue TuHrory or Isostasy Introductory Type of deformation postulated in the theory of isostasy Lines of criticism of isostasy Facts not Accounted for by the Type of Deformation Postulated in the Theory of Isostasy The theory of isostasy as conceived in this paper does not adequately account for the folding of rocks of the earth’s crust The theory of isostasy cannot account for the general uplift of sedi- ments without folding The theory of isostasy does not explain the apparently heterogeneous relation of uplift and subsidence to erosion and deposition Alternative Hypotheses to Account for Hayford’s Geodetic Results The tendency of lateral compression to produce isostatic compensa- tion The automatic compensation of uplifts and subsidences due to expan- sion and contraction SECTION IV. SUMMARY OF CONCLUSIONS 603 604 HARMON LEWIS SECTION I. INTRODUCTION GENERAL According to the present general conception the theory of isostasy consists of two main postulates, first that the elevated portions of the earth are deficient in density, and second that the material of the earth is comparatively weak. It is generally ac- cepted that these two postulates are inseparable, for it is argued on the one hand that, if the elevated portions are deficient in density, readjustment involving deformation and failure must have taken place in order to compensate for the large mass of material eroded from the lands and deposited in the sea; and it is contended on the other hand that, if the earth is weak, it could not support the mountains and continents unless they are compensated by a defect of density below. In accordance with these two main postu- lates it is conceived that the dominant type of earth deformation consists in vertical movements between various segments accom- panied by lateral flowage of rock beneath and possibly crumpling of the rock in the border zones. This conception is not only applied to the major earth segments, the continents and oceans, but to the smaller units of the continents as well. The theory of isostasy is a decided contrast to the alternative conception that the earth is strong enough to support the continents and mountains even though there are no compensating density differences, that changes of weight at the surface do not produce vertical movements of the segments in a weaker substratum, and that the dominant type of deformation is folding and upwarping due to lateral compression. The theory of isostasy if correct would be of fundamental impor- tance to the geologist in interpreting earth movements. Previous to Hayford’s geodetic investigation the conceptions of isostasy were largely speculative. After a comprehensive study of the deflections of the plumb bob carried out by the United States Coast and Geodetic Survey under Hayford’s direction,* «The most complete statement of Hayford’s work which has been published is contained in his two reports issued in 1909 and 1o10 by the U.S.C. and G.S. and entitled, The Figure of the Earth and Isostasy from Measurements in the United States and Supplementary Investigation in 1909 of the Figure of the Earth and Isostasy. THE THEORY OF ISOSTASY 605 the rather startling conclusions were reached that the excesses of mass composing the continents and mountains are completely compensated by deficiency of density below and that this deficiency of density extends to a depth of something like 60 to 150 miles. These conclusions lent strong support to the theory of isostasy. Considering the completeness of the density compensation, there seemed to be no escape from the conclusion that readjustments of the nature postulated by isostasy are continually taking place. With the theory of isostasy apparently on such a firm basis, Hayford and others have elaborated the conceptions of earth move- ments involved in the theory, but these further inferences have not met the approval of many geologists. The possibility that Hayford had made an error in his geodetic work suggested an investigation which is the basis of this paper. The attempt has been made, first, to examine Hayford’s geodetic work apart from any inferences which may have been drawn from it, and second, to examine the theory of isostasy with reference to inferences from geodetic evidence and also on general grounds. This paper has accordingly been divided into two main parts entitled ‘‘The Geodetic Work of John F. Hayford” and “The Theory of Isostasy.” A ‘‘Summary of Conclusions” is given at the end. It seemed highly desirable in connection with the criticism of Hayford’s geodetic work that several of the terms employed should be defined. DEFINITIONS ““*Tsostatic compensation’ is the compensation of the excess of matter at the surface (continents) by the defect of density below, and of the surface defect of matter (oceans) by excess of density below.” ““Tsostatic compensation’’ will also be referred to simply as ‘‘com- pensation”? and an area or segment of the earth will be spoken of as “‘compensated”’ if there is isostatic compensation of the excess or defect of matter over that area or at the surface of the given segment. From the above definition it follows that there will be, in general, a density difference between an average sea level segment t The Figure of the Earth and Isostasy, U.S.C. and G.S., 1909, p. 67. 606 HARMON LEWIS of the earth and a compensated segment, the surface of which is not at sea level. This density difference will be called the “compen- sating density difference.” The ‘depth of compensation” for any segment of the earth 1s the greatest depth below sea level at which there is a compensating density difference. This is different from the definition of Hayford which makes the ‘“‘depth of compensation”’ the depth ‘‘ within which the isostatic compensation is complete.” The former definition allows for the possibility of a compensation which is not complete. The ‘distribution of compensation”’ for any segment of the earth is the manner of variation of the compensating density difference with respect to depth. If the compensating density difference is uniform, the distribution of compensation is uniform; if it is uniformly varying from a maximum at the surface to zero at the depth of compensation the distribution of compensation is uniformly varying. The ‘‘degree of completeness of isostatic compensation”’ is an expression used by Hayford. After defining the depth of com- pensation as quoted above, he says, ‘‘At and below this depth the condition as to stress of any element of mass is isostatic; that is, any element of mass is subject to equal pressures from all directions as if it were a portion of aperfect fluid. . . . . In terms of masses, densities, and volumes, the condition above the depth of compen- sation may be expressed as follows: The mass in any prismatic column which has for its base a unit area of the horizontal surface which lies at the depth of compensation, for its edges vertical lines (lines of gravity) and for its upper limit the actual irregular surface of the earth (or sea surface if the area in question is beneath the ocean) is the same as the mass in any other similar prismatic column having any other unit area of the same surface for its base.” This condition of course follows from Hayford’s definition of depth of compensation, but it would not hold for the definition adopted in this discussion unless the compensation were complete. Hayford continues as follows: “If this condition of equal pressures, that is, of equal superimposed masses, is fully satisfied at a given depth the compensation is said to be complete at that depth. If there is a variation from equality of superimposed masses the differences THE THEORY OF ISOSTASY 607 may be taken as a measure of the degree of incompleteness of com- pensation.’’’ In order to make this idea exact, let A and B (Fig. 1) represent two columns each of horizontal cross-section, a, and extending to the depth of compensation, /,, the upper surface of A being # miles above sea level and the upper surface of B being at sea os level. Let the weight of A equal sea leve/ Wa, and the weight of B equal Wz. y If there were no isostatic compensa- tion, and if the densities of A and B were the same at similar depths, then Wa would be in excess of Wz by the h, amount adh, where 6 is the mean surface density of the earth. If this excess of weight were entirely made up for by a deficiency of density below, compensation would be com- A 2S pee plete. Therefore let the “degree of completeness of isostatic compensation”? for any segment, A, be defined as _ adh—(Wa—WsB) ul aoh The quantity in parenthesis is the amount by which the weight of A is in excess of the weight of B. The whole numerator is, therefore, the weight which has been made up for by a deficiency of density below the surface and the entire fraction is a number expressing the part of the weight, a5, which has been made up for. The above formula holds equally well for land and ocean areas if h be taken positive above sea level and negative below. FOR LAND AREAS Teelit Wa< We then M>1 De AGE Wa=Ws then M=1 3. If Wa—Wp=aoh then M=o 4. If Wa—We>abdh then MWes then M>1 2 ht Wa=We, then M=1 3. If Wa—Ws=adh then M=o 4. If Wa—Wa< adh then Mt. ‘““Complete compensation”’ is such an isostatic compensation that M=1. ‘“‘Under-compensation”’ is such an isostatic compensation that On Ey Isostatic compensation is considered the more complete, the closer M approaches to 1. SECTION II. THE GEODETIC WORK OF JOHN F. HAYFORD BRIEF DESCRIPTION OF HIS WORK AND METHODS On certain assumptions as to the size and shape of the earth and as to the position of a base station on this ideal earth, Hayford, by triangulation and geodetic observations, measured the prime vertical and meridian components of the deflection of the plumb bob from the true vertical at several hundred stations scattered over the United States. He then calculated the deflections which all the topographic features within a radius of 2,564 miles of each station should produce if the density of the earth were the same at similar depths. He found that these calculated deflections, which he called the “topographic deflections,” were universally larger than the “observed deflections.” The only explanation of such widespread observations is that there is some sort of isostatic compensation of the surface excesses and defects of mass. Recog- _ THE THEORY OF ISOSTASY 609 nizing this fact Hayford set out to make a series of least square solutions assuming various kinds of isostatic compensation. He calculated what the deflections should be assuming isostatic compensation by the use of a reduction factor, that is, a factor which when multiplied by the topographic deflection will give the deflection, isostatic compensation considered. In all of his solu- tions Hayford assumed the isostatic compensation to be complete. In his five principle solutions he assumed a uniform distribution of compensation and assumed depths of compensation varying from zero to infinity.7 On these assumptions the conclusion was reached that the most probable depth of compensation is 76 miles since the sum of the squares of the residuals was least for this depth. Subsidiary solutions were made assuming (1) that the compen- sation is uniformly distributed in a ten-mile substratum, (2) that the compensation is greatest at the surface and decreases uniformly with respect to depth until it becomes zero at the depth of compen- sation, and (3) that the compensation is distributed according to the law postulated by Chamberlin.?, The method used in each of these three cases was to find the depth for which the reduction factor was most like the reduction factor for the most probable solution assuming a uniform distribution. Hayford concluded that so far as the geodetic evidence available could test them, any of the three distributions of compensation postulated is as probable as a uniform distribution. The depth of compensation found for a distribution in a ten-mile substratum was 4o miles; for a uni- formly varying distribution, 117 miles; and for the Chamberlin distribution of compensation, 193 miles. A further interesting phase of Hayford’s work is his C-solution3 which was made on the assumption that there is no isostatic com- pensation under land areas but that there is complete isostatic ™ The condition that the depth of compensation is infinite is taken as equivalent to no isostatic compensation. The condition that the depth of compensation is zero is taken as equivalent to the condition that the topographic features do not affect the plumb bob. 2 This law postulates a maximum density difference slightly below the surface. This density difference decreases rapidly at first and then more gradually with respect to depth. 3 P. 168 of 1909 report. 610 HARMON LEWIS compensation at depth zero under ocean areas. It was found that these assumptions were not as close to the facts as the assumption of complete compensation at the depth of 76 miles under both land and ocean. In discussing the C-solution Hayford says: It follows, moreover, that it is an isostatic compensation of the separate topographic features of the continent, not a compensation merely of the conti- nent asa whole. In solution A? it is a compensation of the separate features which is assumed. An inspection of the numerical values of the computed topographic deflections, and of the deflections computed with isostatic com- pensation considered shows that merely to have assumed the continent as a whole to be compensated, not its separate topographic features would have given a solution resembling solution C much more closely than solution A.3 A very vital step in Hayford’s work is his determination of the degree of completeness of compensation. As this is one of the principal points to be criticized, his method will be explained in detail in connection with the criticism. Suffice it to say here that he concluded that the isostatic compensation is on an average nine- tenths complete. CRITICISM OF HAYFORD’S WORK Hayford certainly showed that there is some sort of isostatic compensation; but he did not fully consider all possibilities as to the nature of this compensation. The exact nature of isostatic compensation for any place is determined by three factors, (1) depth of compensation, (2) distribution of compensation, and (3) degree of completeness of compensation. Hayford considered all possible depths of compensation and several distributions of compensation; but all of his solutions involving isostatic compensation were made on the assumption that the compensation is complete. This was a purely arbitrary assumption on Hayford’s part since he gave no reason whatever for believing at the outset that compensation is ‘Quoted from p. 169 of 1909 report. 2 Solution A was made on the assumption that compensation is complete at depth zero over both land and sea. This solution turned out to be nearer the truth than solution C. 3It should be noted that the assumption of complete compensation under ocean areas with no compensation under continents is not equivalent to a compensation of the continents as a whole with respect to the oceans, but only to a compensation for the part of the continents below sea level. THE THEORY OF ISOSTASY 611 complete, and furthermore the fact that he later attempted to find the degree of completeness implies that there is no reason to believe at the outset in complete compensation. In view of this fact the method of determining the degree of completeness of compensation is questionable. Criticism of method of finding degree of completeness of compen- sation.—Hayford’s method is best explained by an example. Sup- pose the topographic deflection at some station is 35.79” This is assuming no isostatic compensation. Suppose that the residual assuming complete compensation at a depth of 76 miles is 3.33”, or in other words, suppose that the difference of the true deflection and the deflection which would exist if the compensation were complete at a depth of 76 miles is 3.33” This value (3.33”) is apparently the part of the topographic deflection which has not been made up for by isostatic compensation. The ratio of 3.33 to 35.79 is, therefore, taken by Hayford as a measure of the incom- pleteness of compensation. In explanation of this method Hayford writes as follows: . The residuals of solution Gt furnish a test of the departures of the facts from the assumed condition of complete isostatic compensation uniformly distri- buted to a limiting depth of 113.7 kilometers. In order to obtain definite ideas let the whole of the residuals of this solution be credited to the incomplete- ness of the compensation. The conclusion as to the completeness of compen- sation will then be in error in that the actual approach to completeness will be considerably closer than that represented by the conclusion—that is, the conclusion will be an extreme limit of incompleteness rather than a direct measure. For by this process of reasoning every portion of a residual of solu- tion G, due to the departure of the actual distribution of compensation with respect to depth from the assumed distribution, or due to the error in the assumed mean depth of compensation, or to regional variation from a fixed depth of compensation, or due to errors of observation in the astronomic determinations and the triangulation which affect the observed deflection of the vertical, or due to errors of computation, is credited to incompleteness of compensation,? The objection to the paragraph quoted is that it apparently is taken for granted that the error in the assumed mean depth of t Solution G, the most probable solution according to the first report, was made assuming a depth of compensation of 70 miles. 2 Quoted from p. 164 of 1909 report. 612 HARMON LEWIS compensation increased the size of the residuals. Is it not very probable that the introduction of an error in depth actually dimin- - ished the residuals? The most probable depth was calculated on the assumption of completeness. If the assumption of complete- ness was wrong, the depth of compensation which would appear most probable would not be the true depth of compensation but a depth which would counteract the effect of the wrong assump- tion in regard to completeness. In other words the error in the assumed mean depth of compensation would be such as to decrease the residuals. Therefore the residuals which would have been obtained had the correct depth been used would be larger than the residuals actually obtained. The degree of incompleteness as measured by Hayford’s method would, therefore, be larger. Tf the depth of compensation were known independently, then Hayford’s method of finding the completeness would be legitimate. To go back to the example cited before, suppose that it is known independently that the depth of compensation is 25 miles and suppose that the residual obtained on this basis and assuming complete compensation is 15’’: this value would be the part of the topographic deflection which had not been made up for by compensation and therefore the ratio of 15 to 35.79 would be an approximate measure of the incompleteness of compensation. The above argument will be made clear by a brief summary. The depth and degree of completeness of compensation are unknowns to be determined. It is claimed that these two unknowns can not be determined by Hayford’s method of assuming complete compensation, calculating the most probable depth, and using the residuals to tell the degree of incompleteness, because this method would only be legitimate for the one case when compensa- tion is actually complete. If compensation were not complete, then Hayford’s calculated depth would be wrong and would furthermore be in error in sucha direction as toat least partially make up for the wrong assumption regarding degree of completeness. The resulting residuals would therefore not furnish a maximum measure of the degree of incompleteness; but the compensation would appear to be more nearly complete than would be the fact. We are forced to conclude that, from the geodetic evidence THE THEORY OF ISOSTASY 613 alone, neither the depth nor the degree of completeness of isostatic compensation can as yet be considered settled. Criticism of C-solution.—The criticism might be made of the C-solution that the assumption of complete compensation at depth zero under oceans obviously does not correspond to the facts and that, by trying several depths, a combination might be found which would appear to be, so far as the information available could test it, as close to the actual conditions as any other hypothesis. From the wide departure of the C-solution it seems, however, rather improbable that such a depth could be found. FURTHER CONSIDERATIONS Possibilities of an incomplete compensation.—Since Hayford only considered the case of complete compensation, it is desirable to see whether or not an incomplete compensation would meet the geodetic requirements as well as a complete compensation. Any test of incomplete compensation based on Hayford’s residuals is apt to be misleading since these residuals may involve two errors that tend to counterbalance each other. By a study of the reduc- tion factor, however, we may be able to tell whether or not an incomplete compensation would be as probable from the geodetic point of view as complete compensation. According to the definition given in this paper the fees of completeness of compensation is abh—(Wa—Ws) aoh ; (x) M= If there is isostatic compensation, there will be a compensating density difference between the material in column A and the material in column B.?_ If at any given depth the density of column A be 64 and of column B, 6,, then the compensating density differ- ence at that depth will be 6,=6,—6, which is of course negative when A is a land segment. It follows that 63=6,—6,. Now tIt should be noted that so far as the nature of compensation is questionable, Hayford’s values for the size and figure of the earth are also open to question. 2 See definition of M, p. 607-8. 614 HARMON LEWIS (uth) Wan a is the weight* which the column A would have if its density were the same as in column B and is therefore equal to the weight of B plus the weight which the material in A above sea level would have if there were no isostatic compensation. Thus (ath) Wa- of =Wp+abh or (Ax +h) a-Wa mobs | ba (2) Substituting (2) in (1), | (Ax +h) aoh— ato fa M= aoh or (Ax+h) dhM = - {a0 (3) The case of a uniformly distributed compensation will be con- sidered. In this case, 6, being a constant, (3) reduces to dhM = —8,(s+h) (4) As stated before Hayford only considers the case where M=t. He further makes the approximation of neglecting / in comparison with /;. This approximation which is permissible for depths of compensation considered by Hayford but which would not be allowable for shallow depths is discussed later. The relation cor- responding to (4) which Hayford uses is dh=—6,h,. If however the unknown quantity, M, is retained, the corresponding reduc- tion factor which we will call Fy is as follows: : P+V ()P+he jy ae ie ta genes (s) log = t As 6; may vary with the depth it is necessary to sum up the product of 6, and the small elements of depth rather than use the product, 6:(/1++h). THE THEORY OF ISOSTASY 615 Adding and substracting M, we have Fyu=1—M(1—-F) (6) where F, the reduction factor obtained by Hayford, is D poy ()?-+h,? pane h+V Real ie p log i (7) Comparing F with Fy we see that, when O< M<1, Fy will be greater than F for the same ring’ and #,._ Also for any given ring F is larger, the larger the depth of compensation,’ /,. It follows, therefore, that Fy for M<1 calculated on any given h, will be greater for all rings than the corresponding factor, Ff, calculated on a smaller h,, for or i ee nae I Now assuming M@=1 Hayford has already shown that the set of factors obtained when /,=76 miles gives a closer result than the set of larger factors obtained when h,>76 miles. It seems probable. therefore that, if a solution were to be attempted assuming O< M <1, nothing would be gained in taking a depth of compensation larger than the most probable depth assuming M=1. The writer would not care to make the preceding statement as a positive fact without an inspection of the data for the calculation of the topographic deflections. For it seems possible, although not probable, that a combination of <1 and h,>76 might yield as close a result as M=1 and /#,=76 on account of the fact that, tIn calculating the topographic deflection the area around any station is divided into concentric rings whose outer radii are 7? and inner radii, 7x. 2See table, p. 70 of 1909 report. 616 HARMON LEWIS though the reduction factor becomes larger, the relative increase in the various rings is not the same as the increase when /, is made larger than 76 but M is kept equal to 1. On the other hand, when o1 were to be considered, the best depth of compensation would probably turn out to be greater than 76 miles. By equation (6) above, it is an easy matter to calculate the factor Fy for any value of , which is equal to the radius of any ring. In order to obtain a typical example, the factors were calculated assum- ing M=.5 and h,=19.29 kilometers (11.987 miles) and are given below together with the factors for Hayford’s most probable solu- tion in which it was assumed that M=1 and h,=113.7 kilometers (70.67 miles).7 It will be noted that, for outer rings, the factor, Py, is approxi- mately .5 while F is nearly zero. From the examples of calcula- tions of ‘‘topographic deflections” given by Hayford it would seem that the outer rings, especially the oceanic compartments, have a considerable effect on the topographic deflection. It might seem, therefore, that M=.5, 4:=19.29 kilometers would not give as close a result as M=1, #;=113.7 kilometers; but without some sort of test this question would remain a matter of conjecture. Furthermore the additional hypothesis might be introduced that the compensation under ocean areas is complete or even that ocean areas are over-compensated. Then when we remember that the assumption regarding either M or h, or both may « The value of the depth of compensation given in the first report as most probable is 70 miles. In the second report 76 miles is given, but the reduction factors for this depth are not published. be varied, there seems to be nothing to show that a combination of a decided under-compensation and shallow depth under land areas with a practically complete compensation under ocean areas would not prove, so far as present information is able to test it, as close an approximation to the actual conditions as a combination of complete compensation with a depth of 76 miles under both And it is not improbable that several combinations involving a decidedly incomplete compensation under land areas could be found which would appear equally as close to the truth land and sea. THE THEORY OF ISOSTASY as M=1 and h,=76 miles. In the case of M>1, h,>76 miles it is probably also true that the compensation under ocean areas would have to be taken com- plete. In the foregoing discussion the distribution of compensation eras kil. epee avakale -996 -994 -9QI .988 .982 -975 .965 .950 .930 .QOO . 860 . 809 747 679 .617 .570 -539 .520 .510 505 502 . 501 . 500 . 500 . 500 . 500 . 500 . 500 . 500 -997 .996 -995 992 988 .983 -976 -965 -O51 -930 evo\e) .859 . 801 afl .618 -493 358 234 - 139 :077 .040 .020 .OTO .005 .003 OO .OOI 618 HARMON LEWIS was assumed uniform. The argument that, if M <1, h, is probably less than the most probable depth assuming M=1 would hold for the Chamberlin compensation or for a compensation uniformly varying from a maximum at the surface to zero at the depth of compensation. For the reduction factors in the subsidiary investi- gations are linear functions of the reduction factor for uniform distribution. Changes in formulae required by shallow depth of compensation.—- Under this head the approximation mentioned above in connection with equation (4) will be considered. dhM = —8,(h, +h) (4) 6, is such a quantity that (W4—a0, [h,+h]) =[Wal,,_, is the weight which column A would have if there were no isostatic com- pensation. Therefore Wa=(Waly— tad +h) Thus the deflection due to W4 may be obtained by adding to the deflection which would be produced if there were no isostatic com- pensation the deflection which would be produced by a6,(h,+h). Therefore the deflection at any station assuming isostatic compen- sation is equal to the topographic deflection (D) plus the deflection (D.) due to the defect or excess of density from the surface down to the depth of compensation. If H be the height of the observing station above sea level, then the deflection due to the defect of density in a compartment whose surface is # miles above sea level is (neglecting the curvature of the earth), 6: ° ° 1D): 44 (sin at'—sin ay) / (H+h:)log rP+YV (7)?+(H4+h:)? nV r+ (Ath)? petal OTE n+V re+(h—H)? + (h—H)log (8) In calculating the topographic deflection Hayford neglects H except when it is necessary to make a slope correction. However, H is introduced in equation (8) in order to make it exact. Whether it would be legitimate to neglect this factor or not can best be told «See A. R. Clarke, Geodesy, 1880, p. 295, and Figure of the Earth and Isostasy, 1909, p. 20, for the derivation of equation (8). THE THEORY OF ITSOSTASY 619 from the resulting expression for the reduction factor. Taking the value of D, given in (8) the reduction factor is D-+-D, t \8(H+h), 2P+V (7)+(A+h)? SS Ss lo Fu ee D B dh : fia V ean h;)2 (9) Ey manvs a BE? eb re (iB From (4) 5: Mh Se oo) Substituting (ro) in (9) og PAV (P+ (H+hn)? ie M(H+h) > ntV re+(A+h,)? Mie hth pa log ~ eV OOO Mis aa Ce aa 2 (11) This reduces to Hayford’s factor if M be put equal to 1 and H and / be put equal to zero. These are the three approximations on which Hayford’s factor is obtained. If h, is large, it is legitimate to neglect H and h; but if 4, were to be taken as 12 miles and H would certainly have to be considered. This fact would increase the length of the computations since for a complete solution of the problem a different factor would be necessary for each compartment whereas, before, the same factor was used for an entirering. Doubt- less, however, devices could be employed which would facilitate the calculations. The necessity of having to use the reduction factor given in (11) serves to make the depth and degree of completeness of com- pensation more open to question than ever. SUMMARY On the basis of Hayford’s work it may be considered settled that there is some sort of isostatic compensation, but so far as Hayford’s investigation has yet gone there are many possibilities 620 HARMON LEWIS as to the nature of this compensation. None of the possible distri- butions of compensation have been eliminated by Hayford’s geodetic work; in fact, so far as the geodetic work is concerned, Hayford has shown that four different distributions of compensa- tion are equally probable. The present possibilities for isostatic compensation may be grouped with considerable certainty under three heads: first, there is the possibility of complete compensation at a depth in the neighborhood of 60 to 150 miles depending on the distribution of compensation; second, there is the possibility of an over-compensa- tion at a greater depth for land areas with probably complete com- pensation for ocean areas; and third, there is the possibility of under-compensation at a shallow depth for land areas with com- plete or over-compensation for ocean areas. SECTION III. THE THEORY OF ISOSTASY INTRODUCTORY In any theory of earth movements it is recognized that the earth is a failing structure in the sense that it has been and is being permanently deformed under the ultimately controlling force of gravity. It is not therefore the essential idea of the theory of isostasy that the earth as a whole is a failing structure; but the characteristic of the theory is the type of deformation which it postulates. This may not be the critical point of isostasy as it was originally conceived, or as conceived today by everyone; but it is the point which has been elaborated by the supporters of the theory and which is of first importance to the geologist; and it will therefore serve as a basis of criticism in this paper. Type of deformation postulated in the theory of isostasy.—The controlling movements of the earth’s crust are vertical movements of the various segments in response to changes of weight produced by erosion and deposition. These vertical movements are brought about by flowage beneath the surface from areas of deposition to areas of erosion or, in general, from areas of excessive weight to areas deficient in weight. This flowage beneath the surface is comparable in speed to the process of erosion and is started under stress-differences so small THE THEORY OF ISOSTASY 621 as to require that all segments of the earth of given area are essen- tially equal in weight. This flowage may be accompanied by folding in the border zones of the segments. Deformation of this kind is not restricted to the major segments, the continents and oceans, but is the type of movement which takes place between the smaller units of the continents. Lines of criticism of isostasy.—In criticizing the theory of isostasy two main lines of argument will be followed. First, the type of deformation postulated by isostasy can not account for certain facts. Second, Hayford’s geodetic results can be accounted for without supposing the type of deformation postulated by isostasy. FACTS NOT ACCOUNTED FOR BY THE TYPE OF DEFORMATION POSTU- LATED IN THE THEORY OF ISOSTASY The degree to which isostasy must be discarded depends on the importance of the phenomena which it will not explain. The theory of isostasy as conceived in this paper does not adequately account for the folding of rocks of the earth’s crust.—Folding is evi- dence of lateral forces of enormous magnitude. On the other hand the controlling movements of isostasy are assumed to be vertical movements. However, it has been suggested by Hayford that folding would be caused by the undertow from an area of deposition to an area of erosion: Horizontal compressive stresses in the material near the surface above the undertow are necessarily caused by the undertow. For the undertow neces- sarily tends to carry the surface along with it and so pushes this surface material against that in the region of erosion, see Fig. 2. These stresses tend to produce a crumpling, crushing and bending of the surface strata accompanied by increase of elevation of the surface. The increase of elevation of the surface so produced will tend to be greatest in the neutral region or near the edge of the region of erosion, not under the region of rapid erosion nor under the region of rapid deposition.* This undertow must exist chiefly below the depth of compensa- tion. If the earth were a perfect fluid the materials of different densities would, if not diffusible, arrange themselves in concentric shells with the heavier material toward the center. There is always t Science, February 10, IQII, p. 205. 622 HARMON LEWIS a tendency for the earth to take this arrangement in the sense that the stress-differences are on an average tending in this direction. It would seem, therefore, as a general proposition, that where the material of the earth is weak, the tendency would be more toward equalization of densities laterally than toward lateral differentiation of densities such as implied by isostatic compensation. Even though the stress-differences due to lateral variations in density were not sufficient to deform the rock so as to equalize the densities, a flowage from beneath an area of deposition to an area of erosion would certainly tend to produce a distribution of density within the zone of flowage itself which would have no relation to topog- raphy. It appears, therefore, that there could be very little isostatic compensation in a zone where yielding occurs as readily as postulated by isostasy. Now, according to the theory of isostasy, compensation would be essentially complete, and if compensation is complete the depth of compensation as determined by Hayford’s geodetic work would be as great as 60 miles. Hence, the undertow postulated by isostasy would exist chiefly below 60 miles. It is decidedly ques- tionable that an undertow even much nearer to the surface than 60 miles would cause the observed folding in the upper few miles of the crust. The theory of isostasy cannot account for the general uplift of sedi- ments without folding.—If the isostatic compensation is complete any deposition of material should cause a sinking of the under- lying segment. Isostasy could not therefore account for the fact that horizontal sedimentary rocks are found far above sea level unless a lowering of the sea level were supposed; but this possi- bility can generally be dismissed because the relative change in sea level is not registered in all parts of the world.* 1In discussing isostatic adjustments (see Science, February 10, 1911) Hayford . suggests that some uplifts are due to expansion and contraction caused by heating and cooling of sedimentation and erosion. These deformations, however, are not a distinc- tive assumption of the theory of isostasy at least as the theory is conceived in this paper; but were suggested to account for certain geological phenomena which the theory of isostasy could not explain. At any rate, expansion due to heating effect of sediments is entirely inadequate to account for known uplifts. In making his estimate that the vertical expansion is THE THEORY OF ISOSTASY 623 The theory of tsostasy does not explain the apparently heterogene- ous relation of uplift and subsidence to erosion and deposition.