ee oat Are ee cs ee fee ee tae A ane . Cree bed dag re apa he ’ babtoe ie “¢ Hey e , 4 ey dew bes Cae eer) a cee re ae Oar ree een et Seca an eeerTy stew to er erent . ear esuane Seer Seer eee ee ee rome Per ee ne Hn ive a eee pen Oo need eh gm Oa . regen ea ea Frere mec eS ee ere Caner eee ee ee oe Panera an atte paw ee ee tweed Parent entrees Pah ieee Parner ar) Pe tenn ee peewee ” . ‘ Arrieta arc an nr ec] Perea oar a Serine ace Seen ie edie swe Fe ae etter ae 7 Ce de get e re Ae A NEN ge ate oe np ad sped iil Wain, @riovebe 016 OFFICERS, 1911 President—Franz Boas, “olumbia University | Vice-Presidents—Gronrcr F. Kunz, Frepertc A. Lucas, . R. S. WoopwortH, WILLIAM CAMPBELL _—- Recording Secretary—EpMunD Otis Hovey, American Museum — seo Corresponding Secretary—Henry HE. CRAMPTON, American ‘SMoeum ‘ ——s- Preasurer-—EMERSON McMituin, 40 Wall Street fie et . ss Librarian Rate W. Towne, American Museum a ANA SECTION OF GEOLOGY AND MINERALOGY Chairman—GeorceE F. Kunz, 401 Fifth Avenue Secretary—Cuar.zs P. Berkey, Columbia University “ees _ SECTION OF BIOLOGY — - Chatrman—Freveric A. Lucas, American Museum. » Secretary—L. Houssakor, American Museum SECTION OF ASTRONOMY, PHYSICS AND CHEMISTRY Chairman—WiLt1am Campsett, Columbia University Secretary—Epwarp J. THATCHER, Teachers College SECTION OF ANTHROPOLOGY AND PSYCHOLOGY ? Chairman—R. S. WoopwortH, Columbia University Secretary—F RepERIc LyMANn WELLS, Columbia University s eet The sessions of the Academy are held on Monday evening at 8: 18 | oa o’clock from October to. May, inclusive, at the American Museum ; Natural History, 77th Street and Central Park, West. | [Annats N. Y. Acap. Scr, Vol. XXII, pp. 1-8, Pl. I. 3 April, 1912.] ON SOME INVERTEBRATE FOSSILS FROM THE LYKINS FORMATION OF EASTERN COLORADO? By Grorce H. Girty (Read by title before the Academy, 5 February, 1912) The fossils which form the subject of the following account were col- lected by Mr. Roy M. Butters and kindly placed by him in my hands for study. They were obtained in the Lykins formation of Colorado and represent a horizon in the Paleozoic higher than any at which fossils have heretofore been found along the eastern flank of the Front Range. A detailed account of the stratigraphic relations and correlation of the Lykins formation has been prepared and will shortly be published by Mr. Butters. To a manuscript of this report, which I have been permitted to read, I largely owe the following data which seemed essential to the understanding of this limited but interesting fauna. The “Red Beds” of the Front Range in Colorado have been variously classified and named. Their nomenclature and synonymy is, therefore, rather complicated, but as a general statement, it may be said that the Wyoming formation of Emmons has been divided into three formations, of which the Lykins is the highest. Below the Lykins, there occurs a series of strata (the lower Wyoming) which are now known as the Foun- tain and Lyons formations, while above the Lykins is the Morrison for- mation. The Lykins, therefore, belongs in the upper “Red Beds” of this area. The Fountain has furnished more or less conclusive paleontologic evidence of Pennsylvanian, or at least of upper Carboniferous, age, while the Morrison has long been known to be Mesozoic. The Lykins forma- tion, from which fossils have not hitherto been known, has usually been assigned to the Triassic, but the evidence herewith presented seems to show conclusively that the formation, or at all events that portion of it from which the fossils were obtained, is Paleozoic. Provisionally, I am assigning the Lykins fauna to the Permian, though more on account of its position at the top of the Paleozoic section than on account of any very close resemblance either to the Permian of Russia, the more or less doubtful Permian of our Western States or the Permian as distinguished from the Pennsylvanian of the Mississippi Valley. The only fauna in 2 Published by permission of the Director of the U. S. Geological Survey. 9 ANNALS NEW YORK ACADEMY OF SCIENCES Colorado which conspicuously resembles the Lykins occurs in the Rice formation of the San Juan region, and the Lykins and Rico are tenta- tively placed in correlation, in spite of the fact that many of the Rico forms are as yet not known in the Lykins and some of the characteristic Lykins forms are not known in the Rico. The character of the Lykins formation, as would naturally be supposed, changes from point to point. A well-exposed section at Heygood Canyon which is fairly representative consists, according to Mr. Butters, of sand- stone and shale with some beds of sandy limestone. The sandstones are pink or red and mostly soft, while the shales are red. The thickness of the Lykins at this point is 816 feet. A little south of Heygood Canyon on the north slope of Table Moun- _ tain were made three collections of fossils (lots 3264, 3265 and 3266). They occur about 300 feet above the base of the formation and about 25 feet above a 25-foot bed of gypsum. They contain the same species, viz: Myalina perattenuata, Myalina wyomingensis, Alula squamulifera and Murchisonia buttersi. The matrix is a compound of fine sand and clay with more or less lime, the color being a rather light brownish gray. Another collection was made near Stout, Colorado, from a red calcare- ous sandstone (lot 3262). Only two species are present, Myalina wyo- mingensis and M. perattenuata. The horizon is about 30 feet above the “crossbedded sandstone” * in the basal member of a local group of cal- careous strata 50 to 60 feet thick consisting of thin limestones, shales and sandstones. The last collection (lot 3263) was made at Perry Park, one-fourth of a mile south of the lake, in a band about 6 inches thick near the base of the “crinkled sandstone.” * The rock is whitish in color and very fine in texture, apparently a mixture of lime and clay and sand. In this collec- tion, I identify Myalina wyomingensis, Myalina perattenuata, Alula squa- mulifera, Alula gilberti? and Pleurophorus sp. — Lot 3263 occurs at a higher horizon than 3262 and recalls, especially by the peculiar and characteristic species Alula squamulifera, the fauna of the northern group of collections (lots 3264, 3265, 3266), from which, however, it is separated geographically by atong distance. The fauna of the Lykins formation is, so far as known, very limited, consisting of only six species, and in addition to describing the two spe- cies, which are new, it has seemed desirable to remark briefly upon the other forms. * Colorado Geol. Survey, First Report, pp. 168-9. 1909. 3U. 8. Geol. Survey, Bull. 265, p. 25. 1905. J GIRTY, INVERTEBRATE FOSSILS FROM COLORADO 3 Alula gen. nov. Shell soleniform, very transverse. Beak strongly anterior, but not terminal. Upper and lower margins contracting posteriorly. Posterior outline obliquely truncated. Umbonal ridge angular, with a second plication on the post-car- dinal slope. Surface elegantly sculptured by fine radial costz surmounted by fine, closely arranged scales or interrupted concentric lamelle. Post-cardinal slope without radial costs, but with similar squamose ornamentation. On the interior, the right valve has a single long, plate-like anterior tooth, a posterior tooth of similar character, with possibly a small rounded tooth at the umbo. Corresponding structures appear to be developed in the left valve. A large anterior scar is indicated. Type, Alula squamulifera. In a general way, these shells suggest a very transverse type of Paral- lelodon, and I believe that they belong to the same family, though clearly representing a distinct genus. In configuration, they differ from Paral- lelodon in contracting posteriorly instead of anteriorly ; in not having the angular and projecting anterior extremity, and in possessing a second plication between the umbonal ridge and the cardinal border. Interiorly, they differ in having a single posterior tooth; in having a linear anterior tooth similar to the posterior one, and probably in lacking the flexuous dental arrangement at the umbones. A certain resemblance to some spe- cies of Plewrophorus exists in the angular umbonal ridge and the post- umbonal fold, but the Plewrophori are not radially striated, and, while they have a similar posterior tooth, the remainder of the dentition is quite different. It is not certain that any American species other than the type can be referred to this genus, but, if so, they are probably to be found among the forms which I provisionally included under Pleuwrophorella. A re- semblance to Alula squamulifera, more or less marked, is found in P. gewmta, P. gilberti, P. lanceolata and Allerisma (Pleurophorella?) re- flecum. Of these, the most similar is P. gilberti. Typical Pleurophorella, as exemplified by P. papillosa, is probably safely distinct, although its internal characters are as yet unknown, because of the deeply introverted lunule and the escutcheon, both characters apparently wanting to Alula, and because of the absence of radiating costz in the sculpture, although a somewhat similar feature exists in the characteristic papille, which show a tendency to radial arrangement. There is, however, scarcely any com- parison in this item of sculpture. As a provisional arrangement, I am removing to the present genus A. gilberti, A. geinitzi and A.? lanceolata. Allerisma reflecum, in spite of a general resemblance to this series of forms, probably has quite different, although indeterminate, relations, distinctly not with typical Allerisma. As a result of a better knowledge 4 ANNALS NEW YORK ACADEMY OF SCIENCES of Allerisma costatum and a renewed consideration of its characters, I believe that my original estimate of its relationship to Pleurophorella papillosa was erroneous. The strong concentric plications which stop abruptly at the umbonal ridge indicate a different type of shell. It is somewhat doubtful whether a papillose surface is a real character of A. costata, which is apparently a much flatter shell, with thin test and pos- sibly different structure in the lunule and escutcheon. It clearly does not belong with Alula, however, but has all the superficial characters of typical Sanguinolites. Alula squamulifera Shell rather small, very transverse. Width about 3.5 times the greatest height. Greatest height near the anterior end at the umbo, which is situated about one-sixth of the entire width back from the anterior margin. Ventral border gently .convex in the anterior half, nearly straight or faintly concave posteriorly. Dorsal outline gently concave or nearly straight, contracting pos- teriorly with the ventral. Posterior outline oblique and more or less sharply truncate. Anterior outline straight above, strongly rounding below. Convexity usually rather high, though variable, sometimes rather tumid in the umbonal region. Beaks large, prominent and incurved, situated relatively close to the anterior extremity. Umbonal ridge prominent, usually strongly angular toward the posterior end, more obscure in the umbonal region. The post-cardinal slope is divided by a second plication about intermediate between the umbonal ridge and the cardinal line, above which the narrow strip of shell is nearly horizontal. Surface marked by fine, radiating ribs which are confined fe the portion of the shell below and in front of the umbonal ridge. This Svu.pr.ure might better be described as made by narrow stris, the elevations between which are covered with closely arranged, fine, flat scales, which recur at equal intervals on adjacent ribs and have also the appearance of interrupted con- centric lamelle. The ribs are more than radiating rows of scales, since the spaces between them are depressed. The scales are sometimes more or less eurved with the convex side uppermost, especially at the anterior end, where they are replaced by two or more rows (the radiating arrangement often not being apparent) of minute spines or papille. Apparently, these spines become more or less compressed toward the middle of the shell and then coalesce at their edges. If they are not quite in alignment, the curved appearance noted above results. The post-cardinal slope, which, as already mentioned, lacks radiating ribs, is nevertheless marked by these flattened scales, which tend to be arranged in concentric rows without, however, becoming connected into continuous lamelle. No radial arrangement is here apparent. The internal structures are imperfectly known. The right valve bears two linear teeth, one before and one behind the beaks. The posterior tooth is long, about two-thirds the entire length back of the beaks. The anterior tooth is much shorter, about one-half the length of the anterior outline. Whether a small cardinal tooth was developed between these at the umbo is not clearly shown, but such a structure is indicated. In the left valve, there appear to be linear sockets corresponding to the teeth of the right. A large anterior scar is indicated, GIRTY, INVERTEBRATE FOSSILS FROM COLORADO 5 Of described species, this appears most closely to resemble A. gilberti, though it is not certain that the two are congeneric. The chief difference of a possible generic character lies in the fact that White’s figure ap- pears to represent A. gilbertt as having a well-marked escutcheon, a structure probably not present in A. squamulifera. Specifically, the latter appears to be a more slender form, more convex, and with a sharper umbonal ridge (these characters, however, may be enhanced by compres- sion in the Colorado form). It is also distinctly, though finely, costate, although A. gilberti in fact is covered with granules arranged in rows, so as to resemble minute radiating lire. Horizon AND Locality: Lykins formation; Heygood Canyon (lots 3264, 3265, 3266) and Perry Park (lot 3263), Colorado. Alula gilberti White? Alula squamulifera is abundant in lot 3263, but specimens are in an unsatisfactory condition of preservation. Many of them show a lower convexity and less angular umbonal ridge than the types. One example is sufficiently shallow, broad and ill-defined as to umbonal ridge to re- semble Allerisma gilbert rather closely. The sculpture is obscure but presents suggestions of radiating coste or of rows of papille. The de- pressed specimens which are provisionally placed with A. squamulifera appear to show a gradation toward but not into the only one referred to White’s species, and the facts which I have been able to observe leave me in doubt as to whether we have three species of not necessarily generically identical shells, or a fairly continuous series of mutations with A. squa- mulifera at one end and A. gilberti (or the form here identified as such) at the other. Horizon AND LocaLity: Lykins formation; Perry Park, Colorado (lot 3263). Myalina wyomingensis Lea Myalinas are extremely abundant in four of the five collections exam- ined, but most of the specimens are small. They vary in specific charac- ter. Some of the larger and more typical specimens agree in every determinable character with forms from the Rico formation of the San Juan region which I identified as Myalina wyomingensis.t. The great majority are of much smaller size, more like the form from Ouray which I somewhat provisionally called MW. cuneiformis.2 They naturally have the anterior lobe less strongly developed than the larger or mature exam- ples which accompany them. They seem as a rule to be less strongly 4U. S. Geol. Survey, Prof. Paper 16, Plate VIII, Fig. 8. 1903. 5 Ibid., Plate VIII, Figs. 16 and 17. 6 ANNALS NEW YORK ACADEMY OF SCIENCES oblique than the type specimens of M. cuneiformis, though some of them have the lobe scarcely more apparent. I am regarding part of these small specimens as being young examples of M. wyomingensis, and this may also be the true relationship of the Ouray specimens referred to cunet- formis. Typical cunetformis should probably be kept distinct for the time being. Horizon AND LocaLity: Lykins formation; Heygood Canyon (lots 3264, 3265 and 3266), Stout (lot 3262) and Perry Park (lot 3263), Colorado. : Myalina perattenuata Meek and Hayden The Myalinas of the Lykins formation in addition to showing varia- tion in the amplitude of the posterior wing vary conspicuously in the development of the anterior lobe. Some specimens have scarcely any perceptible development of this feature. These, although they are not sharply distinguished from the typical J/. wyomingensis, | am separat- ing as a different species under the title M. perattenuata. A similar phe- nomenon was observed in the Myalinas of the Rico formation of the San Juan region, and a similar course was pursued in regard to them. These Lykins specimens, however, are for the most part much smaller than those from the Rico formation and in this character approximate M. cuneiformis, but most of them are distinctly less oblique. A not very considerable breakage along the hinge line of these small shells however, or a concealment of the true outline in that region, makes an appreciable difference in their apparent obliquity. Horizon AND Locality: Lykins formation; Heygood Canyon (lots 3264, 3265 and 3266), Stout (lot 3262) and Perry Park (lot 3263), Colorado. Pleurophorus sp. A very imperfect internal mold showing best the impression of the hinge structures in the umbonal region, where they possess the charac- teristic dental arrangement of Plewrophorus. For the rest, there is indi- cated a transverse, oblong shell of medium size with rather strongly projecting anterior end. Horizon AND LocALiITy: Lykins formation; Perry Park, Colorado (lot 3263). Murchisonia buttersi sp. nov. Shell of medium size, slender, with high, many-whorled spire. Length of the type specimen as restored about 25 mm. Diameter of final whorl 11 mm. Number of volutions 10. Volutions angular with a thick, prominent carina situated considerably below the middle, the height of the upper zone being to that of the lower about as 2 to 1. Upper and lower zones more or less planate GIRTY, INVERTEBRATE FOSSILS FROM COLORADO 7 and standing at approximately a right angle to one another. The lower is gently concave, more so than the upper, although the upper spoons outward as it approaches the carina. Suture deeply depressed. The most conspicuous superficial feature consists of narrow angular cost, leaving between them broad shallow interspaces, which cross the upper portion of the yolutions transversely or in a direction longitudinal to the shell as a whole. They are straight, but are slightly oblique, retrally directed from above downward. These plications are’ perhaps restricted to the three or four older volutions, and there is some irregularity in their arrangement. They die down before reaching the carina. In addition, the whole surface of the upper zone is marked by microscopic transverse and revolving lire producing a more or less cancellated effect. The revolving lire are rounded, closely arranged and prone to be wavy. The transverse lirze are finer, sharper and more irregular, more of the nature of incremental lamelle, and the coste may perhaps be looked on as fascicles of these markings. The lower zone of the volution is marked similarly to the upper, but the angular cost are less strong. They have a slight forward obliquity from the carina. There appear to be two, possibly more, strong rounded revolving lire on the final volution at a point, as it would appear, about half-way down from the carina, and the volutions so embrace as to leave about two of these lire visible above the deeply sunk suture. The final volution is not well shown by the specimens examined, so that the sculpture below these two lire, the relative distance at which they occur below the carina, the shape of the aperture, ete., are not known. The carina is the site of the slit band. The band is occupied by two rather coarse, rounded lirz, separated by a narrow stria and appears to be defined by two delicate lamellose lines, one above and one below, bounded on the median side by slight striz. The two revolving lire which occupy the whole of the band and are more projecting than the edges are rendered nodose by the cost#e described as crossing the upper and lower surfaces of the volution. That is, the swellings occur where the cost would cross them, but the coste are evan- escent on the upper surface near the band, and the nodes are much more prominent than the costze and much more elongated spirally. Tn its specific relations, this shell is most nearly related to Murchisonia lasallensis and M. terebra. It differs from both in the presence of trans- verse plications. From terebra, which seems to be more nearly related than the other, it apparently differs also in having the carina containing two crenulated lire instead of one, in having two revolving lire just above the suture and in other details of sculpture. Generically, this shell can hardly be classed with typical Murchisoma, though it belongs to a group frequently cited under that genus. In some important respects, it is comparable with such representatives of the genus Worthenia as W. tabulata. This is especially true of the structure of the slit band, which seems to be identical in both. Given a much higher spire and more gradually enlarging volutions, with some modifica- tions in the modeling of the whorls, especially the lower part, it is easy 8 ANNALS NEW YORK ACADEMY OF SCIENCES to conceive how such a configuration as that of M. butters: might be evolved from that of Worthenia tabulata. The sculpture also appears to be of the same general character, the most essential difference being the development of transverse costz and of revolving lire more prominent than the rest on the lower half of the inferior zone. Some important characters of M. butters: are still unknown, but if these show no addi- tional differences, it may prove to be a rather extreme form of Worthenia. Horizon AND LocALiry: Lykins formation; Heygood Canyon, Colo- rado (lots 3264 and 3266). PLATE I LYKINS FOSSILS Myalina perattenuata (p. 6) Fig. 1. A large right valve referred to this species. 2. A left valve of more nearly the average size. Lykins formation, Heygood Canyon, Colorado (lot 3266). Myalina wyonrvingensis (p. 5) 8. A large left valve. Lykins formation, Heygood Canyon, Colorado (lot 3265). Alula squamulifera (p. 4) 4, Side view of an internal mold of a right valve. 4a. A view obliquely down on the cardinal margin of the same specimen, showing the impression left by the linear anterior and posterior teeth, x 2. 5. Squeeze of a right valve showing the surface characters. 5a. Same, x 2. Hyen with this magnification, the fine squamose character of the costz cannot be shown. Lykins formation, Heygood Canyon, Colorado (lot 3266). Alula gilberti? (p. 5) 6. Side view of a doubtfully identified right valve. Lykins formation, Perry Park, Colorado (lot 3263). Murchisonia buttersi (p. 6) 7. Side view of a squeeze made from the type specimen. 8. Another squeeze made from the same specimen, x 2. Lykins formation, Heygood Canyon, Colorado (lot 3264). ome | NTN Uo Aa Savy WE ax = aiboas aidt oF bertister. alse “tis gxie ouniavh sdd vlisond Tot jos dose tol} pea Te MIO TAEN hac zo kT PDR ey Te r med tag) PARA IULIT MROIE TE Hela et yo A vind rol 98 tol). ohstoice Hoye Hoos? selina pe -. j ‘ ‘ oe i a. ment seenne Wy siti t zy : Stiae yg #® te. bhoas haieaaaed: 6 30 weary obie “peneriseag ota te edt to: mitten. Lodibess add no iciieb gloupildo WOE b =“ wiieiseg Das sottedcis {HGH i ody vd dtl trofeportepti Bae ariwoner: ae > aise eds oogiage ail eat wits sting Idgt4 % ee 9 ) QROnEbITpE Sih ad ,colisobtinsin eit diiy foyva-& vee ; “aroda sd joamtns atzox off to 4VORE 408) obsitolo’> town hoow eet ottaarot anihy = NE hE ALD RAE tie hse ih be + Sy se ee ovine tdsis befitinabt’ utitidueb FS Yo. walt ‘aiid a ‘ase. toh of RTGS Seige erat mot Paes 08 aril a : = cA uote vin @shiGAWAE ay emiissqe9 eyed ort ni" t oben awaotspe #10 ee abte ra His &2 entisage nine 9) ito abrir ANSON PA, wsltom A: gf POL ¢ tol) ole TOL)’ MOV ee hooneall petiANtO vartalgle ANNALS N.Y. ACAD. Sci. VOLUME XXII, Plate | FOSSILS FROM THE LYKINS FORMATION nes ACADEMY OF SCIENCES ” “* ‘(bron OF Naruran History, 1817-1876) — ublications of the Academy consist of two series, viz.: By r NS Ses oe is led that — shall be devoted to icone are sent fre ¢ to Fellows and Active Members. The . s are sent to Honorary and Corresponding Members desiring them. a Lees oe Inquiries concerning current and back numbers of “Ge LIBRARIAN, New York Academy of Sciences, . eare of American Museum of Natural History, New York, N. Y. ae ne a int Joun D. HaseMan ey i. NEW YORK Tee PUBLISHED BY THE ACADEMY — es 31 May, 1912 : THE NEW YORK ACADEMY OF SCIENCES (Lyceum or NatruraL History, 1817-1876) OFFICERS, 1912 President—Emrrson McMuttin, 40 Wall Street Vice-Presidents—J. EDMUND WoopMAN, FREDERIC A. Lucas, CHARLES LANE Poor, R. 8S. WooDwoRtH Corresponding Secretary—Henry EH. Crampron, American Museum Recording Secretary—Epmunp Otis Hovey, American Museum -Treasurer—Hunry L. Douerty, 60 Wall Street Librarian—RatpPu W. Tower, American Museum EHditor—Epmunp Otis Hovey, American Museum SHCTION OF GEOLOGY AND MINERALOGY Chairman—J. E. WoopMan, N. Y. University Secretary—CuHarLes P. Berkey Columbia University SECTION OF BIOLOGY Chairman—Freperio A. Lucas, American Musepm Secretary—Wi.uiamM K. Grecory, American Museum SECTION OF ASTRONOMY, PHYSICS AND CHEMISTRY Chairman—CHARLES LANE Poor, Columbia University Secretary—F. M. Prprrsen, College of the City of New York SHCTION OF ANTHROPOLOGY AND PSYCHOLOGY Chairman—R. 8. WoopwortH, Columbia University Secretary—F RepEric Lyman WELLS, Columbia University The sessions of the Academy are held on’ Monday evenings at 8:15 o'clock from October to May, inclusive, at the American Museum of Rian Natural History, 7th Street and Central Park, West {Annas N. Y. Acap. Scr., Vol. XXII, pp. 9-112, pll. II-XVI. 31 May, 1912] SOME FACTORS OF GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA By Joun D. Haseman | Presented in abstract before the Academy, 12 February, 1912| CONTENTS PaGcr PLY POT SIGIR DE J ~ o'g.5 5 Sag SIs GS BnIene ono Gas STH One OOH IRE ec 10 Part I. Geology and topography of South America..................... 17 Disipuuion Of c oo 5b sain Sieve Sled GS OTE Oe en Me aE ae ea 83 OPES s cob at UPR SOR Seo eee eee ann 83° (GOMER TOPS. ere po B Cues PCO ean epee 84 FoTON AUB ILE TO ILCSIBMER IEP gS cron exe Siena neky cat hays cc 2 chad algvapiacsasd dev ately quale 95 NHN 5 5,4 WG a be Bee OU OS GP ENOL Sa en et ae rr a irra 99 Scag SRN Tell Te PP eee ee eh se le dase x cpociaivv'c o 6 u'o-bace® ceetsnecuwed 103 TES ONO BYINY 5555.9 BIS aS Oh cr Ot SI ete as es i 106 10 ANNALS NEW YORK ACADEMY OF SCIENCES INTRODUCTION South America has not lacked the labor of scientific explorers; none the less one may safely state that fully one half its surface is still prac- tically unknown. This is due to the fact that its investigators have been relatively few, but even had there been many scores of them they could not adequately have explored such an expanse of land beset with count- less natural obstacles. When attention is drawn to the fact that Brazil, which is one of the best known South American republics, is larger than the United States itself, it becomes evident that the few dozens of explorers within recent times could not have examined more than the regions accessible from the coast, from the few railroads and from the !arger rivers. Pioneer work in the interior of tropical America must often depend on meager and often incorrect information. This is reflected even in the maps of South America, for in every one some of the material has been taken from untrustworthy local sources. It was on account of this lack of decisive knowledge concerning cer- tain parts of South America, even in the best books of reference, that the director of the Carnegie Museum dispatched the writer, in 1907, in charge of the Carnegie Museum Expedition to Brazil. One of the primary objects of this journey was to study the distribution of the fishes; but kindred problems were not to be neglected, and data were collected on every hand and over a far greater territory, partially known or quite unexplored, than had been originally suggested. This was found possible because the writer was fortunate in maintaining excellent health throughout his journeys. (See Plate II for routes followed.) For this excellent opportunity to explore many regions of South Amer- ica, [ am deeply indebted to the founder and the trustees and to Dr. W. J. Holland, the director of the Carnegie Museum. Acknowledgment should be made to Dr. O. A. Derby, the director of the Brazilian Geological Survey, for many favors and much useful in- formation in furtherance of my work; to Prof. J. C. Branner and to Mr. R. Crandall, both of whom assisted me notably on my first trip to Bahia. Mr. Crandall examined my notes and map of the Cretaceous *of northern Brazil. This enabled me to add valuable corrections. He has also kindly submitted me some notes on the trend lines of northeast- ern Brazil. Dr. Schuchert has put me into his debt for many invaluable suggestions, and so too has Dr. David White for corrections and sug- gestions in the discussion of the Gondwana flora. My thanks are alse due to the directors and staffs of the museums at '——<- ANNALS N. Y. AcApD. Scr. VOLUME XNIJI, PuatTe II dnt pagent Bhonsafnaslatin fopese Tonal Celdara Fon te! Moae Vova vier MAP OF PART OF SOUTH AMERICA Showing journey made by J. D. Haseman under the auspices of the Carnegie Museum, 1907 to 1910, in Brazil, Uruguay, Argentina, Paraguay and Bolivia HASEMAN. GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 11 Rio de Janeiro, Sao Paulo, Para, Montevideo and Buenos Aires, and among my other South American friends, to Ricardo Krone of Iguape, John Gordon, Alipio Miranda Ribeiro and Carlos Moreira of Rio de Janeiro, Dr. Jappa Assu of Bahia, Rudolfo von Ihering of Sao Paulo, Dr. Frank Davis ani Feliciano Simon of Corumba and Dr. Snetlager of Para. Among my preceptors, I record gratefully my indebtedness to Dr. Higenmann, with whom I carried on my special studies in the University of Indiana; to Dr. A. E. Ortmann while I was in Pittsburgh, and to Drs. Dean, Grabau. Gregory and Hussakof while at Columbia, where a scholarship and the income of the Dyckman Fund for 1910 were gener- ously granted me.* The living and extinct fauna and flora of South America possess in certain cases, at least, a close genetic relationship with that of the south- ern portion of the 2astern hemisphere. This remarkable fact has led to a prevailing view that South America was cnce connected with some point of the eastern hemisphere. This view assumes that the closely related fauna and flora are descended from common ancestors which existed in the old land mass or continent to which the name Gondwana has been applied. In this thesis, it will be my effort to show that at least certain, if not all, elements in the fauna and flora of South America have evolved from forms which have from time to time been introduced from the northern hemisphere. In this view, I have brought together not only materials from references and from the laboratory but also data obtained during two and one half years of active field work. The amount of this ma- terial altogether will be sufficient, I think, to demonstrate that South America was never connected with the eastern hemisphere by a hypo- thetical southern and sunken mass of land. In the preparation of my paper, I have been obliged to omit numerous data and assorted faunal lists whose bearing has been more or less direct upon the present theme—in the latter cases since I am convinced that the lists do not explain distribution unless accompanied by detailed ob- servations upon the geology and the environmental conditions of the country considered. Hence I have been led to divide my thesis into two parts. In the first of these, we picture the past and present environ- ments in which the fossil and existing animals lived. In the second, we deal with the changes through which these animals and their ancestral stocks have undergone after arriving in these environments. 1The geographical names are spelled as they are in their respective countries. I have also used the old way of spelling such words as “Silurian” instead of “Siluric.”” I have also omitted marks of accentuation and other similar marks. _— oO ANNALS NEW YORK ACADEMY OF SCIENCES | It is in the first part of this thesis that my effort will be to demon- strate on geological grounds that South America has not been connected with the eastern hemisphere. In this connection, I have also been able to map for the first {ime the outline of the Plano Alto (the highlands of South America) which was deposited by the wind and rivers in a dry land and fresh-water basin which I am calling the Permian Inland Basin. This highland, I propose to show, was of the utmost continental impor- tance, from its dip, its lack of Mesozoic and Tertiary marine deposits and the direction of the trend lines, taken in connection with the Tertiary rise of the Andes; in fact, upon this I shall base my doctrine that the Amazon is a reversed river whose headwaters originally flowed into what I have designated the East Andean Sea. This outlining of the Plano Alto is of prime importance, not only be- cause it has given me the key to the correct explanation of the distribu- tion of the aquatic life, but also because it shows that South America has not been cut into islands by east and west invasions of the sea as has been proposed by some of the exponents of the Archhelenis theory of von [hering. In other topographical matters which are of importance from the zoogeographical viewpoint, I will also show that the Paraguay River is not connected with the Guapore, as has been so often erroneously stated, and that Rio Sao Francisco is connected with Rio Tocantins. I will note additional cases of stream piracy and will show that these taken in connection with waterfalls, swamps and certain environmental conditions will aid us in interpreting with a certain degree of accuracy questions in the distribution of the South American fishes. : In the last part of this thesis, an attempt will be made to demonstrate that the fauna of South America has been evolved from the forms which originally lived in North America. In arriving at this conclusion, I have not entirely limited my studies to the fishes, but I have considered carefully the voluminous data derived from other groups of living and extinct animals and plants which have been used to establish connections between South America and the eastern hemisphere. I have considered all of these data, because the facts derived from the distribution of any one group of plants or animals are not sufficient alone to warrant an in- terpretation which involves profound modifications of the earth’s surface as maintained by many authors. In the matter of the distribution of fishes, as bearing upon the greater problem, my effort will be to show 1. That the present distribution of fishes gives no clue to the point of origin of the families, but it does for some of the genera and many species. 2. The point of family origin can only be determined after the living HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 15 and extinct forms have been carefully compared and a sharp distinction is made between paleotelic (old or phylogenetic) and cenotelic (recent, adaptive or physiological) characters. It is necessary to draw this dis- tinction, because only paleotelic characters have been widely distributed. 3. That a fresh-water connection or its absence will not alone explain the present distribution of the fishes. Hence, the most important factor of living fish distribution is not land and water connections, 7. ¢., bar- riers, isolation, intermingling, etc., but it is the organic complex of the ancestral stock and the effects of different environments on this stock.? 4. That much of the similarity and some of the identity of certain species of the fishes of Rio Paraguay and Rio Amazonas are due to simi- lar and identical evolution of the highland ancestral stock after arriving in similar environments, e. g., as produced by the erosion of the high- lands. 5. That the existing highland genera, small in size, are the more gen- eralized types from which the bulk of South American fishes has evolved. 6. That the South American fishes have evolved from primitive forms which originally lived in North America. The above statements do not in the main agree with the views ex- pressed by previous zoogeographers. This difference of opinion is largely due to the fact that these investigators work from the static viewpoint of animal geography and have therefore only considered in some cases the “disconnected graveyard material,” i. e., a few isolated spots where the fauna died out; and in other cases the “hot-bed material,” 7. e., the end result of the greatest cenogenic evolution. For this reason, their static faunal lists do not correctly determine the point of origin of families and orders. As a result of their conception of animal geography, some of these writers have maintained invasions of the sea and land-bridges for which there is no evidence. They have also brought to the support of their views some unnatural environments and unwarranted views of the geology and the topography of South America. Another source of error in former interprctations is, I believe, the ignoring of the possibility of similar evolution of the identical ancestral stock in remote but similar environments. The necessity for the recog- nition of such evolution is due to the fact that in the same river basin there often exist (two or three) distinct fau.a! regions, one of which may show close affiuity with another distinctly separated basin, while another river system, although connected (with the latter), may yet re- tain quite distinct faunas. 2The idea that isolation alone produces new species implies the principle of selection, which is still of doubtful value. If it is not selection, then it must be in some way the direct or indiréct (also debatable) influence of the environment on the germplasm or else it is orthogenesis. 14 ANNALS NEW YORK ACADEMY OF SCIENCES My studies on the distribution of South Aierican animals have also led me to place more emphasis on negative evidence than is usually granted by most writers. This difference of view is primarily due to the fact that a specialist in any one group of animals places too much weight on his positive evidence. Such emphasis at iirst sight appears to be absolutely correct, but on closer analysis it is usually contradicted by positive evidence from other groups of animals. For example, one au- thor working on crustacea and mollusca finds that the alleged connection between South America and Africa had already disappeared in the early Cretaceous. If, however, we consider other groups, we find evidence which does not confirm this view. Thus, the affinity existing between the South American and African characinid and cichlid fishes is as close as that of the mollusca, yet there is no evidence that either of these families of fishes existed during the early Cretaceous. In my conclusions, therefore, I have been led to balance the positive and negative evidence in the cases of many different groups. This bal- ancing has been attempted not only by considering long lists of species, but by taking into due account the influence of various environments on the ancestral stocks (whose points of origin are usually unknown). In fact, we should not, in such considerations, lose sight of the fact that our knowledge of the existing species of any group of South Ameri- can and African animals is still very imperfect. Many species are still to be caught, many are exact synonyms and many are without doubt local somatic changes which are not always inherited. Therefore any positive evidence derived from such lists is, in my opinion, entirely in- adequate to warrant the reconstruction of the earth’s surface, unless supported by strong geological evidence. ‘This is all the more true when there are other means of distribution which do not involve great topo- eraphical changes. In point of fact, the inadequacy of the Seine data is at once re- flected by the number of alleged land-bridges and seas required during various past geological epochs. Each writer has constructed his new set of barriers, seas and land-bridges in his effort to explain the distribution of the fauna or flora in which he is recognized as an authority. Notwithstanding the diversity of views concerning the time of exist- ence and the location of the alleged land-bridges to the South American continent, we may roughly consider them under the following two groups: 1. The Gondwana Land of Suess and others—a late paleozoie conti-- nent traversing the greater part of the southern hemisphere and connect-- ing India, Australia, South Africa and South America. The last re- mains of this old land-mass connecting Africa and South America have HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 15 been designated Archhelenis by von Ihering. Antarctica, another por- tion of Gondwana, is the name generally applied to the south polar conti- nent which is believed by some to have been connected with South America and Australia and perhaps with Africa. 2. The persistence of the continental shelves and the great ocean basins. This view precludes the existence of furmer connections between South America and the eastern hemisphere, but admits the North Ameri- ean connections with Eurasia. These two views involve not only the distribution of plants and animals but the geology of the entire earth. They are not new, dating back in fact nearly to the time (1857) when Sclater first placed geographical distribu- tion in tangible form. In his scheme of distribution, the world was con- sidered in six faunal regions, a scheme in which South America formed a major part of the neotropical realm. _In 1876, Wallace published two comprehensive volumes on the geo- graphical distribution of animals. In these two most excellent volumes, the genus was used more than the species as a means of comparing faunal regions, and it appears that such a comparison is more luminous than one based on the species, for it is found that the generic characters are usually more nearly paleotelic than the specific ones; moreover, the list of species is the less accurate, since a far greater number of species than genera are still undescribed. The extensive data given in these two volumes indicate that the bulk of the ancestral land animals originated in the northern hemisphere. It is worthy of note that the views expressed by Wallace and some of the other earlier writers were far more conservative than those of more recent date. It was during the interval from 1876 to 1890 that zodlogical, paleonto- logical and geological data accumulated rapidly and yielded, especially, the excellent summation of the geology of the face of the earth by Suess and his views of the Gondwana Land (earlier suggested on purely paleon- tological grounds by Neumayer), and his considerations appear to have paved the way for von Ihering’s Archhelenis theory which has more or less dominated most of the later studies on the zodgeography of South America. Von Ihering has made, from time to time, slight changes in his theory, in order to meet the demands of more recent investigations. In 1907, he reconstructed the surface of the earth according to the views which he obtained from a detailed study of the mollusca. At the time, he main- tained that, previous to and during a part of the Tertiary epoch, Brazil (Archibrazil or Archamazonia) was connected by Archhelenis with Africa and by Archiplata with Archinotis (Antarctic continent), which was also 16 ANNALS NEW YORK ACADEMY OF SCIENCES continuous with Australia. The sea traversed northern Brazil and sepa- rated the Guiana highlands plus the West Indies (Archiguiana) from the rest of South America, but it was connected with both Asia and Europe. Africa was not connected with Asia at that epoch. In 1911, he had made many changes in his views. P On the other hand, after a long detailed study of both the mollusca and crustacea, Ortmann has also maintained that an Archhelenis existed, but he differs with von Ihering both in regard to its location and the time of its disappearance. He believes that Archhelenis had already disappeared before the beginning of the Tertiary (perhaps the early Cretaceous) and that it connected Guiana and Africa. Eigenmann (1909) has tested the Archhelenis theory with the distribu- tion of the South American fishes and has found no objections to it. In fact, he states that the theory is quite useful in explaining the distribution of certain families of fishes, especially the Characinide and the Cichlide. D. White (1907) in an excellent paper on the Gangamopteris flora (Gondwana flora) of Brazil does not, however, favor the Archhelenis land-bridge; he believes that an ancient connection (Permian) very probably existed between South America, the Antarctic continent and Australia or Africa. In Part V of Volume III of the Princeton Patagonia reports, Pilsbry has summed up the distribution of non-marine mollusca of South Amer- ica. His figure, page 632, indicates a former connection between the region of Pernambuco, Brazil and South Africa. He is inclined to believe that this connection disappeared by the end of the Cretaceous. He also admits the probability of a connection by way of the Antarctic islands with Australia, but he does not believe in a former isolation of the Guiana highlands from those of Brazil. Schuchert (1911) also believes that the distribution of the brachiopoda shows clearly not only the former existence of an equatorial Gondwana across the Atlantic, but as well that its vanished Atlantic bridge still con- trols the distribution of living forms. He is of the opinion that Gond- wana probably existed until middle Eocene times. In the Age of Mammals (1910), Osborn re-states the widely accepted belief in an Antarctic connection between South America and the Aus- tralian realm, but he rejects a Tertiary Archhelenis. He thinks that this connection is necessary in order to explain the great similarity which exists between some of the fossil marsupials of Patagonia and the mar- supials of the Australian realm. He also states that Matthew has rejected both the Archhelenis and the Antarctica connections and now maintains a northern origin of the southern fauna. ANNALS N. Y. ACAD. SCI. VOLUME XXII, PLATE TEI ry a EE ER A EC LET | " | ARCHEAN AND OTHER PRE-CAMBRIAN AREAS OF SOUTH AMERICA The solid black represents the Archean; thee horizontally lined area represents the basal highland formation, which is generally considered to be pre-Cambrian and which is covered by alluvial deposits along the Amazon (dotted area). HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 17 The view held by Dr. Matthew and the writer (who have independently and from different standpoints arrived at several identical and funda- mental conclusions concerning the distribution of South American ani- mals) in a general way was put forth in 1886 by Haacke as the North Polar theory of the origin of land animals.* Part J. GroLtocy anp TopoGraPHy oF SourH AMERICA # DISTRIBUTION OF GEOLOGICAL HORIZONS It has been extremely difficult to map even the larger divisions of the geological ages found in South America, since the exact age and extent of many of the known formations have never been satisfactorily determined. Archean Archean rocks, as shown in a general way in Plate III, extend in a nar- row belt, more or less broken, from Tierra del Fuego to the Isthmus of Panama. A great depression east of the Bay of Arica hes between the coastal Archean rocks of Chili and the inland Archean rocks of the north- ern Andean region. Another great belt of Archean rocks extends from Uruguay into the Serra do Mar and its various northern spurs of eastern Brazil. Archean rocks have also been encountered in isolated places of Patagonia, and a belt more or less broken extends from near Bahia Blanca through the Cordova Mountains into southern Bolivia. Finally, rocks of similar age have been encountered in northern Colombia, parts of Vene- zuela and Guiana. The exact age of the crystalline schists, gneisses, granites, etc., which ‘underlie the Plano Alto has never been satisfactorily determined, but they are usually considered pre-Cambrian or Archean on account of the entire absence of fossils. The lined portion of Plate III indicates the extent of these rocks. Lisboa, one of the most competent Brazilian geologists, has reported crystalline rocks of pre-Cambrian (?) from near Miranda, Matto Grosso and in the basins of Rio Apa and Rio Aquidauana. The writer observed the basal highland rocks on the Bolivian side of Rio Guapore below the mouth of Rio Verde and on the Brazilian side of the Guapore at the 2 The bibliography gives a list of the publications which have been extremely useful to. me, and I take the opportunity here to acknowledge my indebtedness to the authors. Few references have been given in the pages of this thesis, because it has been deemed ad- visable to omit them for the sake of clearness and brevity. Therefore some common information has been freely used. +In the preparation of this part of the thesis, the writings of Derby, Branner, Suess, Eschwege, Hartt, Hatcher, Steinmann, Phillipi, Stelzner. Hauthal. Katzer, Crandall and numerous other authors given in the bibliography have been indispensable. 18 ANNALS NEW YORK ACADEMY OF SCIENCES waterfalls of Forto de Principe da Beira. They were also seen along Rio Mamoré above Guaja Mirim, where the Serra de Pacas Novas approaches the river. Evans has reported gneisses, etc., from the Rio Beni and the Madeira falls. WKatzer has maintained that this same basal highland for- mation extends under the deep alluvial deposits of the lower Amazon Valley. Asa result of the writer’s own observations, he is convinced that this is in the main the correct view. Hence this basal highland formation is very large, perhaps larger than indicated on Plate III.*° The extent and form of these ancient rocks have given South America its present shape from the very beginning. Silurian © Silurian fossils have been reported from Rios Curnua, Maecurt and Trombetas, which are affluents from the north side of the lower Amazon River; from Bom Jesus da Lapa, of Bahia; from some of the promon- tories of Venezuela; from the eastern chains of the mountains near the headwaters of Rio Bermejo, Argentina; west of San Juan along the eastern base of the Andes of northwestern Argentina; from Sierras de Famatina and La Rioja; from the mountains on either side of Sierra de Aconquija; from Sierra Aguilar, and from Cuzco, in southern Peru, ex- tending past Hlampu and Illimani through western Bolivia toward the Argentine chains of mountains. (?) Silurian (Arthrophycus harlant) was reported from Sierra de Ja Tandil. The locations of these formations are shown on Plate IV, except ‘that no distinction has been made for Ordovician because of the lack of data. Devonian Devonian fossils have been reported from Alameirim to Rio Uatuma at Erere, Rio Maecurt south of Larangal on the south side of Rio Amazonas ; from Lagoinha near Cuyaba; from a belt extending from the State of Sao Paulo into the State of Parana at Ponto Grosso, at Jaguarahyva and in the Ivahy basin ; in southern Peru and about Lake Titicaca, and in the Andes of Chili. Doubtfully from the northern part of Sierra Tandil. The locations of these formations are shown on Plate V. Carboniferous Carboniferous fossils have been reported from Rios Trombetas, Curua, Maecurt, Uatuma to Janary near Prainha and at Alemquer on the north 5 It must be granted that the regions supposed to yield exposures of Archean rocks may with more careful study be materially diminished, as has been the case in North America and Europe. If these rocks mapped as Archean are all older than the Carboniferous formations, then the ensuing views are not in the least affected. 6 The Cambrian has not been mapped because of the lack of data. Deposits of this age have only been reported from a few places, as northern Argentina by Keyser, etc. ANNALS N. Y. ACAD. SCI. VOLUME XNII, PLatTe 1V oF S of, <\ ENOWN MARINE DEPOSITS OF SILURIAN AGE IN SOUTH AMERICA i¥ ANNALS N. Y. AcapD. Sct. 2—_NY VOLUME XXII, PLaTse V : ism Z : GS 7: + ie of ~ e en LP SON KNOWN MARINE DEPOSITS OF DEVONIAN AGE IN SOUTH AMERICA ORT at ANNALS N. Y. Acab. Scr. VOLUME XXII, PLatE VI ie WSS pu? ) KNOWN MARINE DEPOSITS OF CARBONIFEROUS AGE IN SOUTH AMERICA ce AE ey aaa tea = ANNALS N. Y. ACAD. SCI. VOLUME XXII, Pratn VII a A) 2g y ( 5 % 4 < OP NY us KNOWN MARINE DEPOSITS OF PERMIAN AGE IN SOUTH AMERICA ANNALS N. Y. ACAD. SC1. VOLUME XXII, PLaTE VIII KNOWN MARINE DEPOSITS OF TRIASSIC AND JURASSIC AGE IN SOUTH AMERICA HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 19 ‘side of Rio Amazonas and along Rio Tapajos from Itaituba to Aveiro and west to Maueassu at Fructal and Rio Pedra do Barco on the south side of Rio Amazonas; from the Cordillera Oriental of Peru; from Lake Titicaca near Yampopata and extending south towards Cochabamba at Arque, Bolivia; Santa Cruz, Bolivia, and in the Chapoa Valley at la Ligua, Chil. The locations of these formations are shown on Plate VI. Permian Permian fossils have been reported from the State of Sao Paulo, Brazil, and this same belt extends south through Parana and Rio Grande do Sul and ends in Uruguay. Permian fossils are also known from a few isolated places in the Andes. The location of these formations is shown on Plate VII. Triassic and Jurassic The Triassic and Jurassic periods are unknown in Brazil. ‘Triassic and Jurassic fossils have been reported from northern Colombia and Venezuela; from Puerto Puruay and Rio Maranhao of Rio Amazonas, and are widely spread in the Andes of Peru and Chili at Passo de los Patos, Coquimbo, Copiapo and elsewhere, where they are mixed with fresh- water deposits. The locations of these formations are shown on Plate VET. Cretaceous Cretaceous fossils have been encountered in a narrow belt along the Brazilian coast from Ilhéos, south of Bahia, to Pirabas, near the mouth of Rio Amazonas. This coastal belt is never more than about 100 feet above the sea level, and its fossils also show that it must not be confused with the inland Cretaceous of northern Brazil which extends north from the State of Bahia across Rio Sao Francisco at Jatoba and widens out in the states of Piauhy and Ceara. This belt is often 1000 or more feet above the sea level. If the writer is correct in considering a formation (in which he found no fossils, but near where at Curimata fossil fishes, (?) Diplomystus, are said to exist) near Lagoa da Paranagua in the State of Piauhy as Cretaceous, then this belt will with all probability later be found to extend as far south as Serra da Tabatinga, but it does not extend into the Jalapao region of northern Goyaz. The extreme northwestern extension of this Cretaceous is unknown, and consequently it offers alluring opportunities for future study,’ be- 77This belt may extend northwest into the State of Para and as far west as the lower ‘Tocantins basin. Mr. Roderic Crandall of the Brazilian Geological Survey has done the most field work on this inland Cretaceous belt, and his views agree in the main with those of the writer. 20 ANNALS NEW YORK ACADEMY OF SCIENCES cause it will extend the limits of the Cretaceous deposits over a vast portion of northeastern Brazil. Cretaceous deposits have been reported from Bahia, Espirito Santo, Aracaju, Alagoas, Maria Farinha, Jatoba, Riacho Doce, Serra de Ara- ripe and Pirabas in Brazil; from the base of the mountains of Guiana; from Bogota and the region of Lake Maracaibo; from Sierra de Merida of Colombia and Venezuela; from Cordillera Nevada and from both sides of the Andes in the region of Alto Rio Maranhao; from Rio Acre (Mosasaurus) ; from Caracoles, Bolivia; from near Lima, Peru; Cochi- yacu west of Rio Huallaga and north to Celendin; from near Guayaquil ; from Tingo and south along both sides of the Andes toward Chili at Tomé, etc.; from Laguna Argentina to Tierra del Fuego; from Rio de los Patos west of San Juan, Argentina; from Colchagua, Coquimbo and Copiapo in Chili; from Sierra de Zenta east of San Juan, and perhaps from Gran Chaco toward the mountains about the headwaters of Rio Bermejo and Pilcoinayo, but I am inclined to believe that these de- posits are fresh water, as the most common form, Melania, is not typi- cally marine. At any rate, this region needs some more careful study. The locations of these formations are shown on Plate LX. Tertiary Professor Branner is inclined to consider some of the marine forma- tions of northeastern Brazil Eocene. This view has been recently cor- roborated by President Jordan’s studies on the fossil fishes from Riacho Doce, but neither of these authors has entirely excluded the possibility of these formations being upper Cretaceous. As far as the writer has been able to ascertain, from a first-hand knowledge of the region in question as well as from that of Crandall and also from a consideration of President Jordan’s paper on the fossil fishes of the Serra de Araripe, it does not appear that any decisive evi- dence exists which establishes any marine Tertiary in northeastern Brazil. Fossil diplomystid fishes, the subject of President Jordan’s paper, are known not only from the Cretaceous of Brazil, but also from the Cretaceous of other Continents. The fact that most of the diplomystids are found later than the Cretaceous epoch is no evidence that those of Serra de Araripe are Eocene. Furthermore, the peculiarities of the diplomystids of the Serra de Araripe will, with all probability, be found in various other localities of this region, when more exploration is com- pleted. For several reasons, therefore, I have mapped the outline of the Cre- taceous belt on the map of the Tertiary epoch with a mark of imterro- ANNALS N. Y. ACAD. SCI. VOLUME XXII, PLATE IX IP =~ KNOWN MARINE DEPOSITS OF CRETACEOUS AGE IN SOUTH AMERICA ANNALS N. Y. ACAD. SCI. VOLUME XXII, Pratn X i] U \ 4 \ ( ‘ \ . ' \ ~ 1 \ On , ‘ . ‘ i, ~s ‘ ( ee ‘ ‘ =. ‘ . 1 / a oN y sien i ’ és a mS eer ‘ \ . 1 ' ‘ 4 | ’ ' 1 ‘ L “ 0 ‘ KNOWN MARIND DEPOSITS OF TERTIARY AGE IN SOUTH AMERICA HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 21 gation, for later work may establish Tertiary ceposits in at least part of this region, especially if we accept Professor Branner’s view of its stratigraphy.® Marine Tertiary fossils have been reported from Entre Rios, the Pam- pas and part of the Gran Chaco of Argentina; from the eastern base of the Andes, extending into the plains of Patagenia and as far south as Tierra del Fuego; from narrow belts along the coast of Peru and Chili; from the lower Orinoco valley; from Pebas, Peru, down Rio Solomoes as far as Sao Paulo de Olivenca, and from Canama on Rio Javary and prob- ably from the region of Rio Acre, Brazil, and Santa Maria, Catamarca, west of Jejuhy in northern Argentina. The locations of these formations are shown on Plate X.° TREND LINES The trend lines of South America are about parallel to the coast, ex- cepting in the region of Pernambuco and Ceara, Brazil. In this region, they appear to be fan-shaped. In regard to this most interesting region, I can do no better than quote a letter from Mr. R. Crandall, of the Brazilian Geological Survey, who has explored this region during the past four years: “The general trend lines through all of the State of Bahia, as you will re-. member, are northwest-southeast and north into northeast-southwest. These lines get lost as we get farther north, and the trend of the coast itself changes north of Pernambuco. The change in the trend of the coast is accompanied by a Similar change in the direction of the Serra da Borborema. The Serra da Borborema is more properly an eroded mass than a structural line, though it conforms quite closely to the general trend of this region. “I consider the lines in the Ceara, Rio Grande and Parahyba true structural lines, as they are Jong lines of intruded granites and allied or similar rocks which indicate intrusions on or along lines of previous weakness. Part of these lines are indicated by the Ceara series of rocks which I have correlated with the Jacobina series of Bahia and with the Minas or iron-bearing series of Minas Geraes. “T have never properly understood the forces that formed this fan-shape in northeastern Brazil. Just at what age it came is hard to say, as the age of the latest folded rock, the so-called Ceara series, is about Cambrian (for all we know, even pre-Cambrian). I believe that Derby considers the granites of southeastern Brazil to be post-Devonian, and I believe that these northern SIt does not make any great difference, as far as the present conclusions are con- cerned, whether marine Tertiary does or does not exist in the above region. Professor Derby (1907) also expresses some doubt about the existence of marine Tertiary in the said portion of Brazil. ®Tt is beyond the scope of this thesis to attempt to discuss in detail any of these for- mations, excepting that of Alto Rio Amazonas. The necessity of this preliminary sketch will become evident after the reader has considered the ensuing topics. What has been land and sea is important from the standpoint of animal geography. ae ANNALS NEW YORK ACADEMY OF SCIENCES granites are in the larger part of an age younger than the Ceara series and. somewhere along in the Devonian is entirely probable. You see between the Cambrian and the Cretaceous in the northern region we have no record. “Tt is entirely possible that the northern fan is due to local folding, but it is. pretty large for that, that is to say, spread over a very large area.” I believe that these facts indicate that no Paleozoic wedge or land- bridge could have existed between this portion of Brazil and Africa, because the trend lines fade away toward the sea, and there is no eyi- dence of the continuance of the lines across the Atlantic into Africa. The folds are crushed and irregular and do not end abruptly along the coast. Besides, if the wedge had existed, this type of fanlike folding would, I believe, have been almost if not quite impossible. This fan-- hike structure may possibly have been produced by some unknown force: pushing from the interior of the earth at an angle to the vertical and. more or less parallel to the coast, but more strongly in the region of the mouths of Rio Amazonas and Rio de la Plata, if the Brazilian coast was. not connected with Africa. This would have pushed the southern and. northern ends of South America to the west and have made the “Per- nambucan fan” on the east and a somewhat similar structure in the region of Lake Titicaca east of the Bay of Arica. In southern Brazil and more especially in val of the Plano Alto, ero- sion and extensive contine::tal deposits have so disfigured the unexplored. forested surfaces that littic is known concerning the trend lines, but: superficially this region looks like an abruptly chopped-off coast. This abruptness, I believe, is due to erosion of late and post-Paleozoic land deposits and not to post-Paleozoic faulting. The northwest strike from Cuyaba past the Madeira Falls, noted by Evans may be due to ancient erosion, but I hardly think so, because the Plano Alto dips toward the southwest. The old drainage was into the Hast Andean Sea and hence the rivers at that time cut the Plano Alto im a western-southwestern direction. When the Amazon became reversed, Rio Madeira cut these old planes of erosion almost at right angles, producing thereby a very complicated topography. By means of the trend of the Sierras de Tandil and de la Ventana, we: can separate both of them from the Andean complex, in spite of the fact that Suess and Stelzner have been inclined to consider the Sierra de la. Ventana as belonging to the Andean system. In fact, the writer considers that the Sierras de la Ventana and Tandil with all of their side chains are the southern extension of the Cordova Mountains by the way of Sierra de San Luis, where there is a break in the system, just as the Serra do Mar of Brazil breaks up into several es ~ xg — = ANNALS N. Y. Acad. Sci. : VOLUME XXII, PLatTE XI OUTLINE MAP OF SOUTH AMERICA Showing the trend lines about the Plano Alto and the hypothetical outlets of the Hast Andean Sea. The “Pernambucan Fan” has been put in from notes and maps contributed by R. Crandall. 3—NY HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 23 chains in east central Brazil. I also believe that the above sierras origi- nally formed part of the southern and western boundary of the great Permian Inland Basin which will be considered in the following pages. I have relegated both the Sierras de Tandil and de la Ventana (as Hauthal did several years ago) to the Brazilian system for the following reasons : 1. The Sierras de Tandil, de la Vantana, de San Luis and de Cordova, like the Serra do Mar of Brazil, are characterized by an almost entire absence of marine fossils and have no marine Tertiary and Mesozoic, which is characteristic of the Andean complex. 2. In general succession; the Archean rocks of these sierras resemble the Serra do Mar rather than the Andean system. 3. These sierras are well separated from the Andean complex by many elevations and depressions, some of which afford evidence of Meso- zoic and Tertiary invasions of the sea. 4. From the absence of fossils, it may be judged that these sierras have remained almost stationary; and as a result, there is a striking correspondence between their altitudes and these of Brazil and Guiana. It is possible from similar reasons that at least part of the Cordilleras Oriental of the northern Andes will be shown to belong to the old Bra- zilian system, but entirely too little is known about this region to war- rant a consideration at this time. The trend lines of the Pacific side are also parallel to the coast. In the region of La Paz, Bolivia, there is an indentation of the coast back of which the chains of mountains are bent out of line and piled up to several thousand feet of altitude. This region is, in a way, the counter- part of the fan-shaped region of Pernambuco along the Atlantic Coast. In other words, the trend lines of the two coasts exhibit a remarkable similarity, as is indicated in a general way on Plate XI. The only strik- ing difference is thay in the region of La Paz, Bolivia, the coast is bent in, while it is bulged out in the region of Pernambuco. ‘Taylor has attempted to explain part of this by a sliding of the Brazilian mass against the Andes. This sliding was assumed to be due to some force applied parallel to the Brazilian coast but having a greater intensity in the regions of the mouths of the Amazon and !a Plata. The lines of weakness and strength in a general way extend north and south (as Schucheri has shown for North America). This is shown by the maps of the marine deposits. The invasions of the sea as a rule appear to have been from the south toward the north. The Permian of southern Brazil or the Devonian extending from the Amazon Valley past Cuyaba into Parana, and perhaps as far south as Sierra de Tandil, 24 ANNALS NEW YORK ACADEMY OF SCIENCES if the (?) Devonian of Siemiradzki exists there, are good examples of a southern-northern invasion of the sea. In all of this region, an east to west invasion of the sea appears to have been impossible, excepting in narrow belts along the coasts, on account of the intervening Archean mountains, like Serra do Mar, which show no traces of marine deposits. This is of the utmost importance from the standpoint of animal dis- tribution. Some authors have attempted to show that the Patagonian region, and others that the Guiana region, was for a long time cut off from Brazil by arms of the sea. In order to isolate either of the above regions from Brazil, it would require an extensive east to west invasion of the sea for which we have no evidence. On the other hand, the maps showing the location of marine deposits offer strong evidence against such a view. Hither the older rocks of Chil or of the Cordova Moun- tains could afford connections between southern South America and the Brazilian region. Uence the observed difference in the fauna of Pata- gonia must be due in a great part to environmental conditions.*° . BRAZILIAN COAST In some respects, the Brazilian coast appears to be the counterpart of the contour of West Africa. Its abruptness is thought by many to be due to the submergence of a “wedge” which originally connected Brazil and Africa. Soundings have shown that deep sea exists within a comparatively few miles of the Brazilian coast; but thus far soundings have not pro- duced the slightest evidence for a submerged “wedge” or land-mass which is believed by many to have originally connected Brazil and Africa. In fact, there is strong evidence derived from soundings against the submergence of such an extensive land-mass into the abysmal depths. This wedge must have been deeply eroded, forming thereby deep, wide and abrupt valleys. When this surface (sucl as Brazil at the present time) dropped beneath the ocean, few soundings would be needed to show that the bottom of the ocean under such conditions would not be uni- form. Inasmuch as soundings have revealed no evidence in favor of such a rough sea bottom, | take this as strong evidence against such a view. It is true that Murray has found a mid-Atlantic ridge, but the trend of this elevation is parallel to the distant coasts, 7. e., more or less north and south and not east and west. 10 Also the imperfection of the fossil records, exploration, etc., in South America can- not be ignored. I saw part of a Torodon from near Uruguayana, Brazil. Even if an arm of the sea separated Patagonia from the rest of South America, it would have been entirely too narrow to have been an effective barrier. HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 25 Soundings as well as fossils indicate that the Brazilian coast at one time extended slightly farther east than at present. There appears to be little or no doubt that the Abrolhos and Barbados islands were originally connected with South America (? late Mesozoic). To the east of these islands, no shallow sea or numerous small islands exist, as they do in regions like the West Indies, East Indies and Alaska, where we know that elevations and subsidences have taken. place. The abruptness of the present Brazilian coast is not necessarily evi- dence in favor of a post-Paleozoic faulted coast.. In fact, my own ob- servations have convinced me to the contrary. The entire absence of fossils along the coast of southeastern Brazil indicates a great stability of this region. Inasmuch as in all this region nearly all of the rivers flow west and southwest, away from the ocean, it appears that there is a gentle dip towards the southwest. Of course much of the surface of this region is covered by late and post-Paleozoic continental deposits which have been deeply eroded. The erosion of these deposits as well as the older rocks would produce an abrupt appearing coast just as is seen on the north and east sides of isolated mesas, or portions of the old Plano Alto, which will be taken up in the following pages. Therefore, I believe that the abruptness of the southeastern coast of Brazil is due to its stability, as is shown by the absence of fossils, and to the southwest dip of continental deposits, especially of late and post-Paleozoic age which have been deeply eroded, and not to a post-Paleozoic faulting and sinking of a land-mass into abysmal depths, for which no positive geo- logical evidence exists. Much of the abruptness, however, is nothing more than sea cliffs. During Cretaceous times, there was a slight elevation of the coast of central and northern Brazil. In comparatively recent times, the entire Brazilian coast has risen 40 feet.1t This is sufficient to change the earlier coast line and to produce another line which is abrupt, but due to ages of erosion and not to faulting. Moreover, there are analogies for this abrupt coastal condition in which no extensive sinking of a land- - mass has occurred. Therefore, summing up all the geological evidence, the lack of east 11 Hart, Krone and others have stated that the Brazilian coast has been elevated about 40 feet since the Pleistocene or Quaternary. I had ample opportunity to confirm this view while traveling in Rio Ribeiro de Iguape and along the coast at various other points of southeastern Brazil. The location and nature of the sambaquis (shell mounds), the ocean caverns and wave marks along the bases of inland morros (hills) show that the sea extended inland almost to the mouth of Rio Juquia (west of Iguape) even as late as the time of the Indians in this region. In further evidence of this, marine shells are dug up in wells east of Campos along the Rio Parahyba and along the west shore of Lagoa Feia, which is now about eight miles from the seashore and about nine feet above sea-level. 26 ANNALS NEW YORK ACADEMY OF SCIENCES and west structural lines, the deep intervening sea, absence of islands, relative evenness of great ocean depths, the absence of the deposits found in deep seas on the continental shelves and the shape and abruptness of the coast due to ages of erosion assisted by recent elevations, we conclude that South America could hardly have been connected. with the eastern hemisphere.*? ) PLANO ALTO AND THE PERMIAN INLAND BASIN I have used the native term, Plano Alto, to include all of the sand- capped tableland which extends south from the Guianas through Brazil into Uruguay and west into Bolivia. The outlines of this region are shown on Plate XII. All of these highlands appear to have been deposited in a fresh-water inland basin during the Permian epoch. The remains of Permian reptiles (Mesosaurus and Stereosternum), of the Gondwana flora and other plants, of Scaphonyx, a Triassic reptile, of Schizodus, Conocardium, Myalina and a few other marine lamellibranchs found in the highland region indicate that the Plano Alto was deposited by wind and rivers in a fresh-water inland basin of Permian age. The thin layer of intercalated marine limestone indicates only a brief Permian in- vasion of the sea in the Plano Alto of southeastern Brazil. The Permian Inland Basin was almost surrounded by Archean moun- tains: on the east by Serra do Mar and its northern spurs; on the south and west by the Cordova Mountains and their southern spurs; on the north by Archean rocks of Guiana and Venezuela, and on the north- west perhaps by the Cordillera Oriental. The characteristic sandstone found in all of this region was in part deposited in shallow fresh water and in part shuffled about by winds into this Permian basin. The basal Plano Alto formation or the floor of the Permian inland basin is composed of granites, gneisses, crystalline schists and the like, which are generally considered pre-Cambrian or Archean, because of the absence of fossils. On various portions of this basal formation are Paleozoic deposits which have already been mapped, but which will without doubt be greatly extended as exploration proceeds. These maps show that none of the Plano Alto included in Plate XII has been invaded 2 The absence of marine Lower Carboniferous fossils from eastern Brazil is not, in my opinion, nearly as strange as the apparently entire absence of post-Paleozoic marine de- posits of southeastern Brazil, in view of the fact that west of Serro do Mar are found marine deposits of Devonian and Permian age. The first great escarpments in British Guiana, not far from the coast, are due to the erosion of the Plano Alto sandstone in the regions of the Kaieteur Falls; similar conditions exist right along the coast of south- eastern Brazil. So it will take far better evidence than exists to prove that this coast of Brazil is a post-Paleozoic faulted one and not a Paleozoic one which has remained stable and has been changed by the erosion of later land sediments. ANNALS N. Y. ACAD. SCI. VOLUME XXII, PLATE XII Antafagartm Blanco bncalain. Puerta del Noarce \tu, fi Voverer Lor Andes Jo are! Val per iaae MAP OF PART OF SOUTH AMERICA Showing the outline of the Plano Alto, which was deposited in the fresh-water Permian Inland Basin and contains no Post-Paleozoic marine deposits HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 27 by post-Paleozoic seas. Hence continental and not marine deposits are the more extensive deposits of Brazil. The Permian inland basin is not uniformly symmetrical in reference to the altitude of the basal rock in different sections, but its northern and eastern sides are always higher than the southern and western. There is also a considerable elevation of this basal formation between the headwaters of Rio Paraguay (also seen in the iron and manganese de- posits 600-800 meters above the level of Rio Paraguay in the Serra de Urucum) and the Amazon, but this may be due to unequal early Paleo- zoic erosion. This is especially true, if the streams flowed westward before the Amazon was reversed. ‘This difference in elevation of the pre- Cambrian and early Paleozoic or basal formation of the Plano Alto was of the utmost importance for the formation of the Plano Alto. The overlaps near these various higher (?) Archean elevations indicate that much of the basal sandstone of the Plano Alto was derived from place erosion of the higher points of the floor of the Plano Alto. Hence no considerable extension to the east of present coast of South America was necessary. The uppermost strata were deposited from the higher Archean mountains on the east and northeast and hence dip toward the southwest. The outlines of the Plano Alto shown on Plate XII may be slightly extended by future work from southeastern Brazil toward Argentina, far- ther into Paraguay and Bolivia, and farther west from Brazil toward the Cordillera Oriental of Colombia and Ecuador. So much of the Plano Alto has been deeply eroded and some entirely washed away that its exact limits cannot be given at the present time. Tt is evident from Plate XII that the Plano Alto proper is bounded by parts of the basins of Rios Sao Francisco, Orinoco, Mamoré and various other smaller rivers as well as by waterfalls and other changes in the geological structure. In this connection, it is interesting to note that all of the rivers which rise in the Plano Alto have clear water at least as long as they flow on its formations, while many of the bounding rivers which flow from the surrounding mountains often have yellow muddy water. Rio Bermejo of Argentina, Rio Sao Francisco of Bahia, Rio Mamoré of Bolivia, Rio Solomoes of Brazil, Rio Gurgueia of Piauhy and Rio Colo- rado* of northern Patagonia are good examples of such rivers. The yellow mud carried by all of these rivers, excepting Rio Sao Fran- cisco, is produced for the most part by the erosion of Mesozoic and Ter- tiary marine deposits; and inasmuch as these formations are not known 18 HWrom the size of the lower valley of this river and the identity of its fishes with those of Rio San Juan, I am convinced that Rio Colorado was formerly much larger and must have had some headwater from southwest Bolivia. 28 ANNALS NEW YORK ACADEMY OF SCIENCES to exist in the Plano Alto, the yellow rivers assist in a most interesting way to separate the Plano Alto from the rest of South America. The original surface of the Plano Alto, as well as its uppermost strata, dipped gently as a whole toward the southwest. This broad conclusion is based on the following facts, after due allowance is made for ages of erosion, reversal of rivers and the Tertiary rise of the Andes.'* 1. All of the streams and rivers which rise on the Plano Alto, 72. e., on the Permian sandstone, at first flow south, southwest or west, even though they afterwards flow north and east, 7. e., after they are eroded deeply into the lower strata and flow over the older Paleozoic and Archean rocks. For example, Rio Guaporé flows at first about 200 miles south and then makes an elbow bend and flows west, north and lastly northwest into Rio Mamoré. The streams of Jalapao in northern Goyaz, the headwaters of Rios Parana and Paraguay and the streams of the Guiana highlands all exhibit these same conditions in their headwaters. Even if stream piracy is said to be responsible for these conditions, the general dip would still be toward the southwest, because piracy could only be produced by the more rapidly flowing northern and eastern streams robbing the head- waters of the streams which flow toward the southwest, in order to explain the existing conditions. ‘The cases of stream piracy considered in the following pages show that it has been produced in exactly this way. 2. The second fact which supports the southwest dip of the Plano Alto is that the north and east faces of isolated mesas or portions of the original (not secondary) highlands are almost perpendicular, while the south and west sides usually have gentle slopes. This is beautifully shown by the Urucum Mountains near Corumba and by the many isolated mesas in the Jalapao region of northern Goyaz. Also the west face of the Serra de Parecis east of Villa de Matto Grosso is not nearly so perpendicular as the east face of the Serra de Ricardo Franco, west of Villa de Matto Grosso. The high Kaieteur Falls of an east-flowing river of British Guiana and the Rio Branco flowing west from the same region having only rapids support the same conclusion, 1. e., the surface of the original Plano Alto dipped as a whole toward the southwest. In view of all this, there appears to be no doubt that the Plano Alto was previously much larger than generally considered. It has been very stable since the Permian epoch. These facts may indicate a vast center of evolution of plants and animals, but I hardly think so, for even at the present time few plants and animals are able to thrive on this sandy ele- vated region. In Permian and Mesozoic times, perhaps, this region was 14 The dip is so gentle that it is difficult to detect in the strata at exposed surfaces. HASEMAN, GHOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 29 somewhat arid on account of higher, surrounding elder mountains, but even if it had sufficient rainfall, on purely lithological grounds, a luxu- riant vegetation and fauna would be impossible, because a sandy soil does not retain the necessary constituents for a luxuriant growth of plants. Besides, it appears more probable that unstable regions would produce greater changes in living things than stable regions. Hence, not until the Plano Alto was deeply eroded, could we expect to find a luxuriant growth of vegetation and a complex fauna. Also, the Triassic Parana trap which spread over much of the highlands of the states of Parana and Sao Paulo of southern Brazil may have been an important factor in extermi- nating and affecting the plants and animals, such as perhaps the Gond- wana flora and Permian reptiles which are discussed in the second part of this thesis.*® EAST ANDEAN SEA Several years ago, Professor Orton, Barrington Brown and others col- lected marine or brackish water fossils mixed with fresh-water forms along the Alto Rio Amazonas. The following is a list of the fossils which have been ascribed to Alto Rio Amazonas by Gabb, Conrad, Etheridge, Woodward and Boettger: Anisothyris tenuis Gabb Hydrobia dubia Etheridge carinata Conrad Lacuna (Hbora) crassilabra Conrad obliquus Conrad (Nesis) bella Conrad erectus Conrad Hemisinus sulcatus Conrad cuneatus Conrad Melania tricarinata Etheridge ovatus Conrad bicarinata Etheridge hauxwelli Woodward scalaroides Etheridge tumida Htheridge Dreissenia fragilis Boettger amizonensis Gabb Turbonilla minuscula Gabb Neritina (Isea) ortoni Gabb Corbula canamensis Etheridge pupa Gabb Melanopsis browni Htheridge puncta Etheridge Cerithium coronatum Etheridge eiczac Htheridge Pseudolacuna macroptera Boettger Hydrobia lintea Conrad Assiminea crassa Etheridge confusa Boettger Bulimus linteus Conrad tricarinata Boettger Anodonta batesi Woodward (Dyria) gracilis Conrad Fragments of the following genera have been reported by Woodward: Myliobatus, Fenella, Thracia, Lutria, Anodon, Unio, Nautica, Odon- stomia. Boettger has also reported Serpula (Vermes) and Rajidum and 15Jt is to be noted here that similar Permian and Triassic continental deposits exist in other parts of South America (Ceara, Brazil, San Luis, Argentina, etc.) which are not included in my Plano Alto, because they have been separated by marine deposits. 30 ANNALS NEW YORK ACADEMY OF SCIENCES Percidarium (Pisces). Etheridge has reported parts of Chara (Plante). Fossil tortoises and Mosasaurus also are said to have been found along Rio Acre. These fossils are then a mixture of fresh-water, land, brackish water and marine forms which lived apparently in a very special environ- ment. ; In reference to the age of the above fossils, I can do no better than refer to Vol. XLIV of the Bulletin of the Museum of Comparative Zoology, Harvard, pp. 25-27. In this résumé, Professor Branner states: “Tf we grant that the upper Amazon region from Iquitos to Tabatinga is Ter- tiary, there is no evidence that the mottled sediments of the lower Amazon are of the same age, to say nothing of correlating them with similar looking beds on the coast of Rio Grande do Norte, Parahyba, Pernambuco and Alagoas, 2500 miles away. This seems also to express Professor Derby’s view of the subject.” Professor Branner also quotes Dall as saying that: “The Pebas fossils are unique and difficult to determine the age because the characteristic forms are extinct and have no obvious relatives. They may be as old as Hocene or as young as Pliocene.” The maps of the location cf the marine formations show that the entire Plano Alto as herein mapped is Archean and Paleozoic; while the Andean complex has an Archean nucleus more or less covered by marine deposits of Mesozoic, and marine Tertiary is known to exist along both bases of the Andes almost for their entire length. The trend lines of this region are north and south, and they strongly indicate an extension of the Hast Andean sea in the same directions. The deposits along Alto Rio Amazonas are known from Pebas, Peru, down Rio Solomoes as far as Sao Paulo de Olivenca, and south along Rio Javary and Rio Acre. My maps show that this region lies between the Cordillera Oriental of the northern Andes and the Plano Alto. It must also be remembered that nowhere east of Sao Paulo de Olivenca, which is 1400 miles in a straight line from the mouth of the Amazon, have similar or any post-Paleozoic marine deposits been found. This fact, as well as others which have already been considered, strongly indicates that the sea did not invade the Amazon Valley from the east, because the Plano Alto would have been a permanent barrier to such an invasion from the Per- mian to comparatively recent times. The dip of the Plano Alto and the character of the sediment carried by the Rio Negro, as well as the fact that no positive evidence exists which warrants a northern extension of the East Andean Sea into Venezuela, indicate a southern extension of the East Andean Sea, because as far as I have been able to find, no Mesozoic and Tertiary are known to exist east HASEMAN, GHOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 31 of the Cordillera Oriental of Ecuador or in the southern parts of either Colombia or Venezuela, 7. e., as far south as the headwater of Rio Negro in the Brazil-Guiana highlands. This vaguely indicates a connection be- tween these Archean mountains of eastern Colombia (Cordillera Ori- ental) and the Plano Alto, but inasmuch as this region is practically unknown, too much emphasis must not be placed on such data. There is then no positive evidence for either a northern or an eastern extension of the Hast Andean Sea, but there is some positive evidence for its southern extension. The location of the marine formations, the dip of the surface of the Plano Alto and the character of the sediment carried by southern Andean affluents of the Amazon vaguely indicate the same. — The southern extension of Cretaceous deposits from the headwaters of Rio Maranhao south toward Cochabamba is a good example of the south- ern extension of part of this East Andean Sea which finally lost its con- nection with the ocean on account of the Tertiary rise of the Andes. In reference to the connection of this East Andean Sea with the ocean, the following points are important : 1. Stelzner (1873) reported a sandstone containing marine bivalves of (?) Tertiary age from Santa Maria, Catamarca of northern Argentina. 2. Brackenbush has reported marine or brackish water fossils of (?) Cretaceous near the headwaters of Rio Bermejo which flows into the Gran Chaco del Argentina. Melania was the most abundant form found, and it may be that these deposits are fresh water and not marine. 3. Far to the west of the above regions, there is a great depression in the Andean complex east of the bay of Arica and south of Lake Titicaca, but so far there is no good evidence of a ys sea passing over the top of the Andean complex. 4. It appears to be obvious from the characteristic fossils of the upper Amazon Valley that a direct and broad connection with the ocean did not exist. If, as the Cretaceous fossils of the Alto Rio Maranhao indicate, the Hast Andean Sea existed during at least part of the Mesozoic to some time in the middle or late Tertiary, there would have been ample time to evolve this peculiar fauna of Pebas, because the conditions which would have existed in this long, slender inland sea, into which many short rivers carrying sediment flowed, would have been very different from those of Patagonia, la Plata or along the Pacific slope. It appears that the fossils of Alto Rio Amazonas are the last of the fauna of the East Andean Sea which became buried in the mud carried by the rivers into this vanishing sea, because the molluscs died with their valves closed and because there are some fresh-water and land molluscs mixed with the marine forms. Beat ANNALS NEW YORK ACADEMY OF SCIENCES In conclusion, then, we may say that the age of the fossils of the upper Amazon Valley is not definitely known, and consequently we cannot more than conjecture where the East Andean Sea joined the ocean ;*® but this sea must have extended south and have had a narrow connection with the ocean in the region of either Rio Bermejo, Rio Colorado-Patagonia or the bay of Arica. This connection was perhaps not broken before the Mio- cene, but it was broken by the Tertiary rise of the Andes. It is also im- portant to note that no evidence is yet known which indicates that the exit of this sea simultaneously cut east by west both the Archean moun- tains flanking the northeast portion of Patagonia and those on the west- _ern side of Patagonia in Chile in such a way as to isolate Patagonia from Brazil. In fact, the evidence at hand is all against such a view. REVERSAL OF RIO AMAZONAS The following facts indicate that the direction of Rio Amazonas has been reversed : 1. The Plano Alto which it now traverses has a general southwest dip. 2. The entire absence of marine Mesozoic and Tertiary fossils in the Plano Alto. 3. ‘The position of the Hast Andean Sea. 4. 'The Tertiary rise of the Andes. 5. The general north and south direction of trend lines and the loca- tion of marine deposits in western South America. In an interesting treatise on the geology of the lower Amazon, Katzer (1993) concluded that previous to the Miocene the Amazon flowed west from somewhere in the region of Rio Paru. To the east and north of the present mouth of the Amazon, he conceived a vast mass of land which sank beneath the Atlantic Ocean when the Amazon became reversed.1* Katzer also states that, during the reversal of the Amazon, a large lake was formed, which extended eastward from Rio Nauta to the original watershed. It appears that he has assumed the formation of this huge lake in order to explain the formation of the fresh-water deposits of Hreré. Katzer also maintains that no marine Cretaceous or Tertiary beds exist in the lower Amazon Valley (Pirabas Cretaceous being on the 16 J am inclined to believe that it extended south along the east base of the Andes into Patagonia, because there is no evidence of an eastern extension, a western over the Andes, a northern into Venezuela. The connection by way of Gran Chaco is very ques- tionable, for I believe the Plano Alto of eastern Bolivia joins the northern extensions of the Cordova Mountains. Hence a southern extension into Patagonia meets no obvious objection. 17 Pilsbry (1911) has also suggested the same idea in his studies on the distribution of fresh water and land mollusea. It is interesting to note that three individuals work- ing in different fields have quite independently arrived at the same general conclusion. HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 33 coast) and that (?) Archean rocks underlie the alluvial deposits of the Amazon and are continuous with the basal rocks of the highlands on both sides of the Amazon. The writer has independently, and from a slightly different stand- point, arrived at the same general conclusion, namely, that the Amazon is a reversed river, but he differs with Katzer in regard to the following details : 1. The writer does not believe that there is any geological evidence which even vaguely suggests either the existence or the submergence of the hypothetical Jand-mass to the east and north of the present mouth of the Amazon River. In fact, just south of the mouth of the Amazon are marine Cretaceous fossils which indicate a slight elevation of the present mouth of the Amazon from the sea and not a continuation to the east with a hypothetical land-mass. In this connection, no support for the hypothetical old land-mass can be derived from the Barbados Islands, because they are entirely too far to the north and belong to the Antillean complex. Both the absence of islands and the deep-sea soundings are strong evidence that, previous to the Miocene, no land existed east and northeast of the present mouth of the Amazon across the Atlantic Ocean. In fact, there is no necessity for the assumption of the supmergence of any great mass of land to the east of the present mouth of the Amazon, not even to form a watershed, because many rivers, like Rio Paraliyba, Rio Tieté, Rio Iguassu, etc., of southeastern Brazil, rise within a few miles of the Atlantic Ocean and then flow several hundred miles before entering it. 2. The reversal of Rio Amazonas from the region of Rio Part is roughly comparable to the supposed reversal of the Mississippi from the region of New Orleans instead of nearer its headwaters. That is to say, the comparatively recent reversal of the Amazon will have to be. I think, in the region where the most of its large affluents enter it, namely, some- where between Manaos and Santerem; because Rios Madeira, Solomoes and Negro, which make the main stem of the Amazon, come together just below Manaos. There is topographical evidence which shows that the Rio Negro formerly entered the Rio Solomoes above its present mouth, and hence the eastward extension of its mouth indicates that its present mouth is near the region where the Amazon began to cut through the old watershed, 7. e., near the dissected western base of the original Plano Alto which has been washed away by the Amazon. The arrangements of the low secondary sierras in the region of the lower Rio Tapajos on the south side and near Obidos on the north side 34 ANNALS NEW YORK ACADEMY OF SCIENCES of Rio Amazonas, taken along with the fact that the Amazon flows in one channel only in two places (just below the mouth of Rio Madeira and near Obidos), also strongly indicate that this is the region of the previous watershed. Is it not to be expected that some trace of the original watershed would still remain, if the reversal of the Amazon took place during the Miocene? There certainly is not the slightest sug- gestion of such a divide between Santerem and the mouth of the Amazon, because the Amazon is so wide and swampy in all of this region, and this is especially true in the region of Rio Part. Below Santerem, the Amazon has many channels and many islands; between Santerem and Manaos, there are two places where the Amazon flows in one channel, and above Manaos Rio Solomoes (Amazon), there are again many channels and many islands. In view of all this, it appears that the original watershed must have been somewhere between the mouths of Rios Madeira and Tapajos, and I think that Obidos is near the actual point of reversal. 3. IT also do not believe that there is any evidence of a huge lake which extended eastward from Rio Nauta to the old watershed. This old view of Tertiary lakes has been ably combatted by Hatcher, Matthew and ethers, but it will perhaps be worth while to consider the formation of highland deposits in order that the fresh-water deposits of Hreré of the lower Amazon as well as the other supposed evidence for a colossal Amazonian lake will have an explanation. In the following brief con- sideration, I have chosen the headwaters of Rio Guaporé, but those of Rio Paraguay are equally as instructive. Many streams, headwaters of Rio Guaporé, dash down the more or less perpendicular faces of the so-called sierras from the flat surfaces of the dissected Plano Alto into a large semicircular valley. This valley includes the extensive Campos de Matto Grosso. Farther down this valley, the campos are replaced by gigantic forests which encroach upon the Guaporé so much that its channel is almost stopped up in several — places below the Villa de Matto Grosso. In this same region, the Serras de Ricardo Franco and de Parecis (made by the Guaporé dissecting the Plano Alto) are much nearer to the river than in the region of the semicircular valley. This serves as @ block to the exit of the water which falls above this point. The dense forest, fallen trees and water plants naturally assist in storing the heavy rains which fall in the semicircular valley during the rainy season (November to April). In the middle course of the Guaporé, its channel widens very much. Tremendous sand bars are encountered in each bend of the river. This HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 35 condition is replaced in the lower course of the river by a series of rapids. after which it enters the Mamoré. The Guaporé has then two great regions where sediments of different nature are deposited at different altitudes and two great regions where the water runs rapidly and carries away the eroded product. Tt is the highland semicircular valley which, I believe, is roughly comparable to the Serra de Ereré. This region is flooded yearly during the rainy season, at which time much sand is deposited. This deposi- tion of sand and other material (laterite, etc.) will eventually make a secondary deposit of considerable extent, while the regression of the headwaters of Rio Guaporé will eventually carry away all of the original Plano Alto remaining in the surrounding Serras de Parecis, Ricardo Franco and Agoaphey. In fact, the Guaporé has already carried away more than one half of the original highland formation from this region. The Rio Paraguay is naturally assisting in this destruction of the “highest point.” As soon as these two large rivers have obliterated the above mentioned sierras, the semicircular highland valley will become one of the highest points (300 to 600 meters). After years of erosion on such a new high point, iso- lated sections or mesas will be formed and secondary sierras like that of Hreré will be produced which never were associated with colossal lakes. I have observed near Corumba, Brazil, a similar deposition of leaves, snails, ete., at two distinct levels. One of these levels is around the base of the Serra de Urucum, including even deposits of limestone, and the other is in the near-by pantanals (swamps) of Rio Paraguay. The same process is going on near Sao Luis de Caceres, Brazil, and San Matias, Bolivia. Hence there is no evidence for a huge Amazonian lake and no neces- sity for assuming one, for, as the writer conceives the reversal of Rio Amazonas, it was a gradual process. The Tertiary rise of the Andes did not suddenly close the exit of the East Andean Sea. but the water cut deeply at the exit, until by the time this exit was almost closed, stream piracy in the region of the old. divide (near Obidos) had pre- pared a new exit for the water of the shallow East Andean Sea. For some time, then, there were two exits for the East Andean Sea. As the Andes rose higher and higher, the southwestern exit became closed and all of the water rushed eastward. The amount of this water flowing eastward was sufficient to wash away the old divide in a comparatively short time. Hyen the Rio Guaporé has washed away the most of the Permian for- mation of the Plano Alto from a strip 50 to 150 miles wide. Is it not | 4—_NY 36 ANNALS NEW YORK ACADEMY OF SCIENCES plausible, then, that the massive Amazon could have opened up its pres- ent wide swampy valley and built up an extensive delta in a compara- tively short time? That the Amazon has swampy margins is no more an objection to its reversal than the original swampy margins of the Mississippi would be to its reversal. The Amazon, like the Mississippi, if it was reversed, has washed away the most of the old divide. Little or no exploration has been done far away from the forested margins of the lower Amazon and its affluents. But an inland trip in any of the region between Santerem and Obidos will usually reveal sandy campos which are typical of the entire Plano Alto formations. When this region is carefully explored, I feel sure that the old divide will be definitely located. The view which I have expressed concerning the Hast Andean Sea offers a ready explanation of the origin of the peculiar marine-like fauna found in Lake Titicaca. Lake Titicaca was doubtlessly once connected with the East Andean Sea by a stream. When the Hast Andean Sea began to disappear, some of its fauna entered or was cut off in the Titi- caca basin. When the Andes rose still higher, the amount of rainfall became more and more reduced until the exit of Lake Titicaca became ‘severed. Higenmann'* states that Steinmann considers that the Titicaca basin ‘was a fresh-water basin whose southeastern exit was dammed by glaciers. Glaciers may have assisted in closing the exit of Lake Titicaca, if they were active at the time when the rainfall was so much reduced that the precipitation scarcely exceeded evaporation; but it is evident that the reduction of the rainfall due to the rise of the Andes was the more im- portant factor. Otherwise, heavy rains would soon have filled up the basin sufficiently to make a new exit. The Great Lakes of North America were glaciated, and yet they made new exits. | The absence of Manatus, Arapaima and Ostleoglossum, above the Ma- deira Falls and their presence above the falls of Tocantins, Tapajos, and in some of the coastal streams north and south of the Amazon are ex- actly what one would expect to find according to all of the facts which have been considered. ‘These three genera originally lived in the coastal streams. When the Amazon became reversed, the Mamoré, which had previously flowed southwest, changed its direction and suddenly formed the Madeira Falls, which have been barriers to the migration of these three genera. Further zodlogical and topographical data could be given in support 48 Princeton Patagonian Reports, Pt. III, p. 372. 1909. HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 37 of the reversal of Rio Amazonas, but inasmuch as the most of it is not sufficiently decisive to account for such a momentous geological change, it has been omitted; for such data may be said to harmonize equally well with any view of the region in question. Finally, in reference to the time of the reversal of Rio Amazonas, I may restate that Katzer has placed it in the Miocene. It appears that it could not have been before that epoch, but until the exact age of the “Hast Andean fossils’ has been determined, we cannot settle this most important question with any degree of certainty. The determination of the exact age of the fossils from Alto Amazonas and the exact location of the exit of the Hast Andean Sea offer alluring opportunities for future exploration. STREAM PIRACY On the divides between various South American rivers, the headwater streams sometimes approach one another with closest intervals. This is particularly true of the divides on the Plano Alto. For example: Rios Sao Marcos and Bartholmeo of Rio Parana; Rios Bezzero, Jardin and Preto of Rio Paracatu and Rio Urucupa, both of which flow into Rio Sao Francisco, and Rio Parana, an affluent of Rio Tocantins. All of these rivers flow from brejos (swamps or sloughs), Lagoa Feia and other highland lagoons between the villages of Paracattt and Formosa of southeastern Goyaz. In fact, these headwaters of three large river basins Tise in sight of each other from brejos which vary from 1100 to 1147 meters of altitude, and the maximum altitudes of the intervening sand hills are never more than 1177 meters.’® Stream piracy (coalescence of streams) may have existed on a very limited scale between the following rivers—at any rate, their headwaters are not far apart: _ 1. Rio Ribeira de Iguape (? robbed Rio Capella) and Rio Paranapo- nema. ®. Rios Sao Francisco and Doce and Rio Grande of the Parana near Carandahy and Miguel Burnier, Minas Geraes. 3. Rio Parahyba and Rio Tieté, Rio Parahyba having robbed its headwaters flowing south from Rio Tieté. 4. Rios Araguay, Xingu and Tapajos and Rio Paraguay. 5. Rio Ibicuhy of Rio Uruguay and Rio Vaccachy of Rio Grande do Sul. 19 The highland fauna and flora certainly interchange between these three river basins, but the typical Amazonian, Sao Franciscan and Paranean fauna and flora are at least one hundred miles away from this headwater region, 7. e., at a much lower altitude. 