—Since isostasy postulates an adjustment or flowage which is comparable in speed to the process of erosion, a high area which is subject to erosion should be further uplifted as erosion progresses and should not be reduced to sea level until its deficient density is equalized by erosion of the lighter material at the surface and restoration of heavier material below. With a depth of compensation of 76 miles, the theory of isostasy would require greater continuous uplifts than are known to exist. Asa matter of fact, some areas have been uplifted as erosion progressed and others have remained stationary. In some cases erosion to a peneplain has been fol- lowed by subsidence and in other cases by uplift. ALTERNATE HYPOTHESES TO ACCOUNT FOR HAYFORD’S GEODETIC RESULTS Erosion and deposition are assumed to be the principal cause of disturbance of the equilibrium condition of isostasy. Since deposition does not in general extend beyond the boundaries of the continental shelves, the cause and effect of the type of deformation postulated by isostasy would be confined to the continents proper. So far, then as distributions of density are to be made a proof of the theory of isostasy, the critical test is not in the density relation of the continental masses as a whole compared to the ocean basins, but in the completeness of compensation of the topographic features of the continents. Now it has been shown in the preceding section of this paper that, though there is very likely a complete compensa- tion of the ocean defects of mass, yet it is a distinct possibility so one foot for every 33 feet of deposition Hayford neglects the fact that the irregularities in the isothermal surfaces near the surface of the earth flatten out with depth. How- ever, taking Hayford’s estimate and assuming that an area of deposition was covered by very shallow water and that the expansion due to heating took place all at one time, the maximum uplift above sea level could not be more than one-thirty-third of the thickness of the sediments deposited. Subsequent erosion would tend to reduce this elevation and any further elevation caused by relief from eroded material would certainly not more than equal the eroded layer. Hence, wherever the present elevation above sea level is more than one-thirty-third the thickness of the last con- formable sedimentary series, some other factor than expansion due to heating effect of sedimentation must be sought to account for the uplift. 624 HARMON LEWIS far as the geodetic evidence is concerned that the compensation of the topographic features of the continent is decidedly incomplete. The theory of isostasy can not therefore be considered as established by the geodetic work of Hayford. Furthermore, the probability that isostasy exists is lessened by the fact that an incomplete com- pensation can be very plausibly explained without involving the conceptions of isostasy. The tendency of lateral compression to produce isostatic compensa-* tion.—In the folding and overthrust faulting of rocks there is abun- dant evidence of lateral compression. It has already been shown that this folding is probably not caused by an undertow such as isostasy supposes to be set in motion by erosion and deposition. The compression indicated by folding may be due to shrinkage of the earth; it may be due to squeezing of the continental segments by the oceanic segments; or it may be due to other causes; but whatever the cause may be, it is certain that it has produced great uplifts. Suppose that the continent is composed of portions of differ- ent densities, but that the stress-differences set up by these differ- ences in density are not sufficient to cause a deformation of the material and a consequent uplift of the lighter masses. If this were the case, it would seem reasonable to believe that there would be a tendency for the effects of lateral compression to localize in the lighter segments since there is always a tendency for lighter seg- ments to move up even though the stress-differences tending in this direction are not sufficient to produce an actual movement. Other things being equal, folding would probably tend to localize in sedi- mentary rocks since the parallel bedding planes allow slipping to take place readily. Here again there might be a tendency toward isostatic compensation since sedimentary rocks on an average are lighter than igneous rocks. There are undoubtedly other factors which determine the place of folding, but it is entirely possible that uplifts by folding are incompletely compensated. A compensation of areas which have been uplifted without folding may be accounted for in a similar way. It is possible that lateral forces similar in magnitude to those forces which produce folding at the surface, but localized at greater depth should cause a deformation which is registered at the surface simply as a general THE THEORY OF ISOSTASY 625 uplift. In this case also the deformation would tend to localize where the resistance to uplift were least, in other words, in the lighter segments. The type of deformation suggested here is perfectly distinct from that postulated by isostasy. The theory of isostasy supposes that light areas are high because the strength of the material below is not sufficient to support segments different in weight. The possibility suggested in the preceding paragraphs is that high areas are light because the great deforming forces of the earth follow the path of least resistance. The automatic compensation of uplifts and subsidences due to expansion and contraction.—lIt is possible that some uplifts and sub- sidences are due to expansion and contraction of the underlying material. Such deformations are not a distinctive assumption of the theory of isostasy. Changes of volume may be due to changes of temperature or pressure which in turn may be due to a variety of causes. Any changes of elevation caused by expansion or contraction will be automatically compensated since the weight does not change. This would be another factor tending to produce compensation which does not involve the type of deformation postulated in isostasy. SECTION IV. SUMMARY OF CONCLUSIONS Isostasy is a theory of earth movements based on the assump- tion that the lighter portions of the earth are elevated in propor- tion to their defect of density because the earth is not strong enough to support segments of different weights. The principal support for the theory is the geodetic work of Hayford from which it was concluded that the excesses of mass at the surface are com- pletely compensated for by defects of density below, said defects of density extending to a depth of something like 60 to 150 miles. It is believed that Hayford made an error in determining the degree of completeness of compensation which invalidates his conclusions, for he assumed complete compensation in calculating the depth and then used this depth to calculate the degree of com- pleteness. Hence, instead of the single possibility of a practically complete compensation, there are, so far as has been shown from 626 HARMON LEWIS the geodetic evidence, three groups of possibilities for isostatic compensation: first, the possibility of complete compensation at a depth in the neighborhood of 60 to 150 miles depending on the distribution of compensation; second, the possibility of an over- compensation at a greater depth for land areas with probably complete compensation for ocean areas; and third, the possibility of under-compensation at a shallower depth for land areas with complete or over-compensation for ocean areas. Hayford’s geodetic results do not, therefore, constitute a proof of the theory of isostasy. An incomplete compensation of the topographic features of the continents can be plausibly explained without supposing the type of deformation postulated by isostasy. There are many important phenomena which the theory of isostasy will not explain. SPECULATIONS REGARDING THE GENESIS OF THE DIAMOND ORVILLE A. DERBY Rio de Janeiro The recent admirable summary by Dr. Percy A. Wagner’ of what is now definitely known regarding the geological conditions in which the diamond occurs in South Africa suggests certain specu- lative points of view, which, if found worthy of attention, may in turn suggest desirable lines of investigation in the field and the labo- ratory. These inquiries may perchance throw light on the intricate, fascinating question of the genesis of the diamond; or, even in a broader way, on the réle of carbon in eruptive rocks, whether in the form of diamond or graphite, or as gas locked up in carbonates or certain silicates. To the student of the occurrence of the diamond in countries other than South Africa, one of the most significant facts established by the prospecting of the African miners is that, aside from its well-known occurrence in pipes, the diamond-bearing eruptive material, kimberlite, occurs also in dikes, and that these usually have considerable longitudinal extension but only small width, except where expanded into pipes, and even these are frequently insignificant in relative dimensions. This slight prominence of the diamond-bearing bodies, coupled with the extreme susceptibility of the material to alterations which render its identification a matter of great difficulty, suggests at once that the failure to detect such dikes and pipes in districts in which the diamond is found only in sedimentary deposits, modern or ancient, is not a conclusive argu- ment against their existence, nor is it clear evidence that the original matrix was notably different from the South African kimberlite. In countries like India and Brazil, in which the diamonds of the modern alluvial deposits have been definitely traced back to con- glomerates of considerable geological age, the presence or absence of kimberlite dikes should be tested by prospecting operations, * Die diamantfiihrenden Gesteine Siidafrikas, Berlin, 1909. . 627 628 ORVILLE A. DERBY not so much, perhaps, in the areas occupied by the conglomerates themselves as in the neighboring ones occupied by formations known to have been in existence when the conglomerates were laid down, and which have escaped being covered up by them or by later strata. Until such prospecting is done on a sufficiently large and efficient scale the opinion, which a few years ago seemed justified, that a mode of occurrence essentially different from the African must be postulated for these countries, should be held in suspense. The material filling the African pipes shows very pronounced fragmenting and apparently explosive action, which has shattered, and to some extent scattered, the eruptive rock characteristic of both the pipes and the dikes and has mixed it with a very consider- able amount of various other rocks, either brought up with it from lower horizons or detached from the surrounding rock masses. Dis- cussion is still going on as to whether the diamonds contained in these agglomeritic pipe-fillings are to be assigned to the eruptive rock proper or to some of the foreign rocks included in it, but the weight of evidence seems to be in favor of the first hypothesis. A very interesting view that was held for many years assigned the formation of the diamond to some kind of reaction zm situ, between the two classes of rock that occur in the pipes, the necessary carbon being supplied by the carbonaceous rocks through which, in some places, the pipes cut. Subsequent developments have completely disproved this hypothesis, but the essential part of it—the formation of the diamond im situ—is still worthy of consideration 7f another source of carbon can plausibly be brought into the question. So general is the association of the diamond with a fragmental state of the eruptive rock that enters into the composition of the pipes that the question naturally arises whether or not the diamond also occurs in such masses of this rock as have not been subjected to the fragmenting action. From the statements at hand it is clear that there is usually considerable difficulty in distinguishing between the massive and the fragmental forms of kimberlite. Apparently the distinction has only been made in a perfectly conclusive manner by the use of the microscope. The masses that can be thus exam- ined are so small that such negative evidence as they may give has in itself little value. Specimens of diamonds inclosed in fragmental THE GENESIS OF THE DIAMOND 629 material are quite common, but thus far those found in which the rock is clearly non-fragmental seem to be exclusively of the type of the so-called “eclogite nodules,”’ which are regarded by some as segregations in the kimberlite magma and by others as transported fragments of a pre-existing rock. In either case the experimental crushing, reported by Mr. Gardner Williams,’ of 20 tons of these nodules from the Kimberley mine without finding a single diamond, tells strongly against any general hypothesis of genesis based on the sporadic occurrence of these nodules. In the statements at hand relative to the occurrence of diamonds in the parts of dikes that are not expanded into pipes, the impression is given that the rock is non-fragmental; but the evidence on this point is not as clear as one could wish. As the case stands at present, and until unequivocal evidence to the contrary is presented, there is a reasonable presumption that a positive, perchance a genetic, relation exists between the diamond and the fragmental condition of the rock in which it occurs. This in turn may mean that the origin of the diamond can perhaps be assigned to reactions between the original rock, or rocks, of the filling and other elements whose introduction was made possible by the fragmenting of the mass, and which accompanied, or followed, the explosive action, if, per- chance, they did not constitute the actual agency that produced it. According to ideas generally received among geologists, the explosive action, as such, is but the culmination of previous thermal processes in the sudden production of gases, principally water vapor. The thermal processes may be protracted and varied in action and may occur repeatedly and extend into late phases of the eruptive period and to stages subsequent to it. One of the most important effects of the protracted action would be to saturate the fractured mass of rock with gases and with liquids resulting from their condensation. Various observers have expressed the opinion that this saturation under the conditions implied reached the point of establishing a marked degree of mobility in the mass, converting it into a veritable rock brew. Be this as it may, such a saturation, whatever its degree may have been, would establish conditions in which a certain amount of hydration (serpentinization) of the erup- t Trans. Am. Inst. Min. Engineers, 1904. 630 ORVILLE A. DERBY tive rock, composed largely of olivine, would almost inevitably result. We thus have in the formative stages of the pipe-fillings (or at least in early stages of their history) a sufficient agency for the hydration of their eruptive portions. Such a change has been observed down to such extraordinary depths that the usual explana- tion of atmospheric weathering seems utterly incredible.t The hydration, which to a greater or less degree seems to be character- istic of all known occurrences of undoubted kimberlite, whether appearing in pipes or dikes, is accompanied by the formation of a certain amount of calcite, which involves the introduction, in some stage of the history of the rock, of carbon in a condition to form the carbonic acid locked up in the mineral. This introduction may also be most plausibly assigned to the stage of thermal agitation, of which the fragmenting and explosive actions were the climax. The analysis cited in the preceding note, representing the least- altered kimberlite thus far examined, gives 2.54 per cent of carbonic acid, corresponding to about 5,000 grams of pure carbon to the unit of volume (load=o.453 cubic meters) used by the African miners in measuring their material. This amount of carbon, if present in the form of diamond, would give about 25,000 carats, whereas the usual yield of the De Beers load is under 1 carat (t/5 gram). There are thus strong a priori reasons for attributing to deep- seated causes long since extinct a great part of the hydration and carbonation which the eruptive rock, originally free from water carbon, has suffered. If such was the case, it becomes important to distinguish the deep-seated actions from those of the atmosphere, which, acting from above downward, have long been producing similar results. These superficial results would be superimposed on the pre-existing ones, if such existed, down to a certain depth. No question can be raised regarding the correctness of the view, t Dr. Wagner gives an analysis of a specimen of kimberlite collected in the deepest part (2,040 feet from the surface) of the De Beers mine, which had 6.81 per cent of combined water. ‘This, as he expressly states, represents the best-preserved material to be found in the Kimberley group of mines, although in the neighboring Kimberley mine the pipe has been opened up nearly 1,000 feet farther down, or fully 3,000 feet from the surface. From this it may legitimately be inferred that there is little likeli- hood of finding unhydrated kimberlite in the known South African diamond mines. THE GENESIS OF THE DIAMOND 631 very generally received, that atmospheric weathering has trans- formed the “hard blue” ground into “soft blue,” and this in turn into “yellow” ground. If, as here suggested, there had been a previous period in which serpentine and calcite were formed, evidence for or against it should be found in the transition zone between the hard and the soft blue ground. So far as can be gathered from the literature at hand careful search has never been made for such evidence. This seems thus to be one of the crucial points in the study of the genesis of the diamond that is yet to be investigated. On the assumption that future investigation may establish the deep-seated origin of the alteration of the diamond matrix, a basis is found for submitting to discussion the elements of a new hypothe- sis regarding the genesis of the diamond. A pipe filled with rock fragments saturated with hot (possibly superheated) gases, and probably also liquids, would constitute an enormous crucible, in which reactions not as yet detected in our laboratories might take place. In this crucible carbon would certainly be present in the form of carbonic acid and probably in other gaseous forms as well. Thus the material and some of the physical conditions for unusual carbon segregation were present, and we are not yet, apparently, in a position to say that a segregation of a minute portion of the carbon into a solid form is a chemical impossibility. It seems to be well established that in certain industrial and experimental processes carbon does separate in the solid form of graphite from carbonaceous gases, and Weinschenk has presented strong evidence in favor of the introduction in a gaseous form of the carbon of the graphite deposits of Bohemia and Bavaria. From a geological point of view the réle of carbon in eruptive rocks and in eruptive phenomena generally is as important as it is obscure. It thus presents an attractive subject for experimental researches, such as are contemplated in the program of the Geophysi- cal Laboratory at Washington, which is so admirably equipped both in material and personnel for such investigations. The inquiries in this line thus far reported by various experimenters, while extremely interesting and important in themselves, are unsatis- factory, in so far as they postulate conditions that are with diffi- culty conceivable in nature. PRELIMINARY NOTES ON SOME IGNEOUS ROCKS OF JAPAN. IV! S. KOZU Imperial Geological Survey of Japan IV. ON LAVA AND ANORTHITE-CRYSTALS ERUPTED FROM THE TARUMAI VOLCANO IN 1909 Introduction.—The volcano Tarumai is located at a distance of about 42 kilometers south of Sapporo, the chief city of Hokkaido. Though the volcano has long been known as one of the active volcanoes in the district, it has become the object of special atten- tion since the outpouring of lava, which took place in April, 1909, forming a dome of 134 meters in height as measured from the neighboring ground, and adding 4o meters to the pre-existing highest peak of the mountain, which is 1,015 meters above the sea- level, according to Oinoue’s report. A revival of the exhausted volcanic energy, which had remained in the solfataric state since the comparatively great explosion of August 17, 1895, took place at the beginning of the year 1909. After that several outbursts and shocks were reported from the region. At last, in the course of about 24 hours, from the evening of the 17th to that of the 18th of April, lava of about 20,000,000 cubic meters in volume, measured by B. Koto, was poured out of the explosive crater, and a dome was formed which is shown in the accompanying photographs (Figs. 1 and 2). Reports of the event, written in Japanese by D. Sat6 and Y. Oinoue, have been published by the Imperial Geological Survey of Japan and the Earthquake Investigation Committee, respectively. The following brief petrographic description was made by the writer on the specimens collected by D. Sato. Megasco pical characters.—The specimens at hand have in general a glassy and ragged appearance. Those taken from the ejecta are t Published by permission of the Director of the Imperial Geological Survey of Japan. 632 NOTES ON SOME IGNEOUS ROCKS OF JAPAN 633 of light-colored pumice. In their inner part the cellular structure appears well developed, while the outer thin crust is usually glassy and compact, strongly marked by cracks, which are char- Fic. 1.—View of the new dome from the east, May 11, 1909 (by T. Kawasaki, Imp. Geol. Surv. of Japan). Fic. 2.—View of the new dome from the southwest, May 11, 1909 (by T. Kawasaki). acteristic of the so-called bread-crust bombs. This variety con- tains well-formed anorthite crystals of considerable size, with an average length of 13 mm. Beside these, there are not a few small 634 SKOZO, megascopic phenocrysts of feldspar and pyroxene, their sizes vary- ing from 1mm. to 2mm. The other specimens taken from the new dome are dark gray, or reddish dark gray, in color and spongy or ragged in appearance. Generally, they are characterized by heterogeneity in texture due to their variable crystallinity, and by flow-structure, which is visible in the lava-mass, as shown in Fig. 3. Fic. 3.—Lava-block, showing the marked flow-structure Sometimes dark-gray to light-gray cryptocrystalline masses are imbedded along the planes of flow in the rock-mass, their shapes being mostly lenticular. Microscopical characters —The rock is made up of plagioclase, hypersthene, augite, olivine, magnetite, apatite, and microliths, scattered in the abundant glassy groundmass. The prevailing phenocrysts are of anorthite. Hypersthene comes next, and is nearly equal to, or is more than, the augite. Subhedral magnetite is not rare as phenocrysts. Though olivine appears abundantly associated with the large crystals of anorthite, mostly as peripheral NOTES ON SOME IGNEOUS ROCKS OF JAPAN 635 inclusions, it is rarely met with in the general mass, and may be considered as an accessory constituent of the rock. The matrix exhibits different textures, according to different conditions under which the lava consolidated. The crustal part of the ejecta is hypohyaline, while its inner part usually shows typical cellular structure, the glass base being colorless. The specimens taken from the dome are more crystalline than those just described, but there is still abundant residual glass. It is moderately clouded with magnetite dust and pyroxene microliths. The megascopically cryptocrystalline part appears holocrystalline under the microscope and consists essentially of granular pyroxene and feldspar, scattered microporphyritically with skeletal crystals of hypersthene. Feldspar.—The feldspar phenocrysts are of two kinds. One of them is well-formed anorthite of considerable size, 13 mm. in average length. Zonal structure is nearly wanting, or is indis- tinct. These occur in the ejecta and peripheral part of the lava, or even as isolated crystals, suggesting that their crystallization was prior to that of other minerals. The other variety is smaller in size, with an average length of 2 mm., and is commomly sub- hedral in shape, sometimes with a strongly curved outline invaded by the glassy groundmass. This variety differs from the first in possessing zonal structure due to variation in the chemical compo- sition and to the arrangement of abundant inclusions. In average composition, the second variety is slightly more sodic than the first. The most abundant inclusions are light-brown or colorless glass with air bubbles in many instances. Apatite and magnetite are also present, commonly in small quantity. In some crystals pyroxene appears as inclusions, but more commonly the feldspar is abundantly inclosed in the hypersthene and augite phenocrysts and shows a distinct automorphic relation toward the pyroxene. ’ The larger crystals will be more fully described in the second part of this article. Pyroxene.—The hypersthene is easily distinguishable from the augite by marked pleochroism, low double refraction, parallel extinction, and crystal habit. It occurs in crystals of two periods of crystallization. The largest phenocrysts are 2.5 mm. in length 636 | S. KOZU along the axis c. Pleochroism is distinct; a=reddish brown, 8= greenish yellow, y=yellowish green. It is optically negative, the optic plane being parallel to the orthopinacoid. There are abun- dant inclusions of plagioclase, magnetite, glass, and a few crystals of apatite; of these the plagioclase is large and conspicuous. The smaller hypersthene is rather euhedral in shape and is sometimes marked with transverse cracks perpendicular to the axis c. Augite crystals are anhedral to subhedral, and also have abun- dant inclusions, just as the orthorhombic pyroxene. Parallel growth with the hypersthene is common, the hypersthene always being inclosed by augite. Twinning parallel to the orthopinacoidal face commonly occurs, and that parallel to (zor) is rare. Olivine crystals, as already stated, occur in association with the large crystals of anorthite and have well-defined form, elongated along the vertical axis with a length of about 2mm. The pre- dominating faces, easily identified by the naked eye, are m(z10), k(o21), and b(o10). They are usually coated by a dark-reddish colored, thin crust. They frequently occur in groups of several individuals associated with a smaller quantity of magnetite grains. Notwithstanding the noticeable fact that olivine is nearly absent, or very scarce, in the general mass of the rock, it appears abundantly as peripheral inclusions of the large anorthite. Magnetite occurs frequently as phenocrysts in association with those of pyroxene, and varies in size from 0.1 mm. to 0.3 mm., in striking contrast with the same mineral in the groundmass, which appears as dusty grains. A patite usually occurs as needle-shaped inclusions, but in a few instances larger crystals with a violet color, finely striated parallel to the vertical axis, appear in the groundmass. Chemical characters—The analysis of the rock made by N. Yoshioka in the chemical laboratory of the Imperial Geological Survey of Japan is as follows: SIO ay ci he ee See Cea Siete cer een eats eae re ee 60.93 ATS O)gy ato! 2p le neM Uae tlh amet Forte ne eee g 16.46 | Ice © Rare ean Seu eT Cee cuntR Ie eR ad ory BOR 6 3.35 1 Sis, @ WaPRRen rer nee ea Rese Stee dius, irate ie ceid iG SY 5-94 NOTES ON SOME IGNEOUS ROCKS OF JAPAN 637 (CEN) Oa ooh OSM Na aR A oI oR gr 7.84 TNFa © Peay testa yt «Ate nas coe aseeel ant oe 1.44 IK AO) 8 orale lec a Ree oR RRR RP Omi gma nee 0.79 SEO ere eek erty SRE CRG usta i ee gee a n.d MIN O ere at. os he hE ih Sins, Seana, nae 0.42 Jee Oar Seat GA EE 0.13 IMO) 8 Ga weet aa Oa Mend SE eae 0.55 100.78 25 eo Orchoclasepiescqencs cet eas Aches, 5 kia Pace ets: elo) JNM BYNES cota oi Sa Ree a ee 72a PAMOREMIL CMe erie cn tee aun evn in See Ss 36.1 ID OVO RIGNS S ce Mara tre Sa hee ance een oe aa 2a sy ELStMEIM Crys sky aac uN Weak eee Le. sees 13.6 IMIAOMELILE econ nein ho ge Waleh aan ak 4.9 Wlineniterewew aa ute ea Naame. Meret» 0.8 99.7 The ratios are: Sal SSR eT a EEE tara 3.66 Fem Q SNM RA ews Gatco, enter Deh auras cuenta 0.47 F K,0’+Na.0’ At RRC IE ev cane re ee 0.25 CaO’ K,0’ BM cars oa cent eran MARA ROIS Spc, ve cet 0.39 Na.0’ By the Quantitative System the rock would be classified under the name bandose. In this classification, it may be noted that the rock is charac- terized by a high percentage of lime, which appears mostly as modal anorthite, and by the comparatively high silica content. Generally speaking, the mineralogical and chemical characters of the latest lava of Tarumai volcano seem to be representative of 638 S. KOZU those of the modern pyroxene-andesites, which are widely spread over the Japanese Islands, judging from a cursory glance over the volcanic rocks of Japan. For this reason the name bandose appears to be particularly appropriate. ANORTHITE-CRYSTALS IN THE LAVA OF I909 The occurrence of the larger crystals of anorthite is noteworthy. The crystals form large phenocrysts in the lava, and have been Fic. 4.—Cavity with anorthite crystal. Natural size ejected separately also as the so-called ‘‘anorthite bombs,” and are scattered abundantly around the crater; as.is the case with the anorthite on Miyake-jima,’ one of the Seven Izu Islands, Zao-san, a volcano in the province of Rikuzen, and Iwate-san, a volcano in the province of Rikuchu; the oligoclase-andesine? on Naka-i6-jima, one of the Sulphur Islands, may be cited as the parallel examples. t Kikuchi, ‘‘On Anorthite from Miyake-jima,”’ Journal of the College of Science, Imp. Univ. Japan, I, Part I, p. 31. 2 Wakimizu, ‘‘The Ephemeral Volcanic Island in the Idjima Group (Sulphur Islands) ,”’ Publications of the Earthquake Investigation Committee, 1908, No. 22 C, Tokyo. NOTES ON SOME IGNEOUS ROCKS OF JAPAN 639 A black, thin coating of lava, which crusts the Miyake-jima anorthite and the Naka-id-jima oligoclase-andesine, is not seen on the mineral from Tarumai. The crystals, however, have attached to them a small quantity of light-colored pumice. It is evident that the matrix of brittle pumice separated easily from the crystals and that the semi-solidified lava was not so viscous as in the case of the lava of Miyake-jima and of Naka-id-jima. Also, in some specimens the crystal is in a cavity having well-defined Fic. 5.—Well-defined cavity from which the anorthite crystal has been lost. Natural size. walls corresponding to the faces of the crystal, with a space about 4mm. in width between the crystal and the lava. The crystal is attached to the walls by slender, needle-like filaments of glass, as shown in the accompanying photographs (Figs. 4 and 5). There may be several explanations of the formation of these cavities, but the writer believes they were formed chiefly by the differential movements of the crystal and matrix when the blocks of lava were ejected in a semi-solidified state. The common sizes of the crystals are io mm. to 15 mm. in the 640 SmaKOZU: longest diameter, though the largest is 20mm. or longer. Their surfaces are not vitreous, or smooth owing to the presence of pumiceous matrix and inclusions of olivine crystals with a few magnetite grains. The olivine is in well-defined forms, as already described. The roughness of the crystal faces and the twin striation upon them made the use of the reflection-goniometer very difficult. Even the cleavage piece used for the measurement of the facial angle (oor) : (o10) did not give a satisfactory result, as the reflec- tion on (o10) was disturbed by the pericline twin striations. The angle measured lies between 85° 48’ and 85° 52’. Other approxi- mate facial angles measured by the contact-goniometer are as follows: AGE Co} Ss oh do) en ea ene arg Ear le Heb o> 50a 130! F(20T) SP (OOT)i eat tal Wyte ee ee hn eee ee 81° 10’ v.20) Mi(oro) gcse ake ee eee Poe go° 50’ ViC2On PLO) ihe weer oe ME Ee se na tics i Asc. 20) Toh eral ed Covey 8 ha unby MUN A alam ne a Ale tLe 42° M(O2E) ROOT) yee te ewe ael erent Lonmee 46° 40 From the above angles and the relation of the crystallographic zones the crystal-faces which were identified have been determined as follows: P(oor), M(oro), T(1To), /(110), é(201), y(20T), e(021), n(o21), m(11), o(11), p(1tT), and v(241). The faces P, M, y, T, /, 0, p, and ” are always observed, of which P. M, and y are the predominating faces. The face e is very rare, and ¢, f, v, and m are only found in the tabular crystal parallel to P(oor). Some distinguishable crystal habits are formed by the pre- dominance of different crystal-faces, as given below: First type: Prismatic, elongated along the axis a, with the faces P, M, and y predominating, as seen in Fig. 6. Second type: Tabular, parallel to M, its elongation being along the axis c, as seen in Fig. 7. NOTES ON SOME IGNEOUS ROCKS OF JAPAN 641 Third type: Tabular, parallel to P. This type might be sub- divided into two varieties, with gradations between them: a) The elongation is rather along the axis a than along the axis 5, as seen in Fig. 8. b) The elongation is along the axis 6, with specially well- developed y, as in Fig. 9, finally becoming thick tabular parallel to y. Fic. 10 EiGssnr Fourth type: Cubic, or equant, with comparatively well- developed face . This type might also be subdivided into two varieties, showing gradations into each other or to the first type: 642 S. KOZU a) With slight elongation along the axis c, as seen in Fig. ro. b) With slight elongation along the axis a, as seen in Fig. rr. The prevailing habits are the first and third types; the second and the 6 type of the fourth are not rare; the a type of the fourth is the scarcest. Twinning according to the Carlsbad, Manebach, albite, and pericline laws has been observed. There are two or more different types in combination. The albite and pericline types occur poly- synthetically, while the Carlsbad type occurs in combination with one or both of these. The Manebach is only found in the tabular crystal parallel to P, mostly combined with pericline twinning. The specific gravity measured by the Westphal’s balance in Thoulet’s solution is 2.759. Optical characters.—Extinction angles on P(oor) and M(oro), measured on cleavage pieces, are —36° 54’ and —35° 24’, respec- tively. The measurement of the mean index of refraction was made approximately by means of Wright’s solution. The solution corresponding to the mean refraction of the mineral was deter- mined on the Abbe total-reflectometer. The result is 2)=1.5785. The measurement of orientation of the optic axis B was made by the Becke method,’ with single screw micrometer ocular. The values p and ¢ of the axis B were measured on three thin slices parallel to P(oor). The results are as follows: (r1) On +P(oor) Micrometer { p= —60° 44.8’ Micrometer } p=+28° 12.6’ parallel (| r= o. 405 diagonal | r= 0.337 (2) On +P(oor) ° Micrometer p=—6o0 Micrometer p=+25° 5° parallel r= 0.346 diagonal (7= 0.318 (3) On —P(oot) Micrometer ( p=+55° 28.5’ Micrometer | p=—20. 11’ parallel r= 0.354 diagonal (7= 0.354 * Becke, “Bestimmung kalkreicher Plagioklase durch die Interferenzbilder von Zwillingen,” T'schermaks Min. Mitth., 1895, Bd. 14, s. 415-42. a NOTES ON SOME IGNEOUS ROCKS OF JAPAN 643 From the above figures, the following values for the azimuth of the axis B against the edge P/M on P (&) and the true angle- distance (@) are given: g w (1) me 10037 (2) —12.3° IQ 14’ (3) —12.7° tg 42’ reduced on +P —12.3° LO. 30. For calculation of , my=1.5785 and k=o.141 were adopted as the mean index of refraction and Mallard’s constant, respectively. From the mean values of € and @, ¢ and A were given as follows: I II Oa On 3e Oe A= —5.8° —4° 23/ The values under I are results approximately obtained by the construction of the stereographic projection and those under II by calculation. Plotting the latter values on Becke’s diagram, which indicates the relation between -the orientation of the optic axis B of plagioclase and its corresponding chemical com- position, the composition of the present mineral would be i@entitied as Ab,Anjg— Ab;An,;, aS shown in Fig. 12. The mineral is optically negative with r as the acute bisectuxe the optic ‘angle measured in cedar oil (ny= 50 §0 70 30 90 §=©100 Fic. 12.—In the figure, A is the anorthite from Vesuvius, B is the anorthite from 1.515) with yellow light, is Tarumai, and C is the bytownite from 2Ha=go° 11.5’ Naeroedal. and its true angle is Nia Sou esr Chemical characters —The mineral is easily attacked by hydro- chloric acid, and the powdered sample is readily decomposed with 644 Se KOAG. the separation of gelatinous silica in slightly hot hydrochloric acid of the strength of 22 per cent. The chemical analysis was made by W. Yasuda in the chemical laboratory of the Imperial Geological Survey of Japan. The sample for the analysis was taken from the clear and fresh part of a crystal, and powdered to grains one millimeter or smaller in diameter. ‘To remove the impure parts, which contained inclusions of olivine, magnetite, and glass, the grains were separated into three portions by Thoulet’s solution, having specific gravities of 2.747 and 2.760. The analysis was made of the sample with the specific gravity lying between the two values. The result is as follows: NOH LO Pesaran aieiiey ities i mers Gent tend ec URA a GP Kueh 43.51 INU O SF syaet is oats teas ve eta etic tea eee ee BGs | Sed Ramee areas Spee emcee NOU ue ges UN et rs ON trace DY) a @ Riera es Wee amen ise era w Wsee anu etary peldbetsEN os Tak LOE OEE aN eet ee ea ae SO ca oh end a uri tts 19.48 TINIE ES © tesa tre ae Onan tamara cae Ee ek rei 0.61 TE ea aN tata ceag leara gs aN peace Oy a 0.05 100. 53 Subtracting silica and magnesia corresponding to the olivine molecule, and potash as impurity, and calculating the remainder as parts in 100, we have: W (percentage) Mol. prop. SiO secre ane BBS 3 On coer Ge Hee ee 6.73) TaN © Naieme lien nla eect ina BONA Tia aes Laon ies ere cha 0.36 CaOn: rile My diets areata pen eeae Sa 0.35 INacO RE Ree erwaten: ONO Qua eat warn nears 0.01 100.00 From which it is found that the composition of the anorthite may be represented as a mixture of 2Ab and 35An, or Ab, , Any, ;, which corresponds closely to the value determined optically as Ab; Ang; pee ANge. FACTORS INFLUENCING THE ROUNDING OF SAND GRAINS VICTOR ZIEGLER South Dakota State School of Mines CONTENTS INTRODUCTION THe MOLECULAR FoRCEs OF LIQUIDS MOLECULAR FORCES AND TRANSPORTATION METHODS OF ROUNDING Summary of Previous Work Experimental Work SUMMARY INTRODUCTION In 1910, while discussing the rounding of sand grains with Professor A. W. Grabau, it seemed to the author that the influence of viscosity was not sufficiently emphasized in the literature on that subject. Subsequent discussion with Professor C. P. Berkey suggested this investigation. The thanks of the author are due to Professor James F. Kemp, and especially to Professor C. P. Berkey for many kind and valuable suggestions. THE MOLECULAR FORCES OF LIQUIDS For a clear understanding of the forces acting on a particle submerged in water it is necessary that we review briefly a few of the elementary definitions of physics. This can most clearly be done by means of an illustration. If we look carefully at the surface of a glass of water, we notice that it is not horizontal but curves upward at the sides of the con- taining vessel as though attracted by it. If we dip a clean glass rod in water and remove it, we shall see adhering to it a thin film of water. Upon slightly shaking the rod this film will be dis- 645 646 VICTOR ZIEGLER lodged and in falling will assume a more or less spherical form. The smaller the drop, the more perfect its spherical shape. Here we have a homely demonstration of the forces acting on the liquid. The creep-up of the water on the sides of the glass is due to the attraction of the glass for the water; the drop of water remaining on the glass rod is held there by the same force, that is—adhesion. In falling, the water from the rod does not fly off in a series of small particles, but assumes a spherical shape because the component particles of water, or, in other words, its molecules are attracted toward each other. Thisis cohesion. Adhesion is the attraction of unlike molecules for each other; cohesion is the attraction exhibited between molecules of the same substance.: The force due to the cohesion of the molecules of different substances and that due to the adhesion between the molecules of different substances varies. ‘The cohesion of water is less than its adhesion for glass, hence the glass rod is enabled to tear away a certain amount of water.? If, however, we dip a glass rod into mercury and withdraw it, nothing will adhere, because, in this case, co- hesion is the stronger force. The space through which cohesion is active is the “sphere of molecular attraction.”’ Itisasphere about 0.00005 mm. in diame- ter.3 If we now assume that a liquid is made up of a number of layers of molecules, we will see that the top layer, the free sur- face, will be attracted unequally because part of its ‘sphere of molecular attraction” lies outside the liquid.‘ In Fig. 1 xy is the surface of the liquid. A and B represent two molecules in the surface and beneath the surface respectively. The circles surrounding them represent x = _@--. © the “sphere of molecular sattractiomas Q The molecule B is attracted equally in all directions by the molecules falling within its sphere; in the case of the mole- cule A, however, the attraction will be downward, as the attract- ing molecules only occupy that part of the sphere lying within 6 Bigs t Nichols and Franklin, Elements of Physics, 124. 2 F, Pockels in Winkelman’s Handbuch der Physik, I, 882. 3 Duff, Textbook of Physics, 146. 4Ibid., 147. THE ROUNDING OF SAND GRAINS 647 the water. On this account the surface of the liquid is in a state of tension, and in order: to move the molecule B to the surface we would have to overcome this force. We may liken the con- dition of the surface of the liquid to that of the stretched rubber membrane of a ball. We have a pressure at right angles to the surface, capillary pressure, causing a tension parallel to the sur- face, surface tension.? Let us now consider a grain submerged in a liquid and let. us note the action of the different forces upon it. The body will be pulled down by the force of gravity, the magnitude of the pull being determined by the difference in the specific gravity of the solid and the liquid. If we consider water, then the force will be equal to vg (d—1); where vis the volume, g the acceleration due to gravity, and d the density of the solid. In moving through the liquid, the grain will carry down a thin film of water held by adhesion. There is a certain friction devel- oped in this movement which will not be friction between the grain and the water, but friction of water with water. The friction developed by a thin film of water sliding on water is “superficial viscosity.”’ The term ‘‘skin friction” is also applied to it.2 This is the friction especially considered in the flow of water through pipes and conduits. In addition, through the downward move- ment of the grain, the shape of the liquid is disturbed. Any disturbance or change of shape in a liquid calls forth a resistance, “viscosity.’’ But even if the particle were moving in a “perfect fluid,” 1.e., a fluid without any viscosity, its energy would gradu- ally be dissipated in forming waves. To summarize then, a body moving through water must over- come resistance due to three causes; (1) viscosity, (2) skin- friction, and (3) wave-resistance. If we take a case in which the liquid has a definite velocity, the conditions as outlined above will not change. In this case the grain will be acted on by a force which is the resultant of the velocity and gravity, and will have the direction of the diagonal t Ganot, Physics, 122. 2 Basset, Elementary Treatise on Hydrodynamics, 52. 3 [bid., 51. 648 VICTOR ZIEGLER of the parallelogram of forces constructed with velocity and gravity as sides (Fig. 2). The grain will experi- ence no resistance in the direction of the velocity, as it will simply move along with the water. The downward move- ment will experience the same resistance as though the liquid were at rest. MOLECULAR FORCES AND TRANSPORTATION Sediment is transported by water in one of three methods. It is either floated on the surface, or rolled along the bottom, or carried in suspension. Small grains, when carefully sifted over the surface of water, float, due to the fact that their weight is not sufficient to overcome the surface tension of water. Since surface tension may be defined as the ‘“‘force tending to make a liquid contract to the smallest area admissible,” it will have the tendency to drive the floating grains together.t This apparent attraction of grains into patches, although not explained, has been noted by James C. Graham and F. W. Simonds, who described this method of sand-transportation as occurring on the Connecticut and Llanos rivers respectively.? Experiments carried on by Simonds seem to show that if the launching be favorable, about 40 per cent of the component grains of most sandstones will float on water. Floating patches of sand and dust have been noticed by the author on the Iowa and Cedar rivers on still days during the summer, where they look essentially like floating patches of scum or foam, and also on the quiet water along the shore of the North Sea, near Otterndorf and Cuxhaven in Germany. While the condition necessary for the transportation by flotation are somewhat unusual, this method still appears to have more importance than is usually attributed to it. The floating of the grain depends on two molecular forces, viz., cohesion and adhesion. Cohesion causes the tension in the free surface of the water, and resists all attempts to break this surface. Adhesion serves as a modifying factor. If the adhesion t Duff, op. cit., 146. 2 Graham, A.J.S., series 3, XL, 476; Simonds, Am. Geol., XVII, 20. THE ROUNDING OF SAND GRAINS 649 between the grain and water be strong enough to wet the grain, it will sink at once; if adhesion be weak, the grain will remain dry and float. The adhesion between the grain and the water may be entirely destroyed by coating them with oil. The so- called “‘oil-flotation process” of ore dressing depends to a great extent on this principle. The finely pulverized ore is mixed with a small quantity of oil. The metallic sulphides, such as galena, chalcopyrite, and sphalerite, have strong adhesion for oil, and are readily coated, while the quartz and other gangue remain free, unless an excessive amount of oil is used. When the ore is allowed to slide into the settling tanks, the gangue sinks readily, but the coated sulphides float off. Here it seems that molecular forces cause flotation rather than the decrease in specific gravity due to the combined weight of oil and mineral. As the specific gravity of the oil taken is approximately o.8, in the case of galena the volume of oil to mineral would have to be in the ratio of 32 to 1, to bring the density of the combined material down to that of water.t Sharp, angular grains float more readily than those of spherical shape. This is due to the fact that the force due to the surface tension increases with an increase in the surface area exposed to it. The more nearly spherical a grain, the smaller the ratio between the surface area and the mass of the grain, and hence the greater the ratio of weight’ to surface tension. Irregularity of shape increases the ratio of surface to mass, and hence decreases the tendency to break through the surface of the film. The power of water to carry material in suspension depends on a number of factors, some of which are: the shape, size, and composition of the particles; the viscosity, composition, and ve- locity of the water; the presence of colloids; the character of the river bottom; the course of the stream, etc. The size of grain carried depends directly on the velocity. The more irregular the shape, the greater will be the resistance encountered in settling. The presence of colloidal substances causes rapid settling. Again there may be a change in the composition of the water causing an interaction with the sediment, such as the precipitation of alumina t Adams, M. and Sc. Press, May 7, 1904,.etc. 2F. W. Clarke, Data of Geochemistry, 430 (Bull. 330, U.S.G.S.). 650 VICTOR ZIEGLER by the carbonates of calcium and magnesium, and a consequent settling of the silt.t The presence of salts, alkalies, and acids in solution hasten the rate of precipitation. However, Wheeler arrives at the conclusion that there is practically no difference in the rate of settlement of sand and silt in salt and fresh water.” When the particles are very fine, as mud and ooze, the rate of settlement is slightly faster in salt than in fresh water. Others have shown that settling is far more rapid in salt than in fresh water, and attribute this fact to a chemical interaction between the salt water and the sediment, carried in this case as a colloid. There is reason to doubt this explanation, and the more rapid settling in salt water seems to be due to a decrease in the viscosity of the water.4. Rough and irregular river bottoms and swinging meanders tend to keep the water in a stirred condition and hence aid in holding material. METHODS OF ROUNDING Sand grains are reduced in size by collision and friction. Hence we know that the wear of a grain depends on a number of factors, such as hardness, weight, distance of travel, cleavage, tenacity, velocity of movement, etc. The rounding of sand grains under the varying conditions has been ably discussed from the geological standpoint by McKee’ and Goodchild. The movements of solids through fluids have been investigated from the mathematical stand- point especially by Basset? and Allen.’ This feature has also been noted to some extent by Blake,? Walther,’? and Barrell." ‘E,W. Hilgaard, A.J.S., 1873, p. 288; 1879, p. 205. 2W.H. Wheeler, Nature, June 20, toot. 3 See F. W. Clarke, Bull. 330, U.S.G.S., and H.S. Allen, Nature, July 18, 1901, for bibliographies. 4J. F. Blake, Geol. Mag., Decade IV, Vol. X, 12; W. B. Scott, Introduction to Geology, 141; Carl Barus, Bull. 36, U.S.G.S.: Chamberlin and Salisbury, College Geology, 3106. 5 McKee, Edinburgh Geol. Soc., VII, 208. ® Goodchild, ibid., 208. 7 Basset, Elementary Treatise on Hydrodynamics. 8 Allen, Phil. Mag., 1900. 9 Blake, Geol. Mag., Decade IV, Vol. X, 12. to Walther, Das Gesetz der Wustenbildung. ™ Barrell, Jowr. Geol. (1908), XVI, 150. THE ROUNDING OF SAND GRAINS 651 Summary of previous work.—McKee in his work evolves the formula size X specific gravity X distance traveled R« ; hardness where R is the rounding (or the wear). Considering a cube with the edge x, the distance traveled would be roughly proportionate to the number of times the grain D turned over, hence me could be placed instead of distance. The weight of the cube would be «3 Sp. Gr. Substituting in the above equation we have x3 Sp. Gr“ * hardness — reducing to R ae Sp. Gr. d 4h Or in more general terms— a2: op. Grd mh R« where m is a constant depending on the shape of the grain. m is 4 in the case of a cube, 3.1416 in the case of a sphere, etc. If the grain is under water allowance must be made and oe - (Sp. Gr.—1) -d IK mh Goodchild goes farther and determines a general limiting con- dition to the wear taking place. His work may be summarized as follows: Since the sand is completely surrounded by a film of the water in which it is submerged, it will be acted on by surface tension. By decreasing the size of a particle we increase the ratio of area to volume, and hence to weight. Since the surface tension of water will act over the area exposed, its magnitude compared to the weight of the grain will increase with decrease in size. Finally, he assumes that a balance between weight and surface tension will be reached, such that no further rupture of the film of water surrounding the grain can take place, and hence all wear will 652 VICTOR ZIEGLER cease. Thus Goodchild concludes that the factor limiting the amount of wear possible on submerged bodies, is surface tension. Experimental work.—As stated before, in the movement of bodies through water resistance due to three causes must be EXPERIMENTS mm. Diam. Glycerin Water Alcohol Cassiterite (6.4)* Bao eae at at nen Collision Collision Collision De Tanai aa a eee ee Collision Collision Collision Te oie cue cael cua ? Repulsion ? Collision Collision Ca ham ok Renee ie. Repulsion ? Collision ? Collision ae a meme tine pan eee Repulsion Repulsion ? Repulsion ? Chromite (4.5) BD ee eg sh eae RAR Repulsion Collision Collision Peel n a aen this eaineney eeeteey c Repulsion ? Collision ? Collision De etic een ae ee Strong Repulsion | Repulsion Collision CAS Stren ene eon aren Strong Repulsion | Repulsion ? Repulsion ? ae ERI cet ae ile re Strong Repulsion | Repulsion ? Repulsion ? Quartz (2.65) BD Ue aoe aa Collision Collision Collision peer te tie NL Peed Vir Shs ee ? Repulsion ? Collision Collision hee a A nel ee aD eo Repulsion Repulsion ? Repulsion ? le CRE et re WEN inter we Repulsion Repulsion ? Repulsion ? Gypsum (2.35) BO. tensa a eC ura Collision Collision Collision QT a's is gh aa eee ? Repulsion ? Collision Collision Tee ed ae flaps ae ? Repulsion ? Repulsion ? Collision ? Cain ach Gen ete ? Repulsion ? Repulsion ? Repulsion P rae NU pra ? Repulsion ? Repulsion Repulsion Anthracite (1.6) rar ae RA rete ee alle cee aR eatte ia Collision Collision Dea ge cl haere ge on (ea rae rare cree ? Collision ? Collision it el Caer ee ene Sate Nola Slo CEMA eee Repulsion Repulsion ra See renee ciate Wo cnc eite an 4 oe Repulsion Repulsion Raa PAOD Raia! sie tash l= Ras eiea mae ote te orate Repulsion Repulsion * The figures beside the minerals represent specific gravity. DATA | | Sp. Gravity Surf. Tension Viscosity Gly cerinyets sgn atin one Taig 66.5 8.0 Warten it aie ahead else ae TO) FR .10 Nl eoholAeatine nh. ce eerie 887 23.4 .OIL overcome, viz., viscosity, skin-friction, and wave-resistance. Unless these three factors are overcome, grains cannot collide. THE ROUNDING OF SAND GRAINS 653 The effect of surface tension, however, is one aiding wear, since it tends to draw grains together in its effort to force the water to assume the least area permissible under the conditions to which it is subject. Thus viscosity, since it is the most potent of the three factors mentioned, limits the minimum size to which wear takes place. The energy of the particle must overcome the viscos- ity to allow collision. Since the velocity of different grains in water is roughly equivalent, their energy varies directly with the size, the larger grains only having enough power to overcome viscosity. In the case of small grains the water acts as a cushion preventing actual collision, or checking the velocity of contact. To show the action of viscosity in preventing collisions of grains the following experiments were performed. Grains were dropped down long glass tubes filled-with liquids of different viscosities, and the action at the meeting of the grains was observed. Grains of different specific gravities . were taken so as to overcome the difference in the specific gravities of the liquids. 4 Again an experiment was performed (Fig. 3) A in which the glycerin was allowed to run from the reservoir C through the tube AA down which the different grains were dropped. The results were practically identical with those above. It will be noted that the surface tensions of A water and glycerin are nearly the same, but that Fic. 3 the viscosities are in the ratio of eighty to one. In the case of glycerin it was apparently impossible for the grains of small diameter to collide. Whenever a larger grain would over- take a smaller and slower falling one, there was an apparent repul- sion between the two as they were held apart by the viscosity. In small and light grains the repulsion appeared violent so that often a clearing space of a quarter of an inch was shown by grains that apparently were going to collide. As can be seen from the table, in the case of water the protection against collision was much less. Small grains of quartz, less than 1 mm. in diameter showed fairly strong repulsion, but above that size collisions were the rule. Again in the case of alcohol, with a surface tension of 654 VICTOR ZIEGLER 23.4 and a viscosity of o.o11, repulsion was only noticed in the finest grains. SUMMARY The results of these experiments seem to show that viscosity is the factor protecting grains from wear. Viscosity will not only prevent the wear of the smaller grains, but it will also act as a buffer and will greatly lessen the velocity of grains when about to collide with each other or with the bottom of the river. In view of the results it seems improbable to the writer that grains less than 0.75 mm. in diameter could be well rounded under water. Well-rounded grains of about this and smaller diameter appear to be the result of wind work, in which case the protecting factor, viscosity, would be practically zero, so that there would be no limit to the minimum size attainable by wear. THE UNCONFORMITY BETWEEN THE BEDFORD AND BEREA FORMATIONS OF NORTHERN OHIO? WILBUR GREELEY BURROUGHS Oberlin, Ohio In Lorain County of northern Ohio, 30 to 4o miles west of Cleveland, occurs a striking unconformity between the Bedford and Berea formations. In Ohio the Bedford formation is the low- est member of the Waverly group, Mississippian system. It is an argillaceous shale, the lower portion being a dark bluish gray, the upper portion a chocolate or dark red color. The Berea formation above is a bluish-gray, fine-grained sandstone. STRUCTURE OF THE BEDFORD AND BEREA FORMATIONS Dynamic movements of the region have taken place since the laying-down of the Berea sandstone, both formations being uni- formly folded. The general structure is that of a syncline whose axis runs northeast and southwest.- The red Bedford shale comes to the surface on either side of this trough, which averages about two miles in width. A great deal of the sandstone in the syncline itself has been eroded away, exposing the red Bedford shale beneath. The large rock trough contains minor anticlines and synclines, with axes parallel to that of the large syncline. A compressional force from the east and west has folded the axis of the northeast- southwest syncline into a series of anticlines and synclines. At South Amherst, in the region under discussion, the axis of the large syncline is plunging toward the east. LENSES OF BEREA SANDSTONE IN THE HORIZON OF THE BEDFORD SHALE The Bedford forms steep banks where the streams cut against it. As one goes along Beaver Creek, which flows just east of the «The writer wishes to thank Professor G. D. Hubbard, of Oberlin College, for criticism of the manuscript. The work was done in the Department of Geology at Oberlin College. 655 656 WILBUR GREELEY BURROUGHS Berea sandstone quarries at South Amherst, or Chance Creek on the west, he will occasionally find the high banks covered by a mass of Berea sandstone talus. This débris came from a lens of sandstone im situ at the top of the bank, extending for 50 to 100 feet on the horizontal, and flanked on either side by red Bedford shale. In places the sandstone is in thin beds 2 to 3 inches thick, at other places in massive beds 3 to 4 feet thick. The lenses range from to to so feet in total thickness. Their long axes run ina general westward direction. No evidence of slumping is found in connection with the banks at and in the vicinity of these lenses. Neither can they be the bottom of synclinal troughs, for the dips _ of their minor axes are not great enough to bring the sandstone to the top of the bank of Bedford shale. The only answer to the question of their origin is that there were once channels and depres- sions in the Bedford shale which, on being filled with sand, ulti- mately formed (in cross-section) the lenses as they now exist. So far as the writer is aware, nothing has ever been published regarding this unconformity between the Bedford and Berea forma- tions of northern Ohio, save in the Ohio Geological Survey Report, Vol. II, published in 1874. This report mentions lenses in the horizon of the Bedford shale north of Elyria, which is to the east- ward of the region under discussion in this article. On p. 91 we read, referring to the erosion of the Bedford prior to the deposition of the Berea: ‘It is probably due to this fact that the red shale is so frequently found to be wanting in the section.” Mr. H. E. Adams, superintendent of the Ohio quarry at South Amherst, Ohio, states that in the extreme southeast corner of Lorain County, Berea grit occurs in lenses in the horizon of the red Bedford shale exactly in the same manner as at South Amherst. The Bedford-Berea unconformity is not confined to Lorain County, Ohio. Dr. Hubbard is authority for the statements that ‘an unconformity occurs at the same horizon in: northwestern Fairfield County near Lithopolis; and Professor Prosser believes a similar break exists at the same horizon near Cleveland, Ohio, but further work is there necessary.” The sand-filled troughs in the erosion plane of the Bedford formation which are visible along the streams are small and insig- UNCONFORMITY OF BEDFORD AND BEREA FORMATIONS 657 nificant in comparison with the channels and valleys whose exist- ence is made known by the drill of the quarry-men. The deepest of these sand-filled depressions is that in which is located the quarry of the Ohio Stone Company (Fig. 1). This quarry is situated on the outskirts of South Amherst, Lorain BIG Horizontal and vertical scale - - - - line H-S=4o0 feet. ----N=North. Line A-D=elevation of 600 feet above sea-level. B=Bedford shale. Bs=Berea sand- stone. G=glacial drift. O=Ohio quarry. M=Malone quarry. C=No. 6 quarry, Cleveland Stone Co. County, Ohio. The pit has been sunk along the axis of an anti- clinal fold which runs in a southwesterly direction. The anticline plunges eastward with a dip of 3°. The south flank of the fold in the quarry has a dip of 6° to the southeast; the north side dips 7° northwest. The great thickness of the sandstone, 217 feet, is due to the sand-filled channel of the eroded Bedford horizon. That this is true is shown by drillings and the structure of the strata in the vicinity. One hundred feet southwest of the edge of the quarry on the same level as the top of the quarry pit, the drill went 60 feet thorugh glacial drift and came upon Bedford shale without encountering any sandstone, and yet the strata in the pit were dipping in that general direction. In the quarry 217 feet of sand- stone were passed through before striking Bedford shale. Four hundred feet on the horizontal from the north side of the quarry the strata dip toward the southeast. One thousand feet on the horizontal from this north side of the Ohio quarry, and on the same level as the top of the quarry, another quarry, the Malone, has gone down roo feet through massive sandstone to the Bedford shale. Here the strata still dip to the southeast; the dip is 7°. Thus a small syncline lies between these two quarries. The dips of the strata are not great enough to carry the sandstone to the depth reached in the Ohio quarry even though the syncline did not exist. 658 WILBUR GREELEY BURROUGHS Therefore the Ohio quarry is located in a depression of the eroded horizon of the Bedford shale. The Ohio pit is 175 feet wide, yet neither bank of the Bedford channel in the quarry has been reached. By drill and well records, the writer has traced this channel, in which is the Ohio quarry, for a distance of three and one-half miles to the southwestward where it outcrops on the steep valley slopes of a stream known as Chance Creek. Here the lens of sandstone is 50 feet wide and 15 feet thick. On both sides and at the bottom the sandstone lies directly against the red Bedford shale. The decreasing of the channel in depth and width as it went south- westward indicates that the stream flowed from the southwest toward the east. Beaver Creek flows a little less than one-half mile east of the Malone quarry. Here no sandstone is found along the banks, in spite of the fact that the axis of the anticline is plunging in that direction at an angle of 3°. The outcrops of Bedford shale at this place on the creek are 30 to 4o feet lower in elevation than the top of the sandstone at the Malone quarry, where the sandstone is 100 feet thick. Still, if the Malone quarry deposit of Berea grit is not a sandstone-filled depression in the Bedford shale, the sandstone should outcrop at Beaver Creek, which it does not do. This quarry therefore also.is located in a lens of sandstone in the horizon of the Bedford shale. A short distance farther north of the Malone quarry is No. 6 quarry of the Cleveland Stone Company. Structurally this quarry is on the southward-dipping flank of an anticline whose axis runs in a southwesterly direction. The average dip is 8°. The axis itself is folded into a low, small anticline in the west portion of the quarry. Here, as elsewhere in the region, the sudden great thick- ness of sandstone cannot be accounted for save as a sand-filled channel of the eroded horizon of the Bedford shale. Although the long axis of No. 6 quarry does not exactly coincide with the direc- tion of the channel in which it is located, yet, both being in nearly the same westerly direction, the size of the pit gives some idea as to the size of the valley in the Bedford formation. The quarry is 2,632 feet long, has an average width of 460 feet, and a depth of from 100 to 175 feet. UNCONFORMITY OF BEDFORD AND BEREA FORMATIONS 659 BLUE SHALE AT THE UNCONFORMITY In the bottom of all these deep channels in the horizon of the red Bedford shale is a soft, dark-blue shale, three to four feet thick. This blue shale is not found beneath the sandstone of the small lenses in the Bedford horizon, nor is it found at any given horizon. The bottoms of the quarries are at different depths with the dip of the strata too slight to bring this blue shale to all the quarry floors. The outcrop of the Ohio quarry channel on Chance Creek had no blue shale beneath the sandstone, the sandstone resting directly upon red Bedford shale. Yet this blue shale is found underlying the sandstone in the Ohio pit. The reason for the location of this blue shale may be the follow- ing: The lower and deeper portions of the valleys of the Bedford streams became drowned. Sediment carried by the rivers into these quiet bodies of water was deposited and eventually formed this blue shale which occurs between the red Bedford shale and the Berea sandstone. Dr. Hubbard and the writer made a careful search for fossils in this blue shale, but none were found. CONCLUSION Starting a few miles east of Sandusky, Ohio, and extending eastward to Cleveland, Ohio, there is a well-defined unconformity between the Bedford and Berea formations. The unconformity, however, extends over a greater area than the region above defined, as it has been noted as far south as Fairfield County, Ohio. During the period that the Bedford horizon was above the level of the sea, its surface was dissected, streams cutting deep channels and wide valleys. The lower portions of these valleys became drowned. In the quiet water thus formed, the rivers deposited sediment which later became a blue shale, logically belonging to the Berea formation. The entire Bedford land area gradually was submerged, and the Berea sandstone formation was laid down. TDI LOR AY In recent years there has been a notable increase in the desire for the use of photographic illustrations on the part of authors of articles submitted to the Journal. To an increasing extent such use is coming to be more than a merely helpful or ornamental acces- sory; it is often an essential means to an adequate presentation of results. In a like manner there has been a marked growth in the use of maps, sections, diagrams, and other graphic matter, as also of analyses, computations, statistics, and similar matter assembled in tabular and diagrammatic forms. It seems inevitable that the proportion of these classes of relatively expensive matter will con- tinue to become greater. To this imperative increase in the expense of properly illustrating the matter of the Journal, there is added the greater cost of printing and publishing resulting from the general advance in prices. To meet the demands of these changed condi- tions, it has been decided to raise the price of the Journal to sub- scribers, beginning with the twentieth volume, except that current subscribers may renew their subscriptions for one year at the present rate if they do so previous to July 1, 1912. Details may be found in the Publishers’ notice in the advertising section of this issue. LAC ye: 660 REVIEWS Ueber Erythrosuchus, Vertreter der neuen Reptilordnung Pelyco- simia. By F. vON HUENE. Geologische und paleontologische Abhandlungen, X (1911). Pp. 58; plates rr. The genus Erythrosuchus was described five years ago by Dr. R. Broom, from the Triassic of South Africa; it was referred by him to the Phytosauria, from which it differs especially in having terminal nares and short premaxillae. Dr. Heune, after a careful study of the known remains of the genus, reaches, in the above-cited paper, the startling conclusion that the genus represents a new order of reptiles allied to the Pelycosauria; that is, that it is a branch from the root- stem of that group (“Zweig von der Wurzel der Pelycosaurien”’). Aside from the differential characters already mentioned, Erythrosuchus differs from the phytosaurs chiefly in the structure of the limbs, which seem to resemble more those of the pelycosaurs and other primitive reptiles. The skull, as Huene admits, has “viele und auffallende Ubereinstimmungen mit den Phytosaurien,’’ in its two temporal vacuities, the absence of additional temporal bones, antorbital vacui- ties, etc. The vertebrae also, are of the archosaurian type, differing especially from those of the Pelycosauria in the shallow concavities of their centra, the absence of intercentra, and especially in the articula- tion of the dorsal ribs. It is an important fact, which the author does not seem to appreciate, that the mode of rib articulation is highly characteristic of the reptilian orders. It may be set down as a funda- mental taxonomic principle that no related groups of reptiles, or other vertebrates differ materially in the way in which the dorsal ribs articu- late with their vertebrae. All the archosaurian reptiles are alike in this respect—double-headed ribs articulating with the diapophyses of the arches exclusively, at least posteriorly—a character found in no other vertebrates. And this is the condition in Erythrosuchus, a char- acter in itself sufficient to fix its position among the Archosauria, and by Archosauria I mean the Crocodilia, Dinosauria, Pterosauria, and Parasuchia. The Sauropterygia, it is true, also have the dorsal ribs attached exclusively to the diapophyses, but the ribs show no division into capitulum and tuberculum, differentiating the order sharply. Under the Sauropterygia I include only the Nothodontia and Plesio- sauria—the Mesosauria, which are sometimes included in the order, 661 662 REVIEWS belong, I am satisfied, with the Theromorpha. The Pelycosauria, like other primitive reptiles, have the ribs attached invariably to the inter- central space and the diapophysis; that is, they are double-headed throughout, while the Cotylosauria, with like attachments, may have the articulation continuous from head to tubercle. In the pectoral girdle about all the difference that Erythrosuchus presents from the phytosaurs is a distinct supracoracoid foramen— precisely the character that would be expected in the more primitive form; and the pelvis, while agreeing in the main with the phytosaurs, differs very materially from that of the pelycosaurs. The chief differ- ences that the author finds allying the genus to the pelycosaurs, are, as stated, found in the limbs: “ Erythrosuchus kann, trotz der vielen Ahnlichkeit iiberhaupt, kein Parasuchien sein, da das Femur besonders in der Bildung des Proximalendes mit den primitiven und 4lteren Pelycosaurien und Cotylosaurien . .. . vollig iibereinstimmt.” Ad- mitting this ‘‘complete agreement” of the proximal end of the femur between Erythrosuchus and the Pelycosauria and Cotylosauria, can one not conceive that the resemblances have been brought about by adaptation to like conditions, that the characters are adaptive and not genetic here, as so often elsewhere? But I do not admit this complete agreement. There is much variation in the femora of the cotylosaurs and pelycosaurs, as witness those of Dimetrodon, Araeoscelis, Diadectes, Seymouria, and Labidosaurus. The humerus of Erythrosuchus, although it has a large lateral process and greatly expanded ends, differs materi- ally from that of the pelycosaurs and cotylosaurs in the absence of the entocondylar foramen. One does not refer the moles to a distinct order of mammals because of the differences in the humeri from other rodents. The skull structure of Erythrosuchus, with its upper temporal and antorbital vacuities, is so much at variance with the theromorph rep- tiles, that I can see no possible evidence of genetic relationships between them. Unless Huene would make the Archosauria a part of the same branch, from the root of the Pelycosauria, he attempts to prove too much, for he would make the Pelycosimia a distinct branch or phylum of the reptilia and entitled to more than ordinal distinction. He classes the Pelycosauria with the single-arched reptiles and is correct in so doing, but I confess I am not quite clear as to the real distinctions between upper and lower temporal vacuities in such reptiles. Nor does Huene seem to be either, as witness the following quotations: Op. cit. page 41, second paragraph: ‘Da bei Deuterosaurus das REVIEWS 663 Postorbitale den unteren Rand der einzigen Schlafenéfinung begrenzt, ist sie als die oberen aufzufassen, und sie sind, im Gegensatz zu den ebenfalls monozygocrotaphen Pelycosaurien und Therapsida als ‘hypo- zygocrotaphen’ zu bezeichnen.”’ Same page, fourth paragraph: ‘‘Alle Therapsida (mit warscheinlicher Ausnahme von Cynognathus) besitzen bekanntlich nur eine einzige Schlafenoffnung, die der oberen entspricht (italics mine). Darin und in der Forme des Quadratums stimmen sie alle mit den Deuterosaurien,” etc. Page 43, second paragraph: “Da die untere Schlifenéffnung nicht entwickelt, resp. nach unten nicht geschlossen ist, fehlt den Therapsiden das Quadratojugale,”’ etc. Same page, third paragraph: “Da bei den Therapsiden das Postor- bitale und Postfrontale an der oberen Ecke der Schlifenéfinung liegen, ist letztere als untere Schlafendffnung aufzufassen, die Therapsiden sind also katazygocrotaph.”’ From personal conversation with Dr. Huene I know that the last statement expresses his real views; but nevertheless the flat contra- dictions on these two pages indicate an unsettled opinion. As I have already stated (American Permian Vertebrates, p. 92) Broom has figured Tapinocephalus with the postorbital and squamosal in broad contact, but he nevertheless holds that the vacuity above them is the “lower”’ one. One must therefore wait for further light on the subject before accepting their views. And there is much confusion also about the quadratojugal bone. It is known to occur in only one genus of the Therapsida, Dinocephalus, but both Broom and Huene insist that it is present in the Pelycosauria, and Broom has figured it in Dimetrodon. But, a study of the material in the University of Chicago—material in which this region is preserved most perfectly—enables me to say positively that there is no such suture or foramen in the lower arch as Broom gives. That a very small, vestigial quadratojugal bone may occur at the extreme posterior end of the jugal is possible, but I have never seen any satisfactory evidence of it, and I doubt its presence, as does also Professor Case. In brief my own opinion is that Broom was quite right when he referred Erythrosuchus to the Phytosauria, using the term in a wide sense as a synonym of Parasuchia. In any event Erythrosuchus is an archosaurian reptile with no direct affinities with the Pelycosauria. In expressing these differences of opinion I would in no wise depre- cate the value of Dr. Huene’s paper. It is a useful one and may be perused with profit. 664 REVIEWS In conclusion I wish to protest against the restoration Huene has made of my figure of the pelvis of Eubrachiosaurus Will. (p. 49). The outlines as I gave them are essentially correct, and the bones do not belong on the right side. As to the distinction of the genus from Placerias Lucas, I am, however, not so sure. S. W. WILLISTON The Monroe Formation of Southern Michigan and Adjoining Regions. By A. W. GraBAu AND W. H. SHERZER. [Michigan Geo- logical and Biological Survey. Publication 2. Geological Series 1.] This report describes a series of Paleozoic beds and their faunas which have their greatest development in southeastern Michigan and the adjacent portions of Ontario and Ohio. In the past these strata, which constitute the Monroe formation, have been much misunderstood, and their importance in the Paleozoic section of the region has been greatly underestimated. The maximum thickness of the formation is about 1,200 feet. The Monroe as a whole is divided into two series of dolomitic beds, the Lower and Upper Monroe, separated by the Sylvania sandstones, a bed of exceptionally pure, white, and almost incoherent sand in its more typical development, but merging into arenaceous dolomites in its less typical expression. The maximum thickness of the Sylvania is 300 feet, and the peculiar nature of the formation is explained on the hypothesis that it is an aeolian deposit laid down under essentially desert conditions, the original source of the material being the exposures of the Saint Peter sandstone to the northwest in Wisconsin. The Monroe faunas are described in detail and are illustrated by twenty-five plates; 126 species in all are defined, many of them new forms, and seven new genera are proposed. The faunas of the two divisions of the Monroe are shown to be essentially different, there being almost no species in common. The Lower Monroe faunas are all late Silurian in aspect, being more or less closely related to the Manilus and Rondout formations of eastern New York. In the lower divisions of the Upper Monroe a conspicuous coral element appears which was entirely lacking in the Lower Monroe faunas, and among these corals are many strikingly Devonian forms; among the brachiopods are found both Devonian and Silurian types; the pelecypods are Devonian while the gastropods and cephalopods are essentially Silurian in aspect. REVIEWS 665 Lying above the beds carrying the strikingly Devonian fauna of the Upper Monroe, is the Lucas dolomite, the youngest member of the series, in which the fauna is Silurian in aspect throughout. In their correlation of the Monroe series the authors adopt a new arrangement of the North American Silurian formations, as follows: (1) Lower Silurian or Niagaran, (2) Middle Silurian or Salinan, (3) Upper Silurian of Monroan. The Lower Monroe is said to be unrepresented in either western or eastern New York, but is correlated with the so- called “Salina”? and the lower portion of the Corrigan formation of Maryland. The lower portion of the Upper Monroe is correlated with the Bertie waterlime and Akron dolomite of western New York, and with the Rosendale waterlime and Cobleskill of eastern New York. An equivalent of the Lucas dolomite is wanting in western New York but it is represented by the Rondout and Manlius of eastern New York and by the Corrigan formation of Maryland. In a discussion of the paleogeography of Monroe times it is suggested that the faunas of Silurian aspect in the Lower Monroe and in the Lucas dolomite have had an Atlantic origin, while the faunas with the notable Devonian expression in the Upper Monroe below the Lucas dolomite have come in from the north. Se We The Fossils and Stratigraphy of the Middle Devonic of Wisconsin. By HerpmMan F. CieLanp. [Wisconsin Geological and Natural History Survey, Bulletin No. XXI_] The Devonian faunas occurring in the neighborhood of Milwaukee and Lake Church, Wisconsin, are of especial interest to students of Paleozoic historical geology because of their intermediate geographic position between the much better known Devonian faunas of New York and of Iowa. The present report by Dr. Cleland records a complete census of these faunas with detailed descriptions of the species, accom- panied by fifty-three plates of illustrations. Something over 200 species are recognized. Of the total number of species 81 occur in Devonian faunas east of Wisconsin, mostly in New York, while 48 species occur in the Devonian of Iowa and other localities to the west. This mingling of the eastern and western faunas of late Middle and early Upper Devo- nian time in the Milwaukee region has been pointed out before, but here for the first time do we have a full statement of the evidence. S. W. 666 REVIEWS Yorkshire Type Ammonites. Part III. Edited by S.S. Buckman. The scope of this work has been defined in a notice of the earlier parts. The present instalment includes the original descriptions with additional notes by the editor, and figures of the type specimens, of eight species, bringing the total number of species now defined and illustrated up to thirty. 5. W. Report on Traverse through the Southern Part of the N orthwest Territories from La Seul to Cat Lake in 1902. By ALFRED G. Witson. [Geol. Survey of Canada,.No. 1006.] Pp. 21. The district traversed was wholly an area of Archaean rock (schists and granites). Many of the granites were notable on account of the large amount of microcline contained. Schists were mainly basic, biotite, and amphibole schists. Glacial striae indicated a general glacial move- ment S.W. to W.S.W. H*C.-C, Oil Resources of Illinois with Special Reference to the Area Outside of the Southeastern Fields. By RAYMOND 5. BLATCHLEY. [Bull. Illinois State Geological Survey No. 16, pp. 7-138]; Plates13, Figs; 2. In this report the author presents a general review of the geology of Illinois as applied to the petroleum industry. He tabulates and repre- sents graphically a number of well records which are chosen to furnish a series of sections running in different directions across the central and southern part of the state. The No. 6 coal bed furnishes a key horizon, the underlying formations lying generally parallel with it. In a few of the better-explored areas this horizon is mapped in contour. | eRe Meteor Crater (Formerly Called Coon Mountain or Coon Butte) in Northern Central Arizona. By D. M. BARRINGER. Read before the National Academy of Sciences at Its Autumn Meeting at Princeton University, November 16, 1909. Pp. 24; Plates 18, Maps 3. There seems to be no doubt that the so-called crater is the work of a falling meteorite. The author has made a careful and detailed REVIEWS 667 study of the whole region and finds abundant evidence which renders any other hypothesis untenable. The question as to what has become of the projectile still remains unsettled. There are three possibilities: (1) that it was broken into many small pieces and thrown out of the crater; (2) that it has disappeared within the crater through oxidation or some other cause; (3) that it is still somewhere in some form in the depths of the crater. The author concludes that the last is the true explanation and that the remains of the meteorite may yet be found. EUR, Age and Relations of the Little Falls Dolomite (Calciferous) of the Mohawk Valley. By E. O. Utrich AnD H. P. CusHIne. [N.Y. State Museum Bulletin 140, Sixth Report of the Director 1909, pp. 97-140.] To clear up some uncertainty as to the exact stratigraphic relation- ships of the Little Falls Dolomite, a series of sections in the Mohawk Valley were studied by the authors and described and correlated in detail. The formation was found to be in conformable sequence with the Theresa formation and the Potsdam sandstone below and separated by an uncon- formity from Beekmantown beds above. The paper concludes with a strong argument for the adoption of the proposed Ozarkian system of which the Little Falls Dolomite is a member. Baia. Report of the Vermont State Geologist, r909-1910. By G. H. PER- KINS AND OTHERS. Pp. 361; Plates 71, Figs. 31. The report contains the following papers: “History and Condition of the State Cabinet,” by G. H. Perkins, pp. 1-75; ‘‘The Granites of Vermont,” by T. N. Dale, pp. 77-197; ‘‘The Surficial Geology of the Champlain Basin,” by C. H. Hitchcock, pp. 199-212; “Trilobites of the Chazy of the Champlain Valley,” by P. E. Raymond, pp. 213-28; “Geology of the Burlington Quadrangle,’’ by G. H. Perkins, pp. 249- 56; “Preliminary Report on the Geology of Addison County,” by H. M. Seely, pp. 257-313; ‘‘Asbestos in Vermont,”’ by C. H. Richardson, pp. 215-204 Mineral Resources,” by G. H. Perkins, pp. 331-52- Eight plates illustrate the trilobites of the Chazy and ten the fauna of the Fort Cassin beds (Beekmantown) which are found in Addison County. iis 15 IE 668 REVIEWS Iowa Geological Survey, Vol. XX. Annual Report, 1909, with Accompanying Papers. By SAMUEL CALVIN, State Geologist, and Others. Pp. 542; Plates 42, Maps ro, Figs. 42. The report contains the following papers: “Geology of Butler County,” by Melvin F. Arey, pp. 1-60; ‘Geology of Grundy County,” by Melvin F. Arey, pp. 60-96; “Geology of Hamilton and Wright Counties,” by Thomas H. MacBride, pp. 97-150; “‘Geology of Iowa County,” by S. W. Stookey, pp. 151-98; “Geology of Wayne County,” by Melvin F. Arey, pp. 199-236; “Geology of Poweshiek County,” by S. W. Stookey, pp. 236-70; “Geology of Harrison and Monona Counties,’ by B. Shimek, pp. 271-486; ‘Geology of Davis County,” by Melvin F. Arey, pp. 487-524. Shimek’s report on the geology of Harrison and Monona counties contains a detailed description of the mammalian fauna recently dis- covered in the Aftonian interglacial deposits. These are especially important, since in only one other instance in North America has it been possible to determine definitely the age of a Pleistocene mammalian fauna. Preliminary reports and descriptions of this fauna have been published by Shimek and by Calvin in Sczence and in the Bulletins of the Geological Society of America. E.R. E. Practical Mineralogy Simplified. For Mining Students, Miners, and Prospectors. By JESSE PERRY Rowe. New York: John Wiley & Sons, 1911. Pp. 162. This textbook is arranged to give a few special or characteristic properties or tests for the common minerals, that will enable persons unskilled in chemistry or mineralogy to identify them by simple methods. It is readable and well arranged. It will doubtless serve a useful purpose. \Wi al 18, IEGENG. PUBLICATIONS —ApAMsS, F. D., AnD BARtow, A. E. Geology of the Haliburton and Ban- croft Areas, Province of Ontario. [Canada Department of Mines, Geo- logical Survey Branch. Memoir No. 6. Ottawa, ro1o.] —ApaAms, J. H. The Geology of the Whatatutu Subdivision, Raukumara Division, Poverty Bay. [New Zealand Geological Survey Bulletin No. 9 (New Series). Wellington, roro.] —Acassiz, ALEX. Reports on the Scientific Results of the Expedition to the Tropical Pacific, in Charge of Alexander Agassiz, by the U.S. Fish Commission Steamer ‘“ Albatross,’ from August, 1899, to March, 1900, Commander Jefferson F. Moser, U.S.N., Commanding. XI. Echini. The Genus Colobocentrotus. [Memoirs of the Museum of Comparative Zoélogy at Harvard College. Vol. XXXVI, No. 1. Cambridge, 1908.] —ASHLEY, GEORGE H. Outline Introduction to the Mineral Resources of Tennessee. [Extract (A) from Bulletin No. 2, ‘‘Preliminary Papers on the Mineral Resources of Tennessee,’ Tennessee Geological Survey. Nashville, 1o10.] —Ausserordentliche Sitzung der Gesellschaft fiir Erdkunde zu Berlin zur Begriissung von Commander Robert E. Peary am 7 Mai, toto. [Sonder- abdruck aus der Zeitschrift der Gesellschaft fiir Erdkunde zu Berlin, Jahrgang 1010, No. 5.| —Barrows, H. H. Geography of the Middle Illinois Valley. [Bulletin 15, Illinois Geological Survey. Urbana, roto.] —Bonnet, R., uND STEINMANN, G. Die “Eolithen”’ des Oligozins in Belgien. [Sonderabdruck aus den Sitzungsberichten der Niederrheinischen Gesell- schaft fiir Natur- und Heilkunde zu Bonn, 1909] —Bow es, O. Tables for the Determination of Common Rocks. [Van Nostrand & Co., New York, 1910.| —Canadian Department of Mines, Geological Survey Branch, Report of, for the Calendar Year 1909. [Ottawa, ro1o.] —CHILTON, CHARLES, Epiror. The Subantarctic Islands of New Zealand. Reports on the Geo-Physics, Geology, Zodlogy, and Botany of the Islands Lying to the South of New Zealand, Based Mainly on Observations and Collections Made during an Expedition in the Government Steamer ““Hinemoa’”’ (Captain J. Bollons) in November, 1907. Vols. I and II. [Published by the Philosophical Institute of Canterbury, Wellington, N.Z., 1909. Dulau & Co., Ltd., 37 Soho Square, London, W., England.| —CIRKEL, Frirz. Report on the Iron Ore Deposits along the Ottawa (Que- bec Side) and Gatineau Rivers. [Canada Department of Mines, No. 23. Ottawa, 1909.] 669 670 RECENT PUBLICATIONS —Crarke, F. W. Analyses of Rocks and Minerals from the Laboratory of the United States Geological Survey 1880 to 1908. [U.S. Geological Survey Bulletin 419. Washington, 1910.| —Colorado School of Mines, Quarterly of, Vol. V, No. 4. [Golden, 1910-12.] —Crwper, A. F. Cement and Portland Cement Materials of Mississippi. [Mississippi Geological Survey Bulletin No. 1. Nashville, 1907.] —Cross, WuirmMan. The Natural Classification of Igneous Rocks. [Quar- terly Journal of the Geological Society, Vol. LXVI (1910), pp. 470-506.] —Datt, Wm., H. anp Bartscu, Paut. New Species of Shells Collected by Mr. John Macoun at Barkley Sound, Vancouver Island, British Columbia. [Memoir No. 14-N, Canada Department of Mines, Geo- logical Survey Branch. Ottawa, 1o1o.] —Dowunc, D. B. The Edmonton Coal Field, Alberta. [Memoir No. 8-E, Canada Department of Mines, Geological Survey Branch. Ottawa, 19I0.| —EastmMan, C. R. Devonic Fishes of the New York Formations. [Memoir No. 10, New York State Museum. Albany, 1907.] —ELspEN, J. V. Principles of Chemical Geology. [Whittaker & Co., New York, 1910.] —FARRINGTON, O. C. Meteorite Studies III. [Field Museum of Natural History, Publication 145. Geological Series, Vol. III, No. 8. Chicago, T910.| —FILCHNER, W., NORDENSKJOLD, O., AND PeNcK, A. Plan einer deutschen antarktischen Expedition. [Sonderabdruck aus der Zeitschrift der Gesellschaft fiir Erdkunde zu Berlin, Jahrgang 1010, No. 3.] —Geological Survey of Western Australia, Report of the Annual Progress for the Year 1909. [Perth, 1o910.] —GLENN, L. C. Denudation and Erosion in the Southern Appalachian Region and the Monongahela Basin. [U.S. Geological Survey, Profes- sional Paper 72. Washington, rort.] —GraBau, A. W. Guide to the Geology and Paleontology of the Schoharie Valley in Eastern New York. [Bulletin 92, Paleontology 13, New York State Museum. Albany, 1906.| —GRantT, U. S., anp Hiccins, D. F. Glaciers of Prince William Sound and the Southern Part of the Kenai Peninsula, Alaska. [Bulletin American Geographic Society, Vol. XLII, No. 10. Washington, 1910.] —Hazarp, D. L. Results of observations Made at the Coast and Geodetic Survey Magnetic Observatory at Baldwin, Kansas, 1905 and 1906. [Department of Commerce and Labor, Coast and Geodetic Survey. Washington, roro.] —Iowa Geological Survey, Vol. XX. Annual Report with Accompanying Papers. [Des Moines, ro10.] —Jounson, J. P. The Stone Implements of South Africa. [Longmans, Green & Co., New York, 1907.] RECENT PUBLICATIONS 671 —Maryland Geological Survey, Report of, Vol. VIII. [Baltimore, 1909.] —Mason and Dixon Line Resurvey Commission, Report on the Resurvey of the Maryland-Pennsylvania Boundary Part of the Mason and Dixon Line. [Maryland Geological Survey, Vol. VII, 1908.] —MclInnes, Wm. Report on a Part of the North West Territories Drained by the Winisk and Attawapiskat Rivers. [No. 1008, Canada Department of Mines, Geological Survey Branch. Ottawa, 1910.| —MEEK, S. E., AND HILDERBRAND,S.F. A Synoptic List of the Fishes Known to Occur within Fifty Miles of Chicago. [Field Museum of Natural History, Publication 142. Zodlogical Series, Vol. VII, No. 9. Chicago, r910.] —MitteEr, W. J. Trough Faulting in the Southern Adirondacks. [Science, N.S., Vol. XXXII, No. 811, pp. 95-96. July 15, 1910.] —Oscoop, W. H. Further New Mammals from British East Africa. [Field Museum of Natural History, Publication 143. Zodlogical Series, Vol. X, No. 3. Chicago, rgto.] —Pernck, A. Das Alter des Menschengeschlechtes. [Aus der Zeitschrift fiir Ethnologie. Heft 3. (1908).] Die Weltkarten-Konferenz in London im November 1909. [Son- derabdruck aus der Zeitschrift der Gesellschaft fiir Erdkunde zu Berlin. 1910. | North America and Europe; a Geographical Comparison. [Science, N.S., Vol. XXIX, No. 739, pp. 321-29. February 26, 1909.] —PERKINS, G. H. Report of the State Geologist on the Mineral Industries and Geology of Certain Areas of Vermont, 1909-10. [The P. H. Gobie Press, Bellows Falls, Vt., 1910.] —RUEDEMANN, RUDOLF. Cephalopoda of the Beekmantown and Chazy Formations of the Champlain Basin. [Bulletin 90, Paleontology 14, New York State Museum. Albany, 1906.] —Scottish Geological Survey. The Geology of the Neighborhood of Edin- burgh. (Edinburgh, ro1o0.] —Sewarp, A.C. Fossil Plants, Vol. II. [Cambridge, 1o10.] —SHREVE, F., Curysiter, M. A., BLopceEtt, F. H., anp Bestey, F.W. The Plant Life of Maryland. [Maryland Weather Service Report, Vol. III. Baltimore, 1910.| —STEINMANN, G. Ueber die Stellung und das Alter des Hochstegenkalks. [Mitteilungen der Geologischen Gesellschaft, Wien, III, roz10.] —Utricu, E. O., anp Cusutnc, H. P. Age and Relations of the Little Falls Dolomite (Calciferous) of the Mohawk Valley. [From New York State Museum Bulletin 140, Sixth Report of the Director, 1909. Albany, 1910.| —Western Australia Geological Survey. Paleontological Contributions to the Geology of Western Australia. By Dr. Geo. J. Hinde, F.R.S., E. A. 672 RECENT PUBLICATIONS Newell, Arber, M.A., F.L.S., F.G.S., R. Etheridge, Esq., Ludwig Glauert, F.G.S. [Perth, roro.] —Wuite,I.C. Levels, Coal Analyses. [Bulletin 2, West Virginia Geological Survey. Morgantown, roto.| —Witson, A. W. G. Geology of the Nipigon Basin, Ontario. [Memoir No. 1, Canada Department of Mines, Geological Survey Branch. Ottawa, 19I0.| Report on a Traverse through the Southern Part of the North West Territories from Lac Seul to Cat Lake in 1902. [No. 1006, Canada Department of Mines, Geological Survey Branch. Ottawa, 1g1o.] Wotcort, A. B. Notes on Some Cleridae of Middle and North America with Descriptions of New Species. [Field Museum of Natural History, Publication 144. Zodlogical Series, Vol. VII, No. 10. Chicago, 1910.] Three Distinct Benefits To The Merchant To make his store more attractive to customers; to prevent depreciation of his merchandise, and to increase the efficiency of ig’ employes—are three aims which the successful "Sn TAN an before him. az peat bres Sine AW MG CERN SSS accomplishes all these he cae F a = ae and holds deve dust. With the lessened circulation of dust the air in the store where Standard Floor Dressing is used, is kept pure and sweet; customers and employes alike feel the stimulus of the clear, clean, dustless air, and the merchandise, instead of speedily becoming shop-worn and soiled by dust, retains its fresh, new appearance. ‘The treated floor takes on a cleanly, finished look and its uniform color provides a pleasing background for the display of goods. Illustrated booklet sent free—a booklet on “Dust Danger and How to Avoid It” will be mailed free to you immediately upon receipt of your request. Not intended for household use. Standard Oil Company (Incorporated) Baker's || Fevers prevail in the Fall, due to germs developed and during Summer. To prevent sickness and a protect your family, ' purify’ the waste-pipes, iota sinks, closets, and the cellar with Made by a a: | | Platt's chanical from high grade cocoa beans, sci- entifically process lended, it | ele f th diluted according to a7 RET EPRI on the bottle. A is of the little extra care just now may prevent much sick- finest desliey: full strength and ness and expense. It is an odorless, colorless iquid disi i foul odors and solutel ure and healthful. liquid disinfectant which destroys nea in er Ib., 1/2 Ib. and 1 Ib. disease-breeding matter. Sold by druggists cans, net weight. everywhere. Booklet of Choice Recipes Sent Free tf hh @ @ ri | O ls | | te §S Ss WALTER BAKER & CO. yee i f t t Established 1780 DORCHESTER, : al Py 7 : Simnrectant. are now v using & soap t hat. no: makes you clean, but makes yo: tell your friends about it that t ap have been established over 60 YEARS. By ov aN ‘ ze system of payments ree family in moderate ci EEN) cumstances can own a VOS€ piano. We take ol instruments in exchange and deliver'the new pian in your home free of expense, Write for Catalogue D and explanations, VOLUME XIX NUMBER 8 THE JOURNAL or GEOLOGY A SEMI-QUARTERLY EDITED BY THOMAS C. CHAMBERLIN AND ROLLIN D, SALISBURY With the Active Collaboration of SAMUEL W. WILLISTON ALBERT JOHANNSEN WILLIAM H. EMMONS Vertebrate Paleontology Petrology Economic Geology STUART WELLER WALLACE W, ATWOOD ROLLIN T. CHAMBERLIN Invertebrate Paleontology Physiography Dynamic Geology ASSOCIATE EDITORS 4 SIR ARCHIBALD GEIKIE, Great Britain GROVE K. GILBERT, National Survey, Washington, D.C. HEINRICH ROSENBUSCH, Germany CHARLES D. WALCOTT, Smithsonian Institution _ THEODOR N. TSCHERNYSCHEW, Russia HENRY S. WILLIAMS, Cornell University CHARLES BARROIS, France é JOSEPH P.IDDINGS, Washington, D.C. ALBRECHT PENCK, Germany ; JOHN C, BRANNER, Stanford University HANS REUSCH, Norway RICHARD A. F. PENROSE, Jr., Philadelphia, Pa. GERARD DEGEER, Sweden i WILLIAM B. CLARK, Johns Hopkins University ORVILLE A. DERBY, Brazil : WILLIAM H. HOBBS, University of Michigan T. W. EDGEWORTH DAVID, Australia FRANK D. ADAMS, McGill University BAILEY WILLIS, Argentine Republic : \ CHARLES K, LEITH, University of Wisconsin NOVEMBER-DECEMBER, 191 GONTEN TS THE BEARINGS OF RADIOACTIVITY ON GEOLOGY - - - - - T. C. CHAmBERIIN 673 THE WING-FINGER OF PTERODACTYLS, WITH RESTORATION OF NYCTOSAURUS S. W. WILLISTON 606 THE TERRESTRIAL DEPOSITS OF OWENS VALLEY, CALIFORNIA ARTHUR C, TROWBRIDGE 706 ON CORUNDUM-SYENITE (URALOSE) FROM MONTANA- - - - Austin F. RocEers 748 A DRAWING-BOARD WITH REVOLVING DISK FOR STEREOGRAPHIC PROJEC- TION Sater seyias eth 2 ae ose A ee - Bio. 4s, © > > AUBERT, JOHANNSEN: 752 HEE AY SNS Si i) aaa cE a a iar ole iy AA eM i Coc eee 756 Che Anthbersity of Chicago press ‘CHICAGO, ILLINOIS AGENTS: THE CAMBRIDGE UNIVERSITY PRESS, Lonpon anp EpiInBuRGH WILLIAM WESLEY & SON, Lonpon TH. STAUFFER, LEtIrzic THE MARUZEN-KABUSHIKI-KAISHA, Toxyo, Oona, Kyoto The Fournal of Geology Published on or about the following dates; February 1, March 15, May 1, June 15, August 1, September 15, November 1, December 15. Vol. XIX CONTENTS FOR NOVEMBER-DECEMBER, 1911 No. 8 THE BEARINGS OF RADIOACTIVITY ON GEOLOGY- - - - - = = = T. C. CHAMBERLIN 673 THE WING-FINGER OF PTERODACTYLS, WITH RESTORATION OF NYCTOSAURUS S. W. WILLISTON 6096 THE TERRESTRIAL DEPOSITS OF OWENS VALLEY, CALIFORNIA - - - ArtHur C. TROWBRIDGE, 706 ON CORUNDUM-SYENITE (URALOSE) FROM MONTANA esis Oe NG es AUSTIN F. Rocers 748 A DRAWING-BOARD WITH REVOLVING DISK FOR STEREOGRAPHIC PROJECTION Atsert JOHANNSEN 752 REVIEWS: come Sad Seis isp cho RI eee SiR ase eae eR A nea The Journal of Geology is published semi-quarterly. 4] The subscription price is $3.00 per year; the price of single copies is 50 cents. {| Postage is prepaid by the publishers on all orders from the United States, Mexico, Cuba, Porto Rico, Panama Canal Zone, Republic of Panama, Hawaiian Islands, Philippine Islands, Guam, Tutuila(Samoa), Shanghai. {] Postage is charged extra-as follows 2 eor Canada, 30 cents on annual subscriptions (total $3.30), on single copies, 4 cents (total 54 cents); for all other countries in the Postal Union, 53 cents on annual subscriptions (total $3.53), on single copies, 11 cents (total 61 cents). ]Remit- tances should be made payable to The University of Chicago Press and should be in Chicago or New York exchange, postal or express muney order. If local check is used, 10 cents should be added for collection. The following agents have been appointed and are authorized to quote the prices indicated: For the British Empire: The Cambridge University Press, Fetter Lane, London, E.C., and William Wesiey & Son, 28 Essex Street, Strand, London. Yearly SHES EL USILOHS, including postage, 14s. 6d. each; single copies, including postage, 25. 6d. each. — For the continent of Europe: Th. Stauffer, Universitatsstrasse 26, Leipzig, Germany. Yearly sub- scriptions, including postage, M. 14.80 each; single copies including postage, M. 2.60 each. For Japan and Korea: The Maruzen-Kabushiki-Kaisha, 11 to 16 Nihonbashi Tori Sanchome, Tokyo, Japan. Yearly subscriptions, including postage, Yen 7.00 each; single copies, including postage, Yen 1.25 each. Claims for missing numbers should be made within the mont following the regular month of publication. The publishers expect to supply missing numbers free only when they have been lost in transit. Business correspondence should be addressed to The University of Chicago Press, Chicago, IIl. Communications for the editors should be addressed to them at The ates of reteeees Chicago, Ill. Contributors are requested to write scientific and proper names with particular care and in citations to follow the form shown in the pages of the Journal. Separates, if desired, must be ordered in advance of publication. Separates of leading articles will be printed as ordered, of which 50 (without covers) will be furnished grads, the actual cost of the remainder (and covers, if desired) to be paid for by the author. The table below shows the approximate cost of separates consisting of plain text or text with line engravings. The actual cost may vary from the figures given, and will depend upon the amount of work in re-making the pages into forms, press work, paper, binding, etc. Separates containing half-tones may ‘be expected to cost somewhat more than the rates given, the increase depending upon the number of cuts and the amount of work required upon them. Number of copies 50 I00 150 200 Letter-press, for 4 pagesorless . .. . . $1.40 $1.60 $1.80 $2.00 Letter-press, for 8 pagesorless . . . . . 1.90 2.20 2.50 2.80 leetter-press;tor To;pasestor less .¢ ne 3.20 4.00 4.60 5.30 Single plates (x double=2 single). . . : -90 1.10 1.30 1.50 Covers, with title (paper like Journal cover) ; 1.20 1.69 2.00 2.40 Entered March 20, 1893, at the Post- Office at Chicago, Ill., as second-class matter, under Act of Congress, March 3, 1879. S THE ROURNAL OF GEOLOGY NOVEMBER-DECEMBER, ror THE BEARINGS OF RADIOACTIVITY ON GEOLOGY T. C. CHAMBERLIN University of Chicago To the geologist the center of interest in the phenomena of radioactivity lies in the spontaneous evolution of heat attending atomic disintegration. This interest is the more piquant because the source of the internal heat of the earth is one of the oldest of its problems and the discovery of radioactivity brings into the study an unexpected element. During the last century there was a rather general consensus of opinion that the earth’s internal heat was derived from the condensation of the nebula from which the earth was then commonly supposed to have taken its origin. This nebula was usually regarded either as a gaseous body or as a quasi-gaseous meteoritic swarm, and in either case its condensa- tion was thought to have given rise to intense heat. The primi- tive gaseous or quasi-gaseous earth-mass was held to have passed later into a molten globe, and the subsequent incrusting of this to have entrapped in the interior the heat supply of subsequent ages. This older view was still in general possession of the field when the apparition of radioactivity forced a new line of thought. But there was also an alternative view built on the belief that the earth grew up gradually by the slow accession of discrete orbital matter in distinction from the direct condensation of a gaseous or quasi-gaseous mass. In this view, the internal heat arose mainly from the self-compression of the earth-mass as it grew. Vol. XIX, No. 8 | 673 674 T. C. CHAMBERLIN This view had its origin in the grave cosmogonic difficulties that had been discovered in the gaseous and quasi-gaseous theories of the earth’s origin. Of the two rival views thus already in the field, the one postulated a plethora of heat at the outset and a gradual loss in all later time, the other postulated at the outset a more limited supply of heat which was increased as compression pro- gressed. The adequacy of such compression to give a sufficiency of heat was a subject of debate from the inception of the view.! To the interest that naturally attaches to the discovery of a wholly — unexpected agency, already acute because of the agent’s singular qualities, there was thus added piquancy in view of its inevitable bearings on the thermal problem of the earth’s interior and on the hypotheses of the earth’s origin. An even more fundamental though less imminent interest was awakened by the discovery that some of the atoms of the earth- substance are undergoing spontaneous disintegration and that all atoms may possibly be doing so and that even the permanency of terrestrial substance may be brought into question. However, matters of this ultra-radical nature cannot be discussed with advantage as yet, for little light has been shed on the broad ques- tion whether all terrestrial substance is in process of disintegration, and on the complementary question whether atoms are some- where and somehow undergoing integration. If the general tenor of the studies thus far made is to be trusted, nothing in the field of common experience seriously inhibits the dissolution of the radioactive substances. It does not appear that even the greatest heightening or lowering of temperature or pressure that can be brought to bear either stays or hastens, in any material measure, the progress of atomic disintegration. Nor do any known changes of chemical union or disunion, of concen- tration or diffusion, or of freedom or confinement seem materially to retard or accelerate the spontaneous dissolution. There is probably no warrant for an unqualified affirmation that neither temperature, pressure, concentration, exposure, nor combination t The status of the problem of the earth’s heat as it stood near the opening of the twentieth century is sketched more fully in Year Book No. 2, Carnegie Institution, 1903, 262-65, and in Geology, Chamberlin and Salisbury, I (1904), 533-47. BEARINGS OF RADIOACTIVITY ON GEOLOGY 675 affects the progress of radioactive decomposition, but no specific effects of a critical value have been certainly disclosed by experi- mentation. These conditions that so much qualify most geologic _ processes must apparently be regarded as negligible for the present so far as radioactivity in the earth’s crust is concerned. It is thought by the leaders in radioactive science permissible to treat radioactive substances as undergoing disintegration persistently and uniformly under all known terrestrial conditions. In the thermal problem of the earth radioactive particles may be dealt with tentatively as centers of heat-generation whose efficiency and endurance are conditioned simply by their atomic constitutions and their mass values. In so far as these remarkable deductions from experimentation may be thought to fall short of full warrant, weakness in equal degree must of course be held to enter into the geological inferences based on them; and in view of the radical nature of the conclusions to which they lead, we cannot perhaps too constantly bear in mind that the postulate of immunity to conditions is the main basis of the geologic contributions credited to radioactivity. But the remarkable verifications of skill and accuracy that have followed the multiplication of tests furnish an ample warrant for a serious discussion of present deductions. There is strong presumption that future tests will further sub- stantiate present conclusions so far as their main bearings on imme- diate terrestrial problems are concerned, whatever interrogations one may be disposed to indulge in regarding ulterior problems. The clue to this extraordinary tenacity of radioactive disso- lution in spite of conditions that profoundly influence most ter- restrial processes, probably lies in the fact that the action springs from the internal motions of the atomic constituents and that these are of such intense nature and are actuated by such pro- digious energies that the influences of ordinary chemical and physical conditions are relatively insignificant. At the same time, the radioactive substances show a decided aptitude to enter into chemical combination under common con- ditions. None of the parent radioactive metals is known to occur in the earth in a native state. In the form of compounds they have become widely distributed over the face of the globe in the 676 T. C. CHAMBERLIN course of the surface changes it has undergone. Radioactive sub- stances have freely entered into solution in the natural waters and have thus been carried wherever the hydrosphere reaches, and in turn they have been deposited therefrom. Their singular property of passing spontaneously from certain states into gaseous forms (emanations) and then back into the solid or liquid form, on defi- nite time schedules, has caused them to be given forth freely into the atmosphere, and, drifting in this, to be later precipitated in - the solid or liquid form, and this has naturally been dispersive in an extreme degree. Radioactive matter is therefore found in practically all the rocks of the surface of the earth, in practically all the waters, and in practically all the atmosphere. But this highly diffusive distribution has not been uniform. There have been special tendencies toward concentration running hand in hand with the general tendencies to diffusion, and these concentrative tendencies constitute a critical element in this dis- cussion. So far as the accessible part of the earth is concerned, the igneous rocks may be taken as the original source of the radio- active substances. How the igneous rocks themselves came to have their present content will be considered later. Whence the radioactive substances came still more remotely is problematical. There may be even now accessions of radioactive substances from without the earth for aught that is known, and indeed this is prob- able; but, except in the form of meteorites whose content appears, from the few tests made, to be relatively meager,’ such accessions are not yet demonstrated. The cycle of distribution on the earth’s surface is simple. From the igneous rocks the radioactive substances are dissolved and disseminated through the waters and carried wherever they go; while from both the rocks and the waters the emanations are given forth into the atmosphere. From the air and the waters in turn the radioactive derivatives are reconcentrated into the earth, except as their disintegration becomes complete and they pass permanently, in the form of helium, into the atmosphere or are lost from the atmosphere into the cosmic regions outside. 7 Strutt, Proc. Roy. Soc., LXXVII A, 480. BEARINGS OF RADIOACTIVITY ON GEOLOGY 677 The special distribution of the radioactive substances among the different kinds of igneous rocks is no doubt full of meaning, but as yet the determinations have not been sufficient to justify more than a few broad generalizations, and these must be held subject to revision.