38 ANNALS NEW YORK ACADEMY OF SCIENCES 6. Rio Sao Francisco and Rio Itapicurt east of Joazeiro. ?. Rio Agua Branca and Rio Negro of Rio Sao Francisco and Rio Palma of Rio Tocantins. 8. Rio Grande of Rio Mamoré and Rio Pilcomayo of Rio Paraguay. 9. Rio Branco of Rio Negro, Rio Part, Rio Trombetas and Rio Esse- quibo. 10. Corrego da Boa Ventura of Rio Guaporé and Corrego de la For- tuna of Rio Paraguay. Even if these streams, however, were previously connected, the con- nection would have been so small and at such altitudes that nothing but the highland fauna could have interchanged. JBesides, all these streams have and have always had waterfalls in some part of their courses. | Plate XIII shows in a general way that no continuous waterway exists between Rio Paraguay and Rio Guaporé, as has been so often erroneously stated. Several years ago, an attempt was made to cut a canal between Rio Guaporé and Rio Jaurt, but it was given up on account of the intervening clistance (about 20 miles) and the nature of the material to be removed. c. grandson of the man who attempted to make this canal hauls rubbe2 over this divide in an ox cart, but it is far more difficult than by the Bolivian trail from San Ignacio past San Matias to Descalvados. The former trail (Villa de Matto Grosso by the way of Jaurt to Sao Luiz de Caceres) is only 301 kilometers long, while the latter trail ( Villa de Matto Grosso to Bastos, las Encruzijadas, San Matias and Descalvados or Sao Luiz de Caceres) is 488 kilometers, but it is much smoother. While following this latter trail, I had ample opportunity to observe that no connection exists between the Rios Alegrete and Agoaphey. In fact, Rio Santa Rita flowing off this same sierra has one waterfall about 400 feet high. All of these rivers are nothing more than creeks. The writer found that the nearest as well as the lowest approach be- tween the headwaters of Rios Guaporé and Paraguay is in the region of the Corrego de Boa Ventura and la Fortuna; but even though there is a break in the continuity of the Plano Alto in this region, firm hills of con- siderable height separate these creeks, which are not more than ten feet wide and three feet deep and at least four miles apart. Dr. Alipo Miranda de Ribeiro, Secretary of the Brazilian National Mu- seum, was a member of the telegraph commission which has explored the northern portion of the Paraguay River. He explored Rio Jaurti and Sepatuba and found large waterfalls in each of these rivers. On the other side of the divide, the commission found two large waterfalls in Rio Juruena which flows into Rio Tapajos. One of these falls was about 400 We Sy Son SE rpg AWS < “ay S00 Luiz de Gaceres. a Don VicTo S Apts Gipsy Ung Sy HW oN Ley, a HM oy Uh A ll ys Wh Me LAVA are MH folly, Mele, Te Stra de eT EY My ' i Ra aw wl i RS U7 Ot H CO 2 Op. cit., p. 13. HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 51 of the genera are found in the Amazon valley, we have yet absolutely no evidence that the ancestral Cichlide originated in the Amazon. In fact, as Matthew has already pointed out, the presence of a large number of species in a given locality is no evidence as to their point of family origin. Indeed, neither the genus nor the family need have originated there, since, as Matthew has shown, the point of origin is apt to be the first place of extinction. We may note, therefore, that the above rules which have been used to determine the point of origin of any group of animals have only a hmited application. So far, there have been only a few specialists interested in the ichthy- ology of South America. Of these, Professor EHigenmann has certainly given us the best treatise on the distribution of the fishes. In fact, his publications are the only ones which deal in a comprehensive way with the great mass of these fishes, including as they do almost two thousand species. In the problems of the origin and dispersal of Cichlid and Characinide, Professor Higenmann (1906) has given interesting data. He has prepared a hypothetical map which indicates that the Cichlid dispersed from east- ern Guiana and the Characinide from the Amazon. This conclusion ap- pears to be based on the fact that the most of the genera of Cichlide and Characinidz are found in these regions. As I have already stated, how- ever, this is no evidence for the point of origin and subsequent dispersal of a family of animals. If it is, we might erroneously conclude from the present distribution that the deer and tapir originated in the state of Matto Grosso, Brazil, and the camel in the Andes, because more of the species are found there; but in these latter cases, paleontology has shown that the first point of origin was in the northern hemisphere, where the species no longer exist. I hope to show in the following pages that the point of origin of many species and genera of living Cichlidz has been in the Amazon, but that the point of origin of the ancestral Cichlide was not in South America. In another paper, Professor Eigenmann draws the following conclu- sions in his extended discussion of the distribution of the South American fishes : : 1. The fishes of South America exhibit no close affinity with those of North America. 2. The South American fishes, certainly the Characinide and Cichlid, lend support to the Archhelenis theory. 3. The fishes of the coastwise streams of eastern Brazil differ more widely from the Amazonian than do the Paraguayan. 4. The distribution of the fishes indicates that South America was D2 \ ANNALS NEW YORK ACADEMY OF SCIENCES ‘divided into a northern and a southern part, 1. e., Archiplata and Arch- -amazonia of von Ihering, and a connection appears to have existed be- ‘tween Guiana and Africa. There may be little reason to question the data of Professor Higen- ‘mann, but there is, I believe, much reason to question his interpretations. ‘The first conclusion, we believe, is to be accepted for the living affinities of the South American fishes, but when we are dealing with their points of family origin, we are in most cases concerned with ancestral forms “which have been dead for geological ages. In this case, the fossil record, ‘seanty though it be, shows affinities between the hving South American cand the fossil North American fishes.?°@ Conclusion number two is also questionable, because: ‘1. There is no geological support for the Archhelenis theory. This is all the more true for the late Cretaceous, when the Cichlide probably originated. 2. The point of origin and dispersal of the Cichlids, as I propose to show in the following pages, was not correctly determined. In the matter of his third conclusion, it should be said in Professor Eigenmann’s behalf that he did not then know that the Paraguay was not connected with the Guaporé and that the Sao Francisco was connected with the Tocantins. He also did not know which waterfalls were not barriers for fishes and that both the coastwise streams and the Alto Rio Parana have more than double the number of species which he assigned to them. In the same report, Professor Higenmann also states that, of the number of species of fishes, 60 per cent of the Guianan, 40 per cent of the Sao Franciscan, 53 per cent of the Paraguayan, 30 per cent of the coastwise streams of eastern Brazil, 42 per cent of Trinidad and 6 per cent of Cen- tral American are Amazonian. In this static comparison, he has intro- ‘duced a probable source of error due to the environments when he draws conclusions from the above data. His lists include the fishes from the entire basins of the coastwise streams, including Rio Sao Francisco, the entire Amazon basin and only the central and upper Paraguay River, which is only one of the affluents of the great La Plata basin. If we com- pare the massive Amazon with the entire basin of the coastwise streams, as 1t not necessary to compare the entire a Plata basin and not only one part, 7. e., only one environmental complex? For example, if we should compare the fishes from Rio Sapao of the Rio Sao Francisco with those of the mighty Amazon and its affluents, 100 per cent would be Amazonian, 2a See H. F. OsBorn: “The Age of Mammals.” New York. 1910. HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 53 and hence it would be more like the Amazonian than are the Paraguayan ; but in this case we would compare only one environmental complex which is duplicated in the enormous Amazon Valley. (In cases lke Central America, the zodgeographer is concerned more with the ancestral dis- tributed form than with the recent cenogenic modifications of them due to the environment. ) In the same lists, we find 122 species for the lower Parana, only 64 of which, or 52 per cent, are Paraguayan. If this could be true, then the Paraguayan fishes are more like the Amazonian than the Paranean. This, of course, is to a certain degree quite absurd, because small ocean steamers can sail up the La Plata into the central course of either Rio Paraguay or Rio Parana. In this case, according to the old view, fishes would, in the first place, have to find their way overland from the Amazon Valley to the Paraguay (a distance of about 200 miles separating the typical fauna of the two basins), and then, for some unknown reason, remain in the Alto Rio Paraguay and not venture to swim down the Paraguay into the Parana. Professor Eigenmann also states that the Alto Rio Parana had only 31 species, but I have often collected more than this in a single day. I col- lected more than 100 species in the region of the Alto Rio Parana and its affluents, and there is no reason to believe that the list was then exhausted. The point brought out in the above brief review is that far too little is known about the fishes or any other South American fauna to prove any hypothesis by a numerical comparison of the species found in diverse regions. If, for example, we add Cichlasoma bimaculatum and Ovenicichla lepi- dota to the Sao Franciscan fauna and Gymnotus carapo, etc., to the coast- wise streams, and if we compare the entire river basin, there can be no doubt that some of the faunal regions, and especially the cause of the differences in their fauna as explained by Professor Higenmann, do not agree with the actual facts. The alleged support derived from the fishes for an Archiplata, Archi- guiana and Archamazonia needs no discussion, because the geological evi- dence shows that no post-Paleozoic seas have invaded the Plano Alto. In the case of Patagonia, there have always been two possible connections with the Plano Alto, one by the Cordova and the other by the Archean rocks of the Andean region. Inasmuch as the invasions of the sea were usually north and south, there is no evidence that southern South America was completely isolated for a long period from the rest of South America by an arm of the sea.?¢ °6 See PILsBry, 1911. 54 ANNALS NEW YORK ACADEMY OF SCIENCES Use of Characters The problem of defining the characters or the kind of characters which distinguish species, genera, etc., has never been satisfactorily solved. The writer will venture to consider only that part of it which concerns the distribution of South American fishes. In the main, there are two schools of ichthyologists, the American and the English. The American school makes many divisions of the families, genera and in some cases the species. ‘The English school has, as a rule, been more conservative with taxonomic divisions and has therefore fewer but larger groups. In support of the American school may be given the results of the excellent experimental evidence obtained by de Vries, Tower, Johannsen and others, which indicate that the sys- tematic species is a complex one. In other words, this experimental evidence tends to split up the species of the systematist into several ele- mentary species. Much can be said in favor of this finer analysis of species from the experimental standpoint, but little can be said in favor of it from the standpoint of the systematist, because he does not know whether his specimens are hybrids, whether they have a wide range of fluctuating variation, whether they are mutations or whether the pecu- liarities of the observed somatic differences are inherited or not. Therefore, from the standpoint of geographical distribution, it ap- pears that the English system, with its fewer divisions and divisions based on more than single characters, is the better one, at least until we have analyzed our species experimentally. In reference to what characters are important from the standpoint of the fish geography of South America, we are exceedingly fortunate, at least in the case of Priscacara, a fossil cichlid described by Cope from the Eocene of Green River, Wyoming.?’ From this interesting genus and from a comparative study of the South American Cichlid, we are able to state with a high degree of certainty that the ancestral Cichlide had the following characters: Three anal spines; short gill rakers; more than one row of short conical teeth in each jaw; pharyngeal teeth; ctenoid scales; serrated preoper- culum; a continuous spiny and rayed dorsal with more than eight spines; single naris or a tendency for narial coalescence; a rather short, deep body, and a tendency to form a two-parted lateral line.2* *7T have examined some of Cope’s types and believe that Woodward and Pellegrin are correct in considering Priscacara a fossil cichlid. 28 Jt must be granted that it is difficult, if not impossible, to decide in the case of all of the characters of fossil and living forms which characters are paleotelic and which are cenotelic, but we can agree on at least a sufficient number to show that no living cichlid fish could have given rise to those of both Africa and South America. HASEMAN, GHOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 55. These, then, were actually some of the characters of the primitive Cichlide of the eastern and western hemispheres. | These ancestral, primitive or phylogenetic characters may be desig- nated paleotelic, a term which has already been used in a similar sense by Gregory in his book on the orders of mammals. In contrast to these paleotelic cichlid characters are the four to thir- teen anal spines of Cichlasoma and other genera, the long gill rakers of Chetobranchus and Chetobranchopsis, the lobe on the upper branch of the first gill arch of Geophagus and Heterogramma, the long teeth of Ptemia, the chisel or incisor teeth of Uraru, the long and more slender body of Crenicichla, etc. These recent, adaptive or physiological char- acters may be designated cenotelic. ‘There is no evidence that any of these cenotelic characters have been distributed anywhere excepting in South and Central America, because it is these and other characters which distinguish the Cichlide of the western hemisphere from those of the eastern. I cannot overemphasize the importance of paleotelic and cenotelic characters, because many zodlogists and paleontologists have not made any distinction between these two types of characters in their tabulated comparisons of various faunal regions. In the case of cichlid fishes, cenotelic characters have evidently to do with the origin and dispersal of variation, species, etc., while paleotelic characters deal with the ancestral fauna which gave rise to these genera. The paleotelic characters have to do with the ancient distribution, hence theories like Archhelenis; while the cenotelic have to do with the present distribution of a given genus. ‘The cenotelic characters are usually modified by the action of the environment on the ancestral forms of a given genus, while the paleo- telic characters in part have extended down through all of the genera of the cichlid family. Therefore, from the standpoint of the origin and lines of dispersal of the Cichlids, a few paleotelic characters will out- weigh a bookful of cenotelic ones. What shall we learn, then, by a careful compilation of all of the Cich- lide of South America and Africa and by comparing all of those found m one river basin with those found in another? Would this show that a connection had existed between certain points of Africa and South America? Or would it merely be a compilation of cenotelic characters formed by the action of the environments of the different localities on the ancestral Cichlide which possessed paleotelic and not the recent cenotelic or secondary characters ? ‘ From the standpoint of the origin of the ancestral Cichlids, then, the living species or the specific characters alone give us no clue, because 56 ANNALS NEW YORK ACADEMY OF SCIENCES they are based on cenotelic characters. The generic characters are often no better, because they, too, are usually cenotelic. Take, for example, the genus Cichlasoma, which is primarily distin- guished by the presence of four to thirteen anal spines from the genus Afquidens, which has only three anal spines. For all we know, a muta- tion with four anal spines could have easily appeared from the three anal forms, or vice versa. In fact, that is exactly what seems to be the case in Rio Sao Francisco, where I found only two specimens with three anal splines among thousands of the four-anal-spine form, Cichlasoma bimac- ulatum. In southern Brazil, I found just the opposite, namely, only three four-anal-spine forms of the same size, color, etc., as thousands of three-anal-spine forms which are known as Avquidens portalegrensis. Hence all such closely drawn genera must be used with great care in the determination of the point of family origin and dispersal, because they are based on such characters that a mutation or individual or local varia-- tion might readily establish a new genus and species in a very short time. The importance of having a clear idea of the exact status of the ich- thyological taxonomy, and especially the value of generic and specific character and which are paleotelic and which are cenotelic, are of prime importance in the explanation of the distribution of the fishes. For ex- ample, the writer described Geophagus brasiliensis iporangensis as a new variety from the headwaters of Rio Ribeira de Iguape. This variety may be a distinct species, and it may be only a somatic change due to the en- vironment. ‘The adult forms certainly look and measure like a distinct species, but I observed that the young individuals, one and two inches long, could not be told from the lowland young of Geophagus brasiliensis. It is highly probable that this new variety would produce typical Geo- phagus brasiliensis, if it were removed to the lower course of the same river. ‘The same may be true of Crenicichla iquassuensis, also described by me from the Rio Iguasst. This species is closely related to C. lacus- iris and would perhaps produce that species, if subjected to the same ex- ternal conditions. These two examples illustrate the present status of South American ichthyology. One can, I believe, readily separate out a couple of hundred forms of the two thousand catalogued for South America which are probably not species, but merely “ontogenetic species,” showing, for example, somatic changes of a color spot, a few more or less scales and spines or some other trivial physiological difference. Many of these species, in fact, are based on single characters, such as teeth, which are variable structures. The same have been shown experimentally with birds, beetles, butterflies, etc., ving in different temperatures, etc., and until the fishes are better and experimentally known, their present HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 57 static distribution offers only hypothetical evidence of the shakiest kind for any theory. Ichthyological Faunal Regions In the first part of this thesis, the writer briefly outlined the principal environmental complexes in which the sum total of the natural condi- tions was about equal in each complex. In the following pages, I pro- pose to analyze the changes which have been wrought by the environments on the distribution of the fishes: but before making the analysis, it is first necessary to consider the range of the fishes. The Siluride are found everywhere in South America. The Cichlidee and Characinide are found everywhere north of Patagonia and east of © the Andes.. A few Characinide and Cichlid are found on the western slope of the Andes of Peru and Central America, and a few are also found in the West Indies. Two species of Characinidz are also found in northern Patagonia. The few species found in the West Indies diminish in number of genera the farther the islands are from the mainland ‘There is thus a suggestion that the Cichlide and Characinide came directly from South America by way of the sea. I have no doubt that this is true, since by actual experiment I have determined that certain of these genera will live for some time in sea water. The Characinide have never been completely revised, and consequently the genera cannot be of equal value. Notwithstanding this lack of re- vision, the following summary of the genera is instructive. Of the 129 genera of Characinide which have been described, 94 are found in the Amazon Valley, 50 are widely distributed, 60 are found in the Paraguay, 64 in the Guianas, 58 in the Orinoco, 54 in the La Plata, 41 in the Sao Francisco and 37 in the coastwise streams of southeastern Brazil. The _ Paraguay harbors about 60 genera of Characinide, 58 of which are found in the Amazon. The species belonging to these genera are not usually as similar as are the genera, for example, of the 118 species of Characinidee reported by Higenmann from the Paraguay, only 63 are found in the Amazon, while 45 of his 47 Paraguayan genera are found in the Amazon. Do the above data indicate a direct connection between the Paraguay and the Amazon? No, and for the following reasons: 1. There is at the present day no connection and no indication of an ancient connection, at least since the present ichthyological fauna has developed, at such an altitude as to be favorable for lowland forms to cross from one basin into the other. 2. No connection is known to exist between the rivers of Guiana and 58 ANNALS NEW YORK ACADEMY OF SCIENCES the Amazon, and yet there is as great, if not greater, identity of fishes there than in the case of the Paraguay. 3. There are connections between the Orinoco and the Amazon and the Sao Francisco, and yet there is less similarity between their faunas than between these of the Paraguay, Guiana and the Amazon. The key to the explanation of the distribution of the characinids lies in the 50 genera which are not only widely distributed, more or less cos- mopolitan, but the more generalized members of their families. The same is also strongly indicated by the fact that when these widely dis- tributed generalized genera arrived in the different environmental com- plexes, there resulted less conformity between the species than between the genera and a greater similarity between the species which lived in similar environments, even though they are not connected, than between the species in dissimilar environments which are connected. I have already shown that the Plano Alto separates the La Plata basin from the Amazon. This being the case, we have to look for another ex- planation of the distribution of the fishes other than river connections. The most natural way to seek the explanation is first to consider the fishes which live on the Plano Alto. I have found the following genera on the Brazilian highlands (a max- imum list) : Siluride: Callichthys, Pydidiwm, Rhamdia, Pimelodus, Pimeladella, Plecostomus, Doras, Trachycorystes, Auchenipterus, Heleogenes, Loricaria, Hoplos- ternum and Corydoras. Characinide : Erythrinus, Hoplias, Holerythrinus, Characidiwn, Oreatochanes, Pecilu- richthys, Acestorhynchus, Curimatus, Moenkhausia, Astynax, Tetrago- nopterus, Phyrrhulina, Pecilocharaxz, Serrasalmo and Chalcinus. Cichlide: Geophagus, Crenicichla, Aiquidens, Cichlasoma and in two places Hereto- gramma. Rivulus, Symbranchus marmoratus, Hypopomus brevirostris, Gymnotus carapo, Higenmannia virescens and Sternopygus macrurus. All of the above genera are not only widely distributed but also the most generalized types of their sub-families. They are represented in the highlands proper by only a few species. For example, I found only twenty-five species in the highlands of Parana above the big Iguassu Falls. The following is the list which I collected in the highlands of northern Goyaz: HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 59 Hoplias malabricus, Bloch. Acestorhynchus falcatus, Bloch. Hoplerythrinus uniteniatus, Spix. Characidium fasciatum, Reinhardt. Curimatus elegans, Steindachner. Creatochanes sp.? Moenkhausia oligolepsis (?) Gunther. Astaynax bimaculatus, Linnzus. Crenicichla lepidota, Heckel. Cichlasoma bimaculatuwm, Linnzeus. A similar paucity of species has been noted by Eigenmann in the high- lands of Guiana. He has kindly given me the following list which he collected above the Kaieteur Falls: Rhamdia quelens Pecilurichthys bimaculatus Heleogenes marmoratus Astyanax mutator “Pydidium gwianense Hoplias malabricus Callichthys callichthys Hoplerythrinus uniteniatus Lithogenes villosus Erythrinus erythrinus Corymbophanes andersoni Gymnotus carapo Phyrruhlina filamentosa Hypopomus brevirostris Pecilocharazx bovalli Rivulus holmie Moenkhausia oligolepis Ajquidens potarensis Moenkhausia brown Heterogramma ortmanni Creatochanes affinis Crenicichla alta It is at once evident from the above that there have never been found more than twenty-five species in any one locality on the Plano Alto. Compiling the possible lists from various localities, however, we might obtain as many as fifty species as inhabitants of the Plano Alto. If we even doubled this probable list, there would still be about fifty species, many belonging to lowland genera common to the Paraguay and Amazon valleys, to account for in some other way than by a possible overland passage. Hence other factors than mere land-bridges, river connections, etc., are involved in the present distribution of South American fishes. It has already been stated that the highland genera, usually small in size and widely distributed, are the generalized types which have pro- duced the bulk of the ichthyological fauna. In fact, the highland genera which I have enumerated include 33 per cent of the 1917 species which have been reported from South and Central America by Higenmann. The highland genera of Cichlid include 150 of the 187 species of Cich- lide reported by Eigenmann from South and Central America. The above statistics are sufficient to show that most of the purely fresh- water fishes are directly related to the highland genera which continue to 60 ANNALS NEW YORK ACADEMY OF SCIENCES enter practically all of the rivers north of Patagonia and east of the Andes. None of these highland genera need a direct connection between the river basins, because they are found not only above high waterfalls - which have been barriers, but also in all of the river basins, some of which are distinctly separated. In this connection, I may note that I saw young Hoplias swimming during a heavy rain in a trail over the highlands of northern Goyaz fully two miles away from the nearest rill. I have already stated that the highland genera directly account for the distribution of at least 33 per cent of the entire ichthyological fauna without the necessity of direct connections between the different river basins. Neither river connections nor the existing highland genera, however, will account directly for all of the species in common between any of the rivers; nor will they explain the more difficult question why so many of the common species have remained identical in river basins which may or may not be connected. In order to answer this most diffi- cult question, I have chosen the cichlid fishes, because they are the best known of any large family of South American animals. In brief, the question is, Why are there more species of Cichlid in common in Guiana, the Paraguay and the Amazon than in the Parana, Uruguay, the coastwise streams of southeastern Brazil aad the Amazon? The following list gives the genera of Cichlide and their distribution. Those marked with an asterisk (*) have few species and are more closely drawn than the others. The word “general” means everywhere north of Patagonia and east of the Andes. * Chetobranchus, Amazon to North, * Chetobranchopsis, Amazon and Paraguay, * Cichla, Amazon to north, Orinoco and Guiana, * Uraru, Guiana and Amazon, * Herotilapia, Lake Managua, * Neetroplus, eastern slopes of Mexico and Central America, * Acaropsis, Amazon, Guiana and Orinoco, : * Petenia, Lake Peten, * Tomocichla, Costa Rico, Herichthys, Texas to Guatemala, * Astronotus, Paraguay, Amazon and Orinoco, * Vannacara, Essequibo, Ajquidens, general and western Ecuador, Thorichthys, eastern slopes of Mexico and Central America, Cichlasoma, general and both slopes of Central America, Crenicara, Amazon and Guiana, Crenicichla, general, * Retroculus, Amazon, Heterogramma, Paraguay, Amazon and Guiana, Geophagus, general, HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 61 * Symphysodon, Amazon, * Pterophyllum, Amazon, Guiana and Orinoco, * Biotecus, Saraca in Amazon, ; * Paranecetroplus, Rio Sarabia, Mexico. According to the static viewpoint of animal geography, we shouid con- clude from the above data that the Cichlide originated or dispersed from either the central or the northern part of the Amazon Valley, because sixteen of the twenty-three genera are found there. Five other genera are found only in Central America. All of the seven genera which are © found south and east of the Amazon basin are found also in Rio Ama- zonas. Statically, also, we could interpret the entire absence of Cichlidze from Patagonia as meaning that this group had a more northern origin; but inasmuch as the genus A’quidens®® possesses more of the paleotelic characters of the ancestral Cichlid, it is evident from its distribution that it may have originated in many places not embraced by the mighty Amazon. Also the fact that at least eight of the sixteen Amazoniam genera are highly specialized, 7. e.. cenotelic, may be taken as evidence that these forms have evolved in this region from less specialized forms whose center of origin was not necessarily in the Amazon. In fact, the Amazon has only two genera which are not found elsewhere in South America, and both of these genera are closely drawn and contain only one species each.. Therefore the above list of genera and all that is known about them offer no conclusive evidence that the majority of the living American Cichlid originated in the Amazon. They may equally well have origimated, as far as the above evidence shows, in either Guiana, the Ormoco or anywhere on the old Plano Alto. Before attempting to deter- mine the poimt of origi of the Cichlidx, I will first put some of the typieal genera im their environmental complexes in order to explain their present distribution: iA Rro Uruguay anp Rio GRANDE DO Sut, INCLUDING PART OF THE LOWER La PLAtTs This complex. is characterized by medium to low altitudes, sub-tropical to temperate climates, campos, slow flowing water and little or no forests. This region harbors the following species of Cichlide: 23 Pellegrin (1904) considers this the most primitive genus, but Crenicara also has: many paleotelic characters. 6—NY 62 ANNALS NEW YORK ACADEMY OF SCIENCES ( balzanii, Geophagus brasiliensis / brachyurus, gymnogeyns, gymnogeyns, brachyurus, balzanii, Aiquidens portalegrensis — Cichlasoma bimaculatum, haying only one anal spine more, Cichlasoma facetum, Crenicichla lepidota, southern form of C. saxatilis, lacustris, vittata, southern form of C. macrophthalmus. These species are found on both sides of the divide in the State of Rio Grande do Sul. The region is characterized by Geophagus gymnogeyns and brachyurus. ALTO PARANA AND COASTWISE STREAMS OF HASTERN BRAZIL This complex is characterized by higher altitudes, numerous water- falls, rapid water, sub-tropical temperature and forests. It harbors the following species of the Cichlid: Geophagus brasiliensis with two varieties, Bis 3 CO. jaguarensis Crenicichla lacustris i 0 eee & lensis, vittata (C. dorsocellata, southern form of C. macrophthalmus), lepidota, southern form of C. sazxatilis, Cichlasoma facetwin (C. autochthon and C. oblongwm being synonymous), Cichlasoma bimaculatum, northern form of A/quidens portalegrensis. These six species and their varieties are found on both sides of the divides between the headwaters of the Alto Parana and the coastwise rivers of eastern Brazil. This region is characterized by two varieties of Geophagus brasiliensis, two varieties of Crenicichla lacustris and one variety of Crenicichila vittata. Whether these varieties will breed true or whether they exhibit only somatic changes which are not necessarily inherited is not known, but they show, at any rate, changes due to the peculiar environments in which they live. For example, the young of Geophagus brasiliensis var. iporangensis could not be distinguished from the young of the common form, but the adults were strikingly different. The variety lived in the rushing headwaters of Rio Ribeira and the com- mon form, G. brasiliensis, lives more in the lowland sections of the rivers, lagoons and swamps. I observed a similar change from lowland to high- Jand forms near Santos. I also observed that Geophagus brasiliensis HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 63 occasionally is taken from salty water. I put this species directly out of fresh into a bucket full of sea water. It was able to live several hours under these conditions. Sao FRANCISCO AND THE Secca (Dry) REGion or NorTHEASTERN BRAZIL This region is characterized by desert-like flora, no forests, scanty rain- fall and occasional long dry seasons during which many rivers become dry. The altitude is medium (246 m. at Jatoba above the Paulo Affonso Falls). This region harbors the following species of the cichlid fishes : Cichlosoma bimaculatum,® Crenicichla lepidota, Geophagus brasiliensis. It is interesting to note that the above three genera are also the only three which are found in Rio Grande do Sul, where, in place of having the same three primitive species, six new more cenotelic species have evolved from the above three more generalized widely distributed forms. The only explanation is that the environment of the muddy semi-arid Rio Sao Francisco has not been conducive to either the production-or the maintenance of new species, because the original species were identical for both of these environmental complexes, i. e., they came from the same ancestral highland stock. THE PARAGUAY AS PaRT OF THE AMAZONIAN COMPLEX The great swamps, called pantanals, of the Paraguay are in all respects the exact southern counterpart of part of the Guaporé and the central portion of the Amazon basin. Their similarity must be great, because they all lie in the confines of the central portion of the remains of the Plano Alto. This is not true of either the Parana, Uruguay or Sao Francisco rivers, as well as the Andean affluents of the Amazon. The similarity is further very striking in altitude, while the Parana is three or four times higher. Inasmuch, however, as the Amazon Valley is so large, it duplicates several times the natural conditions found in the Paraguay as well as in several other rivers. This duplication includes temperature, altitude, food, volume of water, swamps, nature of currents, humidity, rainfall, nature of sediment and muddy and clear water. °°T collected two specimens with three anal spines which were exactly like Hgwidens portalegrensis. 64 ANNALS’ NEW YORK ACADEMY OF SCIENCES Given then this duplication of environmental complexes and given also the same common generalized widely distributed genera of Cichlid, may we not also expect some similar changes in the common germplasm? The Paraguay harbors the following species of cichlid fishes: Chetobranchopsis australis, southern form of C. orbicularis, Astronotus ocellata, Alquidens paraguayensis, southern form of A. teteramerus, portalegrensis, southern form of Cichlasoma bimaculatum, dorsigera, Cichlasoma festivum, Orenicichla simoni, perhaps synonymous with C. reticulata, semifasciata, perhaps Synonymous with C. cyanonotus, lepidota, southern form of C. saxatilis, vittata, southern form of C. macrophthalmus, Heterogramma teniatum, trifasciatwn, borelli, giving off H. ritense, corumbe, Geophagus balzanii, jurupari. Of the sixteen species of Cichlid, only five are found in Rio Uruguay and Rio Grande do Sul, only three are found in Rio Sao Francisco and ~ only three are found in the Alto Rio Parana and the coastwise streams of southeastern Brazil. These three are widely distributed species. All of these sixteen species excepting five are also found in the Amazon, and of these five, two are Heterogramma, which are connected by intermediate stages in such a way that Heterogramma teniatum can easily give rise to all the species of this genus. Two of the other species which are not found in the Amazon are Crenicichla semifasciata and simon, but these are still questionable species, and they also may even exist in the Amazon. In order to explain this identity of the Cichlide, I might even grant that all of these species have in some unknown way interchanged between the Agpazon basin and the Paraguay. If, however, I should grant so much as that, I should then find even greater trouble in explaining not only why the rest of the Cichlide of the Guaporé were not able to get in the Paraguay, but also why the Paraguayan species have remained identical with those of the Amazon and why the Paraguayan species have not in- vaded the Rio Parana and Rio Uruguay, all of which have navigable channels in their lower courses. In other words, the La Plata basim has a triple cichlid fauna, and these correspond exactly with the natural en- vironments of the Alto Parana and the coastwise streams of eastern Brazil, Kio Uruguay and Rio Grande do Sul, and Rio Paraguay The HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 65 changes wrought by these environments on the more generalized high- land genera is adequate, I believe, to account for the present distribution of the Cichlide, but this does not explain their point of origin. Before taking up this latter subject, it is necessary to produce further evidence showing that it is the action of the environmental complexes on widely distributed genera which has produced the present distribution of the South American fishes and not direct river connections or inter- mingling of species and isolation. Rio AMAzONAS REGION The following is a list of the Cichlid of Rio Amazonas :*+ Chetobranchus flavescens, Guaporé, semifasciata, - Chetobranchopsis orbicularis, Cichla temensis, Orinoco, ocellaris, Orinoco, Guiana, Guaporé, Uraru amphiacanthoides, Acaropsis nassa, Guaporé, Astronotus ocellaris, Guiana, Orinoco, orbiculatus, Ajiquidens teteramerus, Hssequibo, Guaporé, vittata, Colombia, Guiana, paraguayensis, Guaporé, subocularis, Guiana, portalegrensis, Guaporé, dorsigera, Guaporé, duopunctata, zamorensis, guaporensis, Guaporé, awani, Guaporé, Cichlasoma bimaculatum, festivum, Guaporé, severum, Guiana, Guaporé, psittacum, Orinoco, spectabile, . coryphenoides, Crenicara altispinosa, Mamoré, - maculata, punctulata, Guiana, i This list does not exactly agree with my report from the Carnegie Museum, which was slightly changed by Professor Eigenmann. Perhaps somé of the omitted species should be added, but I am inclined to believe more should be dropped, even @ few de- scribed by the writer. In the main, however, this is the most accurate list at hand and is sufficient for its present purpose, including as it does all the genera. 66 ANNALS NEW YORK ACADEMY OF SCIENCES Crenicichla reticulata, Guiana, cyanonotus, lepidota, Guaporé, saxatilis, to the north, lucius, Guiana, macropthalmus, Guaporé, acutirostris, lenticulata, Guiana, strigata, cincta, johanna, Venezuela, Guiana, Guaporé, lugubris, Venezuela, Guiana, Guaporé, santeremensis, Retroculus ladifer, Heterogramma teniatum and a variety, Guaporé, agassizi, Guaporé, trifasciatum and a variety, Guaporé, corumbe, Guapore, Geophagus surinamensis, Guaporé, cupido, Essequibo, jurupari, Guapore, acuticeps. Biotecus opercularis, Symphysodon discus, Pterophyllum scalare. It may at first sight appear strange that the Amazon harbors so many species and genera of Cichlid, but it is exactly what one would expect because of its vast size and tropical location. Of these fifty-three species, at least twenty-two are found in Rio Guaporé, and of these twenty-two, twelve are not found in the Paraguay. This fact alone is sufficient to disprove any wholesale exchange of fishes between the Paraguay and the Guaporé. From the standpoint of phylogeny, I am able to throw a little light on the above distribution of the Cichlide. These conclusions were derived from both the field and laboratory, and inasmuch as the purely systematic data have already been published by Pellegrin, Regan and more recently by myself, I will not repeat them. 1. I consider Geophagus brasiliensis as the most primitive of the living members of this genus. It is interesting to note that this species is not found in the Amazon and that its nearest northern ally is Geophagus steindackneri, which was originally described by Steindachner as Geopha- - gus brasiliensis from Rio Magdalena, which, like the coastwise streams of southeastern Brazil, flows out of Archean mountains and possesses therefore remarkably similar environments. 2. Aiquidens teteramerus appears to be the most primitive of its genus. HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 67 It changes in the Guaporé River into A. paraguayensis, which is also closely related to A. vittata. All of the species of this genus form a natural group about A. teteramerus, excepting A. portalegrensis, and I consider it as giving rise (or vice versa) to Cichlasoma bimaculatum. 3. Heterogramma tematum can easily give rise to all of the species of this genus. In fact, I have reasons to doubt the reality of all of these species, because they may be nothing more than fluctuating variations, principally in color, or somatic changes which may or may not be in- herited. At any rate, there is an almost complete intergradation of all of the species of this genus. Hence experimental work is needed before this genus can be properly classified. 4. There can be no doubt that Cichlasoma bimaculatum is the most primitive of its genus, because it is not well defined from Mquidens portalegrensis, which is the most primitive living cichlid genus. 5. I consider Crenicichla saxatilis as the most primitive of its genus. It is represented in the south by QO. lepidota. C. vittata is the southern form of C. macropthalmus, and they are connected by varieties through C. lucwus to saratilis. C. johanna and the other elongate Amazonian species of this genus also can be linked to C. savatilis. | I will not venture to discuss the relationship of the other genera and species, because the results would be only an opinion with little or no support. The above brief consideration, however, is extremely ‘useful, because Cichiasoma bimaculatum, Aiquidens teteramerus, Geophagus brasiliensis and Crenicichla saxatilis are the generalized types which not only are widely distributed but also have been the origin of the bulk of the Cichlids found in the various environmental complexes. The genus Cichlasoma alone, according to Eigenmann, contains eighty-four of the one hundred eighty-seven known species of American Cichlid. These four genera actually embrace at least 80 per cent of the species of the American Cichlide, and several other genera can easily be derived from them. i These four genera are found from one end of the Plano Alto to the other, and consequently from their present distribution we can explain the origin and distribution of their derivatives, but this has nothing to do with the origin of the cichlid family. To sum up briefly, then, the distribution of the Cichlide, we may say that three highland genera are found in Rio Sao Francisco and have not evolved any new species. The same three genera have produced nine species in the Rio Grande do Sul, 7. ¢., the three old plus six new species,. and six species in the Alto Rio Parana and the coastwise streams of east- ern Brazil. Sixty per cent of the Paraguayan Cichlide are also included 68 ANNALS NEW YORK ACADEMY OF SCIENCES ‘by the same three genera and more than 50 per cent of the Cichlid of the Guaporé are not found in the Paraguay. - The diagram forming Plate XVI indicates the evolution and distribu-_ ction of the-cichlid fishes of the Amazon Valley and the rivers south of it. dt shows that river connections or interchanging of fauna and barriers or isolation are not the important factors of geographical distribution, but that the organic complex of the ancestral stock (three highland gen- era—Geophagus, Crenicichla and Alquidens-Cichlasoma) and the compo- sition of the environmental complexes in which they came to live, pro- duced by the rivers sinking into the Plano Alto, are the important factors. The figure also shows that similar and identical evolution of the common ancestral stock has taken place in similar environments and dissimilar evolution in dissimilar environments regardless of whether the environ- ments are or are not connected. (See Plate X V.**) The phylogeny of the extra Amazonian species, 7. e., more than thirty- three (fifty-three existing in it), is not yet clear, but they will eventually be deduced from the highland stock, because I have shown that the Amazon Valley as we now know it has existed a comparatively short time. The facts as shown on the diagram are almost exactly the opposite to what one would expect, if land and water connections or isolation were the important factors of living animal distribution. The numbers at the end of the arrows show the number of new species which have un- questionably descended from the old highland stock when it entered the rivers which were gradually eroded in the Plano Alto. The diagram also shows the basins connected or not and the identity of fauna in dis- connected regions, such as Alto Parana and coastal streams. I have no first hand knowledge of the Rio Orinoco; hence I do not discuss it. Additional proof of similar evolution in similar environmental com- plexes, even if they are not connected, is offered by the larger species of South American fishes. For the sake of clearness, I have divided a typical abbreviated list of large species of fishes into the following classes : : 1. Large species of fishes found in the upper Guaporé and Amazon, and neither the genera nor the species found in the Paraguay-La Plata basin. A few examples of such fishes are Cichla ocellaris, Phracto- cephalus hemiliopterus, Brachyplatystoma reticulatum and Electrophorus electricus. 'To these and many other fishes may be added the large croco- 82 Whether we should call these isolated species identical or by the same name may be a debatable question. Not any two individuals are identical, but these species are not at present distinguishable. ANNALS N. Y. ACAD. SCI. VOLUME XXII, PLatTE XVI OUTLINE MAP OF PART OF SOUTH AMERICA Showing the centers of evolution and the distribution of the cichlid fishes of the Amazon Valley and the rivers south of .it HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 69 dilian Caiman niger, the large Amazonian turtles Podocenemis expansa and P. tracaxa and the red porpoise Inia geoffroyensis. All of these forms are found as far up Rio Guaporé as Bastos, Rio Alegre. If any direct swamp connections existed, I should certainly expect to find these animals in the Paraguayan pantanals. ‘The smaller species of caiman, Caiman sclerops, is found.all over South America excepting Patagonia and west of the Andes, but it has been seen six miles away from water, and hence an overland trip is not impossible for it. 2. Large species of fishes found in both the Guaporé-Amazon and the Paraguay-La Plata, but never near their headwaters, 7. e., at least fifty miles apart in a straight line between the two basins. A few typical examples of such fishes are Sorubim lima, Hemisorubim platyrhynchus, Sciades pictus, Mylossoma aureus, Charax gibbosus, the giant Paulacea jahu of La Plata and Paulacea lutkent of Amazon (I consider the last two species synonymous). . These and at least fifty other species which are found both in the Amazon and the Paraguay have not interchanged as such between these basins, for the following reasons: 1. No connection has existed. 2. They are not found in the headwaters, 7. ¢., above the waterfalls and at such high altitudes as exist between these rivers. 3. The distance between the headwaters is so great that an accidental distribution is not possible. 4. Hven if a connection had existed, it would not explain why the spe- cies have remained the same and why Rio Uruguay and Parana, belong- ing to the same river basin as Rio Paraguay, do not possess all of these species but, on the contrary, harbor many species not found in Rio Para- guay. Why, also, did not other species interchange, if a connection has existed ? 5. The Sao Francisco River is connected with the Amazon Valley, yet it does not have nearly as many Amazonian species as does the Paraguay. In fact, its common or Amazonian species are cosmopolitan forms. Rio Sao Francisco has an unfavorable environment (dry, hot, high, muddy, fewer swamps, etc.) and therefore has fewer species than the large Para- guay, with its favorable cichlid environments which have produced more cenotelic changes in the ancestral stock. Tn view of all this, it appears that the only answer which can be given to the question why the Paraguay has at least 53 per cent of Amazonian fishes is .,,. las ais 1. About 50 per cent of the similarity is due to the cosmopolitan forms, 1.-€., to-overland distribution of the small generalized highland genera which are widely distributed. 70 ANNALS NEW YORK ACADEMY OF SCIENCES 2. When these highland forms arrived in the same kind of environ- ments, they often underwent identical evolution with that which was taking place somewhere in the massive Amazon. 3. The remainder of the similarity is due to marine immigrants. The first part of this answer needs no further comment, but the second may appear to be absurd, at least to those who are not familiar either with the South American fishes or with the environmental complexes in which these fishes live. The view that the common highland genera of fishes have often undergone identical evolution in similar environments, even if these environments are well separated, is of the same general nature as those given in the following publications. Bateson has shown in the case of Cardiwm edule of the Aral Sea that, as the sea dried up, isolated basins were formed in which the salinity was greatly increased, and under these conditions the cockles so separated show similar variations under similar conditions. These shells of the cockles in the higher to the lower terraces showed a progressive change in regard to the following features: (a) Shells became much thinner. (b) Shells became highly colored. (c) Size of beaks became reduced. (d) Shells became smaller in size. (¢) The grooves between the ribs on the outside appeared on the in- side of the shells as ridges with rectangular faces. (f) A great increase in length in proportion to the breadth of the shells. Are not the changes observed by Bateson as profound as required to make Cichlasoma bimaculatum out of Mquidens portalegrensis. or vice versa? In this case, the loss or gain of one spine makes a different genus and a species. ‘Thus it is with many other genera and species of fishes whose distribution must be explained by identical evolution in similar environ- ments.