* It may be said safely that the igneous rocks carry a higher ratio of radioactive substance than the average sediments. The reason for this is simple. The sediments are derived from the igneous rocks, and in the process of derivation some of the radioactive matter inevitably goes into the waters and into the atmosphere, and this diversion leaves the content in derivative rocks lower than that of the original rocks. If all the radioactive matter that is lost into the waters and the air were gathered into the derivative rocks, their content should equal that of the igneous rocks from which they came, if no account be taken of the loss by dissolution. The earlier determinations of the amounts of radium in the igneous rocks by Strutt seemed to show that the acidic class hold more radioactive matter, on the average, than the basic class, and a portion of the later determinations seem to support this generali- zation, but the determinations of Eve and Joly, which have been important, seem to bring the richness of the basic class into some- what near equality with that of the acidic, and even to make the preponderance of the one class over the other doubtful. The point of special interest here lies in the inference that, if the lique- faction and eruption of the igneous rocks is dependent on the heat derived from radioactivity, the distribution of radioactive sub-. stances in the erupted rocks should be inversely proportional to t The larger number of determinations of radioactivity in rock have been made by Strutt: Proc. Roy. Soc., LXXVI A (1905), 88 and 312; LXXXVII A (1906), 472; LXXVIII (1906-7), 150; LXXX A (1907-8), 572; Eve: Phil. Mag., September, 1906, p. 189; February, 1907, p. 248; August, 1907, p. 231; October, 1908, p. 622; Am. Jour. Sci., XXII, (December, 1906), 477; Bull. Roy. Soc. Con., June, 1907, pp. 3 and g; July, 1907, p. 196; Joly: Nature, January 24, 1907, p. 294; Phil. Mag., March 1908, p. 385; Radioactivity and Geology (1909), general treatment with references; Elster and Geitel: Phys. Zeit., II (1900-1901), 590; III (1901), 76. For the physics of radioactivity see J. J. Thomson: The Conduction of Electricity through Gases; E. Rutherford: Radioactivity; (1904); Radioactive Transformations (1906); F. Soddy: Radioactivity (1904); The Interpretation of Radium (1909); R. J. Strutt: The Becquerel Rays and the Properties of Radium (1904); and the papers of Boltwood, McCoy, and many others. 678 T. C. CHAMBERLIN their temperatures of mutual solution or of fusion. But it must be observed that even if such a casual distribution prevailed in the rock-matter when first it took the liquid form, this distribution might not persist indefinitely, for selective segregation has appar- ently taken place during the later processes. It is quite clear that the radioactivity is concentrated in some constituents rather than others, as for example in zircon, pyromorphite, apatite, and some other minerals, and in pegmatite and some other rocks. The pegmatitic material, in segregating from a granitic magma, seems to have gathered into itself an unusual proportion of the radioactive substance of the parent mass. In the details of final distribution, therefore, the different parts of the segregated rock- material may rationally be expected to differ from one another and from the parent magma in radioactive content. The deter- minations thus far made, though not adequate to demonstrate this, seem to be in consonance with it. Much interest will there- fore gather about the forthcoming determinations as they multiply and contribute their quota of evidence bearing on the radioactive qualities of the various species of igneous rocks. Among the derivative and sedimentary processes it seems clear that there are modes of concentration also which have given to different sediments different contents of radioactive substances. It appears from the determinations already made that the radio- active substances are leached out of the parent igneous rocks faster than the average minerals of those rocks, for weathered igneous rocks are found to carry less radioactive matter than fresh rocks. This is in accord with the aptitude for chemical change already noted; and yet soils which are almost the type of ultra-weathered material still retain notable radioactivity, but a part of this is probably a redeposit from the atmosphere. In general, it appears that the clayey element carries more radioactive material than the quartzose sands or the calcareous derivatives. In the deep-sea deposits radioactive matter is higher than in the deposits of the shallow parts of the ocean. In the red clays and radiolarian oozes of the abysmal depths the content is markedly greater than in the land-girting muds and sands, or the calcareous oozes of mid-depths. This is assigned in part to the removal by BEARINGS OF RADIOACTIVITY ON GEOLOGY 679 solution of the lime from the original matter of the abysmal deposits, leaving them residual concentrates, and in part to the collection in the depths, in relatively high proportions, of phosphate-bearing relics (teeth, bones, etc.) with which radioactive substances are associated. It is a suggestive fact that the phosphatic nodules of the great deeps are highly radioactive compared with ordinary sedimentary material. A part of this is clearly due to the con- centration of the radioactive substances after the phosphates were deposited, for fresh phosphatic material is notably less radioactive than fossilized phosphates." It appears then that the radioactive substances on the surface of the earth are subject to special agencies that lead in part to greater concentration and in part to wider distribution, and that these act co-ordinately with the general dispersing agencies that give radioactivity to the derivative rocks, to the waters, and to the air. If it were permissible to reason from what is known of surface phenomena, particularly from the broad fact that radioactivity increases as we go from air to water, from water to sediment, and from sediment to igneous rock, it might be inferred very plausibly that radioactivity would be found to reach its maximum concen- tration in the heart of the earth, and certainly that the deeper parts would be as rich as the superficial ones. This presumption might very justly be felt to be strengthened by the fact that the atoms of uranium, radium, and thorium are among the heaviest known and that if the earth were ever gaseous or liquid, these — heavy atoms might naturally be expected to be concentrated toward its center unless the viscosity of the fluid mass were too great to permit this, in which case the distribution should be either equable or indifferent to depth. But Strutt? early called attention to the fact that if such an increasing abundance exists toward the center of the earth, or if there were an equable distribution in depth, the heat gradient as the earth is penetrated would be higher than observation shows it to be. By computations on the data then available he «Strutt, Proc. Roy. Soc., LX XX A, 582. 2 Proc. Roy. Soc., LXXVII A (1906), 472; LX XVIII A, 150. 680 T. C. CHAMBERLIN concluded that a distribution of radioactive substance equal to that of the surface rocks for a depth of only 45 miles would give the rise of heat actually observed in wells, mines and other deep excavations. Later data and closer scrutiny seem to confirm the general soundness of Strutt’s inference, and to make the limita- tions even more narrow. Joly, approaching the problem from the geological as well as the physical point of view, and with the advantage of later data, reached the conclusion that radioactivity of the amount observed at the surface, if continued to a depth ranging from 27 to 37 kilometers (17.2 to 23.5 miles), would give rise to heat equal to that implied by the loss at the surface.t_ Accord- ing to Joly, however, a complete concentration of radioactivity in a shell of this depth does not meet the apparent requirements of igneous phenomena if this be assigned to radioactivity. A deeper distribution of a part of the radioactive matter and a less concentration in the outer part of the crust is felt by Joly to be required and he was led to this final statement: “If we said that the richer part of the crust must be between 9 and 15 kilometers deep, we cannot be far from the truth. This appears to be the best we can do on our present knowledge.’ It is to be noted that these deductions are reached on the supposition that all the internal heat given out arises from radioactivity; no margin is left for any original heat or for secular heat from any other source. On the other hand, the computations seem to take no account of loss of heat by means of igneous extrusions. These remarkable deductions raise two questions of radical import: (1) If supplies of heat are generated currently by radioactivity in such abundance that it is necessary to put these severe limits on the distribution of radioactive substances, must we abandon entirely all further consideration of supposed supplies handed down from a white-hot earth or from any other form of the primitive earth ? (2) Is there among the internal processes previously postu- lated any that provides a way in which such a concentration at t Radioactivity and Geology (1909), 175. 2 [bid., 183. BEARINGS OF RADIOACTIVITY ON GEOLOGY 681 the surface might naturally have taken place, or must we find a new geological process to fit the new thermal difficulty ? The rigor of the dilemma is softened somewhat by noting that the deductions of Strutt, Joly, and their colleagues are based simply on comparisons between the heat-generating power of radioactive substances in the crust and the conductive power of the crust. The functions of igneous extrusion as a mode of trans- fer of internal heat do not seem to be taken into account. This is not unnatural since the heat carried out by extrusive matter and by waters heated by igneous intrusions has not usually been regarded as an important factor in reducing the high temperature inherited by the earth under the older view. But the movement of igneous matter and of waters and gases heated by it has been made to play an essential part in the working concepts that have been based on the planetesimal hypothesis. There will be occasion to return to this critical difference of view. When the apparent excess of thermal riches arising from the new source was first realized an escape from the dilemma raised by it was sought in the natural supposition that the disintegration of uranium and thorium was restrained by pressure in the depths of the earth, and that, though present there, their activity was greatly subdued or possibly inhibited altogether. This plausible explanation was diligently tested; but the general tenor of experi- ments on the effects of pressure, notably those of Eve and Adams' in which the pressures were carried to intensities sufficient to cover earth-pressures to the depths supposed to limit radioactivity and beyond, showed no appreciable restraint on the disintegrating process. It seems necessary, therefore, in the present state of evidence, to accept the inference that the radioactive substances are really concentrated toward the surface, and that the radio- active content in the depths of the earth is of a much lower order. It does not fall to me to adjust the new requirements to the older view of the earth’s internal temperatures based on a molten earth, for other considerations led me to the abandonment of this view before the advent of the new issue. I must leave it to those who hold to the molten hypothesis to battle with its new perils. tNature, July, 1907, p. 260. 682 EC: CHAMBERELN, With such a plethora of heat at the start as a molten earth implies and with a new agency whose current production of heat would seem to be excessively great if its prevalence were not construc- tively minimized, it is not with regret that I feel absolved from the task of finding a reconciliation between this venerable view and the requirements of juvenile discoveries. The discussion of Professor Joly,’ though not explicitly based on the theory of a molten earth, is sympathetic with the general tenets associated with such an earth, and his treatment may be taken as offering the best approach to a reconciliation that seems now possible. It is interesting to note, however, that when Professor Joly reached the critical question of a possible mode by which the surface concentration of radioactivity could have come about (Radioactivity and Geology, 184) he turned to the accretion or planetesimal hypothesis. While he indicated the central line of action on which the concentration might have been accomplished he left without elucidation the line of reconciliation between the heat gradient postulated by the planetesimal view and the gradient he deduces from radioactivity. It is the chief purpose of this paper to set forth what seems to me to be the true harmony between the new light shed by radio- activity and the tenets of the planetesimal view as shaped by me before the discovery of radioactivity and to show the co-ordination of the planetesimal and radioactive agencies in jointly leading to the results observed. To this end it is necessary to sketch with some care the thermal features of the planetesimal view in the form to which preference was given from the start so that it may be clear just what part radioactivity plays in the assigned co-operation. On the assumption that the earth grew up by the accession of planetesimals, whatsoever heat arose from the condensation of the nucleus about which the growth took place centered in the inner- most parts and can affect present surface phenomena only by transfer. The infalling matter that is supposed to have built up the earth to its mature size must have generated much heat by ™ Radioactivity and Geology, 154-82. BEARINGS OF RADIOACTIVITY ON GEOLOGY 683 its impacts, but as the infall is held to have been slow and as this heat was superficial, it may be assumed that it was largely radiated away before it became so deeply buried as to be permanently retained, and so the most of the heat of impact may be regarded as negligible.t In the original shaping of the planetesimal hypothe- sis (before the discovery of radioactivity) the main source of inter- nal heat was made to spring from the compression which the deeper parts of the earth underwent by the increase of its mass as the planet grew to maturity. This chief source was supposed to be abetted by heat springing from the rearrangment and recom- bination of molecules within the mass as time went on. Changes in the distribution of the heat after it was developed were supposed to follow by means of conduction and especially by the transfer of hot fluid matter carrying latent heat. It is important to the present discussion to note that the heat generated by pressure did not affect the outer part and that it -began to be sensible only when those depths were reached at which the rocks suffered appreciable compression from the weight of the rock-mass above them. Thus the heat gradient so generated would rise only slowly in the outer part of the earth and faster in a systematic way toward the center for a considerable depth, if the compressibility of the rocks remained uniform to indefinite depths. If the compressibility fell off as compactness increased the rate of thermal rise toward the center would have been slower. Compressibility at the surface seems to be nearly proportional to pressure, but the compressibility of rocks after they have been compacted by such pressures as are attained at considerable depths is unknown, and it is necessary to proceed here by alternative hypotheses. The extrapolation of the curve found under experi- mental pressures is of course entitled to precedence and this alter- native was used as the basis of the first approximation to the heat curve of the earth’s interior. For the other factors, such as specific heat, necessarily taken into account in the computation, assump- tions as near to known facts as possible were made. On these assumptions it was found that the heat generated between the surface and the center of the earth may be represented by a curve t Chamberlin and Salisbury, Geology, I, 533. 684 T. C. CHAMBERLIN which rises at a very low rate near the surface and is followed by a slowly increasing rate for about one-third the distance to the center, beyond which it rises at a decreasing rate to the center; or, if traced from the center outward, this computed curve of temperature declines faster and faster at every step for about two- thirds of the distance and then declines less and less rapidly to a vanishing-point near the surface. Hence if conductivity be assumed to be the same at all depths, the outward flow of heat on such a gradient would increase in rate from the center to the two- thirds point and then grow slower toward the surface, from which it follows that, on these assumptions of uniform compressibility and uniform conductivity taken by themselves, the internal heat should have been progressively lowered in the deep interior and raised in the more superficial parts. The conductivity of rocks is so very slow, however, that its effects at the surface under the con- ditions named cannot have been large up to the present unless the earth is much older than even radioactivity seems to imply. This first approximation to a theoretical curve of heat, even when modified by conduction, has not been supposed to represent the actual distribution of heat at the present time, for reasons that follow. There is ground to think that compressibility falls off as increased degrees of compactness are attained. In working out the curve which was published in Geology, I, 566 (Chamberlin and Salisbury), Dr. Lunn used as a guide the Laplacian law of density which postulates that density varies as the square root of the pressure. This distribution of density harmonizes fairly well with such astronomical tests as are available and gives a mean density for the earth which is near that required by the earth’s total weight. The assumption that the increased density of the interior is all due to compression, however, makes no allowance for the probable transfer of lighter matter to or toward the sur- face by extrusive action which would tend to increase the mean specific gravity of the residue. The curve of Dr. Lunn may be regarded as a second approximation. But this, as noted, does « Year Book No. 3, Carnegie Institution of Washington, 1904, p. 156; also ‘‘Geo- physical Theory under the Planetesimal Hypothesis,” Section II of “Tidal and Other BEARINGS OF RADIOACTIVITY ON GEOLOGY 685 not take into consideratiion the effects of liquefaction and extru- sion and these in the planetesimal view are of the first order of importance. The theoretical curve mathematically deduced by Dr. Lunn is, however, an indispensable basis for a third approxi- mation in which the effects of liquefaction and extrusion are taken into account. Before passing on to consider liquefaction and extrusion, it is well to note that the Lunn curve based on the Laplacian law of density also is low near the surface and that its rate of rise is much below that of the temperature gradient observed in wells and mines. Dr. Lunn, on assumptions carefully specified in his dis- cussion in the paper cited, found the rise in the first 200 miles only; 3307.C. This low development of heat in the outer part of the earth seemed at first thought to present a difficulty of a rather serious nature, but it was believed to be met by the effects of liquefaction and extrusion, and these were made the chief basis of an additional approximation to the actual temperature curve (Chamberlin and Salisbury, Geology, I, 265-67). It was held that the rising heat of the interior would reach the temperatures of fusion or of mutual solution of some ingredients in the mixed material much earlier than that of other ingredients, and that the ascent of the portion that became molten carrying its latent as well as sensible heat into the cooler outer zone would necessarily raise the temperature of that zone. It was held that the continuation of this process served as a constant influence tending to retard the rise of tem- perature in the deeper zone where the partial liquefaction was in progress while it progressively raised that of the outer zone into which the liquid rock was intruded, whether it lodged in the crust or passed through it to the surface. This extrusive process was supposed to have continued to the present day and to have resulted in a permanent adjustable working curve of accommodation between thermal, fluidal, and mechanical conditions. ‘This curve, except in the cool crust, was essentially identical with the fusion- Problems,” Publication No. 107, Carnegie Institution of Washington, 1909, pp. 169- 231; for a summary and figure of curve see also Chamberlin and Salisbury, Geology (1904), I, 566. 686 T. C. CHAMBERLIN solution curve, whatever that might happen to have been for the time being under the local conditions of pressure, state of strain, nature of material, means of escape, and other properties that affected liquefaction and extrusion. It was regarded as essen- tially a curve of equilibrium between solidity and liquefaction accom- modated to the conditions present at each depth and at each stage _ and was maintained automatically. The actual curve as thus assigned continued always to be essentially the liquefaction curve after that was once attained. The view excludes automatically all internal temperatures higher than the local liquefaction tem- peratures and of course excludes all pervasive gaseous conditions except that of the interspersed and occluded gases of the mixed mass. ‘These interspersed gases assisted extrusion and hence were among the parts most freely extruded. All theoretical inferences based on temperatures higher than the temperatures of liquefaction are excluded from consideration under this view by its very terms. Certain structural conditions postulated by the planetesimal hypothesis greatly favored this automatic action. The infalling matter was assumed to have built itself up in a very heterogeneous manner with the result that the mass of the earth was an intimate mixture of all the kinds of material that made up the spiral nebula from which it was supposed to have been gathered. As this mixed matter was heated by compression, some parts of it must certainly have reached temperatures at which they could go into mutual solution or into fusion while as yet other closely associated parts had not reached temperatures that permitted such action, and as the rise of temperature was very slow by the terms of the hypothe- sis the passage of successive parts into liquefaction was widely separated in time. Fluid parts thus came temporarily to be inti- mately mixed with solid parts. These fluid parts, in the act of passing into solution or fusion, absorbed the necessary energy of liquefaction at the expense of the increasing supply. On their ascent into the crust they heated it. If they lodged there and resolidified they gave up their heat of liquefaction. If they reached the surface the residue of heat, both sensible and latent, was lost. By such liquefaction and transfer these portions served OPA Ci nS OMe BEARINGS OF RADIOACTIVITY ON GEOLOGY 687 to protect the residue in the deeper parts from liquefaction for the time being and the continuation of the process extended the protection to such residue as continued to persist. It is not necessary to offer evidence that ascent of liquid rock took place in great quantities in the early geologic ages and has been more or less active in all ages down to the present. One of the extraordinary facts of the Archaean terranes is the extensive lodgment of liquid rock in the crust, and even in later ages batho- litic phenomena have attained surprising magnitudes. The extrusion of molten rock at the surface was a very pronounced phenomenon as late as the Tertiary and is still an active process. As this extrusive action was widely distributed over the surface at various altitudes and at various stages through great lapses of time and yet was never really very massive when measured in terms of earth-volumes at any one time or place, it is of critical value here to note that the view built on the planetesimal hypothe- sis appeals to a special set of conditions of liquefaction and extru- sion which are peculiarly favorable for selective work in small masses and unfavorable for general liquefaction. In this respect the conditions it assigns stand somewhat in contrast with the conditions usually assumed to be the natural inheritances from a general molten condition. The inference that general liquefaction would take place on any general rise of heat is natural enough in a case in which the whole mass has been solidified from a previous molten state, for such a mass might be presumed to return mass- ively into its former state on a reversal of conditions; but the heterogeneous condition of the mixed matter of the interior postu- lated by the planetesimal view is not favorable to a simultaneous fusion of the whole mass or any large continuous part of it unless extrusion be restrained until a high temperature is attained. Such restraint is here held to be dynamically inconsistent with the mechanism and the stress conditions of the earth-body. In addi- tion, therefore, to such a mixed state of material in the interior as _ peculiarly to invite selective liquefaction as the temperature slowly rose, the planetesimal view postulates a set of stress agencies that worked co-operatively to effect extrusion as fast as liquid matter accumulated in workable volume. 688 T. C. CHAMBERLIN In considering stress effects, it is necessary scrupulously to distinguish between hydrostatic stresses which operate equally on all sides of a given unit and so only produce compressive and like effects, and differential stresses which promote movement and change of form. ‘The effect of differential stresses on the solid parts of the earth is primarily to produce strains; the effect on liquid parts is primarily to produce flow and relocalization. And so by reason of this difference of effect, a general differential stress on any large part of the earth is apt to become locally sub- differentiated when solid and liquid parts are intermixed, especially if the liquid and solid states of these parts are partially inter- changeable because their temperatures lie so close to the line of equilibrium between solidity and liquidity. Tensional strains promote liquefaction in bodies constituted as most rocks are; compressive strains resist liquefaction in such bodies. And so general differential strains co-operate with temperature in pro- moting or in restraining the passage of matter from the one state to the other according to the nature of the strain and thus have some influence in directing and facilitating movement as well as in forcing it. Some of the differential stresses in the earth are essentially fixed and constant, such as the direct pressures that arise from the action of gravity. These stresses range from one atmosphere at the surface to about three million atmospheres at the center. Such pressures tend to force lighter bodies toward the surface while heavier bodies seek the center in ways so familiar that we need not dwell on them, nor on the fact that, since molten rock is usually lighter than the same rock in a solid state, this static differential stress of gravity presents a general condition that favors the ascent of liquid rock. So also the incorporation or generation of gases in liquid rocks tends to lessen the specific gravity and increase the mobility and hence the gaseous element adds another general influence that favors ascent. In addition to these very general and persistent stresses, more special differential stresses have arisen at various times from inequalities of accession, from transfers of matter, from loss of heat, and from other varying agencies, and these have been present, BEARINGS OF RADIOACTIVITY ON GEOLOGY 689 in one form or another, at nearly all times in the earth’s history. They have often been cumulative until they reached diastrophic intensity and manifested themselves in impressive deformations. That these have been effective agencies in forcing the movement of liquid parts within the earth in the lines of least resistance and of best accommodation to existent conditions is scarcely debatable. In addition to the simple stresses of gravity and to the dia- strophic stresses, there have been superposed at all times a series of stresses of a rhythmical pulsatory nature acting throughout the body of the earth. The nature and function of these has not been so generally recognized. ‘These stresses are derived from the differential action of the gravity of neighboring bodies, particularly that of the moon and of the sun. ‘Tidal and tidelike stresses and strains have swept through the earth’s body in a constant cycle bringing to bear on each part a perpetual succession of compres- sive and tensional stresses and strains alternating with one another. The effect may be pictured as that of a minute kneading of the earth-body. There is not only a superposition of pulsating strains on the more static strains but a superposition of pulsating strains on pulsating strains. The pulses of the twelve-hour body tides are overrun by tides of longer periods and these are attended by shifts of direction of strain, all of which tend to knead the mixed matter to and fro and promote insinuation of the liquid parts along the lines of escape. Underlying all these rhythmical strains there has been ever present a variation in intensity from center to surface. Sir George Darwin has shown that the tidal stresses generated by the moon at the earth’s center are eight times as great as those at its surface. Each compressive strain squeezes the lower part of each liquid vesicle or thread more than the upper part. The coexistence of these pulsatory and periodic strains with the simple static stresses of gravity and the less constant dias- trophic stresses sufficiently imples their co-operative nature. All these three classes are either differential stresses or have factors or phases that are differential, and so, in specific local appli- cation, they are all transformed into sub-differentiational effects on the liquid and solid parts. 690 T. C. CHAMBERLIN Under the planetesimal view the joint effect of these differ- ential stresses and their resulting strains has been at all times to force toward the surface liquefied rock as fast as it gained work- able volume. Much aid in insinuating itself along liquid lines and in fluxing a more open path until the fracture zone was reached, is assigned to the mixed nature of the material and. to the local strains imposed by the stress agencies. The whole picture centers on the fundamental dynamic proposition that energy in mobile and expansive embodiments seeks the surface, while its fixed embodiments are forced more firmly together toward the center. The extrusion is held to have begun as soon as the susceptible matter took the mobile form. Possible exception is admitted in the case of matter that may have been too dense to be forced to the surface. However, a high density of small masses enmeshed in masses of less density could only contribute to an average effect so long as a high state of viscosity was retained, and a rela- tively high viscosity for the small mobile masses, naturally arose from the close balance between the liquid and solid states. Such a condition seems equally to be implied by the remarkable mixtures of dense and light matter often seen in the igneous rocks.' The matter forced early to the surface is held to have been buried by further accretions to the growing planet, later to have been subject to a second liquefaction and extrusion, a second burial, and so on. Progressive selection and reselection are postu- lated until the growth essentially ceased. Since then a more complete selection and concentration of the eutectic material at the surface has been in progress as far as further generation of internal heat has furnished the actuating agency. Now if this picture in its working details and in its rather sharp antithesis to the older view is clearly in mind, the part which the radioactive substances may be supposed to play in co-operation with this mechanism without changing the general conception is little less than self-evident. The radioactive particles are sources of self-generated heat. Under the planetesimal view the radioactive substances were promiscuously scattered through the mixed mass as it was gathered in heterogeneously from the nebula * Chamberlin and Salisbury, Geology, II, 121-22. BEARINGS OF RADIOACTIVITY ON GEOLOGY 691 by the crossing of the planetesimal orbits. No original segrega- tion of this class of matter more than of any other heavy material is assignable. The relative amount of the radioactive matter, at least of the classes now known to be radioactive, must have been extremely small and its influence on the specific gravity of the matter with which it was mixed must have been negligible. The self-heating effects of these disseminated particles were necessarily expended first upon themselves and next upon adjacent matter, and, other things being equal, this homemade heat should have given these parts precedence in passing into the mobile state. Normally the mixed units that inclosed a radioactive particle should have been as susceptible of partially passing into the liquid state as similar units that were free from radioactive matter. The special source of heat should have turned the balance in favor of the unit immediately surrounding the radioactive particle. Thus the radioactive matter normally became involved in the mobile matter and passed with it to or toward the surface. With every stage in the growth of the earth and with every reburial of the radioactive material a second similar preferential action should have followed. On the essential completion of the growth of the earth a more complete concentration of the self- heating matter should have followed, for additional weighting by accretion had essentially ceased and compression had become essentially static while the self-heating competency of the radio- active matter, though no doubt somewhat reduced by consump- tion, was probably more efficient relatively in the production of heat than it had been during the more active stages of growth. It seems clear, therefore, that at all times after the volcanic process was well under way radioactivity should have been rela- tively most active in the outer part of the earth and should have become especially so in the latest stages of the earth. It is there- fore not too much, perhaps, to claim that a specific basis in favor- able conditions and a definite working mechanism for an effective concentration of self-liquefying matter at the surface was postu- lated in a singularly apt way before radioactivity was discovered, and quite irrespective of the dilemma which its discovery has involved. 692 T. C. CHAMBERLIN Reciprocally radioactivity greatly eases the burden laid on © compression in the outer part of the earth where it is least compe- tent and where resort was had to igneous intrusions from below to give the crust its observed temperatures. With the addition of the new thermal agency the extrusions are presumed to play much the same part as before but more actively, as they must now be supposed to meet the liquefying effects both of compression and of radioactivity. If there was ground before to question the efficiency of compressional heat, aided by such other sources as were formerly assignable, to give rise to the high degree of igneous activity that marked the Archaean ages and to sustain the lesser igneous action of later periods down to the present, this doubt is amply resolved by the combined efficiency of compression and radioactivity. In any case it is certain that a large amount of energy has been brought to the surface and radiated into space. Radioactivity also comes to the aid of other agencies of extru- sion in the peculiar service it renders in opening a path for the outward movement of the liquid matter. In the liquefying pro- cess, as we have seen, the radioactive particles should have been gathered by their self-heating action into the liquid vesicles and have been forced outward with them. The self-heating property thus became an endowment of the liquid and gave to it thermal efficiency in dissolving and fluxing its way. This efficiency was continually renewed by the progressive disintegration of the radio- active atoms. It is not improbable that the liquid threads were thus aided in a very special way in boring upward, for it seems obvious that the part of the liquid which carried most of the self- heating constituent would come to have the highest temperature, the lowest specific gravity, and the largest gaseous factor—for the disintegration produced gas emanation and helium in addition to the gases generated by the heat alone—and hence would take the uppermost position and bring its liquefying influences to bear on the solid matter which lay between it and the surface toward which it was pressed. The very mechanism may thus have kept the most effective part at the point most critical to its ascent. While this outline falls far short of an adequate discussion of the relations of radioactivity to the planetesimal hypothesis, it BEARINGS OF RADIOACTIVITY ON GEOLOGY 693 will perhaps suffice to point out the line of co-operation of the new thermal agency with the new genetic hypothesis. The two seem to co-operate happily. Jointly they seem to furnish a promising basis for a revised thermal geology in harmony with accumulating geologic data in various lines and with the growing evidence of the elastic rigidity of the earth-body as a whole. At least the con- centration of the radioactive substances at the surface seems to be aptly explained, and the mechanism that conserves the solidity of the earth falls into consonance with the new experimental evi- dence of an elastico-rigid body-tide which seems scarcely less than decisive. There is perhaps one further point, among the many remaining, that should be briefly touched here lest there seem to be an out- standing incongruity in the present distribution of vulcanism. If there is a progressive supply of heat in the earth’s crust springing from radioactivity and if it is this that actuates vulcanism, why are not volcanoes more uniformly distributed over the face of the globe? A general sub-uniform distribution is a natural deduction from the postulates. The distribution of pits on the moon, assum- ing that they are volcanic craters, fairly fits the picture that nor- mally arises from the action of such an agency. Especially is this true if vulcanism is effected in so selective and so individual a way as we have indicated. Why has not such a distribution persisted on the earth? It will perhaps be conceded that the prevalence of vulcanism in Archaean times fairly satisfies the terms of the case. But at present volcanoes are rare in the primitive shields that form the nuclei of the continents while volcanoes are concentrated about the borders of the continents and in the deep basins and are par- ticularly abundant where the great segments of the crust join one another. The primitive shields are indeed intimately scarred and shotted with igneous intrusions of the early ages, but they are almost immune now. There seem to be two lines of plausible explanation. These old embossments have suffered denudation from an early date and the matter removed has been carried to the borders of the adjacent basins. According to the hypothesis of concentration at the surface, this lost matter carried a relatively high proportion 604 T. C. CHAMBERLIN of radioactive substance. When this was in the state of a mechani- cal sediment it was chiefly deposited on the borders of the basins; when it was in solution it mixed with the waters of the oceans and was later largely concentrated in the oceanic precipitates. Thus the prolonged process of denudation cut away the radioactively richer part of the shield and added it to the undenuded crust of the continental borders and the oceanic basins, thinning the one and thickening the other in a special radioactive sense. Besides this the lower crust in the denuded area was lifted relatively toward the cold surface, while in the depositional area it was relatively depressed beneath a growing radioactive mantle. The rise of the denuded embossments of the crust was attended by elastic expansion of the whole sector of the earth beneath, since the gravitative pressure was lessened throughout. A lowering of the melting-points indeed attended this and doubtless a change also of the mutual-solution conditions, but this was anticipated by the elastic expansion and its instantaneous cooling effects, a point usually overlooked. In addition to this immediate expansional effect, it is held by some geologists, with whom I am glad to associate myself, that the protruding portions of the continents tend to lateral creep and that this carries with it tensional effects as well as some further elastic expansion. At the same time, the penetration of surface- water is promoted and this aids effectively in carrying off the heat of the outer crust. It may be observed that while meteoric cir- culation penetrates to considerable depths beneath land surfaces there is little reason to think that there is any effective circulation to appreciable depths in the ocean beds. One further agency is believed to co-operate with these at a lower horizon but this can be touched only with reserve as it involves joint studies yet in progress upon which I do not feel at liberty to draw further than may be necessary merely to indicate their bearing on this particular problem.’ In a previous part of « The studies are common to my son, Rollin T. Chamberlin, and myself and in the particular here applicable the junior partner is the leader in pursuance of lines of inquiry growing out of his studies on “‘ The Appalachian Folds of Central Pennsylvania,” Journal of Geology, XVIII, No. 3 (April-May, 1910). BEARINGS OF RADIOACTIVITY ON GEOLOGY 695 this paper the selective influence of strains on fusion and solution was cited. There seems little doubt that a similar influence is exerted by the great zones of strain that are developed in the earth by diastrophic agencies. Among the tentative distributions of these under study, a specific system seems more probable than others and this is of such a nature as to direct fluid matter, par- ticularly any that may arise at considerable depths, toward the lines that are affected by volcanic extrusions. THE WING-FINGER OF PTERODACTYLS, WITH RES- TORATION OF NYCTOSAURUS S. W. WILLISTON The University of Chicago The question whether the wing-finger of pterodactyls is the fourth or the fifth has been disputed for the past eighty years, though for the past forty years authors have been almost unani- mously agreed that it is the fifth. The first writer of credibility who expressed an opinion on the subject was Cuvier, who con- sidered it the fourth. His reasons for so doing, as published in his Ossemens Fossiles, are today, I believe, unanswerable, and to him should be given the credit, and not to H. v. Meyer, for the correct recognition of the finger. I quote his remarks in full: En fin il a ce doigt énormément prolongé en tige gréle, qui caractérise éminement notre animal. Il a quatre articulations sans ongle. Le quatriéme doigt des lézards aurait cing articles et un ongle; mais, dans les crocodiles, il n’a que quatre articles, et il est dépourvu d’ongle comme ici; seulement il n’y éprouve pas ce prolonge- ment extraordinaire. Le crocodile et les lézards ont en outre un cinquiéme doigt qui dans les, lézards a quatre articles, et dans le crocodile est réduit a trois sans ongle. Il parait que dans l’animal fossile il ne reste qu’un vestige de cinquiéme doigt, mais assez obscur et sujet 4 contestation. Le grand doigt est probablement le quatriéme, car c’est aussi le quatriéme qui est le plus long dans les lézards. Les trois autres le précédaient dans l’ordre inverse du nombre de leurs articles. The first author to adopt the other view, that the finger is in reality the fifth, was Goldfuss, who, as Plieninger has shown in his full and reliable review of the subject, thought he saw in the pteroid bone a first finger, accidentally misplaced in his specimen, and in which he thought he recognized an additional phalange even. H. v. Meyer early adopted Goldfuss’ view, as shown in the following quotation: ‘“‘Es zeichnen sich diese Thiere vor allen anderen wirklich dadurch aus, dass der Finger sie zum Fliegen befahigte, 696 THE WING-FINGER OF PTERODACTYLS und zwar nur ein Finger, die Ohr- finger, welche wegen der Kleinheit womit er in der Hand anderen Geschdpfe sich darstellt auch der kleine Finger genannt wird” (Pale- ontographica [1851], 19). But Meyer soon returned to the Cuvierian position, calling the first of the small, clawed fingers the thumb. I can find no independent arguments of Meyer giving the reasons for his views; indeed in various places he is more or less obscure, referring to the ‘Flug- finger’ as the “Ohrfinger,” though there can be no doubt but that as early as 1860 he had, as I think, correctly recognized the digit as the fourth. Owen in his Pale- ontology and Comparative Anatomy of Vertebrates figures four small, clawed fingers in front of the wing- finger, which he calls the fifth. Later he reverted to the Cuvierian view. Goldfuss’s views were fol- lowed by Oscar Fraas and most modern authors, including Marsh, Zittel, Plieninger, and Eaton. In 1904," without at the time having read Cuvier’s remarks on the sub- ject, I published a brief article in the London Geological Magazine giving reasons for the older view, that the finger is in reality the fourth, as based chiefly upon the recognized normal number of phalanges in the hands of reptiles. “The Fingers of Pterodactyls,” The Geological Magazine, 1904, p. 59. 697 Fic. 1.—Restoration of Nyctosaurus gracilis Marsh, by Herrick E. Wilson 698 S. W. WILLISTON Two years later Plieninger’ discussed the subject fully and well, reaching no positive conclusion, though evidently favoring the Goldfuss view that the finger is the fifth. He showed that Goldfuss, and not Fraas, as I had thought, was the first author to suggest the identification of the pteroid with the first finger, and corrected Seeley’s statement that Meyer had so recognized it. We have seen from the quotation that Seeley was really not so far wrong after all, since Meyer did at one time consider the ‘“ Flugfinger” as the “Ohrfinger.” Finally Abel? in a recent paper has restated the problem, adopting the original Cuvierian view. As bearing upon this question we have been fortunate in recent years in determining the intimate structure of the hands and feet of several of the early reptiles, from which I may say with entire assurance that, until the close of Carboniferous times, and prob- ably till the close of Permian times, the phalangeal formula for reptiles was the primitive one of 2, 3, 4, 5, 3 for the front feet; 2, 3, 4, 5, 4 for the hind. Plieninger has raised a question in the cited paper whether the formula 2, 3, 4, 4, 3, aS seen in the crocodiles, was not really the primitive one for the hands instead of 2, 3, 4, 5, 3, as found in the generality of modern lizards and in Sphenodon. In the accompanying figures the front limbs of three of these reptiles, from the so-called Permian of Texas and New Mexico, are shown, made out with certainty in nearly every detail. In Fig. 4 the distal three phalanges of the fourth finger have not yet been positively fixed, but inasmuch as the fourth digit of the hind foot of the skeleton to which the figured hand pertains has definitely five phalanges, there can be no doubt of the number in the same digit of the hand. In Figs. 2 and 4 the bones of the forearm and wrist are shown in a horizontal plane without the foreshortening of the oblique position that they really had in life, and which is shown in Fig. 3. Fig. 2 is that of a cotylosaur, probably belonging in the suborder Pareiasauria, while Figs. 3 and 4, Ophiacodon’ and Vara- 1 “Ueber die Hand der Pterosaurier,”’ Centralbl. fiir Mineralogie, Geol., etc., 1906, p 399; also Paleontographica, LIII (1907), 301. 2“T)ie Vorfahren der Vogel,’ Verhandl. der K.K. zoologisch-bot. Gesellsch., LXI (1911), 163. 3 The full description of this genus will appear shortly in a paper by Dr. Case and the writer. THE WING-FINGER OF PTERODACTYLS 699 nosaurus, are zygocrotaphic reptiles that may be included in the order Theromorpha or Pelycosauria. Fic. 2.—Right front leg of Limnoscelis Williston, a cotylosaur reptile from the Permian of New Mexico. A little less than one-half natural size. 700 S. W. WILLISTON In all these forms it will be observed that the fifth digit is much reduced, more so than in the hind feet of the same animals. The number of phalanges in this finger in each is three and no more; this is positive. Furthermore it will also be observed that the support- ing carpale 5 is reduced or wanting in all; that is, the loss of this . bone, the rule in all later reptiles, had begun even before the close of Carboniferous times." It may therefore be assumed with assurance that the ancestors of the pterosaurs had the phalangeal formula for the hand of 2, 3, 4, 5, 3, with the fifth finger much reduced in size and its supporting carpale 5 greatly reduced or entirely lost. In adaptation to aerial flight the pectoral girdle? and front limbs in the pterodactyls have ~ been greatly modified throughout. In Pleranodon and Nyctosaurus, the most highly specialized, but three carpal bones remain, a proximal one, doubtless the fused radiale, intermedium, and ulnare; a lateral carpal for the support of the pteroid, which may be either the centrale or the first carpale; and a distal one, which in my opinion represents the fourth carpale alone; which, it will be seen, is the largest in reptiles. The carpale bearing the “ Flugfinger”’ is always the larger; in Pterodactylus there is another, smaller one in front bearing the anterior metacarpals. I cannot believe that this carpale is the reduced or lost fifth carpale of the ancestral pterosaur carpus, nor that the wing-finger has migrated from its own vestigial carpale to the enlarged fourth while the fourth has migrated to a more anterior carpale. From the carpus then of pterodactyls it would seem highly probable that the carpale is the fourth and that it supports its proper finger the fourth, and not the fifth. As has been known since the time of Cuvier, the phalangeal formula in pterodactyls, beginning with the first clawed finger, is tT may mention here that evidence is accumulating to prove that the so-called Permian of Texas, or at least its lower part, and of New Mexico, as well as of Illinois, really pertains to the upper part of the Pennsylvanian. 2In my recent work on American Permian Vertebrates, p. 58, fourth line from bottom, occurs an unfortunate error, due to the omission of a qualifying phrase, “absent ‘among nonamphibious reptiles,’’? whereby I say that the supracoracoid foramen is wanting only among Pterosauria, when its absence in the Plesiosauria, most Ichthyosauria, Phytosauria, Chelonia is known to all. THE WING-FINGER OF PTERODACTYLS 701 2, 3, 4, 4, to which there are probably few or no exceptions. The first three of these agree absolutely with the normal and primitive formula of the first three digits. The fourth pterodactyl finger has Sar, SSS SN Fic. 3.—Right front leg of Opkiacodon Marsh, a theromorph reptile from the Permian of New Mexico. Two-thirds natural size. 702 S. W. WILLISTON but four phalanges, one less than the normal number, and quite that of the crocodiles; that is, as I have previously urged, it lacks the claw. In the acquirement of a membrane-bearing function this is precisely what would be expected in any finger, and is what occurs in the bats, as Abel has said. That the claw gradually elongated, changing its function from prehension to supporting, seems highly improbable. This finger then answers all the require- ments for the fourth. If, on the other hand,.in consonance with the Goldfuss theory, it is the fifth digit which acquired the membrane- supporting function, not only must the claw have changed its function and become elongated but a new phalange must have been added to the finger. Although among aquatic reptiles hyper- phalangy is a common characteristic, we know of no instance among terrestrial vertebrates that I can recall where an additional phalange has been acquired, in either the front or the hind feet. And, if the Goldfuss theory be true, not only must there have been hyper- phalangy in the fifth digit, but hypophalangy in the four preceding digits; that is, in the acquirement of a wing function, an increase and loss of phalanges must have occurred concurrently in the hand. I cannot believe that this was the case. Had we not to deal with the peculiar bone called the pteroid, articulating with the carpus and turned backward toward the elbow, the question of the homol- ogy of the wing-finger would doubtless never have been raised. It is the pteroid, then, which has caused all the dispute, from the necessity of accounting for the bone, which, other than a misplaced first metacarpal, seems inexplicable. ‘Two derivations have been imputed to it, as a sinew bone, and as a sesamoid bone. In favor of its being merely an ossified sinew is the fact that, in the remark- able specimen I have described of Nyctosaurus, seven well-ossified tendon bones are seen lying by the side of the forearm and hand, elongated bones with one end flattened and the other attenuated. In favor of the latter view that it is merely a sesamoid bone developed in the tendon of some carpal muscle originally is the fact that sesamoid bones do occur elsewhere in the pterodactyls. In the above-mentioned specimen of NVyctosaurus I found one lying over the end of the radius and another over the outer end of the coracoid; and I have seen them often in Pleranodon. Sesamoid bones have THE WING-FINGER OF PTERODACTYLS’ bursal sacks and synovial joints. as a tendon or sesamoid bone is quite possible and even probable, but that the bone finally acquired another function, at least in the most highly developed forms, would seem to be very prob- able. The function that has generally been ascribed to it is that of a “Spannknochen”’ or tensor of the patagial membrane in front of the elbow. Under the assumed relations of the membrane to the front of the arm I have protested against this theory, since the fact is that there could have been little or no membrane in this region to be rendered tense, provided the membrane terminated, as is usually assumed, at the shoulder. Under the assumption that it really served as a “Spann- knochen”’ I have suggested in an earlier paper that the mem- brane continued beyond the shoulder along the side of the neck to the skull. In the accompanying restora- tion, Mr. Herrick E. Wilson, of the University of Chicago, after careful study, has embodied these views, based upon my Skeletal restoration of Nvycto- sturus. I believe that this restoration comes nearer to the real appearance of a pterodactyl 793 That the pteroid bone originated 7 \\ eet yi i Wt \ a) WE 5 / oN i y Fic. 4.—Right front leg of Varano- saurus Broili, a theromorph reptile from the Permian of Texas. Seven-tenths natural size. 704 S. W. WILLISTON in life than any that has hitherto been published. That the membrane extended on the neck is of course yet a hypothesis based upon the mode of development of the parachute in flying animals of today, and especially upon the structure of the pteroid bone and its relations to the forearm and shoulder. It is a fact that this bone seems to be better developed in Nyctosaurus than in other known pterodactyls, reaching by its pointed extremity pretty well toward the shoulder. If it was divaricated from the arm, as its perfect ball-and-socket mode of articulation with the carpus would indicate, and not inclosed in a muscle at its pointed extremity, its function as a supporter of a membrane in front of the elbow can scarcely be taken into consideration. With the mem- brane extending past the shoulder to the neck it would have had a distinct function as a ““Spannknochen”’ and not otherwise. Objection may be raised against the wide expanse of membrane between the legs. That the membrane extended to the tarsus on the peroneal side of the legs I think now hardly admits of doubt; the animals would hardly have been “‘flugfahig’”’ were the legs wholly free, since the wing membrane would have been too narrow to serve as a parachute, and since the legs with their attached membrane must have functioned much like the tail feathers of modern birds in the control of flight. Rhamphorhynchus gemmingi has been restored by Zittel without membrane between the legs, but such a condition must seem impossible for such a flying creature. With the wings extended and the membrane connected with the ankles, there must have been a constant and considerable abducting strain on the legs, which must have required a constant muscular tension to withstand; and the legs, in the later pterodactyls at least, seem too frail for such tension. The head of mammals in the horizontal position is kept in place, not by muscular action, which would be unbearable, but by the elastic ligament of the neck. Something like this must have been necessary to withstand the constant abducting tension of the legs of pterodactyls in flight, and I assume that this was the function of a tense membrane between the legs, as well as that of directing flight. It has been suggested that the border of this membrane connected with the end of the vestigial tail; possibly that was the case, but, in Nyctosaurus at least, such THE WING-FINGER OF PTERODACTYLS : 705 an excised membrane would have been little better than none at all. That the ribs of the abdominal region extended out into the patagial membrane on the sides I have given reasons for elsewhere; I can see no other explanation for their position and lack of curva- ture in the specimen of Nyctosaurus to which I have referred. THE TERRESTRIAL DEPOSITS OF OWENS VALLEY, CALIFORNIA ARTHUR C. TROWBRIDGE State University of Iowa CONTENTS INTRODUCTION DEPOSITS AT THE BASE OF THE SIERRA MOUNTAINS Location and Extent Topography Fans and Inter-Fan Areas Channels and Ridges Stream Canyons Bowlder Belts Slope of the Bajada Materials Texture Structure TERRESTRIAL DEPOSITS OF THE INYO MOUNTAINS Pliocene Lacustrine Deposits Lake Beds in Waucobi Canyon Lake Beds near Haiwee Older Deposits at the Foot of the Inyo Mountains Distribution as Indication of Age Constitution Origin The Present Fans Distribution Shape and Topography Materials Summary SOME PROBLEMS OF THE TERRESTRIAL DEPOSITS Manner of Formation of the Fans and Bajada Causes of Deposition Forms Taken by the Deposits The Transportation of Large Bowlders Lens and Pocket Stratification The Dissection of the Sierra Bajada Deposits of Two Ages at the Foot of the Inyo Mountains CRITERIA FOR DISTINGUISHING ALLUVIAL FAN MATERIALS 706 TERRESTRIAL DEPOSITS OF OWENS VALLEY 707 INTRODUCTION. From September 9 to December 1, 1909, the writer in company with James A. Lane was in the southern half of Owens Valley, California, studying and mapping the general geology in a semi- detailed manner, and gathering data on the terrestrial deposits. The deposits studied particularly lie in the Mt. Whitney Quad- rangle of the United States Geological Survey, though work was done in the Olancha Quadrangle, and beyond the limits of both these sheets, as problems demanded. ‘The results of this work are used as a Doctor’s thesis in the University of Chicago, this article being one chapter of that thesis. The purpose of this paper is threefold: (1) to describe the char- acteristics of the terrestrial deposits of Owens Valley; (2) to dis- cuss the causes and processes involved in their deposition; and (3) to deduce certain criteria whereby materials so deposited may be distinguished from other deposits such as those of lakes and seas, even after cementation has taken place. The adequate study of such deposits should lead to the establishment of criteria by which terrestrial deposits of earlier ages may with certainty be separated from marine deposits. It should also lead to the establishment of criteria for the recognition of various kinds of non-marine depos- its. It is recognized that such criteria have already been discussed, and to a certain extent established. But many of these are appli- cable only to formations of pronounced characteristics, and there are yet many formations of one age and another whose origins are not yet established beyond doubt. Owens Valley is an area about too miles long north and south, by 12-15 miles broad. It is situated in extreme eastern California, about east of a point on the coast midway between San Francisco and Los Angeles. The valley includes the villages of Bishop, Big Pine, Independence, Long Pine, and Keeler, which can be reached by the California and Nevada Narrow Gauge Railroad, connecting with the Southern Pacific at Mina, Nevada. Physiographically, Owens Valley is located between the Great Basin on the east and the Sierra Nevada Mountain province on the west. The east wall of the valley is the west face of the Inyo Mountains, one of the semi-arid basin ranges, while the west valley 708 ARTHUR C. TROWBRIDGE wall is the steep eastern slope of the Sierras. Owens Valley, between these two ranges, is occupied partially by Owens Lake, and drained by Owens River which flows into the north end of the lake. The surface of the valley is broken in several places by the Alabama Hills, Poverty Hills, and a series of recent volcanic cones and lava flows (see Plate I). The eastern face of the Sierra Nevada Mountains is a precipi- tous fault scarp, probably of late Miocene age, attaining a height of 10,000 ft. above the bottom of Owens Valley. In this slope, streams and valley glaciers have carved numerous deep canyons, whose lower portions are choked with drift and whose upper por- tions are the cirques and bare surfaces of glacially eroded regions. The rock of the mountains in this region is massive, coarse-grained igneous rock, chiefly granite. This rock is weathered chiefly by mechanical processes. Temperature changes and the wedge work of ice cause pieces of rock varying in size from a fraction of an inch to a score or more of feet in diameter to break off and roll down the steep slopes, each piece being broken or worn smaller as it goes. Plants, animals, and ground water are relatively unimportant as weathering agents here, because by reason of the steep slopes, they are not present in abundance. On the other hand, because of these steep slopes, gravity is more than usually important. Oxidation, hydration, carbonation, solution, etc., as usually performed by atmosphere and ground water, do not take place sufficiently rapidly to produce great results on the rocks before these last are disrupted and taken away. That is, the mechanical processes of weathering and transportation take place more rapidly than the chemical processes, and the result is arkose material carried down the moun- tain canyons and deposited in the valley below. These are the materials to be described as the terrestrial deposits of the valley. Unlike the Sierras, the Inyo Mountains contain both igneous and sedimentary rock, in about equal abundance. Ordovician, Carboniferous, and Triassic sedimentary formations have been interbedded with Triassic lavas, and intruded by Cretaceous granite and diorite. Though these mountains are not so high by 4,000 ft. as the Sierras, and the slopes are not so steep, still here also mechanical processes of weathering keep ahead of chemical 199} COT “[VAIOJUL MOJUOD “YOU IY} O} Sop 9a1Y} Noe ‘a[vosg HIONVAAVA() “IVD ‘AGNIIHA ‘LI Nowy ee Jel ((2)oTsseyaL pue snozegtuoqaep) F=——| ° Syoo.r el ksequeutpes ((2)OTSSBTIL) *sygny pua Seavey OISSBTIL-YS0d) 77 ee otppUeAyH WU, ((2)9ued0tTd) *sqytsodoep | aUTIYTSNOBy] * Lis | : zy (£13Ur87 END) dee See Zy|sqtsodep Tetanutd Foal Se Clue uh é EE \ ‘ Sv Shek 4 S (fe { ‘het ar HNO oe An Neon: : aE > AY, AP NELE STA we s Vara thd) AL PASSE A eC CARN bee it Ghul 8 ‘ON ‘XIX “10A ‘ADOIOR) FO TVNANOfL iia we 429} Cor “[eAIDIUI INOJUOD §“YOUT 243 0} saprur 2a1q7 jnoge ‘a]zI¢ aIONVaGVAd “IVD ‘AINITAN\ “LY Koa ee areeso DQ] cD, ((2)oTsseyay pue Azequeutpes ((Z)otsseqazt) *SJJNZ puw seaeT (2ISStBtaz1-3s0g) 7 “syo0r ota Tug WA, ((2)suesotta) *szysodep euyazsnoey ee a (Aapusezend) __ WP a *sqtsodop Tetantd |i" ae, Le pus _TeTAnta —— (aus004STOeTd) *saapuyto pur |y BART OT9TBSEG ag (quedey) *qsnp pue pues ~]| ueyioe fumyAnttTy (SETH VAs *CNHOaT AWE rN Stak \ 3 ea % oni ‘ TERRESTRIAL DEPOSITS OF OWENS VALLEY 709 processes, this being due partly at least to the fact that the Inyo Mountains receive little precipitation, and the moisture necessary for chemical processes is lacking. Here also canyons have been cut by streams, and material has been transported to the valley and . deposited, but unlike those of the Sierras these streams are inter- mittent, and carry material only after the infrequent rains. The Inyo Mountains have not been glaciated. The deposits considered in this paper occur along the east foot of the Sierra Mountains on the west side of the valley, and discon- tinuously along the west front of the Inyo Mountains, which border the valley on the east. The phenomena on the opposite sides of the valley are sufficiently unlike to warrant description separately. DEPOSITS AT THE BASE OF THE SIERRA MOUNTAINS LOCATION AND EXTENT Within the area of the Mt. Whitney Quadrangle, terrestrial deposits at the east base of the Sierra Nevada Mountains cover a belt 1-7 miles wide. In the Olancha Quadrangle to the south, -corresponding deposits extend for many miles in a narrow and more or less disconnected belt. At the north, the plain of the terrestrial deposits is overlain by recent lavas and volcanic cones. Northwest of Owens Lake the alluvial deposits lie against the west edge of the Alabama Hills, and extend around the north and south ends. Two narrow con- tinuations of the deposit extend through gaps in these hills, and deploy slightly on the east side. Elsewhere the plain joins the flat bottom of Owens Valley along a more or less distinct line. On the west side the plain is limited sharply by the foot of the mountains. In the aggregate, the deposits cover about 175 square miles in the Mt. Whitney and Olancha quadrangles. TOPOGRAPHY FANS AND INTER-FAN AREAS Topographically, this plain of pluvial and fluvial deposits takes _ the form of a series of fans joined together at their lateral edges. At first glance, either in the field or upon the topographic maps, 710 ARTHUR C. TROWBRIDGE it seems to be a continuous plain sloping from the mountains; but studied in detail, it resolves itself into low, gently sloping fans separated by broad, ill-defined, shallow depressions. The fans deploy and become less distinct at a distance from the mountains; the depressions are therefore broader, deeper, and better defined close to the mountains. At its outer edge, the topography of the deposits approaches a plain, in which neither fans nor inter-fan areas can be distinguished. The axes of the fans are on lines which are continuations of canyons in the mountains; the depressions are between the mouths of the canyons. From Owens Lake north, the following fans can be distinguished : those of Richter, Tuttle, Lone Pine, Hogback, George, Bairs, Shepard, Pinyon-Pine, Oak, Thibaut, and Sawmill creeks. The last three are small though sharply defined. A few notes taken north of Lone Pine Creek are here copied, in so far as they refer to the topography of the fans: The fan opposite Lone Pine Canyon is sharply set off from the fan of Hog- back Creek to the north. Beginning at the mouth of the canyon, it spreads promptly to the north, a distance of about half a mile at the immediate foot of the mountains, and one and one-half miles within a distance of a mile from the mountains. Farther from the mountains, it joins the fans on either side, and its distinctness is there lost. Its north edge is fairly distinct for two miles from the mountains, being markedly higher than the broad, irregular, linear depression between it and the fan of Hogback Creek. This depression is dis- tinct near the mountains, but becomes gradually shallower and narrower away from the mountains, until the two fans coalesce two miles or so out. . . . . From the depression, the slope of the fan of Hogback Creek shows a distinct TISGa ecors. The south side of the fan of Shepard Creek is not especially well developed, though it is set off distinctly from the fan to the south. The depression between these two fans is about 200 ft. below their tops, and is one- fourth to one-half a mile broad. .... North of the fan of Shepard Creek, the surface declines and does not again reach the high level of this fan as far as the alluvium can be seen. The streams have a distinct tendency to leave their fans for the depressions between. Shepard Creek now flows in the depres- sion south of itsfan. The North Fork and South Fork of Oak Creek have joined in the depression between their respective fans. At some time they were undoubtedly parallel streams. The photo- TERRESTRIAL DEPOSITS OF OWENS VALLEY Tata graph (Fig. 1) shows the two fans, South Fork flowing in the low place, and North Fork leaving its fan for the depression. This shifting of streams to the inter-fan areas is suggested as a common and efficient process in the tying together of fans, making piedmont alluvial plains, or bajadas.* By this shifting, fans are made between fans, tying them together and tending toward the union of the fans into one plain. CHANNELS AND RIDGES Low ridges and shallow depressions on the individual fans con- stitute topographic features of a second order. These are the Fic. 1.—A photograph of the fans of Oak Creek, showing South Fork (ab) flowing in the inter-fan depression, and North Fork (cd) leaving the fan to join South Fork in the depression. channels and depositional features of the streams which deposited the fans. The depressions are more noticeable than the ridges. The depressions are, as a rule, about ro ft. deep and less than too ft. across, though at a maximum they reach a depth of 20-25 ft. and a width of quite too ft. Their bottoms are usually flat and their slopes as steep as the material will permit. The elevations are less numerous than the depressions, and have less relief. They are seldom more than 5 ft. above the surrounding plain, and their height in many cases is only equal to the diameter of the individual bowlders of which the ridges are composed. The ridges consist t The term bajada has been suggested by C. F. Tolman (Jour. Geol., XVII [1909], 141) to replace the longer term commonly in use. It has the advantage of brevity, but lacks the explanatory value of the older term. Te ARTHUR C. TROWBRIDGE of mere divides between channels and of lines of bowlders bordering the channels. In keeping with their origin, the depressions and ridges are radiate in their arrangement. At the head of each fan, these fea- tures are few; toward the outer edge they are numerous; but at the extreme edge they are again rare. Three miles from the edge there are probably 50 channels, for one close to the head, and some- thing like that proportion between the same three miles from the edge and the outer edge itself. Channels which, near the head of the fan, are close together, diverge outward, and each may break up into other channels, each less deep and less broad than the one from which it springs. Quite commonly these channels lead to depressions between fans and disappear. It is clear that these channels on the fans mark the courses of the distributaries from the fan-making streams. The streams branched again and again, some of the distributaries reaching the inter-fan depressions and flowing off through them. It is equally clear that some of the elevations are merely inter-distributary divides. The origin of the ridges bordering the depressions is not so clear. Possibly they are in principle natural levees, built as the waters overflowed their channels. It is understood that these are the streams which deposited last on the surface of the fans. In the building of the bajada, the channels undoubtedly shifted frequently, those of one time being filled up and a new set formed during periods of greater deposition following heavy rains or the rapid melting of snow in the mountains. THE STREAM CANYONS The streams of the bajada do not now distribute over the fans, but flow in deep, steep-sided, canyon-like valleys; that is, the bajada is being dissected (Fig. 2). This is true to a greater or less extent of all the streams which have played a part in the deposition of the plain. The canyons in the bajada vary in depth from 20 ft. to 250 ft., and average about 200 ft. in width. The depth is determined by the size of the stream, the height of the fan, and the position of the stream on the fan. The most pronounced canyons are those of TERRESTRIAL DEPOSITS OF OWENS VALLEY FL Carroll and Lone Pine creeks, which are large streams flowing in the high central part of well-developed fans. Carroll Creek canyon is 250 ft. deep at the head of the fan, and shallows to less than 50 ft. at the edge of the bajada. Shepard Creek is almost as large as elther of the two streams previously mentioned, but it flows on the side of its fan, where the surface and gradient are lower, and its canyon is only 30 ft. deep. Hogback Creek has cut deeply at the head of its fan, but farther out, where the stream has shifted to the side, there is little intrenchment. Fic. 2—The canyon of Carroll Creek in the Sierra bajada.. The dark strip consists of trees 20 ft. bigh. These canyons are the most conspicuous topographic features of the bajada. They clearly follow the building of the bajada, and were excavated under different conditions. They therefore have both an expository and a historical value. Problems connected with them will be discussed later. BOWLDER BELTS A fourth topographic feature of the bajada consists of almost innumerable lines of bowlders which, though primarily a matter of the constitution of the bajada, affect the topography in a minor way. These lines of bowlders have a radiate arrangement similar to that of the channels and ridges. The bowlders are so close together 714 ARTHUR C. TROWBRIDGE as to make low and discontinuous ridges, which by winding his way among the bowlders one may in some instances be able to cross without climbing. ‘The height of the ridges is determined by the size of the bowlders, and is usually less than to ft. They seldom consist of more than two thicknesses of bowlders. THE SLOPE OF THE BAJADA The slope of the piedmont plain away from the mountains varies rather uniformly with distance from the mountains. It also varies Fic. 3.—The slope of the Sierra bajada as seen on the north wall of the canyon of Carroll Creek. Fic. 4.—The slope of the Sierra bajada on the south wall of the canyon of Carroll Creek. irregularly from place to place along the foot of the mountains. Along Carroll Creek the slope at the face of the mountain is 18° and 20° (Figs. 3 and 4). Where it is 20°, the angle decreases to about 12° a quarter of a mile from the mountains; and where it is 18° at the mountains, the slope is 6-8° a mile or so out. The fan of Lone Pine Creek has a slope of 6° at the mountains, which de- creases almost uniformly to a very low slope at the west edge of the Alabama Hills. The difference between the slopes of the fans of Lone Pine Creek and Carroll Creek might be due to a diastrophic tilting, which either did not occur at Lone Pine Creek or did not TERRESTRIAL DEPOSITS OF OWENS VALLEY TUS affect the fan there. However, no other evidence of diastrophism appears at Carroll Creek. Probably the difference is due to varia- tions in the gradient and size of the two streams in the mountain canyons, at the time the fans were built. The average slope of the fans in the valley at the base of the mountains is about that of the fan of Lone Pine Creek, 6°. The slope is greater along the axes of the fans than in the inter-fan depressions. MATERIALS The material of the Sierra bajada is not well exposed, but some idea of its upper portion can be obtained. Nothing is known of Fic. 5.