*8 MacDougal has obtained nearly all of the mutations observed by de Vries in Holland from @nothera lamarckiana obtained in France, England and Holland and planted in New York. He also obtained one mutant, O. albida, from O. lamarckiana Nantucket City. Even if O. lamarckiana is a hybrid, it makes no difference, for at least part of the common highland stock may also be hybrid. Tower has shown, both in nature and by experiments with humidity and temperature on Leptinotarsa, that he could produce changes both in 83 Some of the small differences may be purely somatic, which are not necessarily inherited. Wxperimental evidence is necessary to settle this point. _. HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA “71 the germplasm and somaplasm or in each separately. For example, he was able to increase the temperature in the laboratory at Chicago and obtain forms of the potato beetle which lived in Mexico. Tower’s work shows almost beyond a doubt that when the same species of potato beetle lived in different environmental complexes, similar variations were pro- duced in similar environments. This is all that is required to explain the similarity of certain genera and species of fishes which are found in the Paraguay and Amazon, but not in the Parana, Uruguay or the coast- wise streams of eastern Brazil. Furthermore, the requirement is not ereat, because many of the genera and species of fishes are based on one more or less spine, three to ten more or less scales in the lateral line, color and position of spots and other trivial characters which are subject to a wide range of so-called fluctuating variation. In fact, some of these variations may even exist occasionally in the generalized widely distrib- uted highland species from which the bulk of the ichthyological fauna of the various river basins has evolved. Any one less familiar than the writer with the region in question would not venture to state that identical evolution has taken place on such a large scale in similar environments. As I have already shown, however, no connection exists between the Paraguay and the Amazon. Accidental overland and marine distribution is more absurd than identical evolution from common highland stock, and even if the species got across, we should still have to admit that they have remained almost the same in the case of the Paraguay and the Amazon and have not in the case of the Parana and Uruguay rivers, which belong to the same basin as the Paraguay. To admit the latter is equivalent to admitting either that identical evolution has taken place in the case of many genera and species of fishes, or else to believing in the fixity of species in one locality and not in another. Furthermore, the Sao Francisco has a connection with the Amazon, and yet its cichlid fauna is composed of the three common - highland genera and species only. If a connection is needed to explain the similarity of the Paraguayan fishes with the Amazon, I desire to ask, “Why has the Rio Sao Francisco only three species of cichlids? Why is there a triple cichlid faunal region in the La Plata basin? Why have the Paraguayan species remained identical with the Amazonian? Why did not more of the Cichlid of Rio Guaporé enter the Paraguay? My answer to these questions is that similar environments have produced some similar changes in the same germplasm, 1. e., the highland genera which are widely distributed,** and dissimilar environments have pro- % This does not at all imply the inheritance of acquired characters, for it can also easily be a direct effect on the eggs, sperms or germplasm. 72 ANNALS NEW YORK ACADEMY OF SCIENCES duced some dissimilar changes. Many of these changes are of such a nature that the species are adapted to live only in certain kinds of en- vironments. The cichlid fishes usually live in swamps, lagoons or lakes, and seldom in rapidly flowing water. In the open channels, the chara- - cins would eat them. The Paraguay and the Amazon have many swamps and many cichlid fishes. Alto Rio Parana has few swamps and few eichhids. The marine immigrants which have entered the rivers and become permanent dwellers of the same have also increased the percentage of similarity between certain rivers more than between others. Typical examples of such fishes are as follows: 1. The fresh-water skates, Potamotrygon, two species of which are found both in the Paraguay and the Amazon. We have no evidence that these two species separately left the ocean and became dwellers of these rivers. Other species of the same genus are found only in the Amazon, in the La Plata, the Guiana and the Orinoco. Is it not possible that some of these species are the results of changes invoked in the marine ancestor by their new environments ? 2. Two species of Peeciliide, Rivulus and Girardinus, could easily have gone along the coast from the La Plata to the Amazon. They may also have gone overland. 3. Stoleophorus olidus. 4. Two species of Scieenide belonging to the genus Pachyurus. 5. Several catfishes, and even a few of the characinids and cichlids, might have migrated along the coast, but it is out of the question to as- sume that all of the identical forms in Rio Paraguay and Rio Amazonas did so because they are not found in coastwise streams, Rio Uruguay and Rio Parana. There is no doubt that the marine immigrants have played an im- portant part in the production of a greater similarity between the fresh- water fishes of certain river basins than others. I found only two marine species of fish in Rio Colorado of Patagonia and at least one hundred are found in the lower Amazon. Even the sawfish (Pristis) is sometimes killed as far up the Amazon as Santerem (476 miles). The volume of water always bears a relation to the number and size of marine immigrants, and this is especially true when there are many islands, many channels, plenty of food, tidal effects and much brackish water. : It is evident, then, that the small rapid coastwise streams of eastern Brazil, Rio Magdalena and Patagonia should have fewer species than HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 73 either the Amazon or the La Plata, regardless of any hypothetical con- nection with the eastern hemisphere, for the following reasons: 1. Smaller volume of water, relative higher altitude, excepting parts of Patagonia and strong currents. 2. The original stock, part of which was marine in origin, did not develop into so many species and genera in the restricted environments as it did in the more extensive environments, because the factors active in the evolution and preservation of life were neither as favorable nor as numerous in the restricted environments. 3. Many species are adapted to live in only certain environments. One would not expect, therefore, to find Lepidosiren in the Alto Rio Parana, the coastwise streams of eastern Brazil and Patagonia. Lepido- siren lives in the vast swamps (chacos and pantanals) of Rio Paraguay and Rio Amazonas. 4, Species of more or less recent marine origin have encountered far greater difficulty in entering small, rapid, rocky, shallow streams with a limited supply of food than large rivers, because marine fishes are used to swimming in the sea and not in rapid rivers and because the change from sea to fresh water is less sudden. 5. Some of these rivers have higher altitudes, and the number of fish always bear a relation to the altitude. The last of the environmental complexes which needs a further con- sideration is that of Patagonia. It lies between 40° and 55° south latitude and is characterized by a general paucity of plants and animals, especially tropical forms. Its plains are arid and forestless. Its rivers are not large, because they are for the most part fed by the melting snow on the lofty Andes. A cold Antarctic current flows along the coast, which is devoid of swamps. Besides, vast tracts of Patagonia have been under the sea during part of the Tertiary period. At this time, it is quite possible that nearly all of its fresh-water life was exterminated, and as the land rose again from the sea with the Tertiary elevation of the Andes, the northern portion of it became semi-desert. In part of this region, the rivers flowing down from the Andes dry up on the barren plains. Hence, a southern migration of fishes would have been almost blocked, excepting in the case of the Pygide, which are found everywhere in South America. Notwithstanding all this and the fact that the Patagonian rivers have few marine immigrants, its twenty-six known species of fishes contrast favorably at least in number, either with the secca (dried) regions of Ceara and Pernambuco, Brazil, or with similar latitudes in some parts of the northern hemisphere. It is not at all strange that such tropical 74 ANNALS NEW YORK ACADEMY OF SCIENCES species of fishes as Arapaima, Osteoglossum and Hlectrophorus are not found in Patagonia, because they are also not found in La Plata; but it is strange that the Pygide are found in Patagonia and Hoplias mala- bricus (of the Characinide) is not, because I found Hoplias to be one of the best overland travelers of all the South American fishes. The absence of Hoplias in Patagonia may be due to its being a tropical genus. The fact that Geotria and the Galaride are found in the Australian realm is no evidence that Patagonia was connected with the same, be- cause at least one of these forms is known to enter the sea.*® The absence of Diplomyste and Pygide from the Antarctic and the Aus- tralian realms seems to me to be far more conclusive static evidence that these regions were not continuous than do the presence of Geotria and Galaxide in these two regions indicate that they were continuous. The latter two genera could have extended their limits of distribution by way of the sea. If a connection existed, Pygide, being good overland tray- elers, would have had a chance to enter the Australian realm. If Diplo- myste is the most primitive living catfish, it, too, would have had a chance to extend its limits of distribution. Hence, only marine fresh water and no strictly fresh-water species or genera are common to these regions, and I take this as strong evidence against a former connection between Patagonia and the Australian realm. In fact, Patagonia has no Osteoglosside, no Dipnoi and other forms found in Australia. Origin of the South American Fishes In the Princeton Patagonian report on the fishes, Professor Higen- mann states that fishes probably interchanged before the beginning of the Tertiary epoch between Africa and South America by way of a land- bridge between Guiana and Africa. The following objections can be raised against this hypothetical view: 1. There is no positive evidence that either the Characinide or the Cichlide as such existed previous to the Tertiary, but I grant the possi- bility of their existence in late Cretaceous times. 2. All of the known fossil fishes indicate a northern origin of the living tropical fishes. 3. There is no good geological evidence in favor of the connection. The evidence is all biological and paleontological and questionable in kind. %>It is also probable that formerly both of these genera were able to enter the sea. This is all the more true in view of the fact that much of Patagonia was covered by Tertiary sea. ; HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA %5 4, The alleged support derived from the distribution of the fishes is derived from the static viewpoint: of animal geography. It is also based on several erroneous ideas concerning the topography, geology and en- vironmental complexes. 5. The point of family origin was not correctly determined, because the greater number of species in a given locality is no evidence that it is a center of family origin or dispersal. In the preceding pages, I have attempted to explain the present dis- tribution of fishes as being primarily due to cenotelic changes produced by different environmental complexes on the ancestral stock. It was also noted that only paleotelic characters were widely distributed. Therefore we must look for the point of family origin at a very remote epoch, when few or none of living Cichlid and Characinide existed. In other words, the present distribution has little or nothing to do with the point of family origin. This being the case, we have to look for fossils. We must confess that no absolutely conclusive knowledge derived from either paleontology or any other source exists from which we can defi- nitely determine the point of origin of the cichlid and characinid fishes ; but the analogies which can be drawn from the following data taken in connection with the facts which I have already stated appear to give the most plausible explanation, because it is more in harmony with the known geological data. The distribution of living and extinct Osteoglosside is as follows: Number of spectes Remarks Phareodus (Leidy)....... 2 Eocene of Green River, Wyoming, United States of America, and upper Cretace- ous of Chico formation of western United States of America. PS RUCIO@UUSE. oc cris alone oes 1 Lower Hocene of England. PAGE DOIG, Jo.0.0.s = elise 0 6 ere eres 1 Amazon and Guiana. Osteoglossum ............ 3 Amazon to north, East Indies and Aus- tralia. ERCLCTOUISS Aco ais vise) cscs 5 0.0 il Tropical Africa. Besides these five genera, two more doubtful genera are known from the late Mesozoic of the United States of America and southeastern Eng- land. The above data indicate a northern origin for this family of fishes, which now lives only in the tropics and the southern hemisphere. Diplomystus, a clupeoid, has recently been divided into two genera by Jordan, and at least seven species are known. They are from the fol- lowing regions: 7T—NY 76 ANNALS NEW YORK ACADEMY OF SCIENCES Eocene—Green River, Wyoming, United States of America. Upper Cretaceous—Mt. Lebanon, Asta. Cretaceous—Mediterranean Islands. Upper Cretaceous—Bahia, Brazil. Lower Oligocene—Isle of Wight, England. Upper Cretaceous—Italy. Living forms closely related to these fossil species are said to exist in Chili and New South Wales. Priscacara, with seven species from Green River and Bridger Hocene of Wyoming and Utah are so far the only known fossil Cichlide.** This indicates a northern origin for the Cichlid. One genus with three spe- cies of fossil Pomacentride are known from the upper Eocene and lower Miocene of Italy. One species of Perichthys is known from Tertiary shales of Taubate, Brazil. This genus still lives in Patagonia and Chili. At least seven genera with forty-three species of fossil Labridz are known from the lower Eocene to the lower Pliocene of Europe and Hng- land and the Eocene and Miocene of New Jersey, United States of America. One tooth attributed to a member of this family is known from the late Tertiary of the Argentine Republic. About fourteen genera of Cyprinide, including about thirty-three fossil species, have been reported from Germany, Bohemia, Sumatra, Java and various parts of western United States of America. These fossil forms range from the Quaternary to lower Miocene (?). Living forms are found all over the world, excepting Australia and South America. The absence of the Cyprinide from South America is the most extraordinary fact in the distribution of fishes. They have existed since the Miocene in south- western Idaho and are now found in Mexico but not in South America. Furthermore, if an Archhelenic land-bridge existed between Africa and Guiana and this was the means of dispersal of Characinide and Cichlidex, why did not the Cyprinidz also enter South America? Tetragonopterus avis and T. lignitis .(Hobrycon Jordan) from Taubate, Brazil, from shales of doubtful late Tertiary age are so far the only definitely known fossil Characinide. _ 86 Woodward, 1898, reported a fragment from Taubate as Alquidens (?). HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA “%? The facts regarding the Siluride are as follows: Number of species Remarks \CUTTATIS 9 5a nk ee 1 Lower Pliocene, India ; still living in India. Heterobranchus .......... 1 Lower Pliocene, India; still living in Africa and East Indian Archipelago. ST (?) Upper Tertiary of Europe and India; still living in Paleoarctic realm. Pseudeutropius .......... il Tertiary of Sumatra. Macrones ............+.+. 1 —_ Lower Pliocene of India; existing in Asia. BUD vere erence cece eeeeee 1 Lower Pliocene of India; existing in Asia. AMIUTUS .0 0s esseeeerecees 2(?) Lower Miocene of Canada; existing in .North America and China. (?) Pimelodus ........... 1(?) Late Tertiary (?) of Taubate lignite, Taubate, Brazil, and of Parana River, Argentine Republic; existing in South America east of the Andes and north of Patagonia. BGIUVIVEUSECSI. or <5 0\0) 0: 61 sis's 010-5 7(?) Lower Tertiary, lower Miocene and Bridger Eocene of western United States of America. Bucklandiwm ......es.00% 1 Lower Hocene, England. ALONG (5 OSA 4(?) Middle Eocene, Belgium. Upper Eocene, England. Oligocene and Middle Oligocene, Germany. Lower Hocene, Copenhagen. Pliocene, India. (?) Tertiary, Taubate, Brazil. RUG QGUUS) che o04 6.6 010 siosisiee's 1 Tertiary of Sumatra. Woodward stated that Bucklandium diluvii appeared to be related to Auchenoglanis, which still exists in Brazil. Rhineastes is the oldest known fossil catfish and appears to be related to Phractocephalus of the Pimelodine, which lives in the Amazon. The age of the Taubate shales, found in the deeply eroded Parahyba Valley, is not definitely known, but the surface deposits containing fishes do not appear to be very old (Pliocene), because the three fossil genera of fishes are still living. Jordan’s generic distinction of Hobrycon does not appear to the writer to be well founded, because the species of both Astynax and Tetragonopterus vary considerably in shape. The two fossil species from Taubate may, in fact, fall into the genus Astynax as now defined by Higenmann. If the writer is correct in considering the Taubate shales as late Ter- tiary, it is evident from the above list of fossils that the South American as well as the African fishes have evolved from forms which earlier lived 78 ANNALS NEW. YORK ACADEMY OF SCIENCES in the northern hemisphere. In fact, Osborn has already stated in “The Age of Mammals” that the South American fishes show an ancient north- ern affinity. The dipnoans and crossopterygians, which are now found only in the southern hemisphere, also were northern in origin, as far as we now know. The fact that South America has many species but few families of fishes also vaguely indicates a northern origin, because both plants and animals often rapidly break up into new species when placed in new environments. If the Cichlid were not northern in origin, how were they able to get into Wyoming during the early Hocene, when they are now not able to get north of Rio Grande and have never been able to enter Patagonia? The mere fact that the characinids and cichlids are still “going wild” in making species indicates that they entered South America during or after the late Miocene, 7. e., after South America be- came permanently connected with North America. This view seems to be all the more probable, because it appears to be easier for animals to move from temperate regions to the tropics than vice versa. It must also be noted, before dismissing the subject of the point of origin of the South American fresh-water fishes, that there is some vague evidence in favor of the marine distribution of at least the Cichlide and Osteoglosside. The genus Priscacara is closely related to the marine Pomacentride, according to Cope, and the formation in which they are found appears to have been near the sea level. Hence, in view of the fact that the actual paleotelic or ancestral forms which were distributed are not definitely known, a marine distribution of many primitive forms is not at all impossible. Such a view is made all the more probable by actual experiment as already noted. The present distribution of the osteoglosside can be explained most easily by considering them as northern in origin, at least if the zodgeog- raphers do not entirely ignore the paleontological evidence. It must be admitted that there is no positive evidence to show that the Characinide are directly northern in origin, but the only positive evidence at present known indicates that both the Cyprinide and Nematognathi, the nearest relatives of the Characinide, originated in the northern hemisphere, and it is highly probable that the ancestral Ostariophysi were also northern in origin—if we admit that the Ostariophysi are a homogeneous group. In this connection, the Nematognathi appear to have first split off from the ancestral Ostariophysi. Then the Cyprinide split off, and some of the later Ostariophysi were pushed, after the Miocene times, into South America, and the Characinide, now found in the southern hemisphere, appear to be the lineal descendants of this ostariophysian stock and have HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA %9 therefore retained more of the primitive characters than either the Nematognathi or the Cyprinide. There have been sufficient connections between North America and South America and between Eurasia and Africa to permit exchange of fishes during past epochs. Besides, many fishes and other fresh-water forms may have been able to migrate short distances along the coast or from island to near-by islands. Hence no objections can be raised against such migrations for want of land connections. SUMMARY OF THE MOST IMPORTANT DATA WHICH HAVE BEEN USED TO SUPPORT THE VIEW THAT SouTtH AMERICA AND THE HASTERN HEMI- SPHERE WERE PRIMITIVELY CONNECTED. INVERTEBRATES Crustacea Ortmann has stated that only one family of crabs found in northern South America lends positive support to his view of Archhelenis, 7. ¢., a pre-Tertiary connection between Guiana and Africa. This family, Poto- mocarcinine, are found only as far south as Guiana. They are found in Central America. Why have they not immigrated into the Amazon Val- ley and south? For all we know, they may have immigrated from the northern hemisphere and have only reached Guiana. Their ancestral stock might have been drifted across the Atlantic by the African-West Indian current or their ancestral stock, which was dis- tributed, may easily have been distributed by the way of Europe and North America. This is especially true in view of the fact that no iden- tical forms are found in Africa and South America. What was this paleotelic form which became distributed and gave rise to the different genera of the eastern and western hemispheres, and where did it origi- nate? This family of crabs offers a splendid analogy to the present dis- tribution of camels and tapirs which now live in the tropics, but which were originally found in the northern hemisphere. Besides, the Mesozoic forms of crustacea are very imperfectly known, and Sd is especially true of the crabs in question. Mollusca The writer found a bivalve, Diplodon, in the Iguasst. River above the big falls, which in some form or other appears to date back to the Tri- assic. This does not necessarily mean that Diplodon dates directly back to the Trias, because this genus is widely distributed over the highlands, but it does mean that it is a primitive form. In contrast to the above 80 ANNALS NEW YORK ACADEMY OF SCIENCES genus is Hyria, which was seen by the writer only in the lower Amazon and therefore probably does not belong to the older highland stock. I did not see a dozen species of bivalves in Rio Guaporé, which probably indicates a:rather primitive stock, but they were very abundant in Rio Uruguay, which indicates a great cenogenic evolution in this region. In a general way, the mollusca follow the same rules of distribution as the fishes, but our knowledge, especially from accurate field data, of the fresh-water bivalves of South America is entirely too meager to be used in support of any theory. For example, Ortmann has found that some of the bivalves have palpi, arrangement of gills and siphonal openings like some of the African bivalves, but this may be due in both cases to living in muddy, tropical water. Before such evidence can safely be used to support any theory, much careful field work and experimental evi- dence is needed. That is to say, we need dynamic and not static data. Many of the fossil bivalve shells in North America resemble living forms in South America. Inasmuch as the soft parts are not preserved, it will be very difficult to determine the paleotelic forms of bivalves which were distributed. Von Ihering has listed 581 species of mollusca along the coast of Brazil, 54 of which are known on both the Antillean and African coasts, while 72 are found in common on the African and Bra- zilian coasts. He does not appear to attach much importance to larval distribution of these forms, but it is well known that some larval crus- tacea and mollusca are found on the high seas. Marine turtles also cross the sea. Nichols has recently reported a case in which he thinks that the mid-ocean whirlpool and side currents swept young T’rachurus from the coast of England to the coast of Florida.. These ocean currents are well known to the sailors, who often go far out of a direct route in order to avoid them. In some places, when the winds are favorable, the one off the Barbados Islands is said to run about five miles per hour. Currents, therefore, may have transferred some mollusca between the South Amer- ican and African coast along with sea weeds and other drift. Then, again, a few forms could have been transferred by primitive sailing vessels between the various ports. It is interesting to note in this connection that practically all of the mollusca known from the Brazilian coast are reported from near larger or smaller seaports. In other words, very little of the Brazilian coast has been surveyed, and until regions remote from seaports are carefully studied, too much stress must not be laid on the present list of mollusca. Besides, the above lists of common forms on the African and Brazilian coasts include such widely distributed (cosmopolitan) forms as Mytilus, which could easily have gone south from the European and North Amer- HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 81 ican coasts. The writer looked carefully along the coast south of Iguape for Aporrhais pespelecani and was unable to find this form either living or dead. The few dead shells of this species known from a seaport have _ little significance, because peddlers and sailors are known to be great dis- tributers of shells. The writer picked up one valve of Lucina jamai- censis (?) on a sand bar below the Urubu-punga waterfalls of the Alto Rio Parana, which is several hundred miles from the seacoast, but this shell had evidently been dropped there by an Indian. A very important factor in the distribution of the marine mollusca of the Atlantic Ocean is the tropical condition which existed in the North Atlantic during the Eocene. This would have given excellent oppor- tunities for exchanges of forms between the African-Huropean and American coasts. These ancestors of the existing forms would have been pushed south again when the climate of the North Atlantic became cooler. As a result of this, many resistant ancestral forms living in similar environments and evolving along rather definite lines would pro- duce a great similarity between the coasts of the South Atlantic. Furthermore, it is not impossible that some young forms of land gastropods could have been carried with tropical plants to the eastern hemisphere just as Litorina litorea has probably been imported in some way to the American coast. After a long detailed study of the living and fossil Tertiary mollusca, von Ihering has recently concluded that Archhelenis, the land-bridge between Africa and South America, began to disappear in the Cretaceous but continued to exist in the Tertiary. Ortmann (1910) used the same. data and arrived at a different conclusion, namely, that Archhelenis had disappeared before the beginning of the Tertiary; but neither of these authors has taken into consideration the effects of similar tropical en- vironmental complexes along the African and South American coasts on the ancestral stock from which the existing species have evolved. When this is done and the cosmopolitan forms are eliminated and when due allowance is given possible larval and adult distribution by ocean cur- rents, floating debris and boats, then no land-bridges are needed to ex- plain the distribution of marine mollusca. Only static studies have been made, and until some » dynamic work has been done the evidence derived from the mollusca is not a safe peg to hang a theory on.*” 87 The brachiopoda appear to me to offer even less evidence, because many almost cos- mopolitan genera have existed during various past ages. This is all the more true when certain similar forms are known to exist in similar environments. So it appears that extensive land-bridges like Archhelenis will have to rest on evidence derived from geology and continental faunas and floras. 82 ANNALS NEW YORK ACADEMY OF SCIENCES Recently Pilsbry has given an excellent static treatise of the distribu- tion of non-marine mollusca of South America. He has combatted the separation of Guiana, Brazil and Patagonia during past times. In this far, I agree with his conclusions, but his data do not alone warrant them, because the highland environments in Guiana and Brazil are very much alike, and hence similar forms are to be expected from a common older stock, even if these regions had been isolated from each other during part of the Mesozoic and Tertiary times. His static evidence certainly almost if not completely destroys that of Ortmann for a Guiano-African con- nection, but he has unfortunately located his Brazilian-South African connection exactly in the region where we have given extremely strong geological evidence against such a view. ‘The entire coast of Brazil in - this region is fringed by marine Cretaceous, which alone would force this connection to have disappeared at least in the early Cretaceous. Besides, the “Pernambucan fan” is strong evidence against any Paleozoic connec- tion in this region. ‘There is also so much other geological evidence against the building up of a land-mass across the great ocean depths of the South Atlantic that we may consider Dr. Pilsbry’s view highly improbable, at least until some dynamic and more careful field studies have been made on the non-marine mollusca of the regions in question. His actual paleo- telic forms which were distributed are not yet absolutely known to have originated in this purely hypothetical Gondwana Land. Ants Von Ihering has even used the distribution of the ants to prop up his Archhelenis theory. He states that we still see the ingression of Bolivian ants into Brazil. These ants are supposed to have come by the way of Antarctica from the eastern hemisphere. This evidence appears to me to be of little weight, because the ants have had a highland since the Per- mian, over which they could have crawled or flown into Brazil. Further- more, practically nothing is known of the ants which live in the high- lands of central Matto Grosso. If one may hazard a guess, I would sug- gest that a detailed study of the ants would show an invasion from Brazil toward the Andes and not vice versa, after the elevation of the mountains in Tertiary times. Corals Gregory has shown that the Miocene corals of the Mediterranean re- semble the living corals of the West Indies. This is not, I believe, evi- dence that the West Indies were directly connected with the eastern hemisphere, because the larvee of these corals may have drifted over by HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 83 the mid-Atlantic whirlpool and ocean currents. They may have origi- nally migrated down the North American coast from Europe, and it is - even more probable that the observed affinities are due to similar evolu- tion in similar environments. If this is not true, why have so many species remained similar to the Miocene forms of Europe ever since the supposed land-bridge disappeared or from before Tertiary times? When and where did this bridge exist? Are not the corals of the North and South American coasts also similar? If not, why not??* GONDWANA FLORA Similar deposits of sandstones, clays, shales and bowlder till are found in Brazil, South Africa, India and Australia. These deposits are chiefly ‘Permian and contain among other fossils the characteristic lower Gond- wana flora (Gangamopteris or Glossopteris flora). The identity of the Gangamopteris flora, along with many fossil and living genetically re- lated animals found in Africa and South America, has led to a widely accepted belief that these continents were originally connected. This old, land-mass has been designated Gondwana and is thought by some to have extended across the Atlantic between either Brazil or Guiana and Africa. Others have ignored this connection and have maintained a southern connection by way of the Antarctic Islands. The Gangamopteris flora, according to I. C. White and David White, has been found six meters above the crystalline floor of the coal fields of southern Brazil. At this level, only Gangamopteris obovata was found. The next higher level, 55 meters above the granite floor near Minas, Santa Catharina, contains Rosellinites gangamopteridis, Hysterites bra- stliensis, Phyllotheca griesbachi, P. mulleriana, Glossopteris browniana, Werte0raria ........ ?..., Gangamopteris obovata, Arberia minasica, Derbyella aurita, Noeggerathiopsis hislopi, Cardiocarpon seixas and Cardiocarpon moreiranum. These species belong to the early typical Gangamopteris or lower Gondwana flora. The same genera, and in many cases identical species, are found in the Ecca shales of South Africa, in the coal associated with marine lower Permian of New South Wales and Tasmania and in the Karharbari beds of India. The same flora is found in the lower Coal Measures of Argentina and the Falkland Islands. Only much later, in the upper Permian of the northern part of Russia, are any of these *87f Archhelenis existed, I fail to see how the rivers could have been arranged on it so that only one family of crabs, two families of fishes and a few fresh-water and land mollusca took advantage of it, when the same theory assumes that the coastwise streams of eastern Brazil have been barriers to at least part of this fauna. 84 ANNALS NEW YORK ACADEMY OF SCIENCES gangamopterids known to occur in the northern portion of the northern hemisphere. Conformably underlying this lower Gondwana flora are the Orleans conglomerates of southern Brazil, which are supposed to be related to the Dwyka conglomerates of South Africa, the Baccas Marsh conglomerates and their equivalents in Australia and Tasmania and the Talchir con- glomerates of India. These conglomerates contain some of the evidence for the alleged Permian glaciation of the southern hemisphere. At 135 meters above the granite, 7. e., about 80 meters above the typi- eal lower Gondwana flora, is found the intermingling of this flora with some species of the older northern cosmopolitan flora. In this forma- tion were found Hquisetes calamitinoids, Schizoneura, Sigilaria aus- tralia, Sphenopters hastata (?), Glossopterts indica, G. ampla, G. occt- dentalis, Neggerathiopsis hislopt and Cardiocarpon oliwveiranum. The flora is still primarily Gondwana. At a still higher level, 157 meters above the granite floor or 100 meters below the Iraty shales containing Mesosaurus, are found more of the northern flora, such as Lepidodendron perdroanum, Lepidophloios larcinus and Sigillaria brardw. At this level, the lycopods are again preéminent as coal makers. The Gangamopteris flora is very imperfectly known, but what is known indicates almost beyond a doubt that the Gangamopteris belong to the southern hemisphere. It is not known from North America and is only known from the late Permian of Russia. The question is, then, Are the known facts concerning the Gangamopteris flora indicative of a contin- uous Gondwana Land somewhere in the southern hemisphere? Before attempting to settle this difficult question, it is necessary to consider the origin and the environmental complexes of this flora.*® The Gangamopteris flora belongs, as Professor Arber and Dr. White have well shown, almost exclusively to families already known in the cosmopolitan flora. They constitute genera and species more or less bound to their northern relatives, though often differmg much in form and aspect. In general, they appear simpler in figure, with a tendency to thickness and rugosity of leaves, and on the whole their general aspect suggests environmental conditions unfavorable to luxuriant growth. . This flora suddenly appears in the early Permian well defined from its Car- boniferous ancestors, which lived in the northern hemisphere and sur- vived more or less the profound geological changes produced by the for- ®9It is barely possible that the lower Gondwana flora of Brazil belongs to a later Per- mian than now believed, as Professor Branner states in his Geologia Hlementar that the intercalated marine deposits containing Schizodus, Myalina and Oonocardium also con- tain other lamellibranchs also found in the Triassic. If this is found to be the case, then the appearance of this flora in Russia-Siberia may have been as early as in Brazil. HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 85 mation of the Brazilian Plano Alto. This cosmopolitan flora continued to exist in the Mesozoic epoch, but the Gangamopteris flora, excepting Glossopteris and Schizoneura, vanished with the close of the Permian. In this remarkable fact, we have, I believe, part of the solution of the Gangamopteris flora, 7. ¢., it existed in Brazil only during the formation of the Plano Alto and died out after this was completed (early Triassic). In this profound change of the ancient Brazilian topography produced by the formation of the Plano Alto in the Permian inland basin, we have the production of unfavorable environments which might have produced the alleged glacial effects on the Gangamopteris flora and have caused the absence of the cosmopolitan plants in the lower Gondwana formation of Brazil. This view is further strengthened by the fact that the shales, etc., of the lower Gondwana formation have a different appearance and chemical analysis from those of the higher formations in which the cos- mopolitan flora are found. The cosmopolitan flora is always associated with the production of large coal fields, and such conditions are not met in the Gondwana formations of Brazil. It appears to me that it was desiccation (perhaps not due to a lack of rainfall, but to its disappear- ance in a sandy soil) and the blowing of sand into the Permian inland basin, and not severity of climate (glacial), which produced the stunted appearance of the early Gondwana flora. Several other objections can be raised against the use of Permian glaciation as a factor which affected the distribution of the Permian plants and reptiles of South America. In the first place, the writer does not believe that the existence of glaciers in Brazil has been definitely established. His experience with glaciation in North America and gla- ciers in the Andes, taken in connection with observations on erosion in the highlands and mountains of Brazil, has strongly suggested to him that the Orleans conglomerates were not deposited by glaciers. In the highlands of Piauhy and various places in Brazil, one can see both high- land streams and extensive slanting surfaces over which gravel and bowlders slide during heavy rains. The underlying surfaces and bowlders are often scratched in a way which resembles glaciation. When pieces of the scratched and polished surfaces are detached, segregated or not, as is often the case due to less and greater amount of rainfall, and deposited at a lower level, a “false moraine” and even false bowlder till is formed. When such a mass of gravel, clay, bowlders, etc., becomes covered up and pressed together by later erosion in little stratified or unstratified beds (due to continual deposition and plasticity of the clay), it can easily be mistaken for a glacial deposit. It can only be distinguished from glacial deposits by means of truly faceted bowlders; and inasmuch as faceted 86 ANNALS NEW YORK ACADEMY OF SCIENCES bowlders are not definitely known to exist in Brazil, I take this as a strong evidence against Permian glaciation in Brazil. In fact, not until recently has any one even seen striation in the alleged glacial deposits of the Permian of Brazil. Some of these recent false signs of glaciation in Brazil, as Branner has shown, even deceived Agassiz (the expounder of glaciation), who described vast sections of Brazil as being glaciated. The direction of the striations and the arrangement of the bowlders also offer no conclusive evidence in favor of glaciation. When “false moraine,” composed of scratched surfaces, bowlders and false tillite, be- came covered up by the Permian sandstone found in the Plana Alto, only those deposits and scratches pointing in the direction of the dip of the country were exposed by the post-Permian erosion of the overlying strata. In most of the Gondwana formation, this dip is toward the south and west. Hence only here and there are the strie exposed, and all of them point more or less in the same direction, 1. e., they are only seen along deeply eroded river valleys below the waterfalls. Furthermore, these deposits often cover vast regions. For example, the deposits of such erosion would have covered many miles of width for the entire length of the Serra do Mar during the late Carboniferous and early Per- mian when the climate was favorable for the deposition of false bowlder tillite, which later became overlapped by the Permian sandstone. Later erosion exposes these deposits over a vast area. Woodworth found striations on bowlders, some of which appeared to have been deposited by ice floating near sea level (as is indicated by in- tercalated marine deposits in the Rio Negro basin). This floating ice may have come from Permian swamps, where it gathered up bowlders. This mass may have floated toward the sea the following season, scratch- ing the rocky surfaces along the margins of the swamps. In this case, faceted bowlders would probably not have been formed. I acknowledge that Woodworth has also found much other evidence for Permian glacia- tion in Parana, Brazil, and the evidence is even stronger for it in the eastern hemisphere.*° I grant that glaciers may have existed in Brazil during the lower Per- mian epoch, but in view of the preceding, it appears that faceted bowl- ders must be found before the evidence in favor of it is sufficient to warrant the use of Permian glaciers as a factor in the distribution of plants and animals. Even if such evidence is found in Brazil, it will probably not be found after the lower Permian, at which time the typical Gondwana flora and reptiles were scarce. Furthermore, if glaciers ex- 40 Woodworth had not published his paper on the Permian glaciation in southern Brazil when these notes were prepared. HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 87 isted during the lower Permian in Brazil, they may also have been only local glaciers of high altitude, like those existing in the Andes. Returning again to the fact that the Gangamopteris flora only existed during the formation of the Plano Alto, it appears to me that this is positive evidence that its environment, composition and extinction were directly associated with this great transformation of the South American topography. In other words, the Gondwana flora was an arid highland flora. It is known in marine series in Australia, but it could have been washed there from higher elevations, and the altitude of these sandy highlands was not necessarily great, in view of the fact that stunted plants are found on the Brazilian highlands at comparatively low altitudes. Similar geological transformations also occurred in India, Africa and Australia. Similar deposits of shales, clays, sandstones, coal, etc., occur in these remote regions. ‘The mere fact that all of these regions possessed almost identical environmental complexes, not known to be perfectly duplicated in the northern hemisphere, is one of the most important factors con- nected with the origin, the distribution and extinction of the Gangamop- teris. flora, regardless of whether these regions were or were not con- tinuous. It is a remarkable fact that so many identical species and genera belonging to different families of both the Gangamopteris and the cosmo- politan flora could exist, unchanged, in such remote regions as India and Brazil during most of the Permian epoch. Little short of ortho- genesis or similar evolution in similar environments can account for such data. I have already shown that the Permian inland basin of South America is almost completely surrounded by higher Archean mountains which have apparently remained almost stationary during part of and since the Permian epoch. This is particularly true of southern Brazil. There- fore, the fact that the Gondwana flora, so widely distributed over the southern hemisphere, was able to enter these various Permian formations over and around higher Archean mountains is evidence that its ancestors - were little or not at all affected by barriers. It is true that this flora might have entered the outlets of the basin along the Permian coasts, but if it was a highland flora, it probably did not. I am unable to conceive how and from where enough of the typical deposits of Gondwana Land could be derived in such a way that a homogeneous environment might have existed between either Africa and South America, or between South America by way of the Antarctic islands and either Africa or Australia. The Gondwana formation of 88 ANNALS NEW YORK ACADEMY OF SCIENCES South America can easily be explained by place erosion of the highland pre-Permian floor, by the erosion of the Serra do Mar and its southern spur, Serra Geral, on the east side of the Permian inland basin, and by similar changes of the Serras de Cordova, San Luis and de la Ventana on the west. All these mountains have a general north and south trend, and there is no evidence of an east-west trend in any of this region. The Gondwana formations of Brazil, Africa and India dip as a whole toward the south and west. This has a ready explanation from the location of the Archean and pre-Permian rock, but there appears to be no explana- tion showing how these formations could have been a part of a greater continuous homogeneous Gondwana Land. Its flora being only known from this special type of environment, how did it traverse these obvious Gondwana barriers ? Permian and Carboniferous deposits of coal are laid down along the sides of the Andes-Rocky Mountain system which extends into Hurasia and also on the sides of the Appalachian system, and extending through Guiana down the divide along the eastern coast of Brazil. I am not aware of any similar formations extending through the Antarctic islands toward either Australia or Africa. The Antarctic islands appear to bear the same relation to southern South America as do the West Indies to northern South America. They are well separated from both Australia and Africa by great depths of the ocean as well as by many miles of dis- tance. Hence it appears more probable that the Gondwana flora or its ancestral forms would have migrated along lines where the conditions of the environments of Gondwana Land were at least partially duplicated, 1. €., Where coal was deposited, rather than over about 2,000 or more miles where such conditions probably did not exist. The fact that the northern Permian plants had many identical species in the southern hemisphere, that there is considerable difference between the Permian reptiles found in Texas and those of South Africa, as well as the fact that the Gangamopteris flora is found in the late Permian of Russia, is good evidence that both the cosmopolitan and the Ganga- mopteris flora were resistent forms which (or their ancestral stock) could migrate into remote regions with comparative ease. The presence of ~ Cardiocarpon, Sphenopteris, Psaronius, Sigillaria and Lepidodendron in both North America and South America, as well as the presence of the Mississippian flora at Cacheuta, Argentina, indicates an exchange of plants between North and South America. Some of these plants (Lepi- dodendron) are not yet positively known in South Africa. The cosmo- politan flora is also known from Europe and other parts of the eastern hemisphere, and it is remarkable that this flora could not only become so HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 89 widely distributed but could also remain almost identical in the various corners of the earth. Hyen if it be granted that a vast homogeneous Gondwana Land en- tirely covered the southern hemisphere, this alone would not explain why so many species of the typical lower Gondwana flora remained identical during the greater part of the Permian epoch. There is positive evidence from the intercalated Permian of Brazil and other Gondwana forma- tions that these regions were not continuous during the entire Permian epoch, and yet the flora remained identical. Hence there appears to be no need for a ready and wholesale exchange of this flora, because it re- mained unchanged during the most of the Permian epoch, and therefore any accidental distribution of each genus and for one time only would be adequate. In this connection, it is to be remembered that neither con- tinuous land connections nor barriers nor isolation causes species (unless there are only a few individuals) to evolve new forms or to remain fixed, but the action of the environmental complexes on the ancestral forms, which had a certain composition, and hybridization will change the spe- . cies. The identity of the Gangamopteris flora in such remote parts of _the earth as already noted may be due to the action of similar environ- mental complexes upon the ancestral stock of each group of this flora. It is then even possible, but not probable, that the Gangamopteris flora is only an environmentally changed form of the cosmopolitan flora which could not exist as such in the early Gondwana environment. At least the sporaginous members of the cosmopolitan flora could have migrated into the Gondwana environments, if they had been able to thrive under such conditions. What the direct antecedent types of the different groups of the Gond- wana flora were is not definitely known. David White considers that Neggerathiopis is probably of Cordaitalean origin and Gangamopteris has a common origin with the neuropterid group of Cycadofilices, to which the genus Glossopteris also is related; yet both are far removed from the known antecedent type of the northern hemisphere. What this direct ancestral stock was and where it originated are questions which must be answered before we can definitely hypothesize lines of dispersion. I have repeatedly pointed out in the case of fishes that direct connec- tions are not necessary for the production of identical species and that the greater number of species in a given region is no evidence that the family originated there. I will repeat here again that the fact that the gangamopterids are found only in the southern hemisphere is not con- clusive evidence that their ancestral stock either originated there or that all of the existing Gondwana formations were directly continuous. 90 ANNALS NEW. YORK ACADEMY OF SCIENCES The Gangamopteris flora was already well defined in the early Per- mian. It has no close antecedent types in either the northern or the southern hemisphere, but its actual origin must have antedated the oldest formations where it is known to exist. It may well have been these ancestral types (like the highland fishes of Brazil) which were dis- tributed and not each individual species. These antecedent types may have arisen from the northern cosmopolitan flora and later have migrated into the southern hemisphere where this ancestral stock of the different groups underwent similar evolution in the similar environments of the Gondwana Land. This view is supported by the fact that these ancestral forms would not have necessitated a homogeneous environment, because they were more generalized forms. Two objections may be raised against this view. First, it is conceiv- able that Gangamopteris or various individual genera may have origi- nated independently in different continents or in different parts of the same continent, but the assumption that a flora of characteristic asso- ciation and unity as the gangamopterid originated in this way is highly improbable. It is not so improbable, however, if we include representa- tive types of each group in the distributed stock and if we grant, as the facts indicate, that identical evolution took place in similar environments. Secondly, there is no fossil record of the gangamopterids in North America. This objection to either a northern origin or a distribution of the ancestral genera by the way of North America is, I believe, far from being a fatal one. If the progenitors of the gangamopterids passed through North America, the line of migration may have been along the Appalachian system during the Carboniferous epoch, but the line of migration was more probably from eastern Asia by way of Alaska. At that time, the necessary Gondwana environment for the Gangamopteris flora may have existed in North America-for a brief period. If it did not, we only have to assume that the progenitors of this flora did not ~ necessitate Gondwana environments for a ready distribution. Dr. White has told me that there is a little evidence in favor of a Cordilleran migration, and recent work in the Permian of Texas indicates the same. The flora would not have required a vast period of time to inter- change between Eurasia by way of North America into South America. The chances are that no trace of these typical Gondwana environments or the ancestors of its flora would remain in either the Appalachian or the Cordilleran regions, because they have been leveled to the sea, except- ing in small patches, and re-elevated several times since that epoch. If the flora was a highland flora, we now know where to look for it in western North America. HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 91 Tn this connection, great emphasis must be placed on the stability of the Serra do Mar and the surrounding Gondwana formations of southern Brazil when contrasted with the violent geological transformations which the possible regions of migration through North America have under- gone. In fact, I firmly believe that no traces of the Gondwana flora would remain in southern Brazil, if this region had undergone the same radical changes as the possible routes of migration through North America. The mere fact that the Gangamopteris flora appeared to live under rather unfavorable conditions in which little coal was deposited is very important. Could not similar Gondwana environments have existed dur- ing a brief period in the northern hemisphere, and might not these en- vironments have been obliterated by the post-Permian changes of the North American topography? It appears that the stability of the Plana Alto has saved the Gondwana flora of southern Brazil. Professor Branner states in his “Geologia Elementar” that marine fossils are found associated with the deposits containing Stereosternum in the State of Sao Paulo. These fossils are all lamellibranchs, which include such Permian genera as Schizodus, Myalina and Conocardium, and other genera which are equally well considered as Triassic. This is the only known invasion of the sea that entered the Gondwana region of Brazil during the Permian and subsequent periods. Hence the Gond- wana of Brazil has been extremely stable. This indicates very great altitudes along the eastern side of South America during the Paleozoic epoch. These great altitudes are needed: to build the great Plano Alto and to account for the existing altitude of Serra do Mar after the vast ages of post-Permian erosion. If the Gangamopteris flora entered South America from North Amer- ica, one may ask why it is not found in northern South America. Whether it came from North America or not, I believe that it existed in northern Brazil, because the Permian inland basin was continuous from southern Brazil to the Guianas. In fact, Psaronius has recently been found by Dr. Lisboa at Floriana on Rio Parahyba do Norte. All this part of South America is very imperfectly known and awaits explora- tion. If the Gondwana flora is later found in northern South America, about 3,000 miles of a northern distribution will be abridged and the remaining distance will not be many times greater than a possible south- ern one. Hence due allowance must be made not only for the imperfec- tion of the fossil record but also for the lack of sufficient exploration. When this is done, a northern distribution of the Gondwana flora is not altogether impossible. S—NY 92 ANNALS NEW YORK ACADEMY OF SCIENCES As an example of the imperfection of the fossil record may be given Cryptobranchus, an amphibian, living in Japan and the Mississippi Val- ley. The only known fossil relative is Andrias, Scheuzer’s Homo diluvii testis of the Miocene of Europe. Its distribution, in a way, is like that required for a northern dispersal of the Gondwana flora. As an example of the imperfection of exploration may be given the two new marine horizons in the Conemaugh series of western Pennsylvania, found by Raymond in a region which has been explored by many geologists. As I have already stated, however, there is little hope of finding the necessary antecedent types of the Gangamopteris flora in North America, because the necessary highland deposits which produced this stunted flora have apparently disappeared, or have not yet been found. The idea of a continuous Gondwana Land has little or no support other than indecisive statistical data derived from the distribution of living and extinct animals and plants. Furthermore, it appears to me that even such a vast amount of indecisive data which admit of a variety of interpretations can never outweigh the fact that the Permian reptiles, the ooze and shark teeth, etc., of the great ocean depths, such as exist between Africa, Australia and the Antarctic islands, the geology of the Brazilian coast, the absence of giant east and west trend lines of South America and the Antarctic islands offer strong positive evidence that a continuous Gondwana Land did not exist. Until this vast array of posi- tive evidence in favor of the persistence of the great ocean depths and the continental shelves has been satisfactorily accounted for, it is just as inconceivable to explain the existence of a continuous Gondwana Land as to conceive either a northern distribution of the Gangamopteris flora or that this flora developed orthogenetically from some unknown north- ern ancestors which evolved from the Devonian cosmopolitan flora. This is all the more true when we have to acknowledge that the actual point of family origin of the Gangamopteris flora is unknown. Before, however, the reader forms an opinion concerning the existence of a continuous Gondwana Land, as is indicated by the distribution of the Gondwana flora, he should consider still two other possible means of distribution of the Permian plants which require no continuous Gond- wana Land. The distribution of the Gondwana flora by strong winds like those which blow dust from the pampas of Argentina to Africa is possible but perhaps not probable. The Gangamopteris plants are now generally considered to be seed-bearing, but in most cases no seeds are known; hence wind distribution would be out of the question, unless the seeds were of very light-winged type. Then they might have been blown for HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 93 some distance, such as might have intervened between South Africa and the nearest Antarctic islands. If any of them were spore-bearing, then wind distribution would be an important factor, as it is in the case of puff-balls and ferns. It is then important to settle definitely whether any or all of the early Gangamopteris flora were seed-bearing or sporogenous. The last possible mode of distribution is by the way of the sea. David White has informed me that some of the seeds may have been able to survive marine drift for some time. He thinks that the migration was by the Antarctic, and if by the way of the sea, it would have been with a minimum interruption by water. He suggests that it would perhaps be better to say that migration was probably by several lands and not by a continuous Gondwana Land. As the facts indicate, however, this flora was a highland flora, and hence few or none of the species could have been distributed in this way unless they lived on low coastal sandy high- lands or campos such as exist in parts of Rio Grande do Sul, Brazil. In conclusion, we may safely say that we do not definitely know where the Gangamopteris flora originated, how and which way it dispersed, why it appeared and disappeared in Brazil during the formation of the Plano Alto and why so many species and genera remained almost iden- tical in such remote regions as India and Brazil during most of the Permian epoch. The distribution of the Permian reptiles, the deep intervening sea, the trend of the Archean mountains, the mode of the formation of the Per- mian deposits, the location of marine deposits and the evidence in favor of the persistence of the great ocean depths and the continental shelves offer conclusive evidence that no continuous Gondwana Land has existed between South America and the eastern hemisphere, at least since Car- boniferous time. Previous to this, it may have existed, but many data are needed to prove that it did. In view of the fact that the Gangamopteris flora once formed did not appear to vary, we have only to explain how it got, one time, into such remote regions as India, Africa and Brazil, because a continuous ex- change of the flora would have been unnecessary. Therefore an acci- dental marine drift of the seeds and the wind, if any were sporogenous, by way of the Antarctic islands are possible means of distribution, but I believe that the distribution of the ancestral stock was along the Asiatic- American “backbone” of the earth and a subsequent similar evolution in similar environments or else orthogenesis of this stock agrees better with the known data relating to geographical distribution. The widely accepted view of a continuous Gondwana Land has been derived from the static viewpoint of living and extinct animal and plant 94 ANNALS NEW YORK ACADEMY OF SCIENCES geography, but it is no longer tenable. The separate portions of the Gondwana Land are, however, more interesting now than ever. The only places for various past connections which are needed and almost universally accepted are the following: Southern South America and perhaps the Antarctic islands. South America and North America by way of Central America and perhaps the West Indies. North America and Eurasia by the way of Greenland and the North Atlantic, Alaska and Siberia. Southern Asia and Australia by way of the East Indies. Kurasia and Africa.** PERMIAN REPTILES Does the distribution of the Permian reptiles indicate the existence of a connection between Africa and South America? Only a few specimens of Permian reptiles have been found in South America. Mesosaurus brasiliensis is the best known species. It was described by McGregor (1909) from the bituminous shales of Iraty, Parana, Brazil. Stereoslernum tumidum Cope is a closely related form. It was found in Sao Paulo and comes from the surface of a thin layer of limestone. Many fragments of it were seen by the writer near Piraci- caba at a limestone quarry on the property of the Agricultural School. A few well-characterized marine fossils have occasionally been found in the series of beds in which Stereosternum is found. 'T'wo more species of Mesosaurus are known from the Dwyka beds of South Africa. Mesosaurus is not a diapsid. Its unique vertebre and ribs, as well as the absence of scales, webbed feet, dorsal but no lateral temporal fenestra, slender teeth, long snout, etc., separate the genus from all known reptiles. Von Huene (1910) derives it from some unknown Carboniferous coty- losaurian. So far, not even the antecedent type, which gave rise on one hand to Mesosaurus and on the other to Stereosternum, is known. It is “1 Tt must be granted that of all the evidence in favor of a continuous Gondwana Land, its flora appears to be the best. But in view of the fact that when it was once formed it did not appear to change, we may suggest as a future working basis that this flora offers a special type of orthogenetic development which has been produced from the cosmo- politan older flora by definitely directed changes in the environment during the formation of the highlands where it is found. In Australia, this flora appears to have been the maker of coal during the Permo-Carboniferous. It is also said to be associated with marine drift and glacial deposits. Hence it appears to be a swamp flora in Australia, but the presence of thick beds of coal and glacial drift in the same regions does not appear to harmonize. If this is true, then the Gondwana environments of Brazil and Australia are distinctly different. A continuous Permian and early mesozoic Gondwana land is needed no more than a Tertiary or a recent one. We now know, however, that no Tertiary Gondwana is required to explain the distribution of animals. HASHMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 95 highly probable that this antecedent type, and not Mesosaurus, was the form which was distributed. The exponents of the Gondwana theory must assume that Mesosaurus originated and died off on the Gondwana continent where its ancestors are not known, and that it necessitated a continuous Gondwana Land when its distribution is used to support a connection between Africa and South America. They must also assume that Mesosaurus was the form which was distributed and that Stereosternum evolved from it after it arrived in South America. Two objections can be raised against this view which are based on positive evidence. First, the nearest ancestral cotylosaurians known are from the northern hemisphere, and none are known from South America. Hence the positive evidence in one case becomes negative evidence in the other. In other words, the positive evidence is in favor of a northern origin of the ancestral form which gave rise to Mesosaurus and Stereosternum. Secondly, Mesosaurus was a good aquatic reptile and did not need a continuous Gondwana Land in order to get into Africa and South America. It is evident, then, that the point of origin of the Mesosauria is un- known. We only know the point of extinction of a few individuals of three species of the genus. The genus was already very distinct from other reptiles in the lower Permian, and its ancestral stock could easily have arisen in the northern hemisphere from some unknown cotylo- saurian. In fact, Moodie’s paper on the Carboniferous air-breathing vertebrates of the United States National Museum indicates that this is probable, because Isodectes punctulatus Cope from the Allegheny series and Sauravus costei Thevenin from the Carboniferous of France very remotely point back to Microsauria on one hand and to Mesosauria on the other. At any rate, the Mesosauria must have originated before the beginning of the Permian, and the point of origin could have been in the northern hemisphere just as well as in the Gondwana continent. The fact that the Mesosauria are known only from Africa and South America is in favor of the latter view, while the nearest related antecedent types of Carboniferous Cotylosauria are in favor of the former. Even if we assume that the point of origin of the Mesosauria was in either Africa or South America, their distribution offers absolutely no argument for a connection between these continents, because the Meso- sauria were aquatic reptiles, as is shown by their long snout, long needle- like teeth, lack of scales, dorsal position of nares, unique ribs and webbed, paddle-like feet. McGregor’s reconstruction of M. brasiliensis indicates that it could not have traveled overland, and inasmuch as it appears to me that no river could have flowed from Africa into South America, or 96 ANNALS NEW YORK ACADEMY OF SCIENCES vice versa, during any past epoch, Mesosaurus would have had to enter the sea in order to get into both Africa and South America. On account of the intercalated marine Permian in the region where Mesosaurus is found as well as the underlying limestone containing marine lamellibranchs where Stereosternum is found, I am inclined to believe that Mesosawrus was at least semi-marine, if not entirely marine. Its needle-like teeth strengthen this view, because they are adapted to eating soft animals, which must have been far more abundant on the surface of the sea than in the shallow Permian swamps of Brazil, which became dried up again and again, at which time Mesosaurus would have been pushed down to the coast. There is, then, little or no doubt that Mesosaurus could live both in salt and fresh water just as Manatus, and originally Jnia, which are now found in Rio Amazonas. This being the case, it could easily have extended its range across the Atlantic, because it was a good swimmer. It could also have gone by way of the European- American coast or from the nearest Antarctic islands to southern Africa, where the distance would not have been much greater than traversed by the giant tortoises (Testudo) or the semi-marine Jguamde (Ambly- rhynchus cristatus) of the Galapagos Islands.*? The mere fact that only Mesosaurus, the best aquatic form, out of sixty-nine genera of Permian and early Triassic reptiles recently enumer- ated by Broom for South Africa, has been found in South America, is strong evidence that no connection existed between these continents. The absence of this vast array of land reptiles from the corresponding Permian and Triassic deposits of South America is negative evidence, but it appears to me to outweigh the positive evidence of the marine or semi-marine Mesosaurus.** Scaphonyz fisher. Woodward from the Triassic of Rio Grande do Sul, Brazil, is another form which has been used to support the idea of a connection between South America and Africa. According to von Huene, the known fragmentary data, which have been derived from verte- bre and foot bones, indicate that Scaphonyx is distinct from Hrythro- suchus of South Africa. He also thinks that both Scaphonyx and Erythrosuchus are related to several forms found in the Triassic of North America. A form of Hrythrosuchus is also known from Europe. Here again the ancestral stock, which was widely distributed and gave rise to these genera, is not known. Hence the fact that Scaphonyx of Brazil has a related genus in Africa is not evidence that these continents were connected, because it also had related forms in Europe. The great ab- #T do not believe that the Galapagos Islands were ever connected with South America. 4 It is to be noted that the region of the Permian deposits of southern Brazil has been inhabited longer than southern Africa, and it is nearly always the natives and not the few scientific explorers who first find strange animals and fossils. HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 9% sence of Permian and Triassic South African land reptiles from South America indicates that Scaphonyx evolved from some northern ancestor from which Erythrosuchus also descended. Broom and others have attempted to show that the Permian reptiles of South Africa and Texas are related. There appears to be no doubt that these reptiles are related, but there is still very diverse opinion as to how nearly they are related. In the conclusion of this interesting paper, Broom suggests a possible scheme of distribution of the Permian reptiles. He thinks that it is probable that a Lower Carboniferous land vertebrate fauna existed in northern South America. This primitive vertebrate fauna included among other forms temnospondylous amphibians, primi- tive cotylosaurians and primitive ancestral pelycosaurs. He presumes that this fauna ultimately migrated on one hand to North America and on the other hand over the Gondwana Land to Africa. In order to obviate the absence of all early Permian reptiles in South Africa except- ing Mesosaurus, he further presumes that Permian glaciers of Brazil prevented the other primitive reptiles from reaching Africa until middle and late Permian. The above view is possible, but not probable, because it is based on the following assumptions: (@) a continuous Gondwana Land; (0b) the existence of lower to middle Permian glaciers in the region of the alleged trans-Atlantic Gondwana, and not in region of northern South America, which was with all probability the highest point at that epoch, and (c) the existence of primitive Carboniferous reptiles and temnospondylous amphibians which are not known from South America. We conclude, therefore, that there is little or no evidence in favor of the exchange of Permian reptiles between South America and Africa by way of a con- tinuous Gondwana Land. In view of all this, the only suggestion which appears to agree with the known facts of geology, paleontology and the Permian environmental complexes is that the primitive Carboniferous reptiles, from which the Permian fauna evolved, originated in the northern hemisphere and were pushed south from Eurasia into Africa, where the descendants retained certain primitive characters and evolved along similar lines in such a way that they more or less remotely resemble the descendants from the same primitive stock which lived in Texas. Mesosaurus is an aquatic and at least semi-marine form, and does not lend any positive support to a Permian connection between Africa and South America,** because 44The distribution of the extinct and living side-necked turtles (Pleurodira) offer another case of the same principle. The pleurodirans are now found only in the southern hemisphere, but they were very abundant in the Cretaceous of the northern hemisphere, where they probably already existed in the Jurassic. The two other groups of “shelled turtles’ (Cryptodira and Trionychoidea) also fit into the scheme of a northern origin and distribution of land animals. 98 ANNALS NEW YORK ACADEMY OF SCIENCES none of the typical South African Permian land reptiles have been found in South America. MAMMALS Is an Antarctic connection between Patagonia and the Australian realm needed to explain the distribution of any of the South American extinct mammals? : The best evidence which has been used to support the Antarctica theory is derived from the mammals. It is the best evidence, because slow-moving mammals need land connections more than do either flying or aquatic animals and because the Tertiary record of the mammals is fairly well known. There is, however, a great blank in the fossil record in the entire lack of pre-Oligocene mammals of Asia and northern South America. The absence of pre-Oligocene animals in both Asia and northern South America is either due to imperfection of the fossil record or to the lack of exploration, because the existence of pre-Oligocene mammals in North America, Patagonia and Africa could not be explained unless the mam- mals entered both Asia and northern South America; for otherwise we must assume the separate origin of mammals in two or three different places. The works of Sclater, Wallace, Lydekker, Matthew, Osborn and others indicate that the most of the orders of mammals directly or indi- rectly originated in the northern hemisphere, which has embraced the bulk of the land at least during the age of mammals. It is true that South America and Africa have been separate centers of origin of many ~ mammals, but even many of these can be remotely traced back to the northern hemisphere. The presence of primitive mammals in the Tri- assic of North America and the Jurassic of North America and Europe taken in connection with the geology of Europe is sufficient evidence to show that the pre-Oligocene animals must have existed in both Asia and northern South America. The distribution of mammals, as I see it, in- volves, unfortunately, exactly the above regions from which we have no fossil evidence. These are transitional regions between the northern and southern hemispheres. Until the known fossil-bearing region of Bahia, Brazil, is examined and until mammalian fossils have been found in the early Eocene of northern South America and southern Asia, the distribu- tion of the Mammalia will never be satisfactorily settled. Nevertheless, on account of its profound geological significance, I think that a brief re-examination of the materials of the distribution of South American mammals should be attempted. The only support for the Antarctica theory from the standpoint of the Mammalia is derived from the affinities in the common presence of both HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 99 polyprotodont or carnivorous forms, allied to the existing Tasmanian wolf (Thylacynus) and of the small diprodont herbivorous forms (Ceno- lestes) very remotely allied to the phalangers and other Australian diprotodonts. The researches of Broom, Gregory, Dieder and others seem to favor the view that Cenolestes is an independent offshoot of the polyprotodont type which was present in the Eocene of North America and the Oligo- eene of North America and Europe. It is also to be noted that the Patagonian cenolestoids (Hpanorthus and its allies) show no clear evi- dence of close relationship with Australian diprotodonts. There can be no doubt that the sparassodonts are true polyprotodont marsupials, as shown by Sinclair. They also agree with the Tasmanian Thylacynus in certain characters which have been assumed to indicate that they belong to the same family. Dr. Matthew, however, is now of the opinion that these few characters have probably arisen independently in the Patagonian and Tasmanian genera by virtue of parallel evolution from primitive didelphids of northern origin. The Tasmanian and Patagonian genera are the end result of cenotelic evolution. It was not these genera which were widely distributed, be- cause there are none in common. It was their ancestral stock, if they are genetically related, which became widely distributed. In the case of the cichlid fishes, I have shown that it was not any of the living genera which were distributed into Africa and South America, but it was a primitive form. Could not this ancestral marsupial, from which Thylacynus on one hand and the sparassodonts (Borhyena, etc.) on the other evolved, have originated in the northern hemisphere from some primitive northern polyprotodont during the Mesozoic to -early Hocene? At any rate, it is not yet known from either Patagonia or Tasmania. This primitive ancestor could have been pushed out of Asia into Aus- traha and out of North America during the late Cretaceous to early Eocene into South America. Then similar evolution in similar environ- ments would easily account for the rest of the puaulasity of the Pata- gonian and Tasmanian Thylacynide. Until it has been definitely shown what this primitive ancestor of the Thylacynide was and where it originated, it appears to be useless to reconstruct the surface of the earth from such evidence. It is interesting to note that the evolution of the South American mammals agrees in a general way with Schuchert’s view of the connec- tions which have existed between North and South America. The first connection existed from the late Cretaceous to the early Hocene, and then 100 ANNALS NEW YORK ACADEMY OF SCIENCES a separation ensued until the Miocene, after which there has been a per- manent connection. It is exactly from the Eocene till the Miocene that South America evolved its typical mammalian fauna, whose last remains are the anteaters, armadillos, cenolests, sloths and a few tropical mar- supials. ‘This indicates that the primitive ancestors of these animals along with others entered South America during the Cretaceous to the early Eocene. It was during late Miocene time that the second important change in South American mammalian life took place. This invasion was without doubt from the north. The third wave was also from the north. It was composed of man and his domesticated animals. The replacement of the older fauna by the later invasions is still seen on all hands in different tropical animals, which still retain the old paleotelic northern characters which are, however, more or less masked by the specialized cenotelic characters.*® In view of all the preceding, the writer, while still in the field, changed his previous views concerning all of the hypothetical connections between South America and the eastern hemisphere, and he now believes that all of the South American animals originally came from North American stock. IT am also inclined to believe that the evolution of paleotelic characters, especially of families and orders, has taken place faster in the northern than in the southern hemisphere. This is indicated by the fact that many tropical animals are often a few geological ages behind their north- ern living or extinct relatives. The edentates, monotremes, ratite birds, many South American birds (screamers, seriamos, sun bittern, ete.), the characins, dipnoi, crossopterygians and osteoglossids (fishes), South Afri- can secretary bird, note Aardvark (Orycteropus), scaly anteaters (Manis), tapirs, camels and many marsupials are examples of tropical animals which are a few geological ages behind time. This retarded evolution of paleotelic characters in the southern hemisphere may be due to a greater stability of the vast Plano Alto of South America. It is not, in my opinion, the stable portions of the earth which have produced the bulk of evolution, but it is the ever-changing regions either by elevation and submergence or tremendous changes produced by ero- sion, like the recent formation of the Amazon Valley. These violent changes produced in the environmental complexes appear to pull the trigger of evolution. Inasmuch as geology shows that more radical 45In taking this view, I have assumed that Matthew and Gaudry are correct in con- sidering the Patagonian beds to be later than the upper Cretaceous. Roth, however, considers the Notostylops beds to be upper Cretaceous, and Ameginho considers them to be still earlier. HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 101 changes in the environments, including the climate, have occurred in North America than South America, it appears probable that paleotelic characters evolve faster. This belief may seem absurd, yet if it is only in part true, it has a profound significance in correlation, and especially in determining the exact age of the fossil beds in South America. If it is true, then older-appearing South American beds are in reality much more recent. At any rate, this impression demands careful study, and especially in the case of the invertebrates. I believe that the North Polar theory of the origin of land animals expounded by Haacke and in a general way supported by Wortman, Sharff and more recently by Matthew, is the view which agrees best with all of the known facts of geology, paleontology and zoology. Ruti- meyer, Huxley, von Ihering, Forbes, Ortmann, Hedley, Sinclair, Ame- ginho, Osborn and others have maintained connections between Australia and Patagonia, but their evidence has been derived almost entirely from the static viewpoint of zodgeography, which, as Tower has well said, is a dead and profitless pursuit. Besides, there never has been a general agreement between any of these authors either in regard to exact position or time of existence of the connection. They have also utterly failed to show how and why just certain animals were able to get across the con- nection. Why, for example, did not edentates and other early Tertiary mammals of Patagonia also get into Australia? Would not such a con- nection have had a barrier? - Besides, the distance across this south polar continent is not small. They also do not attach much importance to the strong geological evidence against such connections. The ideal northern marsupial from which we could easily derive both Thylacynus and the sparassodonts is not definitely known to exist, but it is also not known in either Patagonia or Tasmania. In fact, I should expect to find it in Asia and northern South America, both of which places are entirely unknown from the standpoint of primitive mam- malian paleontology; but even if the necessary ancestor is never found, it will not be the only gap left open in paleontology. The indecisive evidence used in support of the Antarctica theory does not appear to me to outweigh the fact that neither the deep-sea sound- ings, the trend lines, the lack of islands, location of Archean rock nor the location of known marine formations even vaguely suggest a Pata- gonia-Australian connection. Besides, such indecisive biological data are not as weighty as the vast array of data in favor of the persistence of the continents and the great ocean beds, so ably defended by Sir John Mur- ray and others. The deposits in the great ocean depths like those be- tween South America plus the Antarctic islands, Africa and the Aus- 102 | ANNALS NEW YORK ACADEMY OF SCIENCES tralian realm have never been found in the whole geological series of the continental shelves. Also, the great number of shark teeth found in abysmal depths indicate vast time for deposition. Until all such data and the theory of isostacy have been satisfactorily accounted for, it ap- pears to be useless to continue hypothesizing land-bridges. At the present stage of our knowledge, we do not need an Antarctic land-bridge, but we do need both dynamical data and more careful field work in northern South America and southern Asia before we can definitely settle the distribution of mammals.** SUMMARY We have seen from the location of the Archean and early Paleozoic rocks that about the present outline of South America has always existed and that the lines of weakness and strength in its crust are usually par- allel to the coasts. Hence, the invasions of the seas have, in most cases, been in a general southern-northern direction and not east-west. The location of the deposits left by the invasions of the sea has forced us to deny the existence of Archiguiana, Archamazonia and Archiplata as maintained by some of the exponents of the Archhelenis theory. The outlines of the Plano Alto, which was deposited in a continental Permian inland basin, has been given, and the general dip of its surface, its lack of past Paleozoic marine deposits, location of surrounding Archean mountains and marine sediments and the Tertiary rise of the Andes indicate the reversal of the Amazon during the later half of the Tertiary epoch. The eastward movement of the mouth of Rio Negro and the single channel of the Amazon in region of Obidos, where remains of the Plano Alto approach the river, indicate that this is near the old divide which has been washed away. The unique marine or brackish water fossils of Alto Rio Amazonas apparently lived in an arm of the ocean (Hast Andean Sea), which probably extended south, lost its con- nection and finally disappeared with the Tertiary rise of the Andes. It was also suggested from the character of the overlap that no great exten- sion of land to the east of the present coast was needed to form the sedi- ments of the Plano Alto and that great altitudes probably existed in eastern Brazil and Guiana during late Paleozoic times. The southeastern Brazilian coast appears to be very old and remark- ably stable. It apparently never extended more than about 100 miles to the east of its present location. The fringe of upper Cretaceous deposits 46 The evidence from the standpoint of the Mammalia for a Tertiary Archhelenis, 7. e., a connection between South.America and Africa, is given in the “Age of Mammals,” which shows that such a connection did not exist. HASEMAN, GHOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 103 along its northern portion, the fan-shaped trend lines west of Pernam- buco which fade away toward the coast and are roughly comparable to a somewhat analogous condition in the region of La Paz, Bolivia, its recent slight elevation, its lack of numerous islands separated by comparatively shallow sea and its abruptness being primarily due to erosion of Paleozoic and post-Paleozoic land deposits and not to post-Paleozoic faulting, indi- eate that South America was never connected with the eastern hemi- sphere. This view is further strengthened by the absence of deep-sea ooze, etc., from the continental shelves and the abundance of sharks’ teeth in the great ocean depths which indicate a vast antiquity for the abysmal depths. We have noted many facts concerning the South American topography, the most interesting of which were the cases of stream piracy existing between Rio Orinoco and Rio Negro, and Rio Sao Francisco and Rio Tocantins. The Paraguay River is not connected with the Amazon. These facts, taken in connection with waterfalls, altitude, swamp produc- tion, erosion and the composition of environmental complexes, have led to some interesting results concerning the distribution of the South Amer- ican fishes. Using the cichlid fishes only as the best known family of South Amer- ican animals, it has been shown that isolation or barriers and interming- ling or river connections utterly fail to explain their distribution. It was found that the present distribution of the fishes is correctly explained by the organic complex of the more generalized highland genera (which are small in size and naturally widely distributed because the Plano Alto was formerly a continuous unit) and by the action of the environmental com- plexes on this stock. In other words, when the common ancestral forms arrived in similar environments, 1. ¢., similar environments were eroded in the Plano Alto, they evolved along similar and identical lines and in different environments along different lines. When we attempted to determine the point of origin and lines of dis- persal of families and orders, it was found to be absolutely necessary to invoke the aid of fossils. In doing this, it was found necessary to use more than single physiological characters and draw a sharp distinction between paleotelic and cenotelic characters. When this was done, the fishes evidently point to a northern origin and not to an African-South American Gondwana origin. When similar methods were applied to the Permian reptiles, Gondwana flora, mammals and other alleged evidence in favor of connections be- tween South America and the eastern hemisphere, the evidence was not found convincing for a single case. Thus the Permian reptiles, if crit- 104 ANNALS NEW YORK ACADEMY OF SCIENCES ically studied, offer evidence against instead of for a continuous Gond- wana Land. In fact, all of the alleged evidence has been derived from the static viewpoint of plant and animal geography which has led to many erroneous views of correlation and geology of South America. There was not found a single case in the evidence for a continuous Gond- wana of any age or location in which the distributed ancestral form was actually known. In view of all this, I have been forced to change my former belief in a connection between South America and the eastern hemisphere, because the geological evidence overwhelms the biological hypotheses. The frag- mentary positive evidence in a few individual cases may not always indi- cate that this view is true, but when both positive and negative evidence derived from botany, zoology, paleontology and geology is carefully weighed and due allowance is made for the imperfection of exploration and the imperfection of the fossil record, the evidence is decidedly against the hypothetical connections. I therefore believe that continental forms have originated and dis- persed over three great tongues of land which have always extended south from the northern hemisphere. These three great tongues of land have been connected and disconnected from time to time, and it is possible that they were connected at the south pole at some time previous to the | Carboniferous epoch, but so far there is little or no evidence for such a view. There have been from time to time possibilities of plants and animals interchanging between these three tongues of land by way of the north- ern hemisphere. In a word, we do not accept the theses of hypothetical land-bridges and invasions of the sea; we fail to appreciate the weight of the evidence in favor of these theses, and we look forward with keen interest to the results of coming years in field work and in dynamical studies especially in the regions of zoddistributional transitions, 7. e., Central Asia and northern South America.*? ‘7 Tt is in a way extremely unfortunate that so much work is done in regions like Pata- gonia and southwest United States in order to prove a theory which can only be proven by hunting the deposits in northern South America. HASEMAN, GHOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 105 BIBLIOGRAPHY. The following bibliography is not intended to be complete but merely repre- sentative. It contains the important references bearing on the subject of this thesis. ADAMS, C. C.: Southeastern United States as a Center of Geographical Dis- tribution of Fauna and Fiora. Biol. Bull., Vol. III, No. 3. 1902. Agassiz, Prof. and Mrs. Louris: A Journey in Brazil. Ticknor and Fields, Bos- ton. 1868. 5 ALLEN, J. A.: The Influence of Physical Conditions on Species. The Radial Review, May, pp. 107-140. 1877. ALLAPoRT, S.: On the Discovery of Some Fossil Remains Near Bahia in South 7 America. Quart. Jour. Geol. Soc., Vol. XVI, pp. 263-266. London. 1850. AMALITZSKY, W.: Zechestein Animals of Northern Russia. Comptes Rendus, Vol. CXXXII, p. 591. 1901. AMEGHINO, FLORENTINO: Les Formations sedimentaires du Crétacé supérieur et du Tertiaire de Patagonia. Ann. Mus. Nac. Buenos Ayres, Vol. XV. 1906. AMMON, L. von: Deyvonische Versteinerungen von Lagoinha in Matto-Grosso. Zeits. der Ges. fiir Erdk. zu Berlin, Vol. XXVIII, pp. 352-356. 1893. ANDERSON, J. G.: On the Geology of Graham Land. Bull. Geol. Inst. Upsala, Vol. VII, pp. 19-71. 1906. Bates, H. W.: The Naturalist on the River Amazon, with Memoir of the Author by E. Clodd. 8 vo, 352 pp. London. 1892. Bateson, WILLIAM: On Some Variations of Cardium edule Apparently Corre- lated to the Conditions of Life. Phil. Trans., London, Vol. 180, B, pp. 297-330. 1889. BortTceR, Oscar: Die Tertiirfauna von Pebas am oberen Marafion. Jahrb. Geol. Reich., Wien, Vol. XXVIII, p. 503. 1878. : Idem. Jabrb. Kaiserl. K6nigl. Geol. Reich., Wien, Vol. XXVIII, p. 485. 1878. Bowman, IsAtAH: The Physiography of the Central Andes. Amer. Jour. Sci., Ser. IV, Vol. XXVIII, pp. 197-217, 373-402. 1909. BRACKENBUSH, L.: Estudios sobre la Formacion Petrolifera de Jejuy. Bol. Acad. Nac. Ciecos, Cordaba, Vol. V, pp. 137-184. 1883. : IBid. Viaje a la Provenica de Jejuy, pp. 185-252. 1883. BraNnner, J. C.: A Complete Bibliography of the Geology, Mineralogy and Paleontology of Brazil. Bull. Geol. Soc. Am., Vol. XX, pp. 1-132. 1909. : The Economie Geology of the Diamond Bearing Highlands of the In- terior of the State of Bahia, Brazil. Engr. and Min. Jour., Vol. LXXXVII, pp. 981-987, 1031-1033. 1909. : Geologia Hlementar. 305 pp. Rio de Janeiro. 1906. : Stone Reefs of Brazil. Bull. Mus. Comp. Zo6l., Vol. XLIV, pp. 8-321. 1904. : The Oil-Bearing Shales of the Coast of Brazil. Trans. Amer. Inst. Min. Engr., Vol. XXX, pp. 537-554. 1901. : Geology of the Northeast Coast of Brazil. Bull. Geol. Soc. Am., Vol. VIII, pp. 41-90. 1901. 106 ANNALS NEW YORK ACADEMY OF SCIENCES BRANNER, J. C.: Cretaceous and Tertiary Geology of the Sergipe-Alagoas Basin of Brazil. Trans. Amer. Phil. Soc., Vol. XVI, pp. 369-434. 1889. : The Supposed Glaciation of Brazil. Jour. Geol., Vol. I, p. 753. 1889. Broom, R.: Contributions to South African Paleontology. Ann. S. Afr. Mus., Vol. VII. ©1909. : A Comparison of the Permian Reptiles of North America With Those of South Africa. Bull. Am. Mus. Nat. Hist., Vol. XXVIII, pp. 197-284. 1910. Brown, C. BarrRIneton: On the Tertiary Deposits of Solomoes and Javary Rivers. Quart. Jour. Geol. Soc., Vol. XX XV, pp. 76-79. 1879. CARRUTHERS, W.: On the Plant Remains from the Brazilian Coal Beds. Geol. Mag., Vol. VI, pp. 151-156. 1869. CASTELMAU, FRANCOIS DE: Expedition dans des parties centrales de Amerique Sud. Histoire du Voyage, Vol. I-V, Paris. 1850-1851. CHANDLESS, W.: Notes on the River of Aquiry, the Principal Affluent of Rio Purus. Jour. Roy. Geog. Soc., Vol. XX XVII, p. 119. 1866. CLARKE, JoHN M.: A Fauna siluriana superior do Rio Trombetas, Estado do Para, Brazil. English and Portugese. Arch. Mus. Nac. Rio de Janeiro, Vol. X, pp. 148. 1899. : Paleozoic Faunas, Para, Brazil: The Silurian Fauna of Rio Trombetas ; The Devonian Mollusca of the State of Para. English Edition, Albany, N. Y. 1900. Portugese Edition, Mus. Nac. Rio de Janeiro, Vol. X, pp. 49- 179. 1899. Conrap, T. A.: Description of New Fossil Shells of the Upper Amazon. Amer. Jour. Conech., Vol. VI, pp. 192-198. 1871. : Remarks on Tertiary Clays of the Upper Amazon with Description of New Shells. Proc. Phil. Acad. Sci., Vol. XX VI, pp. 25-32. 1874. Corr, H. D.: The Carboniferous Genus Stereosternum. Amer. Nat., Vol. XXI, p. 1109. 1887. : Reports of U. S. Geol. Sur., Vol. III, Tertiary Vertebrates, and Book I, pp. 49-100, on Fossil Fishes. 1884. CRANDALL, RopDERIC: Geographia, Geologia, Supprimento d@’Agua, Transportes e Acudagen nos estados Orientaes do norte do Brazil, Ceara, Rio Grande do Norte, Parahyba. Publicacio No. 4, Ser. 1, D. E. por Ministerio do Viacio e Obras publicos, Rio de Janeiro. 1910. CrosBy, W. C.: On the Age and Succession of the Crystalline Formations of Guiana and Brazil. Proc. Bost. Soc. Nat. Hist., Vol. XX, pp. 480-497. 1880. Dersy, O. A.: A bacia Cretacea da Bahia de Todos os Santos. Arch. do Mus. Nac. Rio de Janeiro, Vol. III, pp. 185-158. 1878. : Contributions for Geology of the Lower Amazon. Proc. Amer. Phil. Soc., Vol. XVIII, pp. 155-178. 1879. : The Geology of the Diamond Bearing Region of Parana, Brazil. Proc. Amer. Phil. Soc., Vol. XVIII, pp. 251-258. 1879. : Geology of the Rio Sao Francisco, Brazil. Amer. Jour. Sci., Ser. III, Vol. XIX, p. 236. 1880. : The Amazon Carboniferous Fauna. Jour. Geol., Vol. II, pp. 480-500. 1894. : Nota sobre a geologia e paleontologia de Matto-Grosso. Arch. do Mus. Nae. Rio de Janeiro, Vol. IX, pp. 59-88. 1896. HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 107 Dersy, O. A.: The Supposed Tertiary Sea of Southern Brazil. Science, Vol.. VIII, p. 348. 1901. : Sedimentary Belt of Coast of Brazil. Jour. Geol., Vol. XV, pp. 218-237. 1907. Dot1o, Louis: Expedition antarctique belge; Résultats du Voyage du S. Y. Belgica-Poissons par Louis Dollo, Anvers, Paris. 1904. -EIGENMANN, C. H.: The Freshwater Fishes of South and Middle America. Pop. Sci. Month., Vol. LXVIII, pp. 515-530. 1906. : The Freshwater Fishes of Patagonia and an HExamination of the Archiplata-Archhelenis Theory. Princeton Pat. Rept., Part III, Vol. III. 1909. : Catalogue and Bibliography of the Freshwater Fishes of the Americas South of Tropic of Cancer. Princeton Pat. Rept., Part IV, Vol. III. 1910. KIGENMANN, C. H., DAvip PERKINS Warp and W. L. McATEE: On Further Col- lection of Fishes from-Paraguay. Ann. Carneg. Mus., Vol. IV. 1907. Hscuwece, W. L.: Beitriige zur Gebirgskunde Brasiliens. Berlin. 1832. JTHERIDGE, R.: Tertiary Deposits on the Solomoes and Javary Rivers. Quart. Jour. Geol. Soc., Vol. XXXV, pp. 82-88. 1879. Hvans, W.: The Geology of Matto Grosso, Particularly the Region Drained by the Upper Paraguay. Quart. Jour. Geol. Soc., Vol. L, pp. 85-104. 1893. : Expedition to Canpolican. Geog. Jour., Vol. XXII, pp. 601-646. 1903. : The Rocks of the Cataracts of the River Madeira and Adjoining Por- tions of the Beni and Mamoré. Quart. Jour. Geol. Soc., Vol. LXII, pp. 88-124. 1906. FLOWER and LYDEKKER: Mammals Living and Extinct. London. 1891. Frou, A.: Viage ao Jalapao do Norte de Goyaz. Commissao Geographiea do Estado de Bahia. Bahia. 1907. lorses, H. O.: The Chatham Islands; Their Relation to a Former Southern Continent. Roy. Geog. Soec., Suppl. Vol. III. 1893. : Antarctica, a Supposed Former Southern Continent. Nat. Sci., Vol. III, pp. 54. 1898. (ABB, W..M.: New Cretaceous Fossils from South America. Proc. Acad. Nat. Sci. Phil., pp. 197-198. 1860-1861. : Description of Fossil Clay Deposits of the Upper Amazon. Amer. Jour. Coneh., Vol. IV, pp. 197-207. 1869. Gapow, Hans: Amphibia and Reptiles. Camb. Nat. Hist., Vol. VIIT. 1901. GAUDRY, ALBERT: Etude sur une portion du monde Antarctique. Ann. de paléont., Tome I, fase. III. 1906. : Fossiles de Patagonie. Ann. de Paléont., Vol. IV. 1909. GEINITZ, H. BR.: Uber Rhatische Pflanzen und Thierreste in den Argentinischen Provinzen, La Rioja, San Juan und Mendoza. Sitz. Akad. Wiss. Wien, Math. und Nat., Cl. XCII, pp. 219-245. 1888. GeRVAIS, PAUL: Crocodile gigantesque fossile au Brésil. Journ. Zodl., Paris, T. V, pp. 2382-236. 1876. Greeory, J. W.: Contributions to Paleontology ond Physical Geology of the West Indies. Quart. Jour. Geol. Soc., Vol. Li, pp. 255-312. 1895. GreEeory, W. K.: The Orders of Mammals. Bull. Am. Mus. Nat. Hist., Vol. XXVII, pp. 1-524. 1910. 9—NY 108 ANNALS NEW YORK ACADEMY OF SCIENCES HaackE, WituM.: Nordpol als Schépfungscentrum des Landfaunas. Biol. Centralblatt, Vol. 6, pp. 863-870. 1886-1887. Hartt, C. F.: Geology and Physical Geography of Brazil. 8 vo., Boston. 1870. Hartcuer, J. B.: Geology of South Patagonia. Amer. Jour. Sci., 4th ser., Vol. IV, pp. 327-354. 1897. ' : Sedimentary Rocks of South Patagonia. Amer. Jour. Sci., 4tn ser., Vol. IX, pp. 85-108. 1900. : Origin of the Oligocene and Miocene Deposits of the Great Plaine! Page: Amer. Philos. Soc., Vol. XLI, pp. 118-181. 1902. HautTHat, R.: La Sierra de la Ventana. Rev. Del. Mus. de la Plata, Vol. It, pp. 3-11. 1892. Hawkins, J. G.: Geological Observations on British Guiana. Quart. Jour. Geol. Soc., Vol. X XVI, pp. 488-459. 1871. Hepey, C.: Considerations on the Surviving Refugees in Austral Lands of the Ancient Antarctic Life. Jour. and Proc. Roy. Soc. New South Wales, Vol. XXIX, pp. 278-286. 1895. : A Zobgeographic Scheme of the Mid-Pacific. Proc. Linn. Soc. New South Wales, Ser. 2, Vol. 14, pp. 391-417. 1899. HasEeMAN, J. D.: The Second, Third, Fourth and Fifth Reports of the Car- negie Museum Expedition to Central South America. Ann. Carnegie Mus., Vol. VII, pp. 287-387. 1911. ; Hoex, H.: Exploration in Bolivia. Geogr. Jour., Vol. XXV, pp. 498-513. 1905. HUENE, FRIEDRICH VON: Uber einen echten Rhynchocephalen aus der Trias von Elgin, Brachyrhinodon taylor. Neues Jahrb. ftir Min., ete., Bd. II, pp. 59-62. 1910. : Ueber Hrythrosuchus, Vertreter der neuen Reptil-Ordnung Pelycosima. Geol. und Paliion. Abhan., Neue Folge X, Heft 1, pp. 1-60. 1911. Huxtery, T. H.: Anniversary Address of President. Geol. Soc. Coll. Mem., Vol. III, p. 548. 1870. IHERING, H. von: Ancient Relation Between New Zealand and South America. Trans. N. Zealand Inst., Vol. XXV, pp. 481-435. 1891. : Die Ameizen von Rio Grande do Sul. Berlin Entomol. Zeits., Vol. XXXIX, pp. 321-446. 1894. : The History of the Neotropical Region. Science, N. S., Vol. XII, pp. 857-864. 1900. : : Les Mollusques fossiles du Tertiaire et du Crétacé supérieur de lYArgentine. Ann. Mus. Buenos Ayres, Ser. 3, Tomo VII, pp. 1-611. 1907. : Tertiary Archhelenis, ete., with a map. Revista do Museu Paulista, Vol. III, Sao Paulo. 1907. : Archhelenis und Archinotis; Gesammelte Beitrige zur Geschichte der neotropischen Region. Leipzig. 1907. : Archhelenis Theorie. Neues Jahrb. ftir Min., ete, Beilage Band XXXII, pp. 184-176. 1911. JOHANNSEN, W.: Hlemente der Hxakten Erblichkeitslehre. 1909. JORDAN, D. Starr: The Fossil Fishes of California. With supplementary notes on other species of extinct fishes. HEobrycon, p. 140. Univ. Cal. Pub., Bull. Dept. Geol., Vol. V, No. 7, pp. 95-144. 1907. : Bituminous Shales at Riacho Doce, State of Alagoas, Brazil. Ann, Carn. Mus., Vol. VII, pp. 23-34. 1910. HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 109 Katzer, FREDERICK: Grundztige der Geologia des untern Amazonas Gebietes. Leipzig. 1903. : As Camadas fossiliferas mais antigas da Regiao Amazonica. Bol. do Mus. Pardense, Vol. I, pp. 436-488. 1896. Kayser, H.: Alguns fosseis paleozoicos de Hstado do Parana. Rev. do Mus. Paulista, Vol. IV, pp. 301-311. 1900. JXRonE, Ricarpo: Informacces Ethnographicos do Valle do Rio Riberia de ‘Iguape. Exploracio do Rio Ribeira de Iguape por Comm. Geol. do EH. Sao Paulo, pp. 23-34. 1908. LisBoa, MiguEL ArRoyADO: Notas Geologicas feito por Hstrado do Ferro Noreste do Brazil. Rio de Janeiro. 1909. LUND, PEDRO W.: Cavernes existentes do Calcareo do interior do Brazil. Ann. do Escola de Minas, No. 3, pp. 61-91. 1884. LYDEKKER, R.: A Geographical History of Mammals. Cambridge. 1896. MacpouGaL, D. T., A. M. Vain and G. H. SHuLi: Mutations, Variations and Relationship of the G/notheras. Carnegie Inst. Wash., No. 8. 1907. McGreeor, J. M.: Mesosaurus brasiliensis from the Permian of Iraty, Parana, Brazil. Pt. III of Rept. of Commissao dos Hstudos das minas de Carvao do Brazil. 1908. MatrHew, W. D.: Is the White River Tertiary an Holian Formation? Amer. Nat., Vol. XXXIII, pp. 403-408. 1899. : Fossil Mammals of the Tertiary of Northeastern Colorado. Mem. Amer. Mus. Nat. Hist., Vol. I, pt. 7, Nov. 1901. : Patagonia and the Pampas Cenozoic of South America. A Critical Review of the Correlation of Santiago Roth. Ann. N. Y. Acad. Sci., Vol. XIX, pp. 149-160. 1910. NICHOLS, JOHN TREADWELL: Horse Mackerel of England off the Coast of Florida. Jour. Am. Mus. Nat. Hist., Vol. 10. 1910. MerriaM, C. H.: The Geographical Distribution of Life in North America with Special Reference to the Mammalia. Proc. Biol. Soc. Wash., Vol. VII. 1892. Moopig, Roy J.: Carboniferous Air-Breathing Vertebrates of the United States National Museum. Proc. U. S. Nat. Mus., Vol. XXXVII, pp. 11-28. 1909. Murray, D.: Geographical Distribution of Mammals. London. 1866. Murray, Sir Joun: Bathymetrical Chart of the Oceans. Scot. Geogr. Mag., Vol. XV, pp. 505-522. 1899. D’ORBIGNY, A.: Voyage dans l’Amerique meridionale. 4 to., III, 3, Géologie, pp. 181 et seq. 1842. OrtMaAnN, A. H.: Marine Cretaceous Invertebrates of Patagonia. Princeton Pat. Repts., Vol. IV, Pal. I, Pt. 1. 1901. : The Geographical Distribution of Fresh Water Decapods and its Bearing Upon Ancient Geography. Proc. Amer. Philos. Soc, Vol. XLI, pp. 267-400. 1902. : Tertiary Archhelenis. Amer. Nat., Vol. XLIV, pp. 237-242. 1910. ORTON, JAMES: The Andes and the Amazon, or Across the Continent of South America. 3d Ed. New York. 1876. Oszorn, H. F,; The Age of Mammals. Pp. 64-81. New York. 1910. 