—The largest bowlder seen in the Sierra bajada. The man is 6 ft. tall. This bowlder lies in the yard of the Cerro Gordo power shanty in Lone Pine Canyon, 14 miles from the foot of the mountains. that portion lower than 300 ft. from the surface, as there are no cuts so deep, and well-records are lacking. The material may be seen in three sets of places: (1) on the unaltered surface of the plain, (2) on the sides of the shallow channels, and (3) in the walls of the canyons. Lithologically, the bajada is composed of material from the granitic rocks of the Sierras, disintegrated rather than decomposed. Its components are bits of granite, rather than crystals of quartz or feldspar. Even the disintegration is not complete, for the mate- rialis commonly coarse. It is clear that the source of the material is the mountains, and that it was removed from the parent ledges mechanically, and transported to its present position by streams, 716 ARTHUR C. TROWBRIDGE aided in the upper parts of the canyons by glaciers. There is every evidence of immature weathering of the materials. At the time the fans were being deposited, the mechanical processes of weather- ing greatly exceeded the chemical, and transportation was free and rapid. TEXTURE The fans contain all textural grades from pieces the size of small sand grains and even clay particles, to bowlders more than 20 ft. in diameter, but pieces less than an inch in diameter predominate. The most striking and surprising feature of these fans is the extreme coarseness of some of its materials. Innumerable large Fic. 6.—Bowlders on the surface of the Sierra bajada. Their size may be esti- mated from the horse. Picture taken on the fan of Sawmill Creek about a mile from the foot of the mountains. and small bowlders appear on its surface, on the sides of the shallow channels, and on the walls and floors of the canyons. On the unaltered surface, they occur in radiating lines and belts, roughly parallel with the radiating channels. They are practically confined to the higher parts of the surface, where the main streams flowed. None appears in the inter-fan depressions. More bowlders are scattered near the heads of the fans than toward the outer edges but they are not noticeably larger here. They are arranged in belts or low ridges along the borders of the shallow channels and are scattered more sparsely on the side slopes. Bowlders are numerous on the walls and bottoms of the canyons, but without definite arrangement. The beds of the streams are everywhere choked with them, and they occur in and along the braided channels TERRESTRIAL DEPOSITS OF OWENS VALLEY oiaig | used by the streams in time of flood. Their unusual abundance in the canyons is doubtless due to the bowlders having been sorted out in the process of canyon-cutting, the finer material being carried on and the coarse left. The whole surface of the bajada considered, the average diameter of the bowlders is perhaps about 2 ft. but those 8 ft. in diameter are by no means uncommon. The largest seen are a mile west of Lone Pine, 6 miles from the mountains, and at the Cerro Gordo power shanty, on Lone Pine Creek, 13 miles from the mountains. Fic. 7.—The canyon of Lone Pine Creek in the Sierra bajada. Bowlders appear almost as large as the two-story house. The one west of Lone Pine is 10X 20X30 ft. above ground. The size of the one at the shanty is shown in Fig. 5, the man being 6 ft. tall. With these exceptional bowlders are thousands of others as large as 10 ft. in diameter, as can be seen from Fig. 7. The size and distribution of this coarse material may be seen further in Figs. 6 and 7. Fine material in the bajada is shown in Fig. 8. STRUCTURE Owing to the scarcity of good exposures, the structure of the materials of the Sierra bajada is not readily determined. The only satisfactory exposure is near the mountains on Lone Pine Creek (Fig. 9). Because the canyon walls never stand in vertical bases, but slump down readily to gentle slopes, they show the 718 ARTHUR C. TROWBRIDGE texture only, not the structure. The relations of coarse and fine material can be seen to some extent on the surface of the plain. As has been shown above, the coarse and fine materials are more or less separated on the surface of the plain. There are consider- able stretches, usually the lower areas, where the surface material is all fine. Such areas are interrupted by narrow belts of large bowlders. If the structure of the whole plain were judged by its Fic. 8.—Fine material in the Sierra bajada seven miles from the mountains surficial aspect, the material could be known to be roughly sorted into many narrow radiating belts of coarse materials, and broader belts of fine materials. Presumably these lines would not have the same position horizontally for any considerable vertical section, as the stream channels undoubtedly shifted and distributed often. So far as cuts in the bajada show, the materials consist of a mixture of large blocks of granite, bowlders not so large, angular fragments the size of cobbles, tiny angular bits of rock, sand, and clay. Where any considerable vertical section is seen, these differ- ent grades are sorted into indefinite lenses and pockets. There are no definite layers of great extent. Divisions of material are no- where seen to be continuous for as much as too ft. Some small TERRESTRIAL DEPOSITS OF OWENS VALLEY 719 exposures show material which is apparently unstratified. Even where sorted into lenses or irregular areas, there is a considerable mixture of all sorts of material in each division, each merely averaging a little coarser or a little finer than its surroundings. All materials are clearly water laid, but under conditions which allowed of very poor sorting. The structure and texture of the materials are brought out best by detailed descriptions and photographs. Such illustrations are given below and in Figs. 9, ro, 11, 12, and 13. Fic. 9.—A section in the Sierra bajada from the south wall of Lone Pine Canyon, 3 mile west of the Cerro Gordo power shanty. 1. Two hundred to three hundred yards above the power shanty on the south wall of the valley of Lone Pine Creek, wash and gravity have exposed the material almost continuously for a distance of about roo ft. vertically. The section consists of both fine and coarse material roughly separated from one another. ‘There are two horizons of coarse bowlders, one 20 ft. from the top. and the other about 30 ft. from the bottom. In the upper horizon the bowlders are fairly well rounded, and range up to 4 ft. in diameter, the average being about 1 ft. In the lower zone of bowlders there is greater range in size. There are numerous pieces 6 in. through, and several 6 ft. or so in diameter. The sorting is very slight. Between these two horizons the materials are mostly fine, though large bowlders are not entirely absent. Immediately above the 720 ARTHUR C. TROWBRIDGE lower zone of bowlders is a fairly well-defined bed of gravel, the constituents of which average 3-4 in. in diameter. Between this gravel layer and the upper zone of bowlders, the material differs in different parts of the cut, the structure being decidedly pockety. In one place the fine gravels grade into the coarse bowlders above; in another, there is a body of clay between the two; in another, the gravel layer does not appear, and clay separates the two beds of bowlders (Fig. 9). 2. A six-foot cut a mile from the mountains shows a matrix of clay and sand in which there are angular fragments averaging Fic. to.—Roughly sorted materials of the Sierra bajada on George Creek to in. through, with occasional bowlders 3 ft. in diameter. The bowlders are angular or subangular, and none are well rounded. They are not abundant enough to touch one another. There is apparently no sorting. 3. A hundred yards above the last section, the material is rudely but distinctly sorted. At the bottom there is a fairly uniform layer of angular gravel, averaging 4 ft. in thickness, of which the upper 1 ft. give place at the east end to a projection downward of a pocket from the coarse layer above. ‘There is little clay in the cut, and the fragments of rock are abundant enough to touch one another. The pores are filled with coarse, arkose sand, loosely packed. 4. The last section gives place, within a few feet, to unsorted material entirely similar to that of section 2. From here to the mountains the sorted and unsorted materials occur with about TERRESTRIAL DEPOSITS OF OWENS VALLEY 721 equal frequency, in the same position relative to the stream and to the surface of the fan. 5. Material is exposed on the Mt. Whitney trail through the Alabama Hills. It is composed of rounded or partly rounded granitic bowlders, up to a foot in diameter, with a sparse matrix of granitic pebbles. Though stratification is not apparent, all material is clearly water laid (Fig. 12). 6. On Dietz Creek, just above its junction with Tuttle Creek, material is exposed. It is a mixture of very fine angular fragments ee Fic. 11.—Unstratified material on George Creek. This section occurs within only a few feet of that shown in Fig. to. with material having the texture of coarse sand. The fragments are not so angular as pieces just broken by weathering. They seldom exceed a half-inch in diameter, and the average is about =}, of aninch. The finer material consists of arkose, flakes of mica, grains of pyrite, and of ferro-magnesian minerals, being almost as common as quartz. ‘The coarse and fine materials are unassorted. From these descriptions and photographs, the following charac- teristics of the materials are shown: 1. The material was derived from the rock of the Sierra Nevada Mountains. OP? ARTHUR C. TROWBRIDGE 2. It is the result of immature weathering in the mountains. 3. In texture the materials range from clay-like particles, to bowlders 30 ft. in diameter. 4. Some of the large bowlders are ice-shaped, and some have been shaped slightly by water. 5. Stratified, partly stratified, and entirely unassorted materials occur in something like equal proportions. 6. Where stratification exists, the materials are sorted into lenses and pockets, never into uniform, continuous layers. Fic. 12.—A pocket of stratified gravel in the Sierra bajada seven miles from the mountains on Lone Pine Creek. 7. The materials become gradually finer as distance from the mountains becomes greater, at least so far as sub-surface material is concerned. 8. No fossils were found in the material. These features will be discussed after the deposits at the foot of the Inyo Mountains have been described. TERRESTRIAL DEPOSITS OF THE INYO MOUNTAINS Terrestrial deposits are represented in the Inyo Mountains by two distinct types of materials of two distinct ages. They will therefore be discussed separately. PLIOCENE LACUSTRINE DEPOSITS © No description of the terrestrial deposits of Owens Valley, and no discussion of the older deposits at the foot of the Inyo Mountains would be adequate without mention of certain lacustrine clays and TERRESTRIAL DEPOSITS OF OWENS VALLEY 723 sands in Waucobi Canyon, described by Walcott,’ and in the vicin- ity of Haiwee described by Turner,’ Fairbanks,3 and Campbell,’ even though such mention leads beyond the confines of the Inyo Mountains. Both these deposits were seen by the writer, and they are here discussed in so far as they may be used as a type of lacus- trine deposits. Walcott’ and Spurr® have interpreted these deposits in slightly different ways, but both agree that they are lake deposits, and as such they will be described. Discussion as to whether they record great recent uplift of the Inyo Mountains, as according to Fic. 13.—Unsorted material of the Sierra bajada Walcott, or were deposited in a deep, widely distributed lake, as Spurr contends, is not in place here, though such discussion is given in the unpublished part of the thesis. LAKE BEDS IN WAUCOBI CANYON Waucobi Canyon, or the Waucobi embayment as it is called by Walcott, is a re-entrant in the west face of the Inyo Mountains a few miles north of the boundary.of the Mt. Whitney Quadrangle tC. D. Walcott, Jour. Geol., V, 240-48. 2 Personal communication to Spurr. 3H. W. Fairbanks, Am. Geol., XVII, 60. 4M. R. Campbell, Bull. U.S. Geol. Surv. No. 200, p. 20. 5 Jour. Geol. V, 344-48. 6 J. E. Spurr, Bull. U.S. Geol. Surv. No. 208, pp. 209-10. 724 ARTHUR C. TROWBRIDGE east of Alvord. In this re-entrant are a series of unconsolidated and partly consolidated sands, clays, and gravels. They were traced from wall to wall of the re-entrant, and up the canyon for some 35 miles. Mr. Walcott reports their continuation almost to the.crest of the mountains. There are two more or less distinct phases of this deposit. Near the north and south walls of the re-entrant are interbedded clays, limestones, and conglomerates. These materials are mostly sorted into distinct beds, but are locally arranged in pockets or irregular ak: Fic. 14.—Lacustrine limestones and conglomerates, deposited near shore in the Waucobi embayment. ‘ areas, when seen in sections. Two miles east by northeast of Alvord, a ledge of limestone and conglomerate outcrops under a low hill of angular alluvium. The limestone is white, porous, earthy, and filled with small fossils of gastropods. Other rock has a matrix of calcium carbonate, but contains enough pebbles to make it conglom- eratic, though the pebbles are seldom in contact. The stony matter is fairly well rounded. In size its pieces reach 6 in. in diameter, though the average is about rin. The pebbles are mostly of sedi- mentary rock, but with some granites. All the material is local. A series of exposures along the main road 13 miles southeast of the fruit ranch of J. S. Graham, at the southeast margin of the re-entrant, shows well the constitution of the beds. All the mate- rial is irregularly sorted. In some places it is of light-yellowish TERRESTRIAL DEPOSITS OF OWENS VALLEY 725 clay, which contains enough lenses and pockets to give it a bedded aspect. In other places it is made up of alternating layers of gravel and clay, the latter containing bowlders in many places ilies 14): Just below these coarse, roughly sorted materials on the main road, the constitution changes abruptly to sandy clay, arranged in continuous, uniform, apparently horizontal layers. The change from coarse, poorly sorted conglomerates to fine clays takes place within 500 ft. These clays appear all the way to the mouth of the Fic. 15.—Lacustrine sand and clay in the Waucobi embayment. The layers are continuous and of uniform thickness. canyon, as erosional hills about too ft. high, and in the valley walls. Most of the clay is fine, becoming white dust when powdered. Some layers are sandy and some calcareous. They are not usually firmly cemented. Layers 1-10 ft. in thickness can be traced con- tinuously along the hills (Fig. 15). These are doubtless lake deposits, the coarser marginal con- glomerates being the littoral phase, and the centrally located clays and sands having been deposited in quieter, deeper water, farther from shore. The fossils have been determined by Dr. Dall as Pliocene to recent. As the beds lie unconformably under Quater- nary alluvium, they are probably Pliocene or early Quaternary in age. 726 ARTHUR C. TROWBRIDGE LAKE BEDS NEAR HAIWEE Toward the south end of Owens Valley, in the vicinity of Haiwee post-office, there is a series of calcareous and arenaceous lake beds. They were seen by the writer 3 mile southwest, 1 mile east, and 4 mile northwest of the post-office. They are fine, white, and distinctly bedded. East of the post-office, the beds dip 8° to the northwest. Northwest of the post-office they dip 14° north. They were best seen 3 mile northwest of the post-office, where a hill 175 ft. exposes them from top to bottom. They consist of light-colored, siliceous Fic. 16.—Unconformity between Quaternary conglomerates and Pliocene lake beds north of Haiwee. Some of the bowlders of the conglomerate are composed of the underlying clays. fine clays or shales. In the lower part of the exposure numerous small flat bodies of gypsum occur. Most of the plates lie parallel with the beds, but in some places they appear as secondary bodies along joints and faults. The lake beds here are covered with a hard, coarse conglomerate derived from the Coso Mountains to the east. The lake beds and conglomerates are unconformable. The conglomerate lies on the very irregular surface of the truncated edges of the dipping beds of clay, the surface between the two having a relief of about 15 ft. (Fig. 16). The constituents of the conglomerate are chiefly granite, sedimentary rock, and scoriaceous basalt, but near the contact many large fragments of the underlying clays are also included. The TERRESTRIAL DEPOSITS OF OWENS VALLEY HOT conglomerates must be as old as the early Quaternary. The hill on which they occur is far from the Coso Mountains and separated from them by numerous valleys, similar hills, and more lake beds. The lake beds are then pre-Quaternary, probably Pliocene, and correlated with the similar beds in Waucobi Canyon. No fossils were found here. OLDER DEPOSITS AT THE FOOT OF THE INYO MOUNTAINS Bearing in mind the main characteristics of the deposits described above, and accepting the idea of their lacustrine origin, as we are apparently forced to do, we can proceed to a description Fic. 17.—Low hills of old deposits surrounded by present-day fans, northeast of Mt. Whitney station. A) and interpretation of the older deposits of terrestrial material at disconnected points along the foot of the Inyo Mountains. Distribution as indication of age.—At several places, most notably northeast of Mt. Whitney station and east of Citrus, hills of ter- restrial material rise 100 ft. above fans which are now in process of making about them. The materials of the hills differ from those of the fans in texture, and in the fact that they are cemented. Along the west face of the mountains immediately northeast of Mt. Whitney station, an exceptional series of events is recorded by the nature and relations of the alluvial deposits. Two hills, roo ft. or more in height, occur half a mile apart, and half a mile from the foot of the mountains. Between and around these hills the lower surface is covered with the typical alluvium of the region. The hills themselves are of gravel and sand. Evidently a great 728 ARTHUR C. TROWBRIDGE deposit was laid down here, its remnants being represented by the hills. Conditions changed so that erosion took place, the old deposit being dissected into hills and valleys. Later, as the streams were brought to adjustment again, they deposited new fans among the remnants of the old deposit (Fig. 17). Due east of Citrus, a series of low spurs projects from the foot of the mountains into Owens Valley. At the foot of the mountains they stand 50-75 {t. above the plain, and become gradually lower west- ward. ‘They are not in direct contact with the present fans, though fans occur at lower levels north and south of them. These projec- tions are not at the mouths of present canyons (Figs. 18 and 109). Frc. 18.—Lacustrine beds (light colored) lying against rocks of the Inyo Moun- tains (darker rock to the right), east of Citrus. Stratification may be seen in the left center. Bs) The deposits in Mazourka Canyon, the canyon east of Aber- deen, and on the flanks of the mountains east of Keeler should also be included in this category. In the two canyons, older cemented gravels occur as distinct, flat-topped, but eroded terraces, 50 ft. above the stream beds. East of Keeler, hills of conglomerate rise 200 ft. above present alluvial surfaces. Constitution.—The materials of the older deposit at the foot of the Inyos differ from those of the Sierra bajada in various ways; especially in (1) lithological composition, (2) texture and shape of pieces, (3) structure, (4) cementation. 1. This deposit is made up of fragments of all rocks occurring in the Inyo Mountains, from which they are derived, including both igneous and sedimentary rocks. TERRESTRIAL DEPOSITS OF OWENS VALLEY 729 2. The constituents have not a great range in size. In place of large bowlders, the coarser materials are large-sized cobbles or very small bowlders. Texturally, the fine material is sand or clay. The average size of particles is probably less than half an inch in diameter. Bowlders even as large as 1 ft. in diameter are wanting. In the Sierra bajada the pieces of rock are either ice-shaped, or they are almost as angular as when broken off by weathering. Here the effects of glaciation are not seen. The fragments have been worn to pebbles, few sharp or irregular edges appearing. Fic. 19.—Close view of the older deposits east of Citrus. Note the layered structure and the dip of the beds. They are in general well rounded, having been shaped by the action of water. 3. These materials are arranged in definite, continuous layers of gravel and sand. ‘The layers can be traced the whole length of the various outcrops as beds of nearly uniform thickness. East of Citrus a definite, continuous layer of clean, fine gravel, uniformly 2 ft. thick, overlies a layer which is a mixture of small angular frag- ments, sand, and clay. Northeast of Mt. Whitney station the talus from a deep cut is of uniform-sized cobbles and sand. The stratification of this material may be seen in Figs. 18 and ro. Wherever exposures were seen east of Citrus, the beds have an appreciable westward dip (away from the mountains). Clinometer readings vary between 8° and 18°. The direction of dip is nearly 730 ARTHUR C. TROWBRIDGE constant. No faults, minor folds, or other evidences of diastro- phism were seen. This dip may be depositional, or the beds may have been tilted to their present position by an uplift in the moun- tains. Eighteen degrees is a high dip to be considered depositional when the material is fine. 4. The older deposit shows a tendency toward cementation, and some layers are firmly cemented. In general it is the layers of coarse material, originally more porous, that are cemented. In the beds east of Citrus mentioned under (3) above, the upper layer of gravel is cemented to firm conglomerate. It is so solid as to ring under the hammer and to need more than one hard stroke Fic. 20.—A stream terrace of older alluvium in Mazourka Canyon before it is broken. The material of the finer layer below cannot be picked out by the hand, but yields readily to the hammer. The gravel layers are almost everywhere so indurated that they stand out conspicuously, the determination of dips thus being made easy. The above characteristics hold for all the deposits of older materials at the foot of the mountains, with the exception of those in Mazourka Canyon. The deposits in this canyon belong to the older deposit, for they occur in terraces above the present deposi- tional surfaces, but the materials are in some respects different. Areally considered they take the form of the canyon in which they were deposited, and thus occur in a long strip. They constitute more or less definite stream terraces on the sides of the present valley (Fig. 20). Texturally they are like the deposits along the foot of the mountains, the chief difference being in the strati- TERRESTRIAL DEPOSITS OF OWENS VALLEY 731 fication. Instead of being in definite layers, as are the deposits at the foot of the mountains, the materials in the canyon are in indefinite lenses and pockets, similar to those of the Sierra bajada, though on a much smaller scale. An examination of two detailed sections noted just above Barrell Springs brings out the difference between these deposits and those east of Citrus and Mt. Whitney station: ig Tal Be ao Angular fragments the size of cobble, and clay. Tepeliteeeca ls crs Well-sorted, very fine gravel.. Pinches out in both directions. Ties Aiba. oe Mixture of cobbles and clay. Pinches out upstream. TEP OURAN a Sats Clay with some small angular bits of rock. Peete ee ccs Well-sorted, loose, fairly well-rounded cobbles; average size 13 in.; one-inch clay layer in middle. 3 ft....... Mixture of sand, clay, and gravel; little or no stratification; pockety; contains one bowlder one foot in diameter. De whbte cases Clay, sand, some gravel; poorly sorted. Dwele ns ek Fine angular fragments; little or no clay or sand; well "assorted. 2. Fifty feet down the valley from the last, the following section occurs: Be enhiteiera i a Mixture of coarse and fine angular gravel, with clay in the interstices; average size of constituents 1 in. in diameter; occasional bowlders 1 ft. in diameter. 1 tt eI Moderately fine gravel; little or no clay; no pieces larger than 3 in. in diameter; pinches out in 12 ft. up valley. Dim lites aaa: Clay and pebbles intermixed; rude layer of cobbles in middle. DB Ves a od Ook Fairly well-sorted gravel, coarser at bottom. Pinches out rapidly in both directions. Loosely packed, interstices not filled. Mae LEN erect ai Irregularly bedded bowlders, cobbles, fine gravel, clay. Digg bya eas Indefinitely bedded fine angular gravel. Average 4 in. in diameter. Tame sais tok ard Pockety, coarse gravel, constituents up to ro in. in diameter. In both sections, the materials are slightly cemented. Nota single subdivision of one could be traced 50 ft. to the other. For further details of these materials see Figs. 21 and 22. It is thus seen that the materials of Mazourka Canyon differ from the rest of the older deposit at the foot of the mountains in Wee ARTHUR C. TROWBRIDGE that it includes coarser materials and has a lens and pocket struc- ture (cf. Figs. 19 and 21). Origin of the older deposit at the foot of the Inyos.—It will be seen from the foregoing that most of the older deposit is sufficiently unlike the Sierra bajada to lead one to conclude that its mode or conditions of origin were not the same. On the other hand, if it be compared with the near-shore phase of the lake beds in Wau- cobi Canyon, a strong resemblance will be seen: (1) Both deposits Fic. 21.—Stream deposit in Mazourka Canyon. Compare with Figs. 14 and 15 were formed and eroded before the deposition of the recent alluvium. (2) Both are firmly cemented, at least locally. (3) They are similar in texture, both being fine and having a low textural range. (4) Their stratification is the same, both being sorted into layers. They are dissimilar in that the constituents of the deposits at the foot of the mountains are better rounded than those in Waucobi Canyon, and the former contain no fossils. With such similarities between these deposits and the lake beds, it seems clear that the older materials northeast of Mt. Whitney station and east of Citrus are of lacustrine origin, and belong to the same formation as the lake beds in Waucobi Canyon and at Haiwee. If so, the lake in which they were deposited was shallow, and the shore lay against the TERRESTRIAL DEPOSITS OF OWENS VALLEY 733 mountains immediately to the east. The dissection of the deposit probably took place subsequent to the uplift of the mountains and the draining of the lake. If lacustrine beds corresponding to them were deposited at the foot of the Sierras, they have been covered and concealed by the more recent alluvium. The deposits in Mazourka Canyon are obviously not lacustrine, but of stream origin. They were probably laid down on the floor of a mature valley, which was tributary to the lake east of Citrus. Fic. 22.—Photograph to show the shapes of the cobble in the terrace of Mazourka Canyon. The differences between these deposits and those along the foot of the mountains may be taken as differences characteristic of lacus- trine and fluvial deposits. THE PRESENT FANS DISTRIBUTION Between Mt. Whitney station and Aberdeen there are nine sepa- rate fans at the foot of the Inyo Mountains. On the map (Plate I) ~ they are numbered 1 to 9, beginning at the south. Of the nine fans, I, 2, 4, and 8 are large, each covering more than a square mile, and 3, 5, 6, and 9 are smaller. All of them occur at the mouths of mountain canyons. The largest one, No. 4, is at the mouth of the largest canyon, Mazourka, though it is not so well shaped as the others. Between the mouths of the main canyons and between the main fans are some patches of alluvium too small to map and not important in any way. ‘These patches occur at the lower ends 734 ARTHUR C. TROWBRIDGE of very small valleys. Fans also occur along the mountains outside the mapped area. They were seen northeast of Aberdeen and at Keeler. SHAPE AND TOPOGRAPHY In general there are two controlling factors in the shapes of the fans. At their mountainward edges they are confined by the walls of the mouths of the canyons, from which they take their form. Their outer edges deploy slightly on the plain. Nos. 2 and 8 extend about a mile into their canyons, and No. 4 extends still farther up Mazourka Canyon. Nos. 1 and 3 show deployment on the plain. No. 2 is made up of two smaller fans, with a depression at their junction. The surfaces of these fans are very similar to the surface of the Sierra bajada, except that all features are on a smaller scale, the fans, except No. 2, are simpler and remain separate, and these fans are not now in process of dissection. The individual fans show numerous radiating channels and low ridges similar to those on the Sierra plain, but they give the surface here a relief of no more than 7 or 8 ft. at a maximum. The ridges are almost invariably belts of bowlders. The fans are not dissected as is the Sierra bajada. The streams still flow on the surface in the radiating channels, but they flow only after the infrequent rains in the mountains. The slope of the fans varies considerably from head to outer edge. Fan No. 8 has a slope of about 10° at its head, 5-6° midway of its length east and west, and approaches flatness at its outer edge. Fan No. 1 has a slope of more than 600 ft. per mile in its upper part. The upper part of fan No. 2 slopes westward 800 ft. ina mile. This is steeper than the average. MATERIALS These fans at the foot of the Inyo Mountains have not been dissected, and exposures of the material are therefore few and shal- low. Some data can be collected from surface materials, there are a few shallow cuts along the channels, and one prospect pit affords a good exposure. Each fan is made up of pieces of the kind of rock in which the canyon back of it is cut. For instance, 99 per cent of the material TERRESTRIAL DEPOSITS OF OWENS VALLEY oe of fan No. 8 is granite, the other 1 per cent scoriaceous lava and slate. All the face of the mountains here is granite, which is bor- dered along the edge of the Santa Anita flat by slate. Lee’ maps a volcanic mountain southeast of Aberdeen, which doubtless ex- plains the occasional fragments of scoriae. The fan northeast of Aberdeen is composed of bits of granite, gneiss, scoriaceous basalt, and limestone. All these rocks occur together in the walls of the canyon. Fan No. 1 is made up largely of bits of lava, sedimentary rock, and granite. No. 2 is mostly of sedimentary rock and lava, Fic. 23.—Bowlders on the surface of fan No. 8 north of Citrus. Their size may be estimated. with some fragments of granite. No. 3 contains a mixture of sedi- mentary rock, diorite, and granite, in keeping with the rocks east of it. On the surface of the fans there are both coarse and fine materials arranged as on the Sierra bajada in diverging lines or belts from the head of the fan. Though the bowlders are not so large as on the opposite side of the valley, they are still astonishingly large for the drainage by which they were transported. The largest bowlders seen were near the head of fan No. 8, where there are some 10~12 ft. in diam- eter. On fan No. 2 bowlders are especially numerous. Toward =W. T. Lee, Water Supply and Irrigation Paper, U.S. Geol. Surv., No. 181, Pl. I. 736 ARTHUR C. TROWBRIDGE the head of the main fan the entire surface is so covered with them that a horse cannot travel overit. The average size of the bowlders is about 13 ft. Their number and size can be seen in Fig. 23. The photograph was taken from fan No. 8. These bowlders show little shaping. If fragments were broken from ledges by the wedge work of ice and gravity, and then the sharp and irregular edges dulled during a short period of transpor- tation, their present shape would result. Neither glaciation nor prolonged rolling has affected them. mee Fic. 24.—Angular material in the fans at the foot of the Inyo Mountains The fine material on the surface of the fans occurs in largest areas near the outer edges, where there are few bowlders, and the surface material has about the texture of fine gravel or coarse sand. A half-mile from the edge it is made of fragments more — or less shaped by transportation, averaging perhaps 3 in. through, with some lines of larger bowlders. At the head the surface is practically covered with bowlders. Although exposures of material beneath the surface are almost wanting, a few shallow cuts were seen. The data they afford follow: 1. In the outer edge of fan No. 8, 4 miles south of Aberdeen, is a pit. dug as a placer prospect, 1o ft. deep, 15 ft. long, and 5 ft. wide. The material is all fine, there being nothing as large TERRESTRIAL DEPOSITS OF OWENS VALLEY Tu as I in. in diameter. The pieces are distinctly angular. The section follows: Baie ete Fine clay and gravel, not laminated. Titan sets) clay Di Aty anne: fine gravel Seen ee clay Bliter yes 2. tee fine gravel DaliNe ws eae ClAy 2. Ten feet of material are exposed in the fan northeast of Aber- deen. It is not stratified. Angular bowlders are imbedded in a matrix of clay. 3. Near the upper end of fan No. 2, a gully affords an exposure. The section consists of both stratified and unstratified material. In the unstratified parts, the main constituent is clay, in which are imbedded numerous angular bowlders. One 35 ft. in diameter occurs in a mass of clay. 4. Distinctly angular alluvium is shown in Fig. 24. SUMMARY It is apparent from the foregoing that there are two sorts of terrestrial deposits in and at the foot of the Inyo Mountains, of two distinct ages. The first deposits appear to have been made when the mountains were low and bordered by a lake: Mazourka Can- yon had been cut and brought to grade, deposition taking place in its bottom from its mouth up. The deposits of the time are all fine, fairly well sorted in Mazourka Canyon,.and very well sorted in Waucobi Canyon and along the mountain foot. Conditions so changed that these first deposits were largely removed by erosion. The change was probably brought about by the uplift of the mountains, as the later alluvium is much coarser than the older. After the uplift, new canyons were cut in the mountains, and new fans deposited among the remnants of the old lacustrine deposits. This process is still going on. SOME PROBLEMS OF THE TERRESTRIAL DEPOSITS The fluvial deposits of Owens Valley, as described above, offer several problems. In some cases the solutions of the problems are simple and obviously correct; in others the solution is not so clear, 738 ARTHUR C. TROWBRIDGE and there is some doubt as to the correctness of tentative conclu- sions; in still other cases, the solution is entirely hypothetical. In some cases various lines of explanations may have partial applica- tion to the observed features. These problems are here discussed individually. MANNER OF FORMATION OF THE FANS AND BAJADA CAUSES OF DEPOSITION Material was deposited at the foot of the Sierra and Inyo moun- tains, primarily because of an abrupt decrease in the gradient of the streams. When the mountains were first uplifted, and precipitation fell on the slopes, streams formed and flowed swiftly down the sides, carrying material from their channels. When the base of the mountains was reached, the carrying power was suddenly and greatly decreased, and the first deposition resulted. Once started, other factors tended to increase the process of deposition. Streams lost volume by sinking into loose material. This not only reduced the volume of the transporting agent, but also lessened the velocity of the water remaining at the surface; both these changes caused deposition. The average relative humidity of the Sierra Nevada Mountains is not far from 60 per cent, and that of Owens Valley probably not more than 40 per cent.t_ When the streams reach the plain, evapo- ration is increased; hence loss of volume, loss of velocity, and decrease in carrying power. This would not be so important in the case of the Inyo Mountains, because the difference in humidity between mountains and plains is not so great there. However what little rain falls, is in the mountains rather than on the plains, and evaporation takes place more rapidly in the latter locality. Water taken from streams by irrigation so decreases their vol- ume in some other localities as to aid in causing deposition, but such is not the case in this region. The Sierra bajada has been under- going dissection rather than gaining by deposition since man began to irrigate the lands, and the streams on the Inyo fans, running only after rains, are not used for irrigating purposes. « For details of precipitation and evaporation in the valley, see Water Supply and Irrigation Paper, U.S. Geol. Surv., No. 181, pp. 17-25. TERRESTRIAL DEPOSITS OF OWENS VALLEY 739 Decrease in volume and consequent decrease in velocity took place on the plains for another reason also. It rained heavily or snow melted rapidly in the mountains, and the mountain streams acquired great volume and velocity. When the rain ceased, or the temperature dropped below 32°, the flood on the plains sub- sided from lack of supply from above. This undoubtedly furnished conditions under which a large proportion of the material of the fans was deposited. Glaciation has played a large part in the deposition of the Sierra bajada. Glaciers prepared immense amounts of material in the mountain canyons for transportation by streams. At the same time they furnished great volumes of water to act as the transport- ing agent during the melting-season. The initial volume and the load being at a maximum, deposition took place on the plains at an unusually rapid rate and to an unusually great extent. FORMS TAKEN BY THE DEPOSITS Deposition necessarily took place at the foot of the mountains, at the end of the mountain canyons. ‘The streams flowed on down over the deposit it had made, until it disappeared; hence the fans slope away from the mountains. While the stream was depositing and especially at times when floods were subsiding, channels were filled, distribution took place, new channels were made and filled, and water courses changed constantly. In each new distribution the channels diverged from the mountains. The resulting feature is broader near its edge than near the head; that is, it is roughly fan shaped. In the Inyo Mountains there is little precipitation, the canyons are far apart and small, and the fans are accordingly too far apart and have grown laterally too short a distance to have been joined. The result is a series of separate fans. In the Sierras, the streams issue at sufficiently small intervals and have built fans of sufficient size, so that they have coalesced to make a compound fan, pied- mont alluvial plain, or bajada. When fans first join, the compound fan is a series of fans, with low places between. If Oak Creek, Shepard Creek, and Hogback Creek are considered as types, there is a tendency for streams to 740 ARTHUR C. TROWBRIDGE leave their fans and locate their main channels in the inter-fan depressions. Deposition then takes place in the depression, build- ing it up and making a fan across the edges of two earlier fans. Presumably streams will shift back to their original positions when those positions have become lower than the site of the original depressions, through deposition in the latter. Streams then shift from fans to depressions, make fans there, shift back, build up the old fans, shift again, etc. How frequent and important this may be is not clear. If streams shift freely from higher places to depres- sions, it is surprising that fans of the bajada stand 200 ft. above the low places between them. The general relief of the bajada should be very slight. The water forming depositing streams probably does not adjust itself freely to the low places, though it is clear that it does so in many cases. The origin of the diverging channels is clear. THE TRANSPORTATION OF LARGE BOWLDERS One could hardly travel a mile on the Sierra bajada, or see the heads of the fans at the foot of the Inyo Mountains without asking how the large bowlders came to their present positions. It is essentially a problem of the means of transportation of the largest bowlders farthest from the mountains, for if they can be explained, the smaller bowlders may be considered to have been carried shorter distances by the same methods. Probably the most difficult problems are offered by the largest bowlders, such as those west of Lone Pine, 6 miles from the mountains, measuring 10X 20X 30 ft. above ground, and the one at the Cerro Gordo power shanty, larger than the one first mentioned and 13 miles from the mountains, and several others in that vicinity about as large. It is clear that these bowlders came to their present positions’ through the agency of water. Though their size suggests glaciers as the transporting agents, such an explanation is out of the ques- tion. The lower limit of glaciation is distinctly marked in the mountain canyons above. The glaciers did not descend to the plains. Nor is there anything in the fact of icebergs floating in a lake. The deposits with which the bowlders are associated are not lacustrine, and no lake existed in the valley during glacial TERRESTRIAL DEPOSITS OF OWENS VALLEY 741 times. The surface on which the bowlders lie clearly was made by running water. The bowlders are so clearly a part of the fans, and the fans are so clearly running-water deposits, that the bowlders must be considered to have been transported by running water. It is also clear that the bowlders were not transported according to the common and well-known methods of stream transportation. They have certainly not been rolled along the bottoms of streams in the usual way. They have not the rounded form characteristic of such motion. On the surface of the fans, furthermore, it is impossible for streams to have existed deep enough and strong enough to have so rolled these bowlders. Such streams would have to have a depth about equal to the diameter of the bowlders, be confined in a narrow channel, and flow with a velocity almost incon- ceivable for a stream. As the bowlders occur on the higher parts of the fans, water of sufficient depth to have carried them would have formed a sheet 20 ft. deep over the fans, and 220 ft. deep over the inter-fan depressions. Even then it would not have been con- fined to a narrow channel. Also the gradient of the fans is rela- tively low. Where the largest bowlders are, the slope is not over 6°, and west of Lone Pine not more than 3°. Bowlders much smaller than these are not now being rolled down the mountain canyons above, where the streams are sharply confined and the gradient is very high. The volume of the stream may have been sufficient to carry them im the canyons when the glaciers were there. The problem involves transportation on the fans only. They were per- haps carried to the heads of the fans by glaciers, glacial waters, and gravity. From there to their present positions, some special methods are called for. A clue to a possible manner of transportation for these bowlders is obtained from observations of run-off water at the side of a pre- viously dusty road after a heavy rain. Where the running water ‘is but a small fraction of an inch deep, pieces of rock an inch in diameter are carried down stream. The moving of the large pieces involves the transportation of a very much greater amount of fine material. The movement of the large pieces is accomplished by the removal of fine material from the area immediately down stream from, and under, the lower part of the large piece. By undercut- 742 ARTHUR C. TROWBRIDGE ting in front, and then by gravity and the push of water and sedi- ment from behind, the large piece is pulled and pushed forward into the depression prepared for it. This process takes place over and over again, the large piece being moved down the low gradient in a halting fashion. The depth of the depression into which the piece falls is never so deep as the diameter of the fragment moved. Once started in motion, the piece is sometimes carried many times its own length by its momentum, and by the force of water and gravity. The motion is usually one of sliding rather than rolling. It is conceivable that these same methods might operate on the surface of an alluvial fan, on a scale large enough to transport bowlders even 20 ft. in diameter distances of several miles. The bowlder starts from the head of the fan in company with a rela- tively large amount of fine material. The volume of the stream varies greatly from time to time, with great differences in preci- pitation in the mountains, and with daily and seasonal ranges in the rate of melting of glaciers. Material is deposited and rehandled time and time again. When the volume is great, fine material is removed from the front of the bowlder and from beneath its front edge, while other material is piled against its upper side, and the bowlder falls, or is pushed, or rolled over into the depression. As the flood subsides, the bowlder may be almost or completely buried, but the next flood uncovers it, and the process is completed. With sufficient time, sufficient variation in volume of water, and sufficient rehandling of material, huge bowlders may thus be transported great distances. Would the slope of the fans be sufficient for such transportation? In the roadside rill the piece of rock moves a distance several times its own length, while dropping less than its own diameter. Suppose the bowlder 20 ft. in diameter moves 4o ft. horizontally, with a fall of 15 ft.; this would require a gradient of 1,980 ft. per mile. If it moves 60 ft., with a 1ro-foot drop, the gradient would be 880 ft. per mile. The average slope of the bajada is about 400 ft. per mile. This requires that the bowlder west of Lone Pine, 10X 20X30 ft., move about 120 ft., or four times its own length, in dropping 10 ft., or about its own smallest diameter, if the proportions observed hold. TERRESTRIAL DEPOSITS OF OWENS VALLEY 743 This is conceived to be possible. The process is greatly aided by the momentum obtained by the bowlder when it first moves toward the depression. The depression below the bowlder would not be deep, before the crowding of the material above, and the force of water and gravity would force the bowlder intoit. The depres- sion would play out very gradually down slope, giving a constant gradient down which the bowlder could roll, slide, or creep. Obviously this process would operate to best advantage where there was the greatest volume of water, and where fluctuations of water were greatest; that is, along the main channels of the fans, and on the Sierra fans rather than at the foot of the Inyo Mountains. Bowlders are usually arranged in lines related to the channels, and they are more abundant and larger on the Sierra bajada than at the foot of the Inyos. If this method of transportation of large bowlders is not adequate, methods which are, are not known. LENS AND POCKET STRATIFICATION It was shown above (pp. 717-22 and 735-37) that the materials of the fans of the region are but crudely sorted, and that the different textural grades take the forms of lenses and pockets, rather than definite and continuous layers. No textural division was traceable more than 50 ft. in any cut, before it played out in one direction or another. The explanation of this seems clear. On the surfaces of all the fans in the region are numerous radiating channels and low ridges. In Mazourka Canyon, these channels are braided in almost all directions, though along lines trending generally down valley. These surfaces represent the last deposition on the respective fans. Beneath the present surface there must be many similar surfaces, made and buried as the fan was built. When flood waters flow over a fan, radiating channels are formed. As the flood subsides, or if the waters are overloaded otherwise, deposition takes’ place in the channels. The channels are filled with whatever grade of material the stream finds itself unable to carry, and the stream is forced over the side. It then makes a new channel, fills it, and overflows to repeat the process. The fan grows by the addition of long narrow strips of material, sorted 744 ARTHUR C. TROWBRIDGE roughly into different textural divisions. These strips diverge from the axis of the fan. No straight section can be cut in such a deposit without cutting the filled channels. If the cut is longitudinal, practically all the channels will be cut obliquely and at low angles; a few might be cut at right angles. If the fan is dissected by streams cutting down in the old channels, buried channels will still be cut obliquely, as distribution does not take place along lines exactly parallel with previous distributaries. The filling of a buried channel, when cut along a straight line oblique to the original channel, is exposed as a lens whose length and degree of pinching out depends primarily on the obliquity of the line of cut. A channel filled, buried, and then cut at right angles reveals itself as a pocket in section, the size and shape of which depends on the size and shape of the channel. Continuous layers, uniformly thick can occur only where the depositing dis- tributary was long, straight, and contained uniform material, and where the filling was cut along a straight line exactly parallel to itself. Obviously where exposures are along longitudinal cuts, the result is many lenses, a few pockets, and practically no con- tinuous layers. That this is the correct explanation of the lenses and pockets of the fans of the region is shown by a correspondence in size between lenses and present surficial channels. On the Sierra bajada, the channels are about 8-10 ft. deep on the average, and the lenses and pockets are about 8-10 ft. thick at their thickest parts. The present flood surface in Mazourka Canyon has a relief of about a foot; the lenses in the older alluvium near by have just about that thickness. It is understood that any deposit from distributing or anasta- mosing streams will reveal a lens or pocket structure in straight cuts. The principle probably applies to all alluvial fans, pied- mont alluvial plains, flood-plain deposits, glacial valley trains and outwash plains, and deposits on tidal deltas. THE DISSECTION OF THE SIERRA BAJADA A variety of events might bear causal relations to the dissection of alluvial fans. Among them are changes in climate, uplift of the TERRESTRIAL DEPOSITS OF OWENS VALLEY 745 fans, down-warping of their surroundings, etc. Such events and resulting processes may be complex. The cause of dissection in this case, however, seems to be the cessation of glaciation, and the process seems simple. All the material seen in the bajada, even to the bottoms of the canyons, shows evidence of glacial wear. Before the mountain canyons were glaciated, the fans must have been smaller and lower than now by an amount at least equal to the depths of the canyons. Glaciers were formed, which carved great amounts of material from the heads of the canyons, carried it to their lower ends, and, melting, supplied great quantities of débris-laden waters to flow out over the fan. The fans grew rapidly and became large, out of all pro- portion to those at the foot of the Inyo Mountains, which were not affected by glaciers. When the glaciers in the mountains had melted away, these enlarged fans were dissected, for the same reason that a valley train is trenched. The streams now reach the fans with less material than they carried when the glaciers existed, and are able to erode material from the fan. The matter may be looked at in another way. The pre-glacial fans, being lower than the present ones, had lower gradients and made a sharper break in gradient at the foot of the mountains, and deposition progressed. Now that the fans are higher, their gradi- ents are steeper, and the break in gradient at the foot of the moun- tains is not so great, and the streams flow out over the fans with their velocities less checked than formerly. This means at least that there will be less deposition on the present fans, and, taken with the fact that the streams have less load, plays a part in the erosion of the fans. Presumably the canyons will be deepened almost or quite to the bottom of the glacial material. This depth has nowhere been reached as yet. DEPOSITS OF TWO AGES AT THE FOOT OF THE INYO MOUNTAINS The older deposit at the foot of the Inyo Mountains is here considered to be a lacustrine deposit, coinciding in age with the lake beds in Waucobi Canyon and those near Haiwee. Ii this be 746 ARTHUR C. TROWBRIDGE correct, the explanation of the dissection of these deposits and the later deposition of fans is not complex. A lake existed in Owens Vailey, probably in Pliocene times, and deposits were laid down in it on the flanks of the mountains. The lake was drained or dried up, and the mountains were probably uplifted. Dissection of the deposits thus exposed and uplifted followed. After erosion had removed a large part of the lacustrine deposit, deposition began at the foot of the mountains, and the present fans have been built up among the remnants of the old lacustrine deposits. CRITERIA FOR DISTINGUISHING ALLUVIAL FAN MATERIALS In conclusion we may bring together the distinguishing features of the materials of fans, as seen in this region. The region affords especially good facilities for the drawing of such conclusions, as it contains both running-water and standing-water deposits of similar ages. Deposits on alluvial fans may be distinguished from those in still water, either lacustrine or marine, as follows: 1. In alluvial fans, coarse material has a wide distribution as against confinement to a narrow zone near shore in standing-water deposits. 2. Textural range in single exposures is large in fan materials. 3. Fan materials are not in general so well sorted as deposits in standing water. 4. The beds and surfaces of fans are likely to have slopes of 6-18°, as against o-3° in standing-water deposits. 5. Fan materials are likely to have fewer and different fossils than deposits in standing water. 6. Fan material has a lens and pocket stratification, as against a sorting into more or less uniformly thick horizontal layers, as in lakes or seas. . 7. Huge bowlders widely distributed vertically and horizontally in a deposit indicate that it was deposited by running water, and with a large proportion of fine material; that is, they indicate that the material is part of an alluvial fan deposit, except in cases where glaciers have affected it, or where standing waters could have TERRESTRIAL DEPOSITS OF OWENS VALLEY 747 recelved icebergs, or where basal conglomerates are formed near shore. 8. Theoretically, fan material will be more compact at first than still-water deposits of the same textural grade, as each particle drops to the bottom with greater force and the film of water around each particle is not so thick. After water is drained from both deposits, still-water deposits are likely to be more cracked than fan materials, because they contract more. After cementation, still-water deposits, say of clay, will have more veinlets than fan materials of the same texture. ON CORUNDUM-SYENITE (URALOSE) FROM MONTANA AUSTIN F. ROGERS Leland Stanford Junior University Specimens of a corundum-bearing rock from the property of the Bozeman Corundum Company, fourteen miles southwest of Boze- man, Gallatin County, Mont., were obtained for Stanford Uni- versity by Mr. R. M. Wilke of Palo Alto, Cal. No information concerning the country rock could be obtained except the state- ments of Pratt in his monograph on corundum:’ ‘‘The corundum seams vary from a few inches to three feet in thickness. . . . . Boze- man: Fourteen miles southwest of this town corundum is found in syenite.”’ From this it would seem that the country rock as a whole is a syenite with bands or seams of the corundum rock. These bands, the writer will show, are corundum-syenite. The rarity of this type of igneous rock accounts for the present paper. Corundum-syenites have been described only from the Urals,’ from eastern Ontario,* and from the Coimbatore district, India.4 The corundum-syenite is a medium to coarse-grained, gray- mottled, more or less banded rock, the banding due principally to the fact that the biotite flakes are mostly in parallel position, though the other minerals are occasionally in rough, parallel position. The gneissoid corundum-syenite, as it may be characterized, is composed of microcline-perthite, biotite, and corundum with sub- ordinate sillimanite, muscovite, zircon, and baddeleyite. The feldspar is for the most part a perthitic intergrowth of microcline and albite, though one slide shows plagioclase, orthoclase, and microcline without any perthite. On a section of the microperthite parallel to {oor} the microcline has an extinction angle of 113°, and the albite, one of 43°. In this section the albite shows only very faint albite twinning. On a section parallel to joro} the microcline has an extinction of —3° and the albite, one of +20°. t Bull. No. 269, U.S.G.S:, 133, 144 (1906). 2 Morozewicz, Min. u. petr. Mitth., XVIII, 217 (1808). 3 Miller, Rept. Bureau of Mines, Toronto, Canada, VIII, Part 8, 210 (1899). 4 Holland, Mem. Geol. Surv. of India, XXX, Part 3, 169 (1901). 748 ON CORUNDUM-SYENITE FROM MONTANA 749 The feldspar crystals are sometimes arranged in rough augen. The corundum occurs in grayish-blue crystals with an average size of 5 mm. and a maximum size of about 2cm. The corundum crystals are tabular or prismatic in habit with the common forms: c jooort, a {r120t, r {rorrt, m {2243t, and @ {8-8-16-3t. The most fre- quent combination is acrn. The corundum is often surrounded by a zone of feldspar, which is nearly free from biotite. A fibrous mineral occasionally observed proves to be sillimanite as tested in fragments. Muscovite is often observed in thin, cleavable flakes. It does not appear to be an alteration of the corundum. Thin sections show a very small amount of zircon in minute prismatic crystals. The baddeleyite is a black, submetallic mineral which is usually found between the corundum and the feldspar. It occurs in rounded blebs and in prismatic crystals not over 3 mm. in size and usually only about Imm. in greatest dimension. The baddeleyite will be described by the author in a forthcoming number of the American Journal of Science. Baddeleyite rather than zircon forms in this type of rock prob- ably on account of the low silica percentage. A rock sample weighing 243.6 grams was crushed, and after sizing, the constituents were separated by means of Thoulét solution. It was found that good separations could be made by panning with the Thoulét solution. The following shows the amounts of the various minerals and also the percentages by weight, assuming the loss to be equally distributed among the minerals: Grams Percentage elals inate eerste scnc tecr ses £36010 O2n7 @orundumes vere os 67.53 30.9 BLO RICE MY usenys tee sess T2540 5S iBaddelevite =n ssa oat O25 WOSS Hee tess eaceaie sue: 26.0 MOtale meres ect ee eee: QA 30 909.9 We. may assume the feldspar to be a eutectic of albite and orthoclase. Vogt gives the eutectic ratio for these two minerals as ab=58 per cent, or=42 per cent. We then have 36.4 per cent albite and 26.3 percent orthoclase. The biotite is the only mineral 750° ; AUSTIN F. ROGERS which does not have a fixed chemical composition. We may, however, assume the following percentages which are average values for biotite SiO,=37 per cent, Al,O,=16 per.cent, Fe,O,=6 per cent, FeO=15 per cent, MgO=12 per cent, K,0=10 per cent, H,O=4 per cent. The recalculated chemical analysis of the rock given is as follows: Orthoclase| Albite |Corundum| Biotite Baddeleyite Total SiQaee ney ere alee 17.0 25.0 ee 2aT 44.1 NEO yee sie area ern oe 4.8 Giga 30.9 0.9 43-7 Hes © yeep pare tera pee Sal. Sa eras O.3 0.3 Tie Oar patna neta 0.9 0.9 Vig © aes Per eG tte gee nen On7 Oni (OE O Mirra ha Wal anne nea tiara ae Bisa Mae Seas Na,0. SAG RRA IN Rea Gees 4.3 ee 4.3 KO ieee se aie cate 4.4 Me 0.6 5.0 AO Ne rs ei meant Pah: Sat ee 0.2 sieee 0.2 LO A Ge Rear TOI ae ae se nee raat Dee 0.5 0.5 26.2 36.4 30.9 Boo) 0.5 99.7 Chemically, this is a peculiar rock on account of the high alumina and low silica content. It may be called a corundum-syenite. In order to place this rock in the new quantitative classification it is necessary to convert the percentage compositions of the oxids into percentages of the standard minerals, which in this case are nearly the same as the actual minerals. In other words, the mode and the norm agree closely, biotite being practically the only critical mineral. The calculated norm of the rock is shown in table on p. 751. All the potash goes into the orthoclase molecule, all the ferric iron and an equivalent amount of the ferrous iron go to make the magnetite molecule. The remaining ferrous oxid goes with all the magnesia to form the hypersthene molecule which requires an equivalent amount of silica. The silica remaining after deducting that required for the orthoclase, hypersthene, and zircon would naturally go into the albite molecule, but it is found that there is too much soda for this amount of silica so that the silica and soda must be distributed between the albite and nephelite according to the equations :* x+ y=molecules of Na.O 6x-+ 2y=available SiO. * Quant. Class. of Igneous Rocks, 194 (1903). 2 ON CORUNDUM-SYENITE FROM MONTANA 751 in which « is the albite molecule and y the nephelite molecule. The remaining alumina goes into the corundum molecule. Or Mt Hy Z Ab Ne (e SiO; =44.1 19.1 ae Tey, 0.2 22). TO 5G Al,O; =43.7 a7) Hike side Stelle 6.3 0.8 212 Fe,0;= 0.3 eat 0.3 fee Des Say SHG Bane FeO = 0.9 one O.1 0.8 oy if MgO = 0.7 at as 54) - ae Na,O= 4.3 eke yee a 3.8 Ons K,0 = 5.0 5.0 ne oe ZrO, = 0.5 al 0.5 29.5 0.4 2 0.7 32.2 2.3 Be2 Orthoclase = 20.5 F Albite = 32.2 Nephelite Si PINE nal 6, , Corundum =e aT 2 Salic Zircon =a ong. 12, Magnetite = Onan Vi Hypersthene = 3.2 P Senne Total = 90.5 The classification of the rock according to the new quantitative system is as follows: = aoe sa Class I, Persalone ag = a ; she : Subclass IT, Persalone = =359< : Order 5, Perfelic aes | = “2 =e Rang 1, Peralkalic aa =e — Subrang 3, Sodipotassic The magmatic name of this subrang is uralose and the magmatic symbol I’, 5, 1, 3. Only two rocks have previously been assigned? to uralose, a corundum-syenite and a corundum-pegmatite, both from the Urals. t Washington, Professional Paper, U.S.G.S., No. 14, 217 (1903). A DRAWING-BOARD WITH REVOLVING DISK FOR STEREOGRAPHIC PROJECTION ALBERT JOHANNSEN The University of Chicago Professor Wiilfing recently showed the writer a wall chart for stereographic projection which he has since described.t_ It consists of a ground glass plate back of which is pivoted a 70 cm. Wulff net which is made of pasteboard and projects beyond the glass cover so that it may be turned to any desired position. The advantages of using the Wiilfing chart were so apparent that the writer has constructed a drawing-board, for the individual use of students, on a somewhat similar plan but combining with the Wulff net a half- net with the north pole at the center. The construction is simple and the board inexpensive. In making stereographic projections by ordinary methods, one must either work out his own dimensions, use a Penfield protractor, or a net like that of Fedorow or of Wulff. Transparent nets are an improvement over Penfield’s method, although one must always carefully center the net for each measurement. With the drawing- board here described, the net is revolved instead of the paper and no centering is necessary. The board was constructed from an ordinary drawing-board, 332435 cm. in size. It was placed on a lathe and a recess, 22 cm. in diameter (D-J in the illustration), was turned out halfway through the board which was 2 cm. thick. ASA Requisite Conditions for the Formation of Ice Ramparts. By William H. Hobbs 157 Restoration of Seymouria Banlorensie Broil an Ameren @orylveane By S. W. Williston . : S239 Reviews : ey 180, a6. ae 460, 76. 661, 756 Rich, John Lyon. leravel as a s Rectaicat Rock / » | AOD Richardson, G. B. Reconnaissance of the Book Clifis Coal Field between Grand River, Colorado, and Sunnyside, Utah. Review by Re. : 95 Ries, Heinrich. Economic Theory mel Sneciall Referenee b ie United States. Review by W. H. E. 90 Ripples of the Bedford and Berea Formations of Central and Southern Ohio, with Notes on the Paleogeography of that Epoch, The. By Jesse E. Hyde ; 5 ‘ : : 5) (G7 Roches sodiques du désert praGe: Les. By J. Couyat. Review by BoG, Calkins” 463 Rogers, Austin F. On Commins vente (Grates) fon Monten 748 Rosenbusch, H. Elemente der Gesteinslehre. Review by Edward B. Mathews 284 770 INDEX TO VOLUME XIX Rounding of Sand Grains, Factors Influencing the. By Victor Ziegler Rowe, Jesse Perry. Practical Mineralogy Simplified. For Mining Students, Miners, and Prospectors. Review by W. H. E. Schaller, W. T. Axinit von Californien. Author’s abstract Schmerber, H. La sécurité dans les mines. Review by R. T. C. Sécurité dans les mines, La. By H. Schmerber. Review by R. T. C. Seeding of Worlds, The. Editorial by T. C. C. Seismic History of the Southern Andes (Historia Sismica ae 1s Andes Meridionales.) Por el Conde Fernando de Montessus de Ballore. Review by W. H. H. : : Serpentines du Krebet-Salatim (Oural de Nord), ‘Sur ies By L. Duparc and M. Wunder. Review by F. C. Calkins . Shaw, Eugene Wesley. High Terraces and Abandoned Valleys in Western Pennsylvania Preliminary Statement concerning a New System of @uater: nary Lakes in the Mississippi Basin . § Sherzer, W. H., and A. W. Grabau. The Monroe Hore tion of South: ern Michigan and Adjoining Regions. Review by S. W. : Shimer, Hervey Woodburn, and Amadeus W. Grabau. North Ameri- can Index Fossils: Invertebrates. Review by S. W. Simple Method for Photographing Large Preparations in Polarized Light, A. By Carl Benedicks and Olof Tenow. Review by W. T. Schaller Skeats, Ernest W. The Volcanic Rocks of Victoria. Review by Albert . Johannsen ; Skiddaw Granite and Its Me tamoriicn The. By R. H. Rastall. Review by Albert Johannsen Slates of Arkansas, The. By A. H. Batdue: vith a Bibkoorphy of the Geology of Arkansas by J. C. Branner. Review by E. R. L. Smith, G. F. Herbert. A Camera-lucida Attachment for the Gonio- meter. Review by W. T. Schaller Southerly Extension of the Onondaga Sea in the Allesheny Report The. By E. M. Kindle ; Speculations regarding the Genego of ile Diamond: By Orie A Derby : i : : ; 3 : : ‘ Spencer, J. W. The Focus of Postglacial Uplift North of the Great Lakes : Steidtmann, Edward. The Evolution of Limestone and Delonte: I. The Evolution of Limestone and Dolomite. II. . ’ Structure and Composition of the Chandakapur Meteoric Stone, On the. By H. L. Bowman and H. E. Clarke. Review by W. T. Schaller . : : Sur l’issite, une nouvelle roche omen dine 1 dunes By L. Duparce and G. Pamphil. Review by F. C. Calkins PAGE 645 668 188 Q2 92 175 757 464 140 481 664 47° 181 466 187 IQI 188 97 627 57 323 392 I8t 464 INDEX TO VOLUME XIX Syllabus of a Course of Lectures on Economic Geology. By John C. Branner. Review by W. H. E. Tables for the Determination of Common Rocks. By Oliver Bowles. Review by Albert Johannsen Tectonic Lines of the Northern Part of the North Ameren Cordillers The. By W. Joerg. Review by R. T. C. : : Terminal Moraine of the Puget Sound Glacier, The. By J. Harlen Bretz Terrains primaires du Morvan et de la Loire, Les. By Albert Michel- Lévy. Review by F. C. Calkins : : ; : Terrestrial Deposits of Owens Valley, C nliferniat By A. C. Trowbridge Testing for Metallurgical Processes. By James A. Barr. Review by Wiese: : Theory of Isostasy, The. By Hon Degas : Tilton, John Littlefield. The Pleistocene Deposits in Warten County) Iowa. Review by R. T. C. : : : : : : Topologie. Etude du terrain. Par le Général Berthaut. Review by Re G: : Traverse through the Souther [Pare of the NGrihwrest Terstones from La Seul to Cat Lake in 1902, Report on. By Alfred G. Wilson. Review by W. C. C. Trowbridge, A. C. The Terrestrial Menosits of Owens Valley: G@alvarai Ulrich, E. O., and H. P. Cushing. Age and Relations of the Little Falls Dolomite (Calciferous) of the Mohawk Valley. Review by E. R.L. Umpleby, Joseph B. Geology and Ore Deposits of Republic Mining District. Review by R. T. C. Un mésure du laminage des Gaments (calenites et aus) par asl a leurs cristaux clastiques de tourmaline, Sur. By F. Grandjean. Review by F. C. Calkins . Unconformity between the Bedford and Been Homnacions a Norton Ohio, The. By Wilbur Greeley Burroughs : Use of “‘Ophitic”’ and Related Terms in Petrography. Be Alewandes N. Winchell. Review by Albert Johannsen Valley Filling by Intermittent Streams. By A. E. Parkins . Van Horn, F. B. ‘‘The Production of Phosphate Rock in rogro.”’ Review by A. D. B. ; : ; : : f Variations of Glaciers, The. XV. By Harry Fielding Reid Variations of Glaciers, The. XVI. By Harry Fielding Reid . ; Vermont State Geologist, 1909-1910, Report of the. By G. H. Perkins and Others. Review by E. R. L. Verwendung einer Glashalbkugel zu quantitativen fonueeten Unter suchungen am Polarisationsmikroskope, Ueber die. By Wladimir Arschinow. Review by Albert Johannsen . 666 706 667 772 ' INDEX TO VOLUME XIX Volcanic Rocks of Victoria, The. By Ernest W. Skeats. Review by Albert Johannsen von Huene, F. Ueber Ery amoaucties, Were der neuen Renal ordnung Pelycosimia. Review by S. W. Williston Walcott, Charles D. Olenellus and Other Genera of the Mesonacidae. Review by S. W. 2 “Middle Cambrian ivieroetarantne a ‘Middle C Suan Holothus rians and Medusae,”’ ““ Middle Cambrian Annelids.” Review by S. W. Watson, Thomas Leonard. Annual Report on the Mineral Production of Virginia during the Calendar Year 1to08. Review by E. R. L. Intermediate (Quartz Monzonitic) Character of the Central and Southern Appalachian Granites. Review by Albert Johann- sen : : : : : : 5 , 3 Weller, Stuart. Genera of Mississippian Loop-bearing Brachiopoda . Weltkarten-Konferenz in London im November, 1909, Die. By Albrecht Penck. Review by R. T. C. : : : . White, David, C. H. Gordon, and George H. Gira The Wichita Formation of Northern Texas : ; : Wichita Formation of Northern Texas, The. By C€. H. Gordon, George H. Girty, and David White Williston, S. W. Restoration of Seymouria Baylorenss Broil an American Cotylosaur The Wing-Finger of Pterodactys, oan Rectarrion of Ngee saurus Wilson, Alfred G. “Report on Stray erse ironed rhe! Southern Part af the Northwest Territories from Lac Seul to Cat Lake in 1qo2. Review by H. C. C.. : : ‘ Wilson, A. W. G. Geology of the Ninieon Sct Ontario. * Review by WavAG an : : ; . ; : : : Winchell, Alexander N. Use of “Ophitic’’ and Related Terms in Petrography. Review by Albert Johannsen Winchell, Horace V. Prospecting in the North. Review ie W. H. E. Wing-Finger of Pterodactyls, with Restoration of Nyctosaurus, The. By S. W. Williston . Woodward, H. P. The Geology and Ore Meno a the West Pilbara Goldfield. Review by A. D. B. ; Wunder, M., and L. Duparc. Sur les Serpentines an ieee ante (Oural du Nord). Review by F. C. Calkins Yorkshire Type Ammonites. Part III. Edited by S. S. Buckman. Review by S. W. Ziegler, Victor. Factors Influencing the Rounding of Sand Grains IIo Nu eae ie af 7 sue TN be tert Nh w\ No 4! & ! Fit oe se vase: = 45a > 15 : = os ta} me 8 Tens se Wife. > ay ae Aa te PLT PAAR ~ Sam WD ar” a ; J TRAE TER -- 2 ae = Se 2 a6 Bs i usd ~ ‘ § A. A yy NA, ws ™ ayy ag aa et 9 Pap ~ AOA Ne ae 3 : ht ab a Go Ve \ >> 1s ye ae bbGateats ARS RBAG: PPLE Do OY eee Oe Ay, War anther’ mee Vt LN Ne gee ™ Senha s Aang g ik awe bp) |) We ie Br ere" nary We cate cae” ‘ee maaagl Mea AR, OO ry 2a Maaaa Oneal } ye. 3 we "22 @ ’ Lal. ith 3 nal al WAMa)) (lst tna ann “ , : g A adihay arse 1990 8) Abe i 1 \eleeiel Lm ayacacen cPrese Rae Mitae,, ee ' gegen am saath df Eh ielelh. Pal hertalta! AB t ganngppent {hy apaAbee Oa i ae A ag eA Ry Q Mal Nis 2 pak ; vi qakil La AABN C PRG an niet ae Aa, a a g a we : Muon eet Pare eUCULL LLL Staccsonansoe@) cocamecin NAMCARAR, RA Wiey sob A) by ee matt i’ aa: HH elt ff . am cee mae Mat. Aaa, a ny AL ald “ae ¥ MapeattTee ar qe : “smen %- rains | Ba ghae eee paper r~ “s agit. Bema eat Aye one OMA Me OMe, ge te Witaagays. a, EHH Udell) TLul _ 3 ie : AN . Wrest on. : Bahan’ PAM i taag Ana haa “Aa ‘aes : a dial + na Wi Ra ees anh ANa, ppianaant BAe: Batsata tase Grae nthe Nay, tecaelsll} PPARRRe ait LT get oAAsa 11] mes eas ep ane z dacty Tee. aeshnn Aga Mi phes. ae ey pinnate TA “WE tram (111777 : ‘Ag nual! faoaaae ATE | at un bbb ALAC , hes | x a P nuetienanens” fen 7 rier Pomaca. ES Mad) AL eld ot NX j9beP ame “~ yi ane @ aa . . Renee OARS as Seate anton enctciiaaalt. fastest | AORNES » * a oe . : & b Sea ee Lt B Ee a { Jal iy) parry Her eh bhpp pny eee EEE E A Ber Plt a ; Seema ac: ean, Ade Wipes, 87% A> 5. . CEE > ¥S an UH ud Manne ~ Ma@creeanacee: SAP Aevems, Pea 5 AP hero co \ a ara. _ eee | | gels On | cs Pe pe Anbabhi Bef Op raran aAA,: A ue e Ene Hida | add. fe a ig ee Be Ai pA ree Adi JAS ERAT Vea a Lideeaey toit tet Briaa Ce wine! a poAare rh: Han Sone Nate NM 5 4 eae Cale tag auenegin pembann a? aT ope guaian PTT] | eR. NapsoAhrs AR) ar ey | Pe nteintstint delet Ye ania | coo) ees ‘s, “4 Bane 4 a es | | A Ald Willi di] Ba aan eT PHATE eile A Ae ia tre anne MiNua ia splay | : TH | a Sareer Ra, eee bs 82. § \p yy : : “NAR ny valle a Seaeess euegna.* 1 peas. eaimlah yp nee reals ooh Naan aay . araneiRanmeee,, ARG aaa..@ Dp eddie, an Poe cee Pp Abaanas am eae A will paetirins oman Aue Besta rinty iL AA TTY o gyfa SL eta ang | TT TT plist er ppd CETL ta aaa ta ange aati aay “8 reeae E Diamar lao f TN shan Sal | las rYY Ti 3 TH A AA! ae eeanAl Pree eeren NY eval tad hil dye AQ OA SAAT TH Tee | # ane ET Aly Oat Maat Seggye VON GORY =~ SO~*A)/AisnaAAAal ty Marvy v VV WL buat Bini me Cr : epee ane LTP wugsahen a J ie ‘ 4 4 } b laa8 spannaa AAA A AY, Pee ‘aha fin : : Nb RS Tt nr tAAAR OLA apn aa, ee Thwus 2 Praw>» A MAis _ a = ee a Nia brea Peery Sq@e@ _sseei1kh.d Maateguaa.. as7 aval paar AX, Ap ah! ANY eer CRPT Oy rainaiarcer ame FLAP. NL Aa, mM 3 9088 01367 0088 HPS NY LUE AG yg Na ty m4) Ww 7 o < rv i] = 2 fe} = =) E Ee o = z < z 1 O 7) = E = 7) OR ell Vo Wier eee Ake rene gee AY . rc esnP \ aad? tee AL dems She Va ay ieee ap Ce! Nae as 9 a y La hy PT men Ga at Se ge a 2 “igi ‘ Pp a c od a re aan” ia ‘ am Me 4. [ Ling abet * SE nl anes Qo ena cen, : al ole) iL a Ane ales ses 6 } sh \ det | a wok ab bene 4st lap esar ( a in Ep semana eng a Ny ee NS! ae