110 ANNALS NEW YORK ACADEMY OF SCIENCES PARISH, WOODBINE: Buenos Ayres and the Provinces of Rio de la Plata. 2nd Ed., pp. 127-128. London. 1852. PELLEGRIN, J.: Contribution 4 l’étude anatomique, biologique et taxonomique des poissons de la familie des Cichlides. Paris. 1904. PHILLIPPI, R. A.: Die tertiiiren und quartiiren Versteinerungen Chiles. 4°, Leipzig. 1887. : Piuspry, H. A.: Non-marine Mollusca of Patagonia. Princeton Pat. Repts., Pt. V, Vol. III. 1911. , ; Pissis, A.: Descripcion topografica i jeologica de la provencia de Colchagua. Anal, de Univ. de Chile. 1860. RATHBUN, RICHARD: Preliminary Report on the Cretaceous Lamellibranchs Collected in the Vicinity of Pernambuco, Brazil. Proc. Bost. Soe. Nat. Hist., Vol. XVII, pp. 241-256. 1875. —: Observacees sobre da Ihla ‘de Itaparica na Bahia de Todos os Santos. Archiv. do Mus. Nac. Rio de Janeiro, Vol. III, pp. 159-183. 1878. : The Devonian Brachiopoda of the Province of Para, Brazil. Proce. Bost. Soc. Nat. Hist., Vol. XX, pp. 14-39. 1881. Ree@an, C. T.: A Series of Papers on the South American Cichlid Fishes. Proce. Zool. Soc., pp. 152-168. 1905; Ann. and Mag. Nat. Hist., Vol. XVI, pp. 66— 77, 225-243, 316-340. 1905. Vol. XVII, pp. 49-66. 1906. — RHEINHARDT, J.: Bone Caves in Brazil and Their Animal Remains. Amer. Jour. Sci., Vol. XCVI, pp. 264-265. 1868. RoTH, SANTIAGO: Beitrag zur Gliederung der Sedimentablerungen in Patagonien und der Pampasregion. Neues Jahrb. fiir Min., ete., Bd. XXVI, pp. 92-150. Stuttgart. 1908. RUTIMEYER, L.: Uber die Herkunft unseres Thierwelt. Eine Zoogeographische Skizze. Basel und Genf. 1867. — SCHICHTEL, CARL: Der Amazonen. Strom. Versuch einer Hydrographie des Amazonas Gebietes auf Orographische-Meteorologischer Grundlage. Geo- graphische Dissertation. Strassburg. 1893. ScHompurek, R. H.: Journey from Fort San Joaquin to Roraima, to Esme- ralda and to San Carlos to Demerara. Jour. Roy. Geog. Soc. Lond., Vol. X, pp. 1-267. 1841. : f SCHUCHERT, CHARLES: Review of Katzer’s Geology of the Lower Amazon. Jour. Geol., Vol. XIV, pp. 722-746. 1906. : Paleogeography of North America. Bull. Geol. Soc. Am., Vol. XX, pp. 427-606. 1910. : Paleogeographic and Geologic Significance of Recent Brachiopoda. Bull. Geol. Soc. Am., Vol. XXII, pp. 258-275. 1911. Sciater, P. L.: On the Geographical Distribution of the Members of the Class Aves. Jour. Proc. Linn. Soc., Vol. II, pp. 180-145. 1858. Sciater, P. L. and W. L.: The Geography of Mammals. London. 1899. SEwarp, A. C.: The Glossopteris Flora, an Extinct Flora of a Southern Hemis- phere Continent. Sci. Prog., N. S., Vol. I, pp. 178-201. London. : 1897. SHarrFr, R.: European Animals; Their Geological History and Geographical Distribution. New York. 1907. SIEVERS, W.: Die Sierra Nevada de Santa Marta und die Sierra de Perija. Zeits. Ges. fiir Erdkunde. Vol. XXIII, pp. 1-159. Berlin. 1888. HASEMAN, GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA 111. Sinciair, W. J.: The Marsupial Fauna of the Santa Cruz Beds. Proc. Amer. Philos. Soc., Vol. XLIX, No. 179, pp. 78 ff. 1905. Sprx, J. B. and C. F. P. von Marrius: Reise in Brazil. Miinich. Trans. by H. H. Lloyd in two volumes. 1824. STEINMANN, G: Reisennotizen aus Chili. Neues Jahrb. f. Min., pp. 199-205. Jahrgang. 1884. : Ueber Tithon und Kreide in den Permanischen Anden. Neues Jahrb. f. Min., pp. 130-153. Jahrgang. 1881. STELZNER, A.: Mittheilungen au Professor H. B. Geinitz with a note on ques- tionable Marine Bivalves at Santa Maria, Catamarca. Neues Jahrb. f. Min., p. 728. Jahrgang. 1873. STILLE, H.: Geologische Studien in Gebiete des Rio Magdalena. Festschrift fiir A. von Koenen, pp. 277-358. Stuttgart. 1907. SuEsS, EpwarpD: Das Antlitz der Erde (The Face of the Earth, translated by Prof. Sollas in four volumes). Vols. I, II and IV contain references to South America. 1880—1904. TAYLOR, FRANK BuRSLEY: Bearing of the Tertiary Mountain Belt on the Origin of the Earth’s Plan. Bull. Geol. Soc. Am., Vol. XXI, pp. 179-226. 1910. TOWER, WILLIAM LAWRENCE: An Investigation of Evolution in Chrysomelid Beetles of the Genus Leptinotarsa. Carnegie Inst. of Wash. Pub. No. 48, pp. 1-113. 1906. : Watiace, A. R.: Geographical Distribution of Animals, with a Study of the Relation of Living and Extinct Faunas as Elucidating the Past Changes of the Earth’s Surface. 2 Vols. London. 1876. WELLS, W.: Trip from Rio SAo Francisco to the Tocantins River. Proce. Roy. Geog. Soc., Vol. VII, pp. 518-517. 1886. WuitE, C. A.: Contribucio a Paleontogia do Brazil. Archivos do Museo Na- tional, Rio de Janeiro, Vol. II. 1887. WuHuirte, Davin: Permo-Carboniferous Climatic Changes in South America. Jour. Geol., Vol. XV, pp. 615-633. 1907. : Fossil Flora of the Coal measures of Brazil, Relatario de Commisio de Estudos das minas de Caryao de pedra do Brazil, Pt. 3. Rio de Janeiro. 1908. Waite, I. C.: The Coal Deposits of Southeastern Brazil. Science, N. S., Vol. XXIV, p. 377. 1906. : Relatorio Final do Commisio de Estudos das Minas de Carviio de Pedra do Brazil, Pt. I. Rio de Janeiro. 1908. : Permo-Carboniec Conglomerates of South Brazil. Jour. Geol., Vol. 21, pp. 774. 1910. : WHITLy, H.: Tablelands of Roraima and Kukeman. Proc. Geogr. Soe., p. 458. 1884. WHyYMPER, Hpwarp: Travels Anongst the Great Andes of the Equator. New York. 1892. Woopwarp, A. S.: Vertebrate Paleontology. Cambr. Nat. Sci. Manuals. 1898. : Consideracees sobre alguns peixes terciarios dos shistos de Taubate, . Hstado de Sao Paulo, Brazil. Revista de Museu Paulista, Vol. III, pp. 63- 75. 1898. 112 ANNALS NEW YORK ACADEMY OF SCIENCES Woopwarp, A. S.: Catalogue of Fossil Fishes. British Mus., Pt. IV, pp. 73-86, 514 and 554. 1901. ; Woopwarp, H.: Tertiary Shells of the Amazon Valley. Ann. Mag, Nat. Hist., Ser. 4, Vol. VIJ, pp. 59-64, 101-109. 1871. WorRTMANN, J. L.: Studies of Hocene Mammalia in the Marsh Collection, Pea- body Museum. Pt. IJ, Primates. Amer. Jour. Sci., 4th ser., Vol. XV, pp. 419-486. 1908. ZEILLER, R.: Notes sur la flore fossile des gisements houillers de Rio Grande do Sul. Bull. Soe. Géol. de France, T. XXIII, pp. 601-629. 1895. as PUBLICATIONS | OF THE : NEW YORK ACADEMY OF SCIENCES (Lxceum oF NarturaL History, 1817-1876) | he publications of the Academy consist of two series, Viz. : iy The Annals (octavo series), established in 1823, contain the re distributed i in bundles 0 on an average of three per year. fh : as and may be learned upon application to the Librarian of the my. The author receives his separates as soon as his paper has . BS AE is aa: to eee oneal Memoirs, Volume II to Zodlogical Memoirs, A us | ete. The price is one dollar per part as issued. ! All publications are sent free to Fellows and Active Members. The Rs Annals are sent to Honorary and Corresponding Members desiring them. — 4 Subscriptions and inquiries concerning current and back numbers of _ ‘ oe of the a of the Academy should be addressed to | THE LIBRARIAN, New York Academy of Sciences, care of Lilien American Museum of Natural History, —_— New York, N. Y. ANNALS OF THE NEW YORK ACADEMY OF SCIENCES Vol. XXII, pp. 113-133, Pll. XVIEXIX Editor, EpMuND Otis Hovey SEX-LINKED INHERITANCE IN POULTRY BY T. H. Morcan anno H. D. GooDALE NEW YORK PUBLISHED BY THE ACADEMY 23 JULY, 1912 THE NEW YORK ACADEMY OF SCIENCES (Lyceum or Narurat History, 1817-1876) OFFICERS, 1912 President—Emerrson McMitxin, 40 Wall Street Vice-Presidents—J. EDMUND WoopMAN, FRrepERIc A. Lucas, CHARLES LANE Poor, R. 8. WoopwortH Corresponding Secretary—Henry KH. Crampron, American Museum Recording Secretary—Epmunp Otis Hovey, American Museum Treasurer—HENRY L. DoHERTy, 60 Wall Street Tibrarian—RaLpPx W. Towser, American Museum Editor—Epmunp Otis Hovey, American Museum SECTION OF GEOLOGY AND MINERALOGY Chairman—J. E. Woopman, N. Y. University Secretary—CuHaArLes P. Berkey Columbia University SECTION OF BIOLOGY Chairman—FrReEveErRIc A. Lucas, American Museum ee es K. Grecory, American Museum SECTION OF ASTRONOMY, PHYSICS AND CHEMISTRY Chairman —Crrsgis LANE Poor, Columbia University Secretary—F. M: PEpERsEn, College of the City of New York : SECTION OF ANTHROPOLOGY AND PSYCHOLOGY Chairman—R. 8. WoopwortH, Columbia University Secretary—FREDERIC LYMAN WELLS, Columbia University The sessions of the Academy are held on Monday evenings at 8:15 o’clock from October to May, inclusive, at the American Museum of Natural History, ?7th Street and Central Park, West. [ANNALS N. Y. Acap. Scr., Vol. XXII, pp. 118-133, pll. NVII-NINX, 23 July, 1912] SEX-LINKED INHERITANCE IN POULTRY By T. H.. Moregsan anp H. D. GooDALE (Presented before the Academy, 8 April, 1912) CONTENTS Page SUE ste ( UA Ls1 GENE eee ea near eP ei tacia! Sle) ceo ersten ccve Shove) Peiieca)ven've: 3 eis! en's: o'9:l6 ie) 6) ep! ea 113 Crosses between Plymouth Rocks and Langshans.....,.................-. 114 IDSSCENTMNOINIOE TKS KAN Sooecogacnooeeooe den so dddoooon Tage eet aod ener ak cual ood 11+ EMIS tO IVAO tab MCT CO US ayerdisrar terse syaieier clatlntthcle io. eke le eeayne Sustcesie is a!e ce Seisvey eves 115 Source of breeding stock used.................6., Make Nasheed ea eee ctaice os 115 PUT eaten seater ec te apn = ey atrag ses ac Nicene Sie ube ehe erie Srai'a neers Ga islole! eheveyale es 115 Explanation of the symbols used in interpreting the results.......... 116 Parental matings...........-. cee eee eee eee eee eee eee cece eee 118 1B, TURGUENTEIS. Sb tiesnicios O's See Cae EDO OREO OD a een ncaa er 118 IBAGIR TWNIMNES C56 das Bid an Cle.c BiG oid COC OG Ota DI IO pIOI it ai eno Io ini Carre 120 SITUOMUMMATAY ond SSE GSCI eek ERAS IS OP CR EERE Ree rear Beal et a cater kis 121 Meseripmoncvok By Adult Rlumace.-. o2.5. 2c .ccnssseessdacnes sevens 12. SS eaira Kee (20) Oder epenenecteate oicictarersyaicke cushanciats lie close cro oe skin aid aus'b.@ ww sidie Se Siar 123: ESCO CH re tepeterecet st cert cha arate ay ereiare suse wise sae claves a aisrdnay so ale angome eed alee au ese 1256 Down colors............. > GeO GiCh OC SmNCEO ES CC CRERERACRS CMTe eR ae ee ES Crosses between American Dominique females and Langshan males...... 128° SAE Ti ee CMLETAGI Ol oes crores eperatevatece? acl ers (esere cis Weave elelsvegiaicia vce sie abaule dle ses 128 F, generation from Langshan ¢ by Dominique 9.,.................. 128 F, generation eee R een ace ICIN SEMEN rarer ae aiavare laielars vie accuse a aieisis vais se waa 128 Back cross of F, ¢ (barred) to Dominique 9...................... 128 Back cross of Hi, black 9 to Langsham @.......-........c.2-..00000% 129 Oper teatures’ Of the CrOSSESi oo. obec de ce yee eee eee se cew esses 129 WIKeR heaters IMy WINGS tine cs acc eas ees ccs case sve dens ee mele eet ewses 130 COOP OF WEBI SS Ss Sola Gb EG RO SEES Oe Sra sa ee 131 Color OE TOM Ses Sere Go is. a Secs ecel OEN OREM Sep ence ene er =e ee ort 131 MEO GalerCONSIGCTALIONS. cc. 626s As eevee cc sess scenes ce decsaeceeewsec 131 SSH fom CHOI eed Ju NVM eae em at oe ai larie> hela) a Aiicto ove, bio Gia ge wiaete gh mieinis Biase ee 133 INTRODUCTION In 1908, W. J. Spillman pointed out that, according to a breeder, when Plymouth Rock females are bred to Langshan males all the females are black and all the males are barred. As far as the evidence went, it seemed to show, as he pointed out, that the case was comparable to that of the moth, Abraxas, described by Doncaster and Raynor, and of certain crosses among canaries described by Miss Durham. (113) 114 ANNALS NEW YORK ACADEMY OF SCIENCES With the intention of examining further the report cited by Spillman, and of testing, by further combinations, the offspring of the first genera- tion, we began the following experiments in 1909, using Barred Ply- mouth Rock and Langshan fowls. We undertook also to extend the experiment by using another breed of barred poultry, the American Dominiques.: It is currently stated that Dominiques (but not American Dominiques) occur in the ancestry of Plymouth Rocks. We wished to see whether “sex-limited” or “sex-linked” inheritance is found also in this other race. Plymouth Rock-Langshan crosses have been made by one of us (Goodale) on the experimental farm of Mr. B. B. Horton, to whom we are under many obligations for opportunities to carry on the work. The Dominique-Langshan crosses were made by the other (Mor- gan) at Woods Hole during the summers of 1910-11. In the meanwhile, Pearl and Surface (1910) have described the results of a cross (and its reciprocal) between Barred Rock and Cornish Game. Goodale (1909, 1910) has given briefly some of the results obtained when Barred Rocks are mated (reciprocally) with Buff Rocks and when Brown Leghorns are mated with White Rocks. Hadley (1910) has called attention to similar results published by Cushman in 1893. Davenport (1906, 1909) has described various crosses to one of which certainly (White Cochin by Tosa) and to the others less clearly may be given the same interpretation that apples to the results described in the other papers mentioned above. These crosses all involve the barring factor. Sex-lnked inheritance of other factors in poultry has been noted, not only by several of the above writers, but also by Hagedoorn (1909), Sturtevant (1911) and Bateson and Punnett (1908). To Bateson and Punnett is due the explanation — of the phenomena of sex-linked inheritance for poultry. More recently (1911) these authors have published a complete account of the inherit- ance of a factor derived from Brown Leghorns which affects the patency of the type of pigmentation characteristic of the Silky fowl. CROSSES BETWEEN PLymMoutTH Rocks anpd LANGSHANS Description of the Breeds.—F¥or a detailed description of the breeds under consideration, reference must be made to the various standard works on poultry. In this paper, only a very brief statement of the chief characteristics involved in the cross will be given. The Black Langshans (Plate XVII, figs. 2 and 3, and Plate XVIII, fig. 1) are uniformly black, varying somewhat in brillianey in different regions of the body. The shanks, too, are dull black; the bottoms of the 4 The American Dominique is a younger breed than the Barred Plymouth Rock. MORGAN AND GOODALE, INHERITANCE IN POULTRY 115 feet are gray. The shanks, moreover, are provided with several rows of feathers, or boot, along the outer edge. The comb is single. The Barred Plymouth Rocks (Plate XVII, figs. 1 and 4; Plate XVIII, figs. 3, 4, 5, 6) are black and white,” the two colors being arranged in alternate bars across the long axis of each feather. The bars vary some- what in evenness, width and depth of color from individual to individual, and also in different sections of the same bird. Although the American “Standard of Perfection” requires that the two sexes shall be alike in color, the males vary from a darker to a very light color, often appearing very light gray, while the females, though to a less marked extent, vary toward a darker shade. In other words, the breed tends strongly toward a sexual dimorphism of color, with indications of a secondary dimor- phism within each sex. The comb is single, and the yellow shanks are free from feathers. History of the Breeds —The modern Langshans are the direct de- scendants of a very old race brought from the interior of China. The Rocks, on the other hand, resulted from a mixture of several races of fowls about forty years ago, which have been gradually brought to a high degree of perfection. The history of the barred character with which we are chiefly concerned is obscure, but evidently it is of very great antiquity, for barred or “cuckoo” birds are to be found in many European countries, Russia included. Brown (1906) states that the plumage of the “Siberian Featherfooted fowl is generally white, whilst others have cuckoo plumage.” He notes also that this variety is said to be of ancient lineage. Wright (1902) states that it is probable that the “original Chittagongs, or at least their crossed offspring, were of an ‘owl’ color as described, probably what we now know as cuckoo or barred.” The Chittagongs came from the district of that name in the upper Malay peninsula. An exhaustive search would probably show that barred fowls have been recorded from southeastern Asia. Source of Breeding Stock Used—The Langshans came from P. P. Ives of Guilford, Conn. ‘lwo of the three Barred Rock males and one of the females were of the well-known Latham strain, but obtained from R. C. Goodale. Four of the barred females were the progeny of the Latham hen by a White Rock male, one was an F, from a similar mating, and one was a pure bred female from a Stamford breeder. The White Rock male is known to differ from the barred birds chiefly in the absence of the chromogen factor. Matings—In the majority of these matings, the progenies of the indi- vidual mothers have not been kept separate. The determinations of the 2 Fanciers prefer to speak of both these colors as grayish. 116 ANNALS NEW YORK ACADEMY OF SCIENCES presence or absence of barring, unless otherwise stated, was made on newly hatched chicks, or those of full term which failed to hatch. This method of determination is made possible by the presence of a gray occipital spot on those chicks which will become barred adults. A full discussion of the point, however, will be given elsewhere. Inroads of vermin, largely rats, have limited the number of which the sex was de- termined. A description of the adult hybrids is deferred until after all the matings have been described. Explanation of the Symbols used in Interpr eting the Results.—It was pointed out by Spillman, following Bateson, that sex-linked inheritance in poultry could be accounted for on the assumption that the female is heterozygous for sex and the male homozygous, and that when in the female, the barred factor alone is present, it is repulsed by femaleness. We may give this interpretation a more concrete form, if we assume that the factor in question is not carried by the same chromosome that carries the factor for the female sex; 7. ¢., in the heterozygous female the chro- mosone that carries femaleness also lacks the factor for barring, and its mate that lacks the factor for femaleness carries the factor for barring. No interchange between the chromosomes (if two really exist) can take place, perhaps because they fail to pass through those stages in synezesis when such a process becomes possible. If F = female, f its absence or male; B = barred, b its absence ; N= black, then the formulas for the barred ok will be: asi CRO cwiarever st suet svayatoesaiatitca cs iskehartie 4 Se muewere FENb {NB : Omaha oeds Sie isast aie «Sie aes eA {NB {NB Bacio POC nears eee: eee dmmne ra CELLS 3 FNb fNb CUE ey tl ak es we ae) ee fNb {Nb Whether the female-producing gamete of the Barred Rock really car- ries black or only the absence of barring will not affect the nominal results here recorded, but other experiments to be described by Goodale will show that “black” is probably present. In order to see how these formulas apply to the crosses under con- sideration, let us take first the case of the cross between the Langshan hen and the Plymouth Rock cock (fig. 1). The formule are as follows: amg shan: O03 ae eee orto recess cui sais FNb fNb Barred ‘Rock «8.0. 5 eee Se eee {NB {NB a Lone Qs ls Goer ee DS 85 FNb fNB_ 1 Ee Sen ee fNb fNB MORGAN AND GOODALE, INHERITANCE IN POULTRY 117 IBIENOG ER: O.'5 hg bbe ded GOR Se eC OER Rene eee FNb fNb USD ES crates Ommeveyenee i crstijehe el Sic. waitose: ciclg esto silts se Seis ENb fNB 2 i; MEST Oe Uo Net spel eiaray sun wie eral BALS {NB fNB aren mieten reese echey se Ss0s) sau cajie es Sos esas: wale {NB fNb Barring is dominant to self color, as is shown in the last case, where in the F, generation all the offspring are barred. In the second generation, Fig. 1.—Cross of Plymouth Rock ¢ to Langshan 9 there oceur barred ¢ and both barred and black @. The grandmaternal color, black, appears in the grand-daughters and not in the grandsons. The reciprocal cross between the Barred Rock female and the Lang- shan male (fig. 2) may be represented as follows: Ang C EPENO CRY OMe cose eo): ajlatetste eheus sieht oie: setae FNb {NB HATS IN TTD Meme A Reyer ey ecroncate, sais tel oy sirshoueie io) spt crap-se tiene fous fNb {Nb eae ee ee aia. marie ..... FNb £Nb pee PE MUTT CLC Ret shes eos ni Gil ih cna fa Mn wie one edo eneeelelonene {NB fNb ( SATCU See O rite so) cen stso tes Seicususerace Sie) sus. atvcene maerets ENb fNB I BSI @ Ee Saloval a raat ieseerie soa ack ce a ae ame coe EE VS ones ae FNb fNb 2 4 AEC Me a st aiaie statis Srasasrouslieile edocs ious giao leapsiiomenie os f£Nb {NB ISIC) Gees econ e cucenaro uo uromapa cd see {Nb f{Nb 118 ANNALS NEW YORK ACADEMY OF SCIENCES Parental Matings.—(1) From the five Langshan hens by a Barred Rock male, there were 34 young, all barred: 12 were females and 8 were males (fig. 1): FUND? SGN eecse artes aos iS aceievel eae eee Langshan 9 TINUE SIENYS Sepatcrscarcscca sare ene. wc eeoei sy oh oacaiieballe Barred Rock ¢ BUN fo eee DINU Sf aie senstere seacnisne seater ostiane voneccenrete Barred @ 12 NIB NIB CS. o crate eae iawn ei reeeke e a 8 Fic. 2.—Cross of Langshan g to Plymouth Rock e) (2) From the various barred females bred to a Langshan male, there were 20 barred (15 4 ) offspring and 25 black (162) (fig. 2): BN ENB Cee ee eae eee enter eee ae Barred 9 fNb ENDS 52) Wee eee ee Langshan ¢ NENG re INO 3s ee ee ae eet Black 92 16 PINIB GUND i ate ele eee oer ae Barred @ 15. PF, Matings—(3) Four barred F, females from (1) were bred to a barred male, No. 568, from (2). This was done because the only adult MORGAN AND GOODALE, INHERITANCE IN POULTRY 119 male, No. 784, from (1) did not mature until long after his sisters were laying, while a change of residence on the part of the writer prevented the accomplishment of the inter se mating. From the cross-mating, however, there were 25 barred (12 ¢ and 9 @ ) offspring and 13 black (8 ) (fig. 1). Expectation on the Spillman-Bateson hypothesis is 2814-914. One indi- vidual, a male, with the gray spot reduced to a few plumules was excluded from the count as doubtful. ( IEINID TENIB ceeceoance suceccsacgenode Barred 9? BE OND) cee Shen see tara IRUNG SNUB 25 28 nate ere ele yorseetenctene sce ,0 iotee Barred @ 9 MMBEBN es CEINIDIR. 0 care tc once eases Black 2 8 Fe fNB | fNB Hic. 3.—Cross of barred 4 to Fy black 9 (4a) The 6 black females from (2) were mated to a litter barred brother, No. 569, giving 41 young; 16 of which were black (86, 62 ) and 25 barred (66,109). (40) Later, they were bred to No. 784, giving 22 young; 7 were black (06, 192) and 15 barred (56,192). The combined results of these matings were 63 young, of which 23 were black (86, 72) and 40 barred (116,112) (fig. 2). The departure from the expected ratio of 3114 is considerable. 120 ANNALS NEW YORK ACADEMY OF SCIENCES r } TEND END ke Oe ciones ete eredeters eaisteemn Black 9 ©) Os SfINTES A ethIND sis Sree eaten esc Bee sees Barred ¢ ( FNbB fNB.........---. ee eee eee eee Barred @ 10+1 RUN DS ao aeNiby ee ss catches Mem ameleee Grae Black 9 6+ 1 Ea AO AEN EN se ae oe Barred @ 645 EN De CEN De cd.c.ce-ae cet hee esate he egere Black ¢ 8+6 Back Matings.—(5) The 6 black F, females used in (3) and (4) were bred to a pure Barred Rock male and gave 9 barred young, fulfill- ing expectation (fig. 3). NID aeBIND octane estes eee eee Black @ ENB) ENB aaccstis cy sitet eieisteraioseooete re Barred ¢ TUNIS US ect os clea eee eetre ee aie Barred 9 PINT: SEIN a ercierend Si chaternsee ekercho cael ceeveaaione Ss a Vic. 4.—Oross of Langshan @ to Fy, barred 9 (6) The 4 barred F, females used in (3) were bred to a pure Lang- shan male. Of the 30 offspring, 14 were barred (14,12 ) and 16 black (06,32). This corresponds closely to the expected equality in ratio (fig. 4), except for the possible barred female. The determination of sex im this case was made on rather poorly preserved material. MORGAN AND GOODALE, INHERITANCE IN POULTRY 121 TN) aiN1B Ges jogo aro coated Gh Oenioro Barred 9 (ENTS. ENID cedecan se coos coped ece pone Black ¢ ify JENID goeeeneougcnococbodep oda Black 9 (NIE ENO? G4 cedure sue eoooos Goon Ot en Barred ¢ (7) Several of the parental barred females were mated with No. 568, barred F, @, giving 8 barred (32,14) and 3 black (12 ). EWING Tove ete Nilo oes apie open laut ashy clickel oMelloneers pees Barred 9 NIS. OND coneisouoddqaseouuoupeuabeT aad ones TEIN Loe wae NUS ibe fe ese ye ses nerceriehsy ef elemece pain sretalecays Barred Q 3 IRIN Dy cpa] Ota e's & cuesmrcrche onc eine Ura irae a Black @ 1 if NG ENG ate cicie sieee ib hsloy a ckcie sue. erevese we Barred ¢ FINGES MOEND Le ene aN eA “4 \ = (8) Four of the parental Langshan females mated to No. 569, barred F, male, gave 17 barred (62,26 ) and 13 black (44 ; no record for® ). TRIN OLSEN OSs one eeeorereicio otectorais cea ey oaetrae Black 9@ fNDES eae TING cacerncncre heroes agate sey cqagscls Sree eetens Barred ¢ FUN unmet NES ener echchore are ieraite essen sachs 3 leha vhs Barred 9 6 FUN De eet skarcaec ocgerays 2 stoke cha dae ete ae Black @ ? TEN O%m COENY uke Oe sea ere rh eR a Barred ¢ 2 ifSIN JO ate aN Dicre ates iene sie eee ni aie a aes sens Black ¢@ 4 (9) Two of the barred F, females used in (3) were bred to a Rock male, not, however, of the Latham strain. There were only 5 chicks, all barred. TEIN Dean PINE eae aes Pie fers la) > a iets evare aieleta'e Barred 9 ESTEE FENTR OL E IO eee ee a nr HINDU ENIBe Eee... eat en eee -.. Barred 9 NTE PEES1S So oe “ Summary.—Expectation in all these matings has been closely fulfilled (with the exception of No. 4 and perhaps No. 6), on the assumption that the barred female is heterozygous for both barring and sex, both female- ness and barring being dominants and that the two factors do not occur in the same gamete. ; Description of F, Adult Piumage.—The males (Plate XVII, figs. 5 and 7) resemble one another closely and together with the barred F, females are very suggestive of the Coucou de Malines, a Belgian breed. The barring of the individual feathers of the males (Plate XVIII, figs. 10, 13) is less sharp and regular than that of the parental Rocks (Plate XVIII, fig. 8), while the dark bars tend to run together, particularly in wings and tail, and at the same time, the light bars become more or less 122 ANNALS NEW YORK ACADEMY OF SCIENCES smoky. The primaries, indeed, can be called barred only by courtesy, for the light bars are only represented by white splashes along the shaft (Plate XVIII, fig. 14). This region, however, is one in which the fan- ciers have found great difficulty in producing even and regular barring. One barred male is particularly interesting in that a few feathers show distinctly the Jungle coloration (Plate XVIII, fig. 12) which probably exists as a cryptomere in the Langshans. All the males have numerous feathers wholly or partly black (Plate XIX, figs. p-w) and this is true also for the barred females. ‘The last, except for the black feathers, are well barred (Plate XVII, fig. 6) and can scarcely be distinguished from the parental stock. Even the indi- vidual feathers, except the remiges and retrices in which the bars run together, conform closely to the pattern shown by many thoroughbred Barred Rocks. The color of the F, black females (Plate XVII, fig. 8) is indistin- guishable from the parental Langshans. Comparatively few members of the F, generation reached maturity. The only points of particular interest are the appearance of very light as well as dark males, of both black and barred males having a few feathers showing the Jungle fowl coloration and of dark-colored females with the barring somewhat blurred. The non-appearance of game (Jungle) colored birds in F, is due pre- sumably to the fact that black is duplex in both Rocks and Langshans, and thus the Jungle fowl color is concealed. There are, however, indi- cations that black may sometimes exist in a simplex condition among Rocks; so that, if suitable matings were made, the Jungle fowl color might appear. Unless the occasional feather showing Jungle fowl color is due to a simplex condition of black, its appearance may mean that hybridization in some way has upset the usual complete dominance of duplex black over the Jungle fowl coloration.* As already stated, Pearl and Surface have published their results in crossing Plymouth Rocks and Cornish Indian Game. Our results en- tirely accord with theirs, as far as inheritance of barring is involved. They classify their birds as barred and non-barred, ignoring intentionally the differences among the non-barred birds. Our results are simpler, in so far as all our non-barred birds are black, but the principle involved is the same in both cases. Pearl and Surface have also made all possible back crosses between the parents and the F, generation. Our results are in entire harmony with theirs, but they have the advantage of a larger number of offspring in their matings. 3 DAVENPORT, 1909, p. 72. MORGAN AND GOODALE, INHERITANCE IN POULTRY 123 Shank Color.—The color of the shanks of all chicks hatched was re- corded, but the color often changes as the birds become older, so these records prove to be of small value. The change in shank color is particularly characteristic for the class called yellowish or flesh-colored black. This class may give rise to all three of the adult shank colors. recognized, black, gray and yellow. Black-shanked chicks seem always to develop into black-shanked adults, and while yellow-shanked chicks probably do not produce black-shanked adults, they may give rise to. either yellow or gray-shanked adults. The infantile shanks among the F, not only show the expected classes, but these classes pass by imperceptible grades into one another. Fre- quently, one part of the shank, particularly the toes, differs from the remainder ; while, in many cases, the distribution of color forms a mottled pattern. The distribution of color upon the toes is likewise extremely variable and often asymmetrical. In almost every case, however, some part of the toes is flesh or yellow. This variation is due, presumably, to some extension or restriction factor. Similar variation in the distribu- tion of color in F, was also recorded. The shank color of the F, adults falls into two classes, black and gray. The term gray is used rather loosely to cover a particular though some- what variable coloration of the shanks. At a distance, the shanks do indeed appear gray just as a Barred Rock appears gray, and just as the “eray” of the Rocks resolves into a pattern on closer inspection, so the gray of the shanks is not a single or uniform color. For convenience of description, we may say that the ground color is steel gray, variously mottled with patches of darker gray or of black. Parts of the shank often have a bluish cast. The posterior side and particularly the bottoms. of the feet are somewhat flesh colored. Mottling does not as a rule occur on the bottom of the feet, so that though the term gray is applied to them in a later paragraph, it is to be understood that they do not have the same appearance on the shanks proper but rather are a grayish flesh, self color. The three classes of black, gray and yellow do not grade into each other. The six F, black females had black shanks. The three males and the four barred females had gray shanks. Apparently, we have here a case. of sex-linked inheritance. This, however, may not be the case but may be due rather to the black spreading over onto the shanks just as it often spreads over onto the comb. - In the barred birds, we may suppose that. the barring factor operates to prevent the spreading of black over the shanks, just as it also produces the characteristic barring of the feathers of the boot. Thus, the colors hypostatic to black are revealed. However- 124 ANNALS NEW YORK ACADEMY OF SCIENCES this may be, in F, the black birds again have black shanks, but the bot- toms of their feet, which are usually incompletely covered by black, are either gray or yellow. The allelomorphs involved, then, are gray versus yellow (or Gray,, Yellow,, X, No gray., Yellow,), the latter being re- cessive to the former. Moreover, among the barred birds, only gray or yellow shanks appear, or in other words, gray-shanked birds always have gray soles, yellow shanked birds yellow soles, but black-shanked birds may have either gray or yellow soles. Since, then, the black-shanked condition is due to an extension of the general black color of the body, we need consider further only the rela- tion of gray to yellow, the determinations being made, of course, only on the bottoms of the feet and when the birds were several months old. In F,, there were only gray or pinkish gray feet, and, therefore, there is no evidence that gray is sex-linked. Moreover, since no other color than yellow appeared in F,, yellow is probably common to both Langshan and Rocks, so that absence of gray in this case means yellow. In F,, not all the adults were available for study, as the importance of foot color was not realized until after many of the birds had been disposed of, but in 17 cases, 13 were gray and 4 yellow. The back mating of F, gray male to P, gray (Langshan) female gave 6 gray. The back mating of P, yellow (Rock) male to F, gray (black plumage) female gave 6 gray to 2 yellow. ‘These results indicate, then, that gray and yellow feet (or shanks, leaving out of consideration the supermelanic coat) behave in ~ simple Mendelian fashion. We have suggested that black individuals have black shanks, because a restriction factor is absent from these birds, so that the body color spreads out as a self color over the shanks. Such a “restriction” factor would be sex-linked. Is it, then, the same as the barring factor? If it were a separate factor, we should expect that, in F., a certain amount of segregation would take place. This has not been observed, so that it seems probable that the black shanks of the black birds are due to the absence of the barring factor and the mottled shanks to its presence, unless some “association” exists. Thus, the presence of the barring fae- tor results in two (perhaps three) distinct somatic conditions, viz.: barred feathers and mottled shanks, and, as a possible third, the gray occipital spot of young chicks. In other words, we have two or more unit characters resulting from the operation of a single factor. There are some considerations of a practical nature resulting from the relations between shank color and sole color which should be mentioned. If the black color covered the entire foot, we should be unable to deter- mine what color underlay the black, except perhaps by long-continued MORGAN AND GOODALE, INHERITANCE IN POULTRY 125 breeding tests. Gray would, therefore, appear to be a sex-linked charac- ter. In F,, however, the results would appear peculiar, for while we should have the three classes of black, gray and yellow shanks, the black shanks would always appear associated with black birds, while gray and yellow shanks would go with barred birds. This conclusion does not agree with the results expected when two independent sex-linked char- acters are involved. In F,, the observed results would be very compli- cated. A discussion of the various possible explanations which might be devised to meet the situation would hardly be profitable here, but a com- parison of the results expected when the color of the soles of the feet is taken into account with those when they are omitted may furnish the _key to similar cases. Booting—The Barred Rocks are typically clean shanked, but occa- sionally a bird is found with a few “stubs.” The boot of the Langshan corresponds approximately to that shown in many of the older pictures of Cochins and Brahmas and may perhaps be regarded as the primitive type from which the modern highly developed boot of Cochins has been developed. For the F, generation, booting was recorded on the chicks as “present” im all cases but two. These two occur among the first four recorded, so that it is possible that, if only a few stubs were present, they may have been regarded as slightly atypical clean shanks. In one other case, boot- ing was nearly absent. Of the 13 adults, the three males and four barred females were alike in that the amount of booting was decidedly scanty, being reduced to about two or three imperfect rows of rather short feathers. The six black females were more variable, due apparently to greater variation in length of feather rather than to variation in the number of rows, the result being a greater variation in amount of boot. A much larger range in the amount of booting appeared in the next generation. The following relative grades of boot in the chicks were - recognized: A, B, C, D, E and absent. No emphasis is to be laid on these degrees, except in so far as they show the general distribution of boot. A and B correspond approximately to that of the parental Lang- shan, and C and D to that of the F, hybrids. Among the adults, not only were there some birds heavily booted like the Langshans, some like the hybrid and others clean-shanked like the Rocks, but one bird had two rows of rather long feathers and one bird four rows of short feathers, indicating that there is more than one component to boot. 126 ANNALS NEW YORK ACADEMY OF SCIENCES TABLE I Distribution of Booting in F, and F,.; MUN aka B c D E | Absent} Total Remarks 3 1 i te elope ome 0 6 | 29 | F, females from 1 x F, male from 2. da 0 10 2 8 3 4 27 | F, females from 2 xX litter brother. 4b 0 6 3 4 5 3 21 | F, females from 2 X re- | ciprocal litter brother. 0 0 0 0 2 6 8 | F, females from 2 x Rock male. 6 0 a 10 il Z) 0 30 | F, females from 1 x Lang- shan male. 7 0 0 0 3 4 3 10 | Female Rocks X male | from 1. 8 3 13 4 10 0 20) 30 | Female Langshans X male from 1. TaBLeE II Expectation ; Observed Ce rg rege oa Total Remarks clean eo Sy Clean | Booted| Clean | Booted i F, 0 0 all 2 24 26 | Langshan females x Rock male. 2 4 0 0 all 0 32 32 | Rock femalesx Langshan male. 3 >» | 18.75] 5.4 | 23.6 6 23 29 | Females from 1 X male from 2. 4a & b) F, 125 6. 42. a 41 48 | Females from 2 Xx males from both 1 and 2. 5 F,., | 00 4, 4 6 2 8 | Females from 2 X Rock male. 6 ee Pe O 0 all 0 30 30 | Females from 1 X Lang- shan male. 7 ies Olea onto amoncoll lic i 10 | Female Rocks, male from I. Suni wenn | CO 0 all 0 30 30 | Female Langshans, male from 1. The results are in entire agreement with Davenport’s and confirm his theory of an inhibitor. The back matings suggest that the amount of boot varies with the increase or decrease in the amount of booted “blood”. There are, however, one or two other theoretical ways of accounting for the observed facts. If we assume that booting is common to both Lang- shans and Rocks and is recessive to a pair of complementary factors, both 4See above in text. MORGAN AND GOODALE, INHERITANCE IN POULTRY 127 of which must be present and one of which must be duplex in order to bring about a complete suppression of the booting, the outcome approxi- mates the observed ratios of booted to non-booted. By assuming that the factor which exerts its effect in either the duplex or simplex condition is sex-linked, the results shown in Table II are obtained. The distribution of the sexes is not given, because the numbers available are inadequate for the solution of a problem as complex as the present one. While the correspondence between theory and observation in this case is close, an attempt to apply it to Davenport’s data resulted in only partial success. This may mean only that more or different fac- tors are involved in the production of boot in Brahmas, Cochins and Silkies than in Langshans, or that the factors causing the inhibition of boot development in Plymouth Rocks are different from those of Tosa, Minorca and other smooth-shanked birds used by Davenport. Among possible factors concerned in boot production should be included those general factors which affect feather growth, in the same way as barring or other color factors control the color of the feathers of the boot as well as those of the body. Down Colors.—The Langshan chick is black dorsally but yellowish white beneath and has white wing tips. The white ventral area often extends upwards, particularly on the head, so that in some cases in this Tegion only the crown and nape remain black. The white area of the wing tips at the same time increases in size, so that the black dorsal surface becomes reduced in amount. The Rock chick, however, though black dorsally except for the gray occipital spot, is usually dark gray beneath, but very often there are several light gray or white areas, which occasionally become more or less confluent, and in extreme cases most of the ventral surface is white and to a limited degree overlaps the Langshan type. In classifying the chicks, all were called “black,” 7. e., of Langshan type, in which at most the breast region was partly pigmented. This region in the Barred Rock chick is the last to lose pigment. All others were classified as “barred”. While this mode of treatment proved to be inadequate for the entire solution of the inter-relations of these charac- ters, 1t was found, first, that both types appear in F,, but that the “blacks” are far more numerous than the “barreds” ; second, that “blacks” | F, interbred or backmated throw some “barreds”, but not in simple Men- delian proportions. 128 ANNALS NEW YORK ACADEMY OF SCIENCES CROSSES BETWEEN AMERICAN DOMINIQUE FEMALES AND LANGSHAN MALES Parent Generation.—Both the hens and the cock were purchased from breeders of these strains.> The one peculiarity calling for notice is the occasional occurrence in the Dominique hens of black or partly black feathers (Plate XIX, figs. 6, d, e). One of the four hens used had sev- eral such feathers. The other hens were free from them. The American Dominiques have barred feathers (Plate XIX, figs. a, c), essentially like those of Plymouth Rocks. F, Generation from Langshan 6 by Dominique 2 —About 15 offspring were reared; the hens were black and the cockerels barred. Of these, five hens and two cocks were bred from. The black hens were like the father as to color; the males were barred like the mother, except that a large number of black feathers were present—some feathers entirely black (Plate XIX, figs. r and ¢) and others barred and black (Plate XIX, figs. p, q, S, WU). F', Generation.—In the second generation, there were recorded 15 blacks and 14 barred birds. Three of the latter died or were killed by animals. Of the remaining, there were 11 male and 15 females tabulated as to color as follows: | 2 } Barred eo e250 Baa eee ea eee 8 4 Blackie ko ee eet eae 7 7 The barred birds were fairly uniform. They were kept for about two months, when their feathers were well developed. A few birds were dis- tinctly darker than the rest, and one bird was much lighter. Certain details regarding white feathers in the wings will be spoken of later. Back Cross of F, & (Barred) to Dominique 2 .—One of the sons was crossed to the four hens that had produced his generation. A first census of the offspring, when the birds were small, gave 19 barred and 4 black — birds. A later count when the birds were older gave 14 barred and 4 black. Five barred birds had disappeared. The distribution of color and sex of 16 of these birds was as follows: 3} IBETVed ss eas ohne ee ee eee 7 5 BLA GI eit e eo creiaed oon ee ee nee 4 0) 5 The Dominiques came from W. H. Davenport, Colrain, Mass. ‘The source of the Langshans is given on page 115. MORGAN AND GOODALE, INHERITANCE IN POULTRY 129 There were also two barred birds whose sex was omitted by mistake in the records. The expectation is three barred to one black, which is closely realized. Back Cross of F, Black @ to Langshan 6 .—Five black hens were bred to a Langshan cock of the same strain but not the actual father of these hens. Another black hen that came from a similar cross with a Plymouth Rock was also present in the same pen, so that some of the offspring may have come from this hen also. There were 18 black young, of which 11 were males and 7 females. In addition, however, there was one barred chick. Now, the black hens had been with a barred cock to give the F, generation. They had been for three weeks with the Lang- shan male before the eggs fertilized by the black cock were kept for incubation. There can be little doubt that one of the spermatozoa of the original male had carried over and produced this bird. If this case is thrown out, the results are consistent. Other Features of the Crosses ——The Langshans have feathered tarsus (booted) ; the Dominiques have clean shanks. All of the F,s recorded were booted, though not strongly. In the F, generation, there were 14 booted and 11 clean shank, distributed as follows: Booted Clean 2 } & $ 1 SIZ EIN EY6 le geeeelene eee arora 7 , 1 2 BS EC Kaa eer euateaeroctonee 1 4 5 3 It is clear that booted shanks dominate® but imperfectly in this cross, as in other crosses of poultry. Some of the F, offspring had heavily booted legs; others were like the F, generation. No sharp line between the classes in the F, generation could be drawn. There is no evidence of any association here between black and booted (the paternal combination) and barred and clean-shanked (the maternal combination). When the barred and booted F, male was bred to four Dominique hens, the results are shown in the next table: Booted Clean Q 3 Q i) 1 BEWTEIENG aS otaeptovotors Bree meieneLS 6 1 4 TRNSVSE callietees ane il 0 3 0 ®TIn the sense that an inhibitor is present in clean shanks. 130 ANNALS NEW YORK ACADEMY OF SCIENCES In this case, the male was heterozygous for condition of tarsus; the hens pure and recessive. The result calls for equal distribution of booted and clean shanks unless “association” occurs. The numbers are too small to have any significance. Even as they stand, however, they have no meaning, if coupling be made responsible for the distribution of the characters. When the Langshan male was crossed to the black hens (both sexes booted, but the hens heterozygous) all of the offspring were booted, which is in accordance with dominance of booted shanks. — White Feathers in Wings.—In the F, young birds, the presence of white and partly white feathers in the wings was noticed (Plate XIX, figs. f-n; 0, v and w). They were most obvious in the black birds, perhaps because of the sharper contrast. These feathers are some of the primaries and a few of the coverts at the base of the primaries. As shown in Plate XIX, figs. f to k, they are rarely pure white, but often mottled or splotched. They were not recorded in the F, birds, and if present they were overlooked. The records of birds without and with these white feathers were as follows: Q }$ Barred: no: whites sie see een 6 4 REA Meminer ek eo Cho orn co Gra'G d-olcu om a OG 2 6 Black, no ‘white:2232 23s. eee eee 6 Rete 4 eee (AU Ua Ter BAA Wabis,Gccud o.0 6800 a Le 3 In all, there were 20 chicks without and 6 with white feathers. This looks like a case of Mendelian inheritance, but it may be purely a coinci- dence.. We do not know how often such feathers occur in chicks of the original breeds, or whether they are only juvenile, or physiological effects of the condition of the bird. Probably they would have disappeared in later molts, had.the chicks been kept longer. When the Langshan cock was bred to the black F, hens, 4 of the chicks had no white and 14 had white in the wings. If the black male is hetero- zygous for this condition, the result is not in accordance w ith the assump- tion that this is a Mendelian recessive character. When the Dominique hens were bred to the F, barred males, hare was no white in the 15 recorded offspring. This result is not in harmony with the same supposition, but the black male used in the last experiment was not the same father as for the barred males of the first cross. The father of the barred male in the first case was a brother of this one. It is still possible, therefore, that one male was homozygous and the other heterozygous for the white-feathered condition. Without, howeyer, fuller information, not much weight can be given to these results. MORGAN AND GOODALE, INHERITANCE IN POULTRY 131 Color of Legs—It has been stated by Bateson and by Pearl that yel- low and black shanks in certain breeds of poultry show “sex-linked” inheritance. This is not apparent in the Langshan-Dominique crosses, ex- cept in so far as black shanks accompany black color of feathers. For ex- ample, in the F, generation, there are recorded 13 black birds with black legs. Of these, 5 were deep yellow on the under side of the feet. In addition, there was one male that had yellow shanks and yellow under the feet. There were recorded 12 barred F, chicks with yellow shanks. Of these 12 birds, 4 are recorded as having very pale yellow or whitish legs. It would appear from this case that black and yellow shanks accompany black and barred plumage, at least as a rule. Tn the back cross of the F, barred male to the parent Dominique hens, in which there were barred males and females and only black females, all 4 of the black birds had black legs, while all 12 barred birds had yellow or pale legs. Among these 12 barred birds, there were 5 with pale legs; in 2 of these and in one yellow, there were spots of black or dark color, at least on the tarsus. These. rather meager figures, as far as they go, show that shank char- acter and color of plumage go together, and that black shanks and yellow shanks are only an accompaniment to sex-linked inheritance of plumage. The data are manifestly few, however, and it may well happen that the two characters may appear disassociated. Color of Bill—The color of the bill seems to run a parallel course. Full records for the back cross given above were kept. Here, 13 barred birds had yellow bills and 5 black hens had black bills, but one of the latter had much yellow on it, and two of the former had black: one was black with yellow tip and the other was yellow and black. There is much variability in the color of the bill, and the above statements are insuffi- cient to warrant any generalizations. THEORETICAL CONSIDERATIONS The current formula for sex inheritance in fowls represents the female as heterozygous for sex, F-O, and the male homozygous, O-O. If F is identified with a special chromosome connected with sex determination, the formula calls for one more chromosome in the female than in the male. At present, evidence on this point is conflicting and insufficient. It is true that Guyer has described two kinds of spermatozoa in the male, one with an X and one without. If this X is the same as in other ani- mals, then the spermatozoa containing it must be female producing, and the female should contain one more chromosome than the male. This means that the male and not the female is heterozygous for sex. The 132 ANNALS NEW YORK ACADEMY OF SCIENCES experimental evidence is flatly opposed to this latter interpretation, and, therefore, until Guyer’s evidence is confirmed or refuted, the case must be left open. On the other hand, if, as the experimental evidence shows, barring is “repulsed” by femaleness and if both of these factors are carried by chromosomes, the formulas are deficient in having no chromosome to carry barring,—a contradiction of terms. It may be, however, that the X-chromosome in fowls has a mate which we may call Y which would carry barring but not femaleness. The formule would then be: NOTA Oxerc tere crsccneheee eae scr ne oer A eee xX —Y JY RET) SRA nS ee A aren hee gar by ee Y— Y XY Female YY Male. On this interpretation, the factor for femaleness would be contained in X but absent from Y, while barring is contained in Y. ‘This scheme is compatible with the experimental evidence and gives consistent results for all combinations. The irregularities that have been observed in the “reduction division” both in birds and in man suggest the possibility that the sex chromo- somes are united to other chromosomes as in some other animals. If the union is variable, as in the nematodes, it may be that the X and the Y (if Y exists) may sometimes pass to the poles of the spindle during reduction in conjunction with other chromosomes and sometimes be free to pass to the poles independently. If further study should establish this view, it will have a very direct bearing on the relations discussed above. If the factor F for femaleness is carried by chromosomes attached to one member of another pair, the mate of this member may be the chromosome that carries the factor for barring. If this were the case, however, interchange between these two members would lead to the barring factor being transferred to the chromosome attached to the sex chromosome. This is in contradiction to the experimental evidence which would lead rather to the conclusion that a Y element lacking the factor for barring is present. The Y may be attached to the mate of the chromosome carrying the sex factor. At present, only a few cases have been discovered in which a sex-linked character is dominant, viz.: in fowls and in one character in Drosophila. The only other cases, besides the one in poultry in which sex-linked in- heritance occurs and sex is heterozygous in the females, is that of Abraxas and that of canaries. In both of the latter, the sex-linked factor is re- MORGAN AND GOODALE, INHERITANCE IN POULTRY 133 cessive. There are no a priori grounds why a character of this sort may not be dominant, if some other Mendelian characters may also be domi- nant. The factor for black, N, is treated in our formule as present in all of the gametes both of the female and of the male. It is not allelomorphic to barring, B, although its presence in the female-producing egg when barring is present in the correlated male-producing egg may appear to bear this interpretation. From the chromosome point of view black may be, so far as we know, in other chromosomes than those carrying barring ; hence its more general distribution. BIBLIOGRAPHY American Poultry Association, The American Standard of Perfection. 1905. Bateson, W.: Facts Limiting the Theory of Heredity. Science, N. S., Vol. XXVI. 1907. : Mendel’s Principles of Heredity. 1909. : The Inheritance of the Peculiar Pigmentation of the Silky Fowl. Jour. Genet., Vol. I. 1911. Bateson, W., and PuNNeETT, R. C.: The Heredity of Sex. Science, N. S., Vol. XXVII. 1908. Brown, H.: Races of Domestic Poultry. London. 1906. DAVENPORT, C. B.: Inheritance in Poultry. Carnegie Pub. No. 52. 1906. : Inheritance of Characteristics in Domestic Fowl. Carnegie Pub. No. 121. 1909. GoopALE, H. D.: Sex and its Relation to the Barring Factor in Poultry. Science, N. S., Vol. XXIX. 1909. : Breeding Hxperiments in Poultry. Proc. Soc. Exp. Biol. and Med. Vol. VII. 1910. Hanviey, P. B.: Sex-limited Inheritance. Science, N. S., Vol. XXXII. 1910. HaGrepoorn, A. L.: Mendelian Inheritance of Sex. Archiv. Ent. Org., Bd. XXVIII. 1909. PEARL, R., and Surrace, F. M.: Studies on Hybrid Poultry. An. Rep. Maine Agric. Exp. Station. 1910. : On the Inheritance of the Barred Color Pattern in Poultry. Archiv. Ent. Org., Bd. XXX. 1910. : Further Data Regarding the Sex-limited Inheritance of the Barred Color Pattern in Poultry. Science, N. S., Vol. XXXII. 1910. SPILLMAN, W. J.: Spurious Allelomorphism; Results of Some Recent Investi- gations. Am. Nat., Vol. XLII. 1908. STuRTEVANT, A. H.: Another Sex-limited Character in Fowls. Science, N. S., Vol. XXXIII. 1911. TEGETMEIER, W. B.: The Poultry Book, etc. London. 1867. WaicutT, L.: The New Book of Poultry. London. 1902. PLATE XVII ED PLYMOUTH ROCKS, LANGSHANS AND THEIR CROSS-BRED OFFSPRIW ji iy iis a Tae NaC ‘1. Barred Plymouth ‘Rock cock. — 2. One of the parental Langshan hens. 8. Langshan cock. ‘Stock of Mr. Ives. 4, One of the parental barred hens. This particular hen is ey White Rock male X Barred Rock female. _ Fes. 5 and 6. The F, from Barred Rock cock by Langshan hen. A Figs. 7 and 8. The F, from Langshan cock by Barred Rock hen. > sedi) ’ = : i) RGN 1O (iie-aaod9 eel oa aig He sod sLblaothed ate Ae joleina? vba halves x ‘hkera ‘ase sit coy al A doo ood ode ago mn al ANNALS N. Y. Acap. Sct. VOLUME XXII, Pharm XVII 4 PLATE XVIII FEATHERS OF LANGSHANS, BARRED PLYMOUTH sacs AND THEIR Se 4 J ‘ " Hackle feathers (except 11 and 14) ‘from the various types. of pirde used i in Best ny _the experiments. Basa and 4. fou a pure bred Barred Rock female. ‘Fie. 5. From a second Barred Rock female. Note the differences in the Bie ness of the barring. Cg 6. From the hen shown in Plate XVII, Fig. 4. i et 7. From F, barred female. — ( Fig. 8. From a pure bred Barred Rock cock. . j 9. From same bird, illustrating a partially black feather occurring in pure bred stock. Fies. 10, 11, 12, 14. From F, bird shown in Plate XVII, Fig. ay, Fig. 10. Hackle feather. Fie. 11. Breast feather. Fig. 12. Shows the Jungle fowl coloration. Fie. 13. From bird shown in Plate XVII, Fig. 7. Wie. 14. Primary, to show reduction in barring. Lig fe Rea ees ® PUIG Ah. GEA Get ra EURUNAME it canta an tii Oem wbittd ths ang) evoliny, off mort CM Dane i sao) minus Orel a Beate Jtaenet Aooh base bene hla th ‘aie ve GAs Te eopusto Dib sii shod aisinel obo05 alae bine 4s ie iL ‘warritioo0 fed deo} oat wise i gute baie Summa s Bai AYA aks 14 ua vie bald 1 port bs ’ bis rain : siniteiotoy | ‘wor yen gilt Pie ae ohn ths Ge ies gah bata ” +4 PLATE XIX FEATHERS OF AMERICAN DOMINIQUE FOWL OLB , a. Dominique hen. Barred feather. b. Dominique hen. Black feather. / . €. Dominique hen. Barred feather. ee Domini me hen. Black feather. Black feather. ae " feathers. Figs. in, Three white covert feathers from mines me ‘ oped feathers at their base. Fic. w. ee of barred F,, showing a nearly barring: on some other feathers. Zeb fe VLE OFT VLPeoey, OR: AR OUS >» oh Sat peng a oii ior BOe: ithe ods cat ballet ‘he aarunlt sq ae gatwods etiur parent ie ass arioge Neehteb ae’ uel Bara 29 mea ‘oy sie aed Bore Sppligead il Pinter HGH CE BPI iE hinon o VOLUME AATI, PLATH AIX are iGaaiuted | in bundles on an average of three per year. The price of pete articles aes upon their length and the number of illus- ae intervals. It is ed that each volume shall be devoted to monographs relating to some particular department of Science. Volume > li is devoted to Astronomical Memoirs, Volume II to Zodlogical Memoirs, ete. ‘The price is one dollar per part as issued. Al publications are sent free to Fellows and Active Members. The Annals are sent to Honorary and Corresponding Members desiring them. ‘Subscriptions and inquiries concerning current and back numbers of oe of the publications of the Academy should be addressed to THE LIBRARIAN, New York Academy of Sciences, _ care of EN : are American Museum of Natural History, . ~ . New York, N. Y. EMA- FAUNA OF NAVY THE NEW YORK ACADEMY OF SCIENCES (Lyceum or Naruran History, 1817-1876) OFFIcERS, 1912 President—EMERSON McMixuin, 40 Wall Street Vice-Presidents—J. E>MUND WoopMAN, Freperic A. Lucas, \ CHARLES LANE Poor, R. S. WoopwortH : Corresponding Secretary—HeEnNnry EH. Crampton, American Museum Recording Secretary—EpMuND Otis Hovey, American Museum Treasurer—HeEnRY L. DoHERTY, 60 Wall Street Librarian—RaLPH W. ToweEr, American Museum Editor—Evmunp Otrs Hovey, American Museum SECTION OF GEOLOGY AND MINERALOGY Chairman—J. E. Woopman, N. Y. University Secretary—CHARLES P. Berkey Columbia University SECTION OF RIOLOGY Chairman—FREpDERIC A. Lucas, American Museum Secretary—Wittiam K. Grecory, American Museum SHCTION OF ASTRONOMY, PHYSICS AND CHEMISTRY Chairman—CuHARLES LANE Poor, Columbia University Secretary—F. M. PEDERSEN, College of the City of New York SECTION OF ANTHROPOLOGY AND PSYCHOLOGY Chairman—R. 8. WoopwortH, Columbia University Secretary—F repertc Lyman WeLLs, Columbia University The sessions of the Academy are held on Monday evenings at 8:15 o’clock from October to May, inclusive, at the American Museum of Natural History, 77th Street and Central Park, West ape [ANNALS N. Y. Acap. Sct., Vol. XXII, pp. 155-160, Pll. XX-XXII, 25 July, 1912] ON THE DICTYONEMA-FAUNA OF NAVY ISLAND, NEW BRUNSWICK By F. F. HAHN - (Presented by title before the Academy, 13 May, 1912) CONTENTS Page IGDEFO CLC THIGIN ey nee ered alten Gididle a-cra Oc. & choiena: cor ORI crete RCC a neers maces eer uere 135 Gury. WBA UNO YS BI PB igs 6 Sinton pis a oo charg oo 6 Heels Se eC Ick hs cat rcs 136 Astogeny of Dictyonema flabelliforme in comparison with that of Stawro- GirUPUUSs. 05.6504 60 6000 OOOO UID b O00.0'0 0.0 00-00 0100 6 CO CICIEIG DOING Olte aiCeutClc aD Ero 144 On the structure of Dictyonema flabelliforMme... 2.1.2 cece cece cee eee cee 146 Mode of life of Dictyonema flabellifOrMme. ... 0... cece cece eee ee eee eens 148 On the occurrence of stratigraphical range of the varieties of Dictyonema MOGI ORTIGsias ce Sooceidoatadcuesucdcobe 0 oOo COC Ud uc oUOO oO coo COCO CUO 151 Significance of the varietal range and its phylogenetic value............. 153. INTRODUCTION One of the older modes of paleontological investigation, the study of the range and significance of variability, seems now to be somewhat overtaken by and neglected on account of the excellent results obtained in the study of the stages of earliest youth, growth and senescence. My own studies on Jurassic invertebrates, which made me familiar with a vast number of individuals of many species, brought me to the convic- tion that a combination of these two modes of treatment would, in every case where the given material is sufficient, lead to an advance greater than would be possible by either method alone. Thus, when I worked over in the Museum of Columbia University the wonderful collections from the Dictyonema-shales made by G. van Ingen and W. D. Matthew at Navy Island, N. B., from Division III, b, ec, commonly regarded as Upper Cambrian,* of course my first thought was to see whether I could find any relationship between the range of variability and the history of the races among the Dictyonemas. And this animal association seemed well adapted to this kind of study, for though the Dictyonema-shales contain a vast number of individuals, there are only four species com- mon, two of which are graptolites. 1 According to the trilobites found in Sweden and BHngland, it must apparently be placed at the base of Ordovician. (135) 138 ANNALS NEW YORK ACADEMY OF SCIENCES Description: Generally infundibuliform, branches more or less parallel, fre- quently dividing, width of tissue 9-10 mm., cross-threads somewhat irregular, of medium number; open rectangular meshes; thecze 15 to 17, projecting, of medium size. Notes: Slight changes of the characters are fully shown on the diagram, Plate XX. Greatest length observed about 200 mm., width 190 mm.; approxi- mately 2x 150 branches. The variation of the number of thece is marked, 16 being present on the American examples, but only 10-15 in 10 mm. on the European ex- amples. This fails to be of general value, as G. F. Matthew has pointed out, since the figures of Brégger (5) and Tullberg (8) also exhibited 16 thece and Westergard states recently a usual number of 15-16 thecz on the Swedish examples ((4), p. 58). When comparing the form, col- lected by Prof. A. W. Grabau from Skane (region of Fogelsang), with the American specimens, a more delicate character of the whole tissue of the European type is apparently the one observable difference on which a separation of var. acadica Matth. and “forma typica’ of the Swedish authors can be based. Var. conferta Linrs (ms.)™ Apotypes : (6) 1860. GOpprrRT, pl. 36, fig. 11, 4. (9) 1861. Satrer, Geol. Survey Gr. Brit., Mem. 3, pl. 4, fig. 1. 1881. SALTER, Geol. Survey Gr. Brit., sec. ed., p. 535, pl. 41, figs. 1, la, 1b. ($) 1882. TuLteerre, p. 20, pl. 3, fig. 3 (1, 22). (4) 1909. WESTERGARD, pl. 3, fig. 7. Description: Characterized by Brégger ( (5), p.36) as having very fine and close network and cross-threads of the same kind. Matthew” added “commonly vasiform, cross-threads more frequent than in acadica (5-7). 15-17 thece are usually met with in 10 mm.” I have observed even 18 on a typical specimen ; young specimens often show only 14-15 thece. The rigid aspect of the branches and their regular branching, which commonly takes place in the same level, seem rather important additional features. Here, too, a little more delicate structure of the whole tissue distinguishes the Swedish specimens, so that Westergard found 13 branches in 10 mm. of width, while 8-10 branches may be usually numbered on the American variety. The largest ob- served colony measures 22 mm. in width, occupied by 19 branches, and 82 mm. in length. ei Broégger (5) furnished as long ago as 1882 an undoubted description of this form, while Pocta created his species “conferta’ not earlier than 1894 (Systéme silurien, Bryozoaires, ete., Vol. WIII, t. ler, p. 194). The last ought to be renamed, if this is necessary at all. 2 Loe. cit. HAHN, DICTYONEMA-FAUNA OF NAVY ISLAND, VN. B. 139, Var. norwegica Kjerullf. Apotypes: (7) 1865. Ksrrur, p. 2, fig. 4 (1-3). (2) 1906. Mosere, pl. 1, fig. 7. (4) 1909. WesrercArp, pl. 30, figs. 8, 9. Description: After Broégger ( (5), p.36), thick cross-threads, close network ; small, short, rounded meshes. I observed only 15 thece. Young stages like var. acadica. Crossings of all kinds into true acadica occur. A shght distinction between my specimens and the Swedish ones is in so far recognizable as the first are furnished with a little more elongated meshes, but they can be precisely compared with Westergird’s figures. My largest colony is 40 by 40 mm., while Westergard cited a specimen 115 by 30 mm. in size. Var. ruedemanni nom. nov. Typical figures: (1) 1904. R. RuzpEMANN, pl. 1, figs. 16-20, 22 (12-15, non fig. 21). Apotypes: (10) 1895. G. F. MatrHew, Trans. N. Y. Acad. Sci., Vol. 14, pl. 49, figs. 1, 2. (The figures are not quite correct, showing a too open tissue. ) Compare (5) 1882. BroeeEr, pl. 12, fig. 19. Description: In nepiastiec stage strongly branching at an almost acute angle. Full-grown specimens with crowded, more or less parallel branehes up to 14 in 10 mm. width. Dissepiments appearing early, but afterwards irregularly scat- tered. Thece small, not much projecting. Adhesion organs relatively common. Notes: Generally funnelform, sometimes of intermediate shape (pyriform) towards the vasiform variety “conferta.” Measurement of a typical pyriform specimen : Length Width in mm. 10 10 20 25 25 35 48 65 The specimens obtained from Navy Island (N. B.) fully agree with Ruedemann’s figures, but several specimens considerably surpass the given characters in the same direction. In Europe it is rare, but abun- dant in America, at Navy Island as well as at Schaghticoke. Largest measured colony 165 mm. in length and 140 mm. in width. Var. desmograptoidea var. nov. Description: Infundibuliform. Open meshwork of irregular, mostly sub- oval fenestrules. Branches undulating and sometimes coalescent. Dissepi- 140 ANNALS NEW YORK ACADEMY OF SCIENCES ments quite irregular, partly thickened. Thece 15 in 10 mm., not much pro- jecting, rarely visible. ; Note: Sometimes an old stage of var. acadica has these characters. J. Hall’s!® Dictyonema irregularis resembles most this variety, but has a closer arrangement of the branches (25-28 in 1 inch) and occurs in higher Ordovician of Canada and Great Britain. Dictyonema hom- frayi, described by T. Hopkinson and Ch. Lapworth,* seems to be a closely related species. Desmograptus intricatus, as shown in fig. 30, p. 609, by Ruedemann (1), exhibits exactly the same type in early stages, though belonging to the Chazy and having appressed thece. No well- preserved, large specimens are known, but several pieces are found indi- cating a width of the colony of more than 60 mm. Staurograptus dichotomus Emmons var. apertus Rued. (1) 1904. R. RUEDEMANN, pp. 612-614, pl. 2. (11) 1891. G. F. Marruew, “Bryograptus kjerulfi,’ Trans. Roy. Soe. Canada, Sect. IV, No. VI, p. 35. 1892. G. F. Matruew, “Bryograptus patens Matth.,”’ No. VII, p. 95, figs. la, 1d (excl. lower figure of 16). (10) 1895. G. F. MatrHew, id., p. 268, pl. 48, fig. 4 (?). 1895. G. F. MatrrHew, “Bryograptus lentus Matth.,” id., p. 270, pl. 48, fig. 2. 1895. G. EF. Matruew, “Clonograptus proximatus Matth.,” id., p. 265, pl. 48, fig. 1. In 1895, G. F. Matthew tried to keep separate five species of Clono- graptus and Bryograptus, occurring in Division 3, band ¢c on Navy Island, N. B. In 1903, Ruedemann’ followed him, believing that four of Matthew’s species were again recognizable, but a year later, in his Mono- graph on the Graptolites, he came to the conclusion that practically all these species and even genera present only definite stages of growth and preservation of Staurograptus dichotomus, a species established by Em- mons in 1855, but rather dubious up to 1904. Ruedemann, indeed, furnished such a description and figures of this species, especially of the variety “aperta,’ that complete identity with the species of Matthew enumerated above becomes apparent, if we look over the types of Matthew, which fortunately were in part accessible to me. 18 Canadian Organic Remains. Geol. Sury. Can., Dec. 2, p. 136, pl. 20, figs. 1, 2. 1868. 144 Graptolites of the Arenig and Llandeilo rocks of St. David's. Quart. Journ. Geol. Soc. London, Vol. 31, p. 668, pl. 36, fig. 13. 1875. “The Cambric Dictyonema fauna in the Slate Belt of Eastern New York.” N. Y. St. Pal. Ann. Rep., Bull. 69, p. 938. HAHN, DICTYONEMA-FAUNA OF NAVY ISLAND, N. B. 141 The type specimen of Clonograptus proaimatus (1895, pl. 48, Figs. la, 1d) proves to be almost identical with Ruedemann’s figure 23 on plate 2, besides a slightly greater elongation of the thece. Of these, I have found distinctly 8 in 8 mm. of length of the branches, the latter not surpassing 25 in number. The fork-lke mode of branching seems highly characteristic. On the same slab, the type of Bryograptus lentus Matth. (1895, pl. 48, fig. 2) is found, appearing at first glance very distinct from that of “Clonograptus proximatus ;” but here, too, 9 to 10 thece of a breadth up to .75 mm. are seen, showing the same aspect as Ruedemann’s figs. 19 and 21. Moreover, this slab exhibits another crowd of branches which in part are completely identical with “Clonograptus proximatus,” while a few branches turned to side view show the real aspect of “Bryograptus lentus.” Here again, 9 thece are recognizable in 10 mm. of length of a branch which is 1.6 mm. thick and of the same shape as Ruedemann’s fig. 15. This group of Staurograptus has still greater importance, since three individuals of medium sizé have their sicule preserved and present the characteristic side view of Ruedemann’s fig. 6. As a great many other specimens of the collection which formerly were determined as “Bryo- ' graptus patens,’? and indeed readily conform to Matthew’s fig. 4, on plate 48, really belong to nepiastic stages of Dictyonema,'® a similar reference of the type of Matthew seems quite possible, but I have not seen the type specimen, and thus this explanation may still be ques- tionable. A few additional notes can be added to the description by Matthew and Ruedemann. The preservation showing the thece in full size and the sicula is rare on examples surpassing a diameter of 5 mm. The thece, scarcely more or less than 9 in 10 mm. length of the branch, are sometimes irregularly exposed, 6 smaller ones being visible in 4 mm. of the distal end and 4 a little coarser ones in 4 mm. of the proximal part of the same branch. Young stages, especially when crowded together, are almost inseparable from those of Dictyonema flabelliforme var. ruedemanni, while the bryo- graptoid neanastic stages of D. flabelliforme var. acadica and conferta are easily mistaken for the “Bryograptus lentus” aspect of laterally com- pressed Staurograptus. The great thinness of the periderm of Stawro- graptus, which never shows such a tubulose structure or wrinkling as in the case of Dictyonema (compare the above), is a very remarkable fact and helpful in distinguishing these two genera. In the same respect, the 16 Compare Westergard (4), pl. 3, figs. 5 and 6. 142 ANNALS NEW YORK ACADEMY OF SCIENCES more rigid characters of the mucronate apertural margin of the thece in Staurograptus may be noticeable. Unfortunately, I was not successful in finding any remains which can be compared with “Bryograptus spinosus Matth.” ((10), p. 269, pl. 48, fig. 3) and “Clonograptus spinosus Matth.” ((11), p. 9%, pl. 7, fig. 2), which is said to present only 8 thece in 10 mm. of length of the branch and a distinct “axis or virgula.” Nevertheless, here, too, the identity with Stauwrograptus dichotomus seems to me not quite impossible. Nor -could I observe any specimens belonging to “Bryograptus ? retroflexus” (Matthew (10), p. 271) or to Callograptus (1. ¢c., pl. 48, fig. 5), which therefore must be at least extremely rare. On the contrary, with an abundance of individuals there is found Monobolina refulgens Matthew (11) 1891. G. F. MatrHew, p. 44, pl. 12, fig. 6. (12) 19038. G. F. MattrHew, Report on the Cambrian Rocks of Cape Breton. Geol. Surv. of Canada, Ottawa, p. 210, pl. 11, fig. 4, pl. 16, fig. 2. Since generally separated valves and even broken pieces occur, nothing can be added to the careful description of the author, besides the obser- vation that this species is seen on the same surface associated with all varieties of Dictyonema flabelliforme and with Stawrograptus dichoto- MUS. Lingulella nicholsoni (?) Almost the same is true of a linguloid shell, cited by Matthew in 1895 (p. 273) as Lingulella nicholsoni (?) and in 1903 ((12), p. 204) doubtfully called Lingulella “lepis Salter.” The specimens of the Co- lumbia collection usually have a length of 4 to 5.5 mm., a width of 3 to 4.8 mm. and exhibit a distinct ridge on the interior of the pedicle valve. They differ from the English type as described by Salter ((9), second ed., p. 538, fig. 11) in a relatively greater feebleness and scarcity of the lines of growth. It may be of interest to note that a very similar “Lingulella” of Scandinavia is referred by Brogger ((5), p. 44, pl. 10, fig. 5, from the Tremadoc shales) and by Moberg ((2), from the Bryo- graptus zone) to L. lepis Salter. In addition to this detailed discussion of the Dictyonema fauna, a callograptoid graptolite of the upper Beekmantownian may be described, because of its interesting phyletic relationships. Callograptus grabaui sp. nov. Flabelliform or shrub-like, not more than 15 mm. in length and 10 mm. in width of the dendromes observed. Short, non-celluliferous, basal stem (1 mm. HAHN, DICTYONEMA-FAUNA OF NAVY ISLAND, N. B. 143 long, .5 mm. wide) with terminal expansion to an adhesive bulb (of .8 mm. in width). Branches 10 to 13 in 10 mm. width, .3-.4 mm. wide, closely arranged, sub-parallel, sometimes slightly flexuose. Sicula of one specimen 7 mm. long, determinable as being a very thin tube about 1 mm. in length, with a minute basal disk .5 mm. in width. Within the first 5mm. of the length of the dendrome, a very frequent branching, sometimes of monopodial aspect, with a common angle of 50-60° takes place. Branches of the third and fourth order are usual, those of the sixth still observable. The interspaces be- tween the branches vary from .83mm. to 1 mm. in width. Most of the specimens are dorsally de- BiG, i—@S. Ge = Ma Za ce nee oD a Fig. 3.—Range of variability of Dictyonema fiabelliforme Finally, any separation by time is significant only between the yar. con- ferta and most of the other varieties, which all have apparently grown together. For the separation of types, I regard the foregoing as of varietal, but not of specific or mutational value. For this last, we may regard either the saltative character or the interval of time as the essen- tial point. As to the cause of those changes, we may find it by starting from a mechanical point of view. How do sessile organisms react to a more or less continuous pull? If the pull is small, there results elongation of 27 The constantly increasing number of Dendroidea, found in the Ordovician of North ‘America, aS compared with their sporadic appearance in Huropean localities, makes it probable that the continuous evolution was more likely confined to the American seas. : Ii y j gy MMI Characters HAHN, DICTYONEMA-FAUNA OF NAVY ISLAND, N. B. 155 the organs combined with elasticity and flexibility; if the pull is too great, only concentrated forms may exist. Near the boundary line of the trees of Alpine regions, we find species with slender, movable branches (Betula, Larix), while in places relatively protected from storms, we find dwarf trees and bushes, low and closely clinging (Pinus, Picea) or extremely thickened stems on the slopes of hills exposed to windy squalls. Water-plants lengthen their stalks or build a concentrated mass accord- ing to their standing places. And the same reaction must be true in benthonic animal life. The whole dendrome of Dictyonema was sub- jected to the pull in water, 7. e., the mechanical stress exerted against the fixed body by the action of currents, waves and breakers and thus the animal mass proves to be influenced. As a result of this we have: I. Var. acadica. This middle type of all the varieties shows very great progress over the small bryograptoid ancestors of Dictyonema, as is proved by the neanastic and nepiastic stages, for the elastic cross- threads enable single branches to grow to such an extent that, for ex- ample, a colony 160 mm. long could have 20,000 theca and yet have sufficient firmness for existence. The varieties given here show the complete possibility of improve- ment, but that acadica has not reached the highest place is proved by the fact that of all perfectly preserved examples of later ephebastic stages, acadica has less than 10 per cent, the pyriform paratype up to 50 per cent, ruedemanni up to 50 per cent, conferta 70-80 per cent, norwegica less than 10 per cent, desmograptoidea less than 10 per cent. Forms like acadica take the middle line of all true Dictyonemas through the Silurian, Devonian into the Carboniferous (compare reti- forme, scalariforme, stenactinotum, spenceri, leroyense, brairt). II. Var. norwegica endeavors to progress by a heavy thickening of the rhabdome net, but it causes great weighting down of the body and much profusion of organic matter, so that this mode of building up will be possible only when food is abundant. Compare D. quadrangulare, murrayi, crassum, arayi. Result: A few sporadic species up to Devonian time; no generic evolu- tion; one-half of one per cent of our forms, more or less broken in pieces.8 IIL. Var. desmograptoidea has closer and even elastic branches join- ing with fewer cross-threads. 28 Also Westergard mentions the scarcity of unbroken specimens of this variety (4), p. 60. 156 ANNALS NEW YORK ACADEMY OF SOIENCES Result: Through forms like irregulare, homphrayi, generic evolution into Desmograptus up to Devonian; one-half per cent of our forms, because of somewhat later starting of evolution. Good stolon in later species, sessile; finally, thinning of rhabdomes (tenuiramosus, a sessile pseudo-Dendrograptus). IV. Var. rwedemanni. Heavy branching and a successive diminution of the elastic cross-threads produce a thick brush and dense crowding of nourishing thecz, while still allowing movement of the rhabdomes. Result: 35 per cent of whole fauna; 50 per cent perfectly preserved. Half of the observed attaching organs belong to this variety. Hvolu- tion: (a) persistence of type in a few Ordovician and Silurian species; stolons and roots common. Compare delicatulum, pereaile, rectilimea- tum, subretiforme, tenellum, filiramum. (6) Further reduction of cross-threads, enforced reduction in size. Intermediate forms: Callo- graptus, Rhizograptus, Odontocaulis, Callyptograptus; sessile, in early stages vagrant; then Dendrograptus; stolons and roots. Further evolu- tion of Dendrograptus leads to thin flexuose types (1.- e., under slight pull) or to succulent types (7. e., under heavy pull) ; latter forms some- times with central axis. End of this evolution: Ptilograptus, Acantho- graptus, ete. V. Var. conferta, cylindrical growth and close network, giving me- chanically the best type. Result: On one side 6 per cent of our forms and 70-80 per cent per- fectly preserved; on the other no specific evolution at all, probably on account of very unfavorable conditions of food supply. VI. Last type of change by thickening and dividing of the cross- threads (groups of peltatum, cervicorne, cavernosum, tuberosum) not represented in our fauna. Appears first in American Ordovician. Sto- lons and roots well developed. Considering this mechanical starting point, it becomes of interest that the Gorgonias of equivalent habitat (disregarding the funnel shape of the Dictyonema colony) present the same lines of changes as those sketched in the foregoing for Dictyonema flabelliforme. Thus we find among the Leptogorgias, L. eximia and media with a network like that of var. acadica, Leptogorgia agassizi with close fine meshes comparable to those of var. conferta, while Leptogorgia rigida looks like a callograptoid type and Hugorgia multifida like a Dendrograptus with beginning central axis. The latter, still further developed, gives Pterogorgia acerosa, while Gorgonia flabellum has a typical desmograptoid appearance and Gorgonia quercifolia resembles closely the norwegica-murrayi line. I cannot help thinking such a conformity indicative of parallelism in HAHN, DICTYONEMA-FAUNA OF NAVY ISLAND, VN. B. 157 evolution and that it represents one of the usual lines of development of such benthonic forms under a given environment. Now, with all the foregoing statements, we may touch upon the final question of the phylogeny of the dendroids as indicated by the range of variation in the earliest Dictyonema. A certain branch of the planctonic ancestors common to both Graptolites and Dendroidea was evidently pushed forward in the lines of directed evolution by the formation of dissepiments as a supporting mechanism of the elongating rhabdomes. This represents the Dictyonema stage, beyond which a group of retarded species and genera (e. g., Desmograptus) never passed to any consider- able extent. While in the Dictyonemas of the early days, fixation was realized only by means of a thin, fragile nema, the adhesive organs were now brought into vital and ever increasing significance. For gradually thickened stems with basal expansions, with stolonial ramifications, with ability of independent budding of thece and of colonies, were built, while the original planctonic period of life becomes shortened to its final disappearance (Dendrograptus-stage). From this point on, an extreme widening of the main stipe, on one side, gave rise to the Galeograptus, Discograptus, Cyclograptus, Rodanograptus-group, while a thickening of the central axis led to forms lke Inocaulis, Acanthograptus, Cacto- graptus, Paledictyota, to which even Chaunograptus, Corynoides and Thamnograptus may be related, as held by Ruedemann. Finally, in types lke Mastigograptus, a striking approach to the present hydroids has been revealed. When considering the various races of Dendroidea on such a broad basis, the various genera do not of course mean anything else than stages in development; and every line of separation seems an arbitrary one, aS 1s shown in a comparison of the species thus far assigned to “one genus” by the different authors (e. g., the Callograpti and Dendro- graptv). Nor is the difference between the forms in the early Ordovician with free sicule and those with unknown sicule a reliable one upon which alone to base the natural classification. Within the last five years, two papers of such importance regarding the differentiation and evolution of the Dictyonemas have been published that they must be considered with great care. In 1907, W. S. Fearnside (8) made some striking suggestions. He believes that “in the earliest Dictyonema, the cells are very indistinct and rarely project more than about a quarter of the diameter of the common canal; cross-threads thin and numerous; stipes close together, parallel, branching at all levels; elongated rectangles of meshes. This supposed Dictyonema diverges in two distinct families, one approaching the true graptolites (Dictyograptus), the other seems more nearly related to the Dendroids (Dictyonema, sensu stricto). 158 ANNALS NEW YORK ACADEMY OF SCIENCES Dictyograptus Dictyonema “Cells: Well-marked; tend to become “Cells: Small, generally disposed at uniserial. angles of about 120°. Crinkly lon- gitudinal ornament appears. “Shape: Like a fisherman’s net from “Shape: Basketlike, starting from a a sicula of no great length with long narrow tube or nema and di- primary branches, diverging at an verging at angles which in the later angle rarely greater than 90°. forms approach 160°. “Dissepiments: Their development “Dissepiments: Their importance, in- ever more and more delayed until crease and the general aspect of they become practically abortive. the later forms is that of a square or rhomboidal mesh in which cross bars and stipes are of approxi- mately equal importance. “Type: var. acadica forma typica.” “Type: var. norwegica.” There is one point in which I completely agree with this author, viz: the importance of the differentiation among Dictyonema flabelliforme. We both believe in this as the point of generic divergence; but in detail, I cannot help stating that all my observations run along the opposite direction. I merely recur to the facts that Var. conferta, restricted to a lower horizon, according to the sugges- tions of Matthew, Fearnside and the majority of the Scandinavian authors, shows clearly the tubulose structure and mode of growth and dissepiments of “Dictyonema” and the projecting cells of “Dictyo- graptus.” Var. desmograptoidea has cells, mode of growth like “Dictyonema,” cross-threads like “Dictyograptus.” Var. acadica has cells, sicula, like “Dictyograptus,” tubulose structure hike “Dictyonema.” Var. norwegica has cells like “Dict, POTS dissepiments like “Dictyonema.” Var. ruedemanni has sicula, Cements, structure ike “Dictyo- graptus,’ cells and growth like “Dictyonema.” Hence, as all the known varieties are in part distinguished by the features of “Dictyonema’ and in part by those of “Dictyograptus,” such a separation, of course, seems impossible. Furthermore, the ancestors were doubtless bryograptoid with very distinct cells, without numerous. dissepiments. Nor did Westergard agree with Fearnside’s assumption,, so that against this hypothesis there appears to be no further objection necessary. On the other hand, Westergird, from whose detailed work my own observations do not differ widely, starts with the following suggestions = HAHN, DICTYONEMA-FAUNA OF NAVY ISLAND, N. B. 159 Among a hundred examples of Dictyonema flabelliforme examined, he failed to observe nemas or disks or any adhesive organs in spite of dis- tinct proximal parts with free sicule. He regards the wrinkling of the rhabdomes as produced merely by pressure, and the early appearing divisions of the nepiastic thece as due to rapid budding of nourishing theeew. He holds it probable that the early Dendroidea like Dictyonema flabelliforme when derived from graptolites must have possessed thece hike those of Bryograptus kjerulf. From this point on, he considers that there existed a great difference between the Dictyonema flabelli- forme and the “Dictyonemas” of the later Ordovician and Silurian, as described partly from sections by Wiman and others, and he introduces “Dictyodendron” for the later forms. The following objections, however, must be made after careful con- sideration: 1. Young individuals of Callograptus (salteri in Ruede- mann’s monograph, grabaui as described in this paper) exhibit distinct sicule and even nemas. 2. True adhesive expansions occur among the Dictyonemas of Navy Island. 3. The tubulose structure is by no means referable to-any kind of stress, because this, when exerted, produces a fine striation, running in parallel fashion over all specimens of one slab, but cutting, of course, the tubes at quite different angles. The associated Stawrograptus, moreover, never show a similar structure. | 4. The young stages of Dictyonema jlabelliforme, as fully discussed in the foregoing, are found to be clearly distinct from all stages of Bryo- graptus, as they exhibit a more primitive character (dependent growth) than Bryograptus (declined or horizontal growth of the first branch- lets). Hence the ancestors of Dictyonema flabelliforme were not at all true Bryograpti, but simpler types with features bridging over those of Dictyonema and Bryograptus. Thus I feel quite certain that the progress of evolution which Wester- gard believes to have existed between the early Dictyonema flabelliforme and the so-called “Dictyodendron,’ did really take place among the varietal series of the Dictyonema flabelliforme. Finally, there is one point left which thus far I have deliberately disregarded in order to simplify its consideration, for it adds merely some complications without modifying the preceding conclusion. There is no doubt of the fact that most of the dendroid genera have the charac- ters not only of stadial, but also of collective groups. Of this James Hall was partly convinced even in the ’60’s, but it has only been recently fully substantiated by the masterly works of Wiman and Ruedemann. Among the Dictyonemas, one part has thece with sharply prominent 160 ANNALS NEW YORK ACADEMY OF SCIENCES lips, another possesses obscure pits and grooves in place of thecal aper- tures; one group is distinguished by an uniserial, the other by a biserial arrangement of the thece; a few species belonging to widely separated divisions are proved to be heterothecal, while others are strongly sus- pected to be homeceothecal. We find similar differences among the genera of Callogeapinel ap- pearance, particularly among the Dendrograpti, the Ptilograpti and in the Galeo-Disco-Cyclograptus group. Those differences are not confined to a certain geological age. Early as well as later species may show the same kind and arrangement of the thece, while congeners reveal the difference as great as possible in this respect, e. g., the highly variegated Niagaran dendroids of America. This difficulty of a heterogeneric but parallel evolution seems still further increased in complexity by the repeated development of features either retrogressive or intermingled. For we observe the strongly differentiated genus Dendrograptus already at the base of the Ordovician (if not earlier) and yet we see that the path of varietal evolution is in the same direction in much later strata. We observe true Dictyonemas, the neanastic stages of which suggest cer- tain ancestral relationship to members of the Desmograptus or the nor- wegica-murrayt group. That the same is true among the Callograpti is proved by a comparison of Pocta’s, Bassler’s and Ruedemann’s species herein referred to. Even the characteristic bifurcation of the apertural margins and dissepiments has now been found in quite different species and genera, e. g., Ptilograptus, a very doubtful and not unlikely retro- gressive genus, to which some Callograpti described by Pocta bear an undeniable similarity. However complicated the paths of evolution of the Dendrordes may be, paths which can be traced back only after the manner of Wiman’s and Ruedemann’s keen-eyed and careful investigations, I am convinced that, generally speaking, they followed directions similar to those mani- fested pre-figuratively in a small but equivalent ratio by the range of variation in Dictyonema flabelliforme. | PLATH XX RANGE OF VARIABILITY IN DICTYONEMA FLABELLIFORME AND ITS EVOLUTIONAL SIGNIFICANCE ‘ - Sa! Varieties of Dictyonema flabelliforme and Callograptus grabani n. sp. cor- er species and. genera sketched after Hall, Ruedemann and Wiman. bata (Reduce all given magnifications by three-fourths. ) i ’ 4 f : n een & eae ier SAP RRD ELOY OD CATE bh? BILBAO DASE ASE OTD BEE 1! ESE, aie §; > VLAD TA SOL \ he me AMI (2 a Tato, ala VON aOT Hae AAC GW, Bats wO ys We gui raid ridatidaw to soissllor oot ti anentisod, Leute tite Rae? Dek stirsuipsheun Hidhtievlsdsiede esonce bak fl. ‘aes - os as ming |‘ ¥ ’ ot ae Oa A TE IO, A rue SRS Sp at as SHR Caw 59 ont oT Be ie a Annats N. Y. Acap. Sct. Ms | | a ——— ? oa - x I IN ; 3 4.15 ATH Ha \ : H 1 HAG HUH ' varconterta x4. Crassum XS vadrangularis o / x4 a a | — DICTYONEMA. -——--==— a J DICTYONEMA leroyense | peltatum tuberosum v N a x20. / X4 — WE GZ Zz = ZZ gat co —— mt acacia --canterta x4 /] cermoorne 4 CQVeINOsum \ hal i » /FLABELLIFORME | WV aca C7 rectifineatum\\i{}} 14 Wi We DICTYONEMA __ subretiforime ‘4 Octal PTILOGRAPTUS poct X70. Ay XA oes ACANTHOGRAPTUS SMO pane TUS * SUCCICUS X20. D xzo "“f flvit tans P \ succulentus AF: . \f tenviramosus ©... gicus DENDROGRAPTUS x4 DENDROGRAPTUS : os Z WSS OAD GA a PLATE XXI BASAL ORGANS OF DICTYONEMA FLABELLIFORME 1 a, b, var. conferta 2 a, b, c, acadica (0, c, juv.): 3 a, b, acadica m. f. ruedemanni 4 a-d, ruedemanm All specimens are preserved in the collection of Columbia University and are enlarged five times in the figures. The measurements given refer to the length of the entire colony, which is not always represented. 49 iu aa ae det Om Gt ak wy : WAL (Gate ; BREESIIOY, oiCl « é > foileriathie Yiodae 2aawel & ey ren: eros jaa et tpadastgs ? 1 sto I ee Seu Bs i eek ret bes (8 fo. Sica sem | sdial Yo; She 8 4 (BLEE #, avicage awoud test soxnond : 6 RTO rotsa) ywol bey} gore Bee a nhy 10 ‘a aor's ighisiac t Li@iss PHO lati $i19ds laligune hie 1-6, rsacrid -)- . brisitiosG wok oe dalinunc’ ct nn a ee ee + a UH >. OOOO), BIBS & nanan 254 ab.gu eotqer | —bellersadase jcmioesetea len : "Beoxiatt } Reiniot aT ESE di ¥ - abl th wise sca Jeeae! nce ) Annas N. Y. Acap. Sct. ee eee cn 0 VoLtuMB XXII, Puatr XXII Rhabdomes Dissepiments Thecee Notes Length’ Per cent Names Shape and size ee Fi Se | | Number Diverg. | Number Basal organs individual a a nies width i in Character se on . in Gharacter Meshes Number Character Passing into Char. of stage Evolution and age | Preservation Dictyonema flabelliforme | infundibuliform and : early stage bryo- | center of variabili ey pyriform 16:1 9-10 4 (.8-.5) mm. 45° 45 thin quadrang. 16 sicula, nema conferta penta Me GEV sites 90% of lat 60% projecting : apache true Dictyonemas e FaveRHG: acadica large—very large 9:1 (8-12) subparallel 80° (3-8) 2mm. (elongated) | (15-17) ?1 stolon (young stage) | esmograptoidea | tate stage with F Me ruedemannij} t stages broken mixed features | up to lower Carbon- | iferous infundibuliform 8:1 4 mm. Ais 4 isp nga , brairi 8-10 25 4-5 .25 mm. CLES Menger le ct mocKe..! i) = onpactvoginte el -- otene all <2 Sc nesoac Species r medium size (?) subparallel lower Carboniferous DOken | | infundibuliform 6 45° basal adhesive neanastic leroyense 1:1 8-10 s 2-4 .25 mm. subquadrang. | ..... | 9 «ss... Fea tee a | ee eT with thickened Onondaga broker large ? ee 80° - disk branches a Cy vasiform—cylindri- 2.5:1 4 (.38-—.5) mm. 45° 7 thin atarie 7 delicate atari Gh early stages: cal (3.7:1) gy, 2 quaoreng: Sua BOGIES bryograptoid; a 20% of lat y. conferta 6% 8.5-10 BuBDAY 1 basal disk ruedemanni 2 not any evolutio % of later parallel— = 4 1-2 mm. z . + oti : . 4 ; late ephebastic y » h. st: medium size (large) | (1.5:1) parallel ie (8) regular (GERM) | (UGE) | apsaktegtas ?1 branching stolon | (pyriform link) contraction iis Seen 5 are .5-.7 mm. a5 thick— a ; char. of var. a few species in ee infundibuliform re um ars 50 6 irregular rectang. re pues a nCadiCn usually restricted Ordovician; Aen : large * robust 90° (4-8) subcircular acute (conferta) to late stages of yarallel evolution up 8 + parallel (.1)-.6 mm acadica 1 F +1) =: p to lower Devonian 2 e 1.5 mm. 50° rigid murrayi , 2 4 sient hae 4 7 rectang. 9-10 MCuUber we PSA TS Sree atetel ate poeeema |b eke Madsen et | Beekmantown = broken arge ‘ .4-.9 mm. parallel | | infundibuliform 1:1 Bb alr 50- irregular rounded distall crassum | 9 Rone CG Yara ee ea emma aot 4 eR ol lee cee ||.8 stance lt. MIeDROUGnDDTOT ae wll ~ sabaroAe. denmig mantel New Scotland + broken | large (?) + parallel (90°) 4-9 mm. subquadrang. rap! cba of var. ri .4 mm. usually restricte infundibuliform 8 + 90 3-4 very irregular irregular delicate to late stages of to genus Desmo- v desmograptoidea 1:1 YG subparallel— IGT] | ee oe a? adn onepoodoeds acadica acadica ; graptus, up to + broken medium size (7-9) flexose— (45°-90°)} (1-5) .1-4 mm. subovate less projecting middle Deyonian joining acceleration be- ginning | infundibuliform tent very rare elongated appressed basal parts with y 8 F o a + well intricatus 2 ¢ 1:1 15-20 fils (64ND) 45°-90° (0-1) tet agama aes ered S85 : siculay?’- - —mylia 3 Bares Dictyonema char- Chazy preserved a medium size coalescent) irregular subovate circular acter Z ae = a f eg 8 chitinous basal 2 infundibuliform 21 ; very rare elongated appressed . = cancellatus A 1214 | adulating (thick) | = 90° | (0-1) frequent Seo Sai Airey gale Al). uenee mbar t + broken medium size (?) yw ; sae 5 irregular subovate circular aOTiatann sicula, emir aiamen Gah similar species up infundibuliform— Segall ess FI SAS ESS to Devonian y. ruedemanni OLY (J) 33% xz Be + 45° : 7 pars | eee 15-16 delicate | Dee orstolon\withidisk, see nee! ete lati 4070 or Tess 6 = ‘ * urther coolution roken large—very large (1.1:1) (10-15) subparallel (1-5) -lmm. elongated branching stolon in (pyriform link) | old Biased parely atin (OFiboe Func ; neanastic stage RoR BEER’ Dendrograptus zi 2 4 irregular old stages partly rectilineatum 2? 12-14 ai) Hel st 45° 3-5 quadrang. | ..... irregular ?shortstem | .seeee desmograptoid Chazy broken medium size au parallel “et —.16 mm, ° eyatiform 1:1 4 45° delicate; quadrang. 13-14 3 A partly well gracile etek 10-12 (6 at the base) 14 BSE projecting Dasalestem and dickoue | eeanmesee eee | ERE Niagaran preserved : a J ] 700 palette (HE) (@) subparallel 10, -1 mm. elongated (207) - = B 3-5 . a infundibuliform irregular : partly dendro- Beekmantown— partly well salteri 5 11 14-15 subparallel st 45° O(-1) very few off case stec 14-18 thin hydrorhiza | ...... graptoid Chazy preserved 3 small—medium size (flexose ) appressed Oo a Sw ls Se ei a lisey les ‘. —fividl iauibana arotilsd ibaritess © sovnl open ete ctte tee 5 acotifdibantai satel z arrdtified ibastat: 2 ogiel uel she a Steers ; as] idee ce : = eae 1 & ee i: | a: ! 3 =. j H Ps) ~ Pees bs 2 poet EY | | 4 H 5 — } : | : : : } ; { SScannES aeiaet CalsiarienaericemeaeE a rel u oe AS ee a Cl Ce PRON ew 8 TN Se tel Le ew Fa eat | | | | | 8 ~ | | | | roe ; | CI OO cael ; | i Y : —~ Rio * Se —. i s en] = i x & i rig Fo t= 9 =< % f Red Oe | gr: Be 4 x 7 AS ; ia Shia . te | } haw ws } hand — | } . | cag | ‘ ys ; i Se { { j | | st f | i | : : } - | : | aly ' & : 3 : 7 <. { . . | me. a ’ SOMA eae ae 2 ese: age tieninndnmiots Pah SK Te EN ane et hg nt ota am si:iapap ted | I omposing the aes are wcated separately, each in its own cover, ee ce cs in bundles on an average of three per year. The price of : irregul r intervals. It is intended that each volume chalk be devoted to mon graphs relating to some particular department of Science. Volume __ ee devoted to Astronomical Memoirs, Volume II to Zodlogical Memoirs, ep ‘The price is one dollar per part as issued. shes The All publications are sent free to Fellows and Active Members. . THE LIBRARIAN, New York Academy of Sciences, care of American Museum of Natural History, New York, N. Y. é oe o ANNALS OF THE NEW YORK ACADEMY OF SCIENCES Vol. XXII, pp. 161-224 Editor, EpmMunp Otis Hovry METAMORPHISM OF PORTLAND CEMENT BY ER ape) ae eae oe eee Be” eee 4 ALBERT B. PACINI NEW YORK PUBLISHED BY THE ACADEMY 10 SEPTEMBER, 1912 ee Se Oe Pay RED Ee PP ee oe ae a re. . y THE NEW YORK ACADEMY OF SCIENCES (Lyceum or Narurat History, 1817-1876) OFFICERS, 1912 President—Eurrson McMituin, 40 Wall Street Vice-Presidents—J. EDMUND WoopMAN, FREpDERIC A. Lucas, - CHARLES LANE Poor, R. 8. WoopwortH Corresponding Secretary—Hunry HE. Crampron, American Museum a | Recording Secretary—EpDMuND Otis Hovey, American Museum Treasurer—HENRY L. DoHERTY, 60 Wall Street Librarian—RAuPH W. Towsr, American Museum ~ Editor—Epmunp Otis Hovey, American Museum SECTION OF GEOLOGY AND MINERALOGY Chairman—J. EH. WoopMaAn, N. Y. University Secretary—Cuares P. Berxuy. Columbia University SHCTION OF BIOLOGY Chairman—Freperic A. Lucas, American Museum Secretary—Wi.LuiaAmM K. Gregory, American Museum SHCTION OF ASTRONOMY, PHYSICS AND CHEMISTRY ~ Chairman—Cuantes Lane Poor, Columbia University Secretary—F. M. Prpursen, College of the City of New York SECTION OF ANTHROPOLOGY AND PSYCHOLOGY Chairman—R. 8. WoopwortH, Columbia University Secretary—Freprric LyMAN WELLS, Columbia University P bg Se ee ig ee te Sg es en Seley oy Fae tes ae s, ee Tree a: SA CRAY eg eee en ae eee he IPM lle Bee heehee Eien Steg The sessions of the Academy are held on Monday evenings at 8:15 . : o'clock from October to May, inclusive, at the American Museum Natural History, 77th Street and Central Park, West. of [ANNALS N. Y. Acap. Scr., Vol. XXII, pp. 161-224. 10 September, 1912] METAMORPHISM OF PORTLAND CEMENT? By Apert B. Pacini (Read before the Academy, Part I on 8 January, 1912; Part II, 1 Apri, Introduction........ é Nature of the problem CREM AMCOMNMPOSMMOMerts seis. 5 slic sca es oecle cies hae ce gals se cetncan ce ee ee. 164 liltmerealoecall COMSAT 6605 6 Sabon oe Doo oe lo Oeibo- co Dom cco mn ic pico ood t5 Ios SHAMS [PROEESS. o ceccwls 6 pe GORE CO DO De BO CIG Ore Dirks o.coie ner aeiricne eno. clarion 165 TEIAIRCWAINTING?: (AIROVGES Sc Gc Glo.c co GIES b CIn IRIS DIOIGs DIC EnDICIC ISI oi oleic CioloiCi ina ie erciicicioae 166 Influence of water upon metamorphism. ...............0.- 2 eee cece eee .. 168 Temperature of the water at first added......................02005- 169 Temperature of the water that may eubseutenuly come into contact Brats Na TehT CRSA LTA eearee ty eters US at craia celle level e le -eversifelsigiesaiele ois ic elaeid.w ele 171 Ant VyAOkeWwAlel: ab MES Ad Med so.5% 22 cnelaels © oe cece se sacle s dec ebisces 172 SUAS Ol CAmSMe [INANE Gonobo somos odUn Oe CbOUmoSlnd oon GoIOoccs 172: BAAN COs ees! 25, ce cis aryans Ses 204 5 SE cs ROE RE LS OE Cs ae 173 EDV CODY SISt UNC Ol yeeicna cel etetety orate ae) ooo ate ala ccy2)ci/dici ceils sta elae oda eee Cues 174 Mechanical agitation when water is added...................... 174 Total quantity of water at first added.............. Sethe eiele vs ae 175 Quantity of water that may subsequently come into ponte with the SV SECIS ie sreter ciel eteiere a fata oes Ao Cee SHESat a CORE GOR Oe ae Ee 175 Steel COM ERE ALM CIES: sarece lateveva eretetae ale Senses oer aipieve fa Hine uvenacs sae erets T/T Membranes........ Hed coosobodooeaubscese soc ORD ES ooage spec soaoo UIT MAIS S eG eat Gl USivers.srttereeusirct apiece hoc tencheeanteeo fave Ua aitezeilel-cilenuieveraavere rele Sateen eli Quahity of water at. first added. es ee dee ela a ee ah ate Tendele us excole eee 178 EwinemMmaterial: iW SOlIiONNs 932522 o us cleo oe seasons coe wees 178 Quality of water that may subsequently come into contact with the SWSUCMM.c.0, e205 < ele 5S EOC O RO es OR A DS See rune ree ENC UEP AER USE Mr oa RA Sige OLS a 180 Senne TMA Th ODIO Goo ou doonomoc cuss cos obo db donee ous eso SS Ge NPC Te syeira cea crete Me cape\vetna. ours nc eeSueare araiabevonehal eapeuaueaicanstaets PEO eco 180 Alailicangd! deep TOck WATS? Hiss cocis shGesiete = bes oats secceperene Sree leo Having material in suspension 1912) CONTENTS eee ee eee see eee eee ese eee 1A thesis submitted in candidacy for the degree of Doctor of Science at New York University, 1912. Acknowledgments are due to Prof. J. Edmund Woodman and Mr. Raymond B. Earle, of the Department of Geology, New York University, and to Engineer Inspector Ernst Jonson, Board of Water Supply, City of New York, for valuable suggestions made dur- “ing the preparation of this paper; also to Mr. Fred H. Parsons, Assistant Engineer, and Messrs. James E. Jay, Charles M. Montgomery and Charles EH. Price, Inspectors, of the Board of Water Supply Laboratory, for material assistance during the experimental work. (161) 162 ANNALS NEW YORK ACADEMY OF SCIENCES Part II Page Experimental investigation. 0.5. s he cs ele Oo ees See Oe 184 Temperature of the water at first added.................. ...cccceee 186 Temperature of the water that may subsequently come into contact With: the Systems. 2). 2005.04 00s. shia cele aeete ole eee 186 High-pressure ‘steam: .. 6. 20025 Sosa a eters seins eee 186 Cold: storages io sae accisie's Rae SE Oe U CE een eee A 187 Quantity of water at first added...................--00+- ceccceccce 189 Size of cement particles... 5. 22.26.5 5. acanc uses Bae 189 Mechanical agitation when water is added...................... 191 Setting time of cement in laboratory air and in damp closet...... 192 Effect of excess of mixing water on strength of concrete......... 193 Effect of excess of mixing water on permeability of concrete. .... 194 Effect of excess of mixing water on strength of neat cement..... 195 Effect of the presence of clay and dissoived substances.......... 198 ‘Quantity of water that may subsequently come into contact with the SY SLCIIE ahaa eh ate een etc isalSila bien Seen en eee 200: PerMe aD ity. ccsyé.ois5.0 oeananes oslo koe ioe sto wee DS ne ee 200 Coneretes containing different aggregates.....................-- 204 Concretes containing different cements...................-+eeeee 205 Effect of the direction of flow through concrete.................. 206 Quality of water at first added). . .....0.6 2. tc eee eee we eles | ole ene 208 Compressive strength of neat cements gaged with various solutions 208 Effect of gaging with various solutions upon the strength of mortars afterward stored in water.............6....----+-ee> 210 Effect of gaging grout with rock waters.................-.---+: 211 Quality of water that may subsequently come into contact with the SV SUC. wc. d05 aie: dite ais loves. « ecanai a wh cea eysh w re tetane ne mite tevlerele teee ee one eI I eke neem 213 Theoretical considerations. .............- 2c cee eee e ee tweet ee ees 213 Effect of storage in various saline solutions upon the strength of GW TINOZNRS Sagoo das losSoncouseceoodcedonsobes oes es ioe ee 214 Effect of storage in rock water upon the strength of lean cement LINOLEIC ESPRIE See eI AIas REN ones ence AIO OAc ACO O COOKS 216 Summary of experimental results................... “beets ia Cae Lee Se RS 218 General GCOMCIUISTONG!. c.5 clos voce okeies e cee cle eile aos sei) le iovedenslololejehckers i) tea Reems 219 ISN HON Nhs onldeoGakcomdoocedonGbOos Bou Guu DOOgnUe Oo DOIUc OSG Cb O00 settee 220 PART I INTRODUCTION The important field of investigation covering the changes which take place in the setting and hardening of Portland cement and in Portland cement which may be considered to have attained the greater part of its maximum hardness calls for the services of experts in various branches of science. Many of the general problems can, as a whole, be relegated PACINI, METAMORPHISM OF PORTLAND CEMENT 163 to the petrologist and hydrologist; and this paper is an attempt to treat cement as a rock, differing from other rocks only in being artificial, but subject to the same internal and external influences as other components of the earth’s crust. _ A training in geophysics and geochemistry is, perhaps, the most valu- able asset in surveying the field of Portland cement. If no other end is achieved by the following pages, the mere representation of the question as a problem in applied petrology will, it is hoped, help future investi- gators in a more systematic inquiry. Part I of this paper is devoted to a necessarily brief review of the present status of the subject, and no attempt is made to discuss data quantitatively. Experimental results in elaboration of the various points discussed are presented in Part II. The experiments described in Part II were made at the laboratory of the New York Board of Water Supply by the writer, and in part by his associates, in the course of the investigations of the Board. The most modern and complete equipment was available, thanks to the prudent foresight of the gentlemen at the head of this great engineering enter- prise. The data are reproduced by permission from the periodical bulle- tins of the Inspection division and from the annual report of the Board iow alae NATURE OF THE PROBLEM Portland cement is a finely ground artificial rock, whose essential constituents are silica, alumina and lime. In it are found a number of component minerals recognizable by definite optical properties, but the individual constitution of which is not yet clear. The percentages of these minerals vary somewhat according to the method of manufacture and the purity of the raw materials, but there is, on the whole, a fairly stable proportion in a series of normal cements. The method of manufacture of Portland cement will not be discussed here further than to state that it consists essentially of the calcination of a mixture of calcareous and argillaceous rocks at high temperatures. Usually, about 2 per cent of gypsum or of plaster of Paris is afterwards added to retard the set. By varying the proportion of these rocks, the temperature and duration of calcination, the fineness of grinding, and also by the addition of foreign substances products are obtained having a wide range of hydraulic properties. The hydraulic properties are setting and hardening. Setting is the attainment of rigidity by the plastic mixture of cement and water and begins immediately after mixing, requiring several hours for completion, 164 ANNALS NEW YORK ACADEMY OF SCIENCES Hardening is the progressive increase in strength acquired by the mass, and it attains the greater part of its ultimate value in about a year. Even after this period it is subject to a small progressive increase (42). In general the properties of setting cement are to be found both in mortars, or mixtures of cement and sand, and in concretes, or mixtures of cement, sand and broken stone, these chemically inert materials added to the cement exerting a physical influnce on metamorphism. CHEMICAL COMPOSITION The chemical composition of normal Portland cement is shown in the following tables: Average of 300 Normal American Portland Cements, Representing 20 Brands of All Types (Analyses By the writer for the Board of Water Supply) Maximum Minimum Average Osc siases srs oho ksucavlin ce coreneaey en aolene nce ahetioe erence: 25.89 19.85 22.70 1 EH O en aceon ntens areca tees MRR Ur eA oe 4.08 1.23 2.73 NTO ee apeaie Mopaeds os tees wee atm catdene Aree, 9.16 3.43 6.17 CAO R etars aniitecets eacedercoe tay seensieeetee eae ans 64.91 59.06 62.67 MES ©) iain 2.2 cde ae ty on eaten eae 4.00 0.30 2Zald SOs sctacercasut tater a acre taruelens iene pace 1.75 0.84 1.37 COREG O; faillcalieSs. Sc .cseacy oe cree eke cacao oie es ea ee nase Pas lerl Average of 100 German Portland Cements (Burehartz, (12) ) STO esas a ceie te catacter a vonblesoraniane ay teats eeebone SPO touell oceania rece eRene, Sane 20.87 1 er! 0 aaron mane rnin ace Sri er eae cinein a Sects bod Ss 2.98 U2 O eae Cees a BIER, eM onl NN Gan entire cin Same m emote’ o Oe 7.63 C621, etre Anse ernment ena Taree Mn RES frre. ce cea BRA Gc 62.99 MeO asic, At Besealereredtates nade eG) ae cere eect aarate nea ete ac ne emepesie ave 1.55 SO hace a Bhs aie a toe ocean ata seanie acy raat cet ta ores aps aaaviopesallatariey amet aeeas 1.85 The ultimate chemical composition of a cement is only, however, a rather indirect clue to its hydraulic properties, just as the ultimate analysis of a composite rock may only give a faint idea as to its con- stituent minerals or possible products of metamorphism. For example, it would be quite possible to synthesize a mixture which would, on analy- sis, correspond exactly to the chemical composition of an excellent Port- land cement, yet which, when gaged with water in the ordinary way, would develop practically no tensile strength, in fact would possibly fail to set at all. Cement, therefore, must owe its hydraulic -jseifbill 2s to a particular grouping of its constituent compounds, quite analogous to a series of 2 Numbers in parentheses refer to the bibliography at the end of this article. PACINI, METAMORPHISM OF PORTLAND CEMENT 165 minerals; and looking to the identification and classification of these minerals, a great deal of investigation has been done. By trial burnings of simplified mixtures, such as lme-silica melts, and by microscopical examination of sections of the resulting clinker, the problem is gradually being clarified, but, owing to its great complexity, much controversial literature thereon has been issued on both sides of the Atlantic (52, 69, 80, 64, 65, 88). The theories put forth have so far had little practical effect upon the manufacture and composition of the commercial product (63). No complete and final enumeration of the chemical duipotiine result- ing from the burning of such a mixture of clay and limestone has yet been accepted as authoritative. The microscopical identification of the individual chemical compounds which go to make up the mineralogical entities is at best somewhat unsatisfactory, especially because of the minuteness of the particles of raw materials necessary to secure thorough and uniform calcination, and consequently the extremely small size of the resulting crystals and aggregates. It has been proposed, in this con- nection, to secure these of a size available for study by the expedient of fusing the clinker in an electric furnace; and, by this means, a partial clarification of the system has been obtained (103). MINERALOGICAL CONSTITUTION The minerals which are recognized in cement clinker have been named alit, belit, felit and celit (101), and a metamorphism? of these occasioned by the action of water is the cause of the setting and hardening of Port- land cement. Alit has been reported a solid solution of tri-calcic silicate in tri-calcic aluminate, and celit a solution of di-calcic aluminate in di-calcic silicate (61). Other investigators have reported alit and celit to be silicates of different silicic acids (26). Belit is probably a calcium aluminum silicate of the composition Ca,Al1,8i,0,,, a form found in nature as the mineral gehlenite (27). SETTING PROCESS Precisely what chemical reactions and physical transformations take place in the setting and hardening processes is not yet definitely settled. It may, however, be stated that by modifying the proportions of clay to limestone through a certain range, we obtain a product which varies in its speed of setting and of hardening. In general, cements high in silica 3 Metamorphism: Any change in the constitution of any kind of rock, Van Hise (104). 166 ANNALS NEW YORK ACADEMY OF SCIENCES are found slow setting and slow hardening, while those high in alumina are quick setting and quick hardening. An increase of lime in the latter retards the setting (63). The calcium aluminates are probably the main factors in the setting of cement, while the hardening is due to the calcium silicates. The mag- nesium compounds are inessential to the hydraulic processes (105). Upon the addition of water to cement, the equilibrium in the system of solid solutions and chemical compounds is destroyed, and a series of changes is inaugurated tending towards the production of a system which will be stable under the new conditions. The first effect resulting from the solutions and reactions brought about by the presence of water is the setting of the plastic mass. Under ordinary conditions of practise, the quantity of water used is about 22 per cent in the case of a neat cement, being less in the case of a mortar, and still less in the case of a concrete. When this proportion of water is used, it is probable that the setting of cement is mechanically analogous to the setting of plaster of Paris and is caused by the growth throughout the mass of a network of crystals, deposited from the satu- rated solution formed by the first stage of hydro-metamorphism. Owing to the low solubility in water of the original component sub- stances, the attainment of final equilibrium is a matter of considerable time, and is further delayed by the automatic protective action of films of insoluble substances coating the active particles (23). 'These films in some cases are semi-permeable, and exert a selective influence upon the solutions osmotically penetrating them. Under normal conditions, that is under those conditions which have been found in practise to yield the densest and strongest product, this attainment of equilibrium considered apart from the setting process at first proceeds rapidly, but the rate of increase of strength grows smaller, tending to a minimum. A. Erskine Smith has shown (90) that there has been no permanent retrogression in the strength of cement in the case of specimens kept under observation for 21 years. Of course, this relates to laboratory specimens protected from weathering, but shows one of the directions which this metamorphism may take. HARDENING PROCESS The hardening of cement has been ascribed variously (48) 1. To the fineness of grinding, 2. To the increasing stability of calcium compounds due to combina- tion of part of the silicic acid as the silicates grow less basic, 3. To the action of free lime upon calcium compounds, PACINI, METAMORPHISM OF PORTLAND CEMENT 167 4. To the decomposition of basic products present in the freshly set cement, 5. To equilibrium of calcium hydroxide with the siliceous constitu- ents, and 6. To the hydration of the double silicates and anhydrides of lime and alumina. The two theories that have at present the greatest claim upon con- sideration are that the strength of set cement is due to the progressive erystallization of calcium hydroxide (80), and, in some respects diamet- rically opposed, that this strength is due to the formation of a dense complex colloid, soft at first but gradually adsorbing calcium hydroxide and thus becoming harder and harder (64, 65). According to the latter theory, cement consists of a mixture of fused compounds of silicic, aluminic and ferric acids with lime, together with an excess of lime, partly dissolved and partly enclosed. Upon the addi- tion of water to this system it is decomposed, and the water becomes a supersaturated solution of salts, which react between themselves. The: compounds resulting from these reactions crystallize about the cement: grains in needle-shaped crystals. So far, the process is analogous to the setting of plaster of Paris (45), and silica takes no part in these pre- liminary reactions. A hydrogel begins to form about each grain, in which the crystals become embedded. This hydrogel consists essentially of calcium hydro- silicate, and to a minor degree of calcium hydroaluminate and calcium hydroferrite. At first it is soft and plastic, but gradually becomes dense and rigid by the adsorption of calcium hydroxide. The strength of cement is mainly due to this process of coagulation. The calcium hydroxide may of course erystallize and lend additional strength; but its crystallization is rather more hkely to burst the har- dened cell walls about each grain of cement, and thus admit liquids later in the process which may be fatal to the integrity of the structure, either by undesirable chemical reactions, or simply by dissolving away the lime, with the formation of soft hydrates of silica, alumina and iron oxide, instead of the desired hardened colloid (64, 65). - Much corroborative evidence has been offered by supporters of this view, and similarly by the exponents of the crystallization theory in defense of that. The question is still at issue, and the main difficulty is the microscopic recognition of the constituents of set cement (34, 78). Unquestionably, colloidal materials result from the action of water on silicates of this type, when the particles have been ground to the fineness of Portland cement (23, 21,95). This has been directly observed in the 168 ANNALS NEW YORK ACADEMY OF SCIENCES case of cement and reproduced with synthetic mixtures. What binding power colloidal material may develop is strikingly seen in the case of conglomerates and sandstones in which hydrous silicic acid, aluminie hydroxide or ferric hydroxide has been the cementing material, so that the theory is attended by a high degree of probability. On the other hand, it is also quite conceivable that the interlocking of crystalline masses between the grains of cement may account in some measure for the strength. There is definite evidence that calcium hy- droxide does crystallize, and its mineralogical and crystallographic con- stants have been determined (24). The two views are not entirely irreconcilable, and it is possible and even probable that, mechanically, the strength of cement acquired by hardening is due to both processes. Whatever be the chemical reactions in detail by which these elements of the structure are produced, the main condition for their occurrence is the presence of water. This paper is devoted to an enumeration of the factors which influence the metamorphism caused by water in Portland cement, and the varia- tions in the physical properties of the resulting rock, brought about by varying these factors. INFLUENCE OF WATER UPON METAMORPHISM The action of water upon Portland cement is a resultant of 1. The temperature of the water A. At first added B. That may subsequently come into contact with the system 2. The quantity of water A. At first added * a. Size of cement particles b. Mechanical agitation when water is added ce. Total water added B. That may subsequently come into contact with the system 3. The quality of water A. At first added a. Having material in solution B. That may subsequently come into contact with the system a. Having material in solution b. Having material in suspension * Owing to the peculiar autoprotective reaction of cement against the action of water, before alluded to, the quantity of water coming into contact with cement is a function of the size of the particles and of mechanical stripping of protective films, as well as of the ratio of cement to water. PACINI, METAMORPHISM OF PORTLAND CEMENT 169 The final effects of geological processes do not differ in the main, whether these operate upon natural substances or: upon the products of human industry. The agent whose activity is responsible for the majority of terrestrial changes, namely water, is also the main factor in the meta- morphism of the artificial rock, cement. By intelligent control of the action of water upon this rock, the desired results are obtained, and its value as a material of construction is inestimable. Lacking this insight, the action of water may result in catastrophe, or at least loss of time, money or efficiency. Geology, then, through hydrology (59), is enabled to give substantial aid to the engineer. TEMPERATURE OF THE WATER AT FIRST ADDED In construction, the water at first added to cement, known as the gaging or mixing water, is subject to the entire range of variation of atmospheric temperature. The lower limit is far below the freezing temperature of water and of course, in this phase, water is useless for the purpose. Within the possible range of temperature under working conditions, it has been established that as the temperature of the gaging water used is higher, the set becomes more rapid. Considering the setting due to the deposition of a network of crystals from the supersaturated mixing water, the beginning of this deposition would be sooner attained, if the water reached its condition of supersaturation more quickly; and this condition would be brought about by a higher original temperature, provided, of course, that the solutes increased in solubility with the tem- perature. With a higher temperature, the volume of the water would be greater and the viscosity less, and consequently its range of activity would be increased ; that is, it Would be enabled to reach a larger number of cement particles and thereby more quickly arrive at its saturation point, and the deposition of the crystalline network hastened in consequence. If the temperature of the mixing water be above about 37° C., the setting, instead of being hastened, begins to be delayed. If the deposition of this network were a simple case of precipitation from a hot solution, it would be logical to state that the solubility of the compounds concerned was so high at this temperature that they were not deposited from solution. The problem, however, seems chemical rather than physical, and it is more probable that this effect is due to hydrolysis. Hydrolysis increases with the temperature. In the case of the weak salts that must exist in the system we have under consideration, the ulti- mate products of hydrolysis are the gelatinous materials—silica, in the hydrated form, aluminic hydroxide and ferric hydroxide. The adsorp- 170 ANNALS NEW YORK ACADEMY OF SCIENCES tive and coagulative properties of these materials unquestionably do not compare with the coagulative powers of the complex colloid which Michaelis postulates (64, 65). If, therefore, the hydration of cement does not proceed in a properly regulated manner, it is conceivable that it may become a hydrolysis, with deleterious effects. If the mixed cement is allowed to freeze, the setting will not take place, but on thawing out the mass, setting is resumed. Obviously the transition of the water to the solid phase hinders solution and diffusion, and upon resuming the liquid form, water promotes these processes as before. A slow setting has, however, been observed in frozen mixes (94), and it is quite possible that the phenomenon of regelation may account for this. Smoke gases have been found to have a disintegrating effect upon cement setting at a temperature lower than 7° C.; this is attributed to the formation at these temperatures of a hydrated calcium carbonate, having the formula CaCO,,5H,O by the action of the carbon dioxide of the smoke gases upon the lime of the cement. At slightly higher tem- peratures this hydrate is transformed to pulverulent calcium carbonate, with consequent disintegration of the structure of which it forms a part (107). The effects of moderate variations in the temperature of the mixing water upon ultimate strength are practically of no great moment; even mixes that have been frozen and afterwards allowed to resume their set are not materially affected in their ultimate strength, if the set has not proceeded too far at the time of freezing (11). More than one repetition of the freezing process upon the same mix, however, will be quite de- structive to the final hardening. If the hardening be considered a process of. crystallization, repeated freezing may be assumed to destroy the strength by the formation, through rapid temperature changes, of relatively small and non-adhesive crystals of the calcium hydroxide during the critical foundation period of growth of the crystalline structure, so impeding and misdirecting consequent interlocking that a weak structure results. If, on the other hand, the colloidal theory is adhered to, it is only necessary to point out that the colloidal cell walls about the cement grains may be ruptured by the expansion of the contained water in freez- ing. This would result in discontinuity of the internal structure, and if sufficiently widespread, as would be the case in repeated freezings, would alone account for weakness. Studies have been made of the ultimate resistance obtained from frozen mortars by varying the amount of gaging water, with the view of estab- PAOCINI, METAMORPHISM OF PORTLAND CEMENT 171 lishing whether “wet” or “dry” mixes best resist the disruptive effects of frost during setting. The results reported are discordant. An excess of water has been found by one investigator to enhance the effects of frost (85), while by another it has been found to diminish them (11). Theo- retically, the disruptive effects of freezing should be enhanced by the presence in the mass of larger quantities of gaging water. On the other hand, it can be assumed from the colloidal standpoint that an increase in the amount of water present will result in the formation of a greater quantity of colloids and a greater elasticity of the resulting mass, to- gether with a smaller total breakage of cell-wall material. TEMPERATURE OF THE WATER THAT MAY SUBSEQUENTLY COME INTO CONTACT WITH THE SYSTEM The action of hot and boiling water upon set cement is strongly marked in the case of cement which contains free lime, producing after a few hours, swelling, distortion and cracking and even total disintegra- tion. A normal cement so treated, however, preserves its original form and volume after short periods of exposure to the boiling temperature. The viscosity of water at high temperatures is greatly diminished, and the liquid is thereby enabled to penetrate more rapidly the capillary and subeapillary voids, thus reaching more quickly a larger internal area. If, as in the case of an unsound cement, free lime is thereby reached, this is slaked much sooner than it would be under normal conditions, and moreover with great violence, owing to the higher temperature of the water, producing internal disruption, and perhaps thus opening up fur- ther avenues to the penetration of water, with a repetition of the slaking process. The boiling test here described is a very important one in the testing of cement for construction, but it is perhaps less reliable in the case of unsoundness from the presence of excess of magnesia. In cements stored in waters of relatively high temperature, it is prob- able that the processes of solution act more rapidly, from the two reasons mentioned above; but evidence is lacking to show that any significant decrease in ultimate strength is thereby occasioned. Data as to the storing of cement in waters of low temperature, yet not subjected to the action of frost, are not available in the literature, but they would be interesting. In the case of exposure to the action of frost, the process is quite similar to that which goes on in the disintegration of natural rocks and depends, in like manner, upon the initial mechanical resistance of the 172 ANNALS NEW YORK ACADEMY OF SCIENCES mass, upon the total volume of the voids and upon the ratio of capillary to subcapillary voids. The disruptive effect is, of course, due to the ex- pansion of the water during freezing. Consequently there is a possibility that during the earlier stages of the history of the mass this effect may be to a great extent neutralized by the presence of soft colloidal material (45), because of its lack of rigidity. Voids are undoubtedly present even in neat cement mixes, and they are more common in mortars and in concretes; when, therefore, these have attained a sufficient hardness, they are in all respects similar to a natural rock and subject to the same katamorphic processes. The effect of frost increases in intensity as the mass ages and loses elasticity. As water permeates the cement, even after hardening has progressed to a considerable extent, it becomes charged with various electrolytes, and its freezing point is consequently lowered. ‘To some extent this immu- nizes the mass from frost action. On the other hand, as we have seen before, cryohydric compounds may be formed at these low temperatures, and the separation of these from solution is a factor in the opposite direction. QUANTITY OF WATER AT FIRST ADDED Size of cement particles—tThe finest particles in cement, provided that they are chemically identical with the remainder, are the most active cementitiously, because of the ease of reaction and of the greater proba- bility of this action being uniform throughout the mass of each particle. ‘This is recognized under the microscope by the ultimate disappearance of these particles as individuals upon the addition of water. Owing to the relative insolubility of the constituents of cement, both before and after metamorphism, each particle becomes covered to a certain depth with the reaction products, which in this case take the shape of gelatinous films (2) in such manner as to offer hindrance to the further action of “water. The particles whose diameter is smaller than or equal to the thickness of this zone evidently are the most efficient chemically. The larger par- ticles are less so, as the passage of water through the enveloping film is a slow matter, and some particles may be so large as to remain internally unchanged. It is probably this fact that gives a hydraulic quality to previously set cement that has been reground and retempered with water ; in fact, this process may be repeated a number of times with the same sample of cement. Not all of each particle, therefore, can take part in the setting and hardening, and sometimes this proportion of inert material is consider- PACINI, METAMORPHISM OF PORTLAND CEMENT 173 able (86). The coarser particles are comparatively inert and might be replaced by grains of foreign material of the same size without ma- terially influencing the ultimate strength of the resulting mass. This has been demonstrated experimentally (17). It does not follow, how- ever, that a cement consisting entirely of uniformly very fine particles would be a desideratum, since such a cement would not pack as well as one containing a greater variety of sizes, and the increase in chemical activity would be markedly overbalanced by the imperfection of struc- ture of the mass. Considering each particle to be spherical, and of equal size with every other, when packed in the most compact manner possible the pore space would be nearly 26 per cent (89). The points of contact of the adjacent spheres, notwithstanding the tendency of the gelatinous envelope to spread, would be relatively few. If, however, this pore space were filled with finer material, the structure would develop more strength, The function of part of the cement is to remain passive and to add to the strength of the structure merely by its action of void-filling. Extremely fine grinding has been found to decrease the ultimate strength, if the cement is used neat, but to give greater strength, if the cement is used in a sand mortar (62). As might be expected from the above considerations, the fineness of erinding has an accelerating effect upon setting. Cement ground in a tube mill until only 1 per cent remained on a sieve having 5000 meshes per sq. cm., was so quick setting that it could not be restrained even by the addition of 10 per cent of gypsum (47). When cement is relatively coarsely ground, the ultimate strength is not so quickly attained, but its acquisition is regular and uniform! Laitance.—In concrete construction under water, especially salt water, there gathers about the freshly deposited concrete a milky white cloud of suspended matter, technically known as laitance. This material is also formed when concrete is mixed very wet, though not deposited under water. An analysis of laitance by the writer, made for the Board of Water Supply, practically coincides with an analysis made by Richardson (97) and leads to the same conclusion as that reached by him; namely, that laitance represents an actual loss of cement and consists of the finest par- ticles of cement which have been washed out of the concrete. The addi- tional conclusion is justified that this portion of the cement, by reason of the small size of its units, has been so acted upon by an excess of water that it has undergone complete hydrolytic decomposition, before the col- loidal enveloping film had adsorbed sufficient electrolytes to completely coagulate it and so render it largely impermeable. This is substantiated by the fact that laitance possesses neither setting nor hardening qualities. 174 ANNALS NEW YORK ACADEMY OF SCIENCES Hydrolysis theory.—The formation of such a protective film upon the surface of a coarse particle will so regulate the access of water to its interior that the contents will be slowly and normally hydrated. If the entire mass of the particle were at once accessible to an excess of water, the weakly acid and basic compounds at first formed would soon be hy- drolysed and shorn of their binding power, and instead of the normal complex colloids described by Michaelis (64, 65), capable of adsorbing electrolytes and so coagulating into a dense rigid mass, simpler colloids such as hydrous silicic acid and aluminic hydroxide would form, which have not these powers to so high a degree. Finally, the rate of setting and hardening of a cement may be con- sidered a function of the proportion of fine particles present. Mortars set and harden more slowly than neat cement, and concretes more slowly than either. This is simply a development of the fact that coarsely ground cement sets and hardens more slowly than that which is finely ground. It may be considered, from another viewpoint, that the inactive material interferes with the liberation of heat from the system, and that chemical reaction is consequently delayed in proportion to the amount of inert material present. Mechanical agitation when water is added.—lf cement in the state of a plastic mass be worked and kneaded, the ultimate strength will benefit thereby, up to a maximum time of working. It is legitimate, a priori, to surmise that the setting is hastened, within limits, although no record of this is found. After the maximum time referred to, which in experiments made at the Board of Water Supply laboratory has been found to correspond roughly with the time of initial set, continued working will cause a fall- ing off in the strength. Up to this time, mechanical agitation with the proper amount of gaging water will cause an inerease in the ultimate strength. The formation of the crystalline network, which constitutes the setting of cement, and which is responsible for the primary strength by holding the plastic mass rigid and in place, while the more important elements of hardening make their appearance, is unquestionably facilitated by agitation. Stirring is a means of hastening chemical reactions by bring- ing the agents into more intimate contact. The compounds that go to make up this network, being sooner brought into solution, perform their function more quickly, and the crystals begin to form. Instead, however, of forming a continuous rigid network, the crystals will be smaller and less cohesive than if undisturbed in their growth, and the set can be delayed and even prevented by continuing the agitation long enough. PACINI, METAMORPHISM OF PORTLAND CEMENT 175 The ultimate resistance of cement which has been thus treated is decreased as well. The formation of the coagulated colloid, or of the interlocking crystal units, whichever may be the cause of hardening, is rendered imperfect and discontinuous, and the structure reflects the weakness of its component units. It may moreover be supposed that more cement has been brought within the range of hydrolysis by this agitation, and so converted into - laitance, even the larger particles being stripped of their protecting films by the attrition, Tests made at the Watertown Arsenal (36) showed that after one hour’s working, cement had gained 4 per cent over the normal strength, but that after 10 hours’ working, it had lost 24 per cent from the normal, in 20 hours 38 per cent, in 50 hours 56 per cent and in 100 hours 69 per cent. Total quantity of water at first added.—Under certain conditions, the entire range of particles of a cement might be destructively hydrolysed, resulting in what is termed “drowned” cement.’ The effect of an increase in the quantity of mixing water is known to result in a diminution of strength, and, bearing in mind what has been previously said regarding hydrolysis, the reason is clear. If, before the cementing of contiguous particles, an excessive amount of water is admitted to contact with the cement, colloidal material will form in increased amount: It has been shown that an increased amount of mixing water results in an increased volume of the paste produced (39). This indicates that a larger amount of the products of hydrolysis is formed. Owing to difference in composition between these hydrogels and those formed under normal conditions, they are incapable, as has been before observed, of adsorbing electrolytes in such degree as to attain to the density and rigidity of the latter, Admitting, on the other hand, that colloids so formed do not differ in composition from those formed in the normal hardening of cement, there still remains the abnormality of the structure formed in this way. Being discontinuous, it would not offer the same total resistance, in the form of connected films, to the passage of water. Moreover, in the presence of an excess of water the working ratio of electrolytes to colloids would be less because of the greater dilu- tion in proportion to the volume of colloid. QUANTITY OF WATER THAT MAY SUBSEQUENTLY COME INTO CONTACT WITH THE SYSTEM The effect of water upon cement after it has completely set rapidly diminishes to a negligible quantity at ordinary temperatures, if the water is reasonably free from dissolved or suspended impurities. There is a 176 ANNALS NEW YORK ACADEMY OF SCIENCES leaching out of calcium hydroxide from the mass of the cement; but this diminishes as the mass grows more and more impermeable, by the coagu- lation of the colloidal cell walls and by the carbonation or other precipi- tation of lime salts in the pores. This deposition of lime salts in the pores is evidently the cause of higher strength in specimens which are allowed to dry out a few hours before testing. It is analogous to the higher strength developed by sea- soned stone than by freshly quarried stone, occasioned by the evaporation of the “quarry sap.” In addition, the carbon dioxide conveyed to the material in a gaseous form is absorbed by the hme and may be considered a positive factor towards strength, while that conveyed in solution (where the cement is under water) is a negative factor, in that it accelerates the solvent effect of the water coming into contact with the cement. On the other hand, cement specimens which are entirely air-hardened are un- questionably weaker, by reason of the absence through evaporation of the requisite amount of water for proper hydration. When the action of water upon set cement is intermittent, the solvent effect manifests itself by unsightly incrustations and discolorations (3), caused by dissolved material brought to the surface through capillary action and there deposited by evaporation. When the mass is perma- nently under water, these salts are merely washed away. The danger from these incrustations, although slight, is the disintegrating effect pro- duced by their increase in volume, through crystallization or efflorescence, and the consequent disruption of the denser surface skin, rendering easier the action of frost upon the entire mass. This surface skin is improved by troweling the semen while in a plastic state, and consists of a closely packed layer of fine particles, which offers high resistance to permeation by water and comparative immunity from the solvent action favored by a rough, porous or fractured surface. If the mass be placed in water before setting, it is more lable to hy- drolysis, as evidenced by the copious formation of laitance; and if greatly exposed, as by agitation under water, it may fail to develop the greater portion of its normal ultimate strength. To prevent this, care is taken, in laying conerete under water, so to convey it that it offers the least possible surface to water action during its descent; and to this end it is. either lowered in cloth bags, or filled in through a chute, so as to escape all avoidable exposure to hydrolysis. If the water which comes in contact with a cement structure be under considerable pressure, so that its tendency is to percolate through the mass, the solvent effects will of course be magnified, proportionally to the porosity of the mix; and experiments made by the Board of Water Supply Ss " PACINI, METAMORPHISM OF PORTLAND CEMENT 177 have shown that concrete subjected to such percolation has been shorn of the major portion of its ultimate strength. In this case, the solvent effect of the water is only part of the influence at work, purely me- chanical factors entering largely into the destructive process, as will be shown later. Stalactitic growths of lime salts form as the result of water percolating through concrete. Micro-organisms of the algal type frequently lodge in the pores of concrete and by their growth may act as a protective influ- ence against the permeation of water. The effect of their products of metabolism and decay upon the concrete structure has not been studied. Numerous waterproofing materials and processes have been devised (40, 73). They may be grouped conveniently under three heads. Surface treatments.—The application to the surface of concrete of a coating similar to a paint has the disadvantage that concrete is not a thoroughly dry material. Where the vehicle is a liquid immiscible with water, the paint will not therefore come into contact with the concrete proper. If the vehicle is miscible with water, unless insoluble products are at once formed by reaction with the constituents of cement, the active agent is quickly leached out. Membranes.—These are layers of waterproof tissue interposed between two layers of the concrete. There is strong probability that these never actually form a bond with the concrete, and thus they necessarily intro- duce an element of weakness and heterogeneity. Mass treatments——The active material is incorporated with the con- crete at the time of mixing, either by dissolving or suspending in the gaging water, or by intimately mixing with the cement or sand. These treatments are many and differ widely in the agents employed. Sub- stances of a waxy or fatty nature, triturated to a great fineness, are the most generally offered, but the incorporation of these in a mass of con- crete is generally followed by weakness of the structure. The general problem of cement waterproofing has been conceded to be simply a ques- tion of void-filling, yet this must be accomplished without the addition of inert material that will weaken the resulting structure. The addition of more colloidal material has been suggested. This is ingeniously effected in a recent process by the use of hydrolysed cement, obtained by treating cement with an excess of water (99). The paste so obtained is added to the cement during mixing. The still unclarified state of our knowledge of the chemistry of the setting and hardening of cement is the great handicap which has thus far prevented the devising of a satisfactory waterproofing agent. A large number of the waterproofing preparations on the market are therefore 178 ANNALS NEW YORK ACADEMY OF SCIENCES purely empirical, and not applicable to the practical waterproofing of large masses of constantly wet concrete. In the interests of efficiency, it is probably more economical to expend money destined for waterproof- ing in the purchase of additional cement to be used in making a richer concrete. QUALITY OF WATER AT FIRST ADDED Having material in solution.—On adding water to cement, heat is evolved, the temperature of the mix rising in some cases to above the boiling point of water. It is the custom to look with suspicion upon cements in which an excessive rise of temperature is obtained, as being liable to develop unsoundness. The abnormal rise is attributed in some instances to the presence of free lime, in others to an insufficient propor- tion of lime. The volume changes caused by a rise in temperature have ‘been given as the reason of the difficulty encountered in joining fresh ‘cement surfaces to old, causing weakness at the plane of juncture, the ‘contraction of the mass on cooling breaking the joint before it has devel- oped sufficient strength to resist the strain. To prevent this, it has been suggested to coat the surface to which fresh cement is to be appled with a retempered mortar; that is, with a cement which has been treated with water after partial setting. This provides an intermediate course of material in which the temperature changes are not so rapid, and upon this course the fresh cement mixture is applied (35). Upon the same principle may be explained the use, for a fresh course of cement which is to be joined to some which has previously set, of mixing water in which a quantity of cement has been stirred, thus retard- ing the chemical reaction and consequent temperature changes. In both cases, the active water is already charged with the soluble portion of cement, its solvent power for the same material is thereby diminished and the chemical action moderated, so that heat is more gradually evolved and violent expansions and contractions avoided. The influence of dissolved electrolytes in mixing water has received much careful study. Through the addition of a small percentage of some soluble salt to the mixing water, many have tried to influence the properties of the completed structure and to produce a mass that would develop greater strength or a higher degree of imperviousness. Unfor- tunately, the panacea has not as yet been discovered that is suitable for practical application. The addition, similarly, of a soluble powder incorporated in the mass of the cement comes under the same category. In this connection, our PACINI, METAMORPHISHM OF PORTLAND CEMENT 179 attention is drawn to the effect of the usual addition of ground gypsum or of plaster of Paris to the ground clinker, for the purpose of retarding the set. There are other salts whose retarding influence on the set of ground clinker is comparable and probably superior to that of gypsum, but their use is not so practical, consequently, it has been adopted as the restrainer for general use. Tt has been shown by Rohland (83) that the salts which respectively accelerate and retard the setting of cement are the same as those which accelerate and retard the hydration of quicklime. From this it is con- eluded that their influence is “catalytic.” A detailed explanation of the mechanism of the action of gypsum has been put forth (79), holding that the presence of calcium ions in the mixing water, resulting from the solution of gypsum therein, decreases the solution of other calcium ions, thus retarding the solution of hme and the hydrolysis of the aluminates, which in turn retards the set. It seems probable, upon this basis, that the presence of certain elec- trolytes in the mixing water acts upon the set by influencing the solu- bility of calcium sulphate therein, and consequently increasing or dimin- ishing the number of calcium ions present in the mixing water as a result of the solution of calcium sulphate. For example, sea water has been found to retard the set of cement (83). Gypsum, although a relatively insoluble salt, may be regarded as fairly soluble in moderately strong solutions of sodium chloride or of other salts having no common ion (14). In the presence of sodium chloride, then, the: calcium ion concentration in the mixing water is raised, and the solution of the calcium aluminates diminished, with the effect of retarding the set. Sulphates have been found, when dissolved in the mixing water, to have the property of retarding the set, with the exception of aluminum sulphate and calcium sulphate when in low con- centration. In view of the latter fact, it is evident that the above expla- nation is perhaps only a partial one. A large number of other electrolytes and miscellaneous compounds have been investigated and the results are recorded (83). The effect of soluble constituents in the sand used for making concrete is by no means negligible (4) and may offer an explanation for many instances of puzzling behavior of the mixture. Sea water has been and is, in many instances, still used for mixing con- crete, and to the best of our knowledge, no cases of failure can be attrib- uted to this cause alone. Apart from the influence upon setting, the presence of dissolved electrolytes in the mixing water seems to increase the strength of cement in the early periods, as far as reported results have 180 ANNALS NEW YORK ACADEMY OF SCIENCES shown (4). This may perhaps be due to an increase of coagulation of the colloidal constituents, by reason of the presence of salts of greater ionization than are generally present. On the basis of the crystallization theory, this phenomenon is rather difficult to interpret. QUALITY OF WATER THAT MAY SUBSEQUENTLY COME INTO CONTACT WITH THE SYSTEM Having material in solution.—A large number of failures in concrete structures have been attributed to the disintegrating action thereon of water impregnated with various salts. Inasmuch as all ground water is charged to some degree with salts which it has accumulated in its passage through the soil and rocks, this problem is worthy of the most careful attention. For our purpose, such mineral-laden waters may be divided into 1. Sea water 2. Alkali water (from western alkali soils) 3. Deep rock waters. The mineral constituents are common in all these cases, and vary only in the prominence of one or more of them. ‘Thus in sea water the chlo- rides of sodium and magnesium, in alkali water the alkaline carbonates, and in deep rock water the chlorides of calcium and magnesium and the sulphate of magnesium are the distinctive constituents. Whether the effect of these electrolytes is cumulative, so that the continued action of solutions of low concentrations will work harm, or if not, what are the limiting concentrations to assure safety to the structure, has not been worked out. Obviously, it is not a laboratory problem, since the factors which obtain in nature are impossible to duplicate on a small scale. The solution lies in careful inquiry into the mechanism of the action and in observation of the instances of failure in construction work, with a study of its causes. Sea water.—The effects of sea water upon set cement have been sum- marized in the statement by Feret, “No cement has yet been found which presents absolute security against the decomposing action of sea water” (97). Le Chatelier, after a series of experiments extending over ten years, confirms this conclusion (53). Poulsen concludes, however, that the chemical action of salt water is not alone sufficient to cause Portland cement mortars to deteriorate (76). The diversity of results reported in the observation of the action of sea water upon cement indicates that there are varying factors at work that so far have not been clearly recognized. Whether the precise nature PAOINI, METAMORPHISM OF PORTLAND CEMENT 181 of the action is physical or chemical is not quite settled. There are not lacking investigators who assert that the destructive action is mostly physical and is due, among other causes, to intermittent submergence and consequent deposition, by evaporation of crystals in the pores of the structure, which, either by their pressure of formation or by expansion during efflorescence, have a disruptive effect similar to that of frost (98). There are those who hold that the action is entirely physical, and is due to this factor and the effects of frost (91, 102), although probably the latter is seldom the case in sea water, owing to its low freezing point {50). The effect of direct sunshine has been found deleterious when alternating with that of tidal action (20). Undoubtedly, all of these factors contribute to the total effect, and there is as well a marked chemical action. The chemical effects of sea water upon cemient are capable of various interpretations. They are summarized as the formation of complexes by the action of the dissolved sulphates and chlorides in the water upon the calcium silicates and aluminates of the cement (74). It has been stated that sodium chloride solutions have the power of dissolving calcium sili- cate with the formation of an unknown salt (58, 70), and also that the sodium chloride enters into combination in the mass, the chlorine ion entering into the combination calcium chloro-aluminate, and the sodium ion combining with lime, silica and alumina, to form compounds of the nature of the zeolites. Working with strong solutions of the individual salts of sea water, it has been found that the chief harmful constituent is magnesium sulphate, and it has been suggested that this salt reacts with the lime of the cement to form calcium sulphate and magnesium hydroxide. The calcium sul- phate further reacts with calcium aluminate to form a calcium sulpho- aluminate, which by swelling causes the disruption of the mass. The magnesium hydroxide formed has been regarded as a restraining agent, by virtue of its filling up the pores of the cement and preventing further ingress of sea water (70). Again, the disruption has been directly at- tributed to the increase of volume caused by the formation of this mag- nesium hydroxide (46). It has been calculated that, apart from the formation of hypothetical sulpho-aluminates, a molecularly equivalent amount of calcium sulphate replacing the calcium hydroxide of the ce- ment occupies 2.08 times as much space and is, therefore, the cause of the disintegration (13). Alkali and deep rock waters——Burke and Pinckney (13) have formu- lated a working theory of the action of the various salts common to all natural waters, They attribute the disruptive action to more rapid re- 182 ANNALS NEW YORK ACADEMY OF SCIENCES moval of the calcium hydroxide, and in some cases to its replacement by material occupying greater volume, as before shown, and consequent disintegration of the structure. That some such reactions occur is indubitable, and that the mechanical factors are a large influence in the disintegration is equally certain. An additional cause which may be of great importance has hitherto been neglected. The electrolytes in these natural waters may act as acceler- ators of hydrolysis, and, in effect, cement which is in contact with sea water is subject to the same action as that of an excess of water from any cause. By the presence of these electrolytes the hydrolysis of a larger proportion of the cement is effected; and the results are increase in the volume of the hydrolysed portion, and production of a larger proportion of inert colloids. It has been found that a larger amount of cement can- be converted into colloidal matter by the presence of an electrolyte in the water with which it is treated (99), and also that the speed of hydra- tion of cement is affected by the presence and proportion of electrolytes present (84). The fact that a larger amount of laitance appears to be formed in sea-water construction also seems to bear out this theory. Besides the reactions mentioned, set cement is subject to the replace- ment of silicic acid by carbonic acid, as are the natural rocks. Especially is: this true in cases where the cement comes into contact with marsh and peaty waters and waters containing ferrous carbonate, which by transfor- mation to the hydroxide liberates carbon dioxide (24), which has been found to act, not only upon the calcium hydroxide but also upon the silicates and aluminates (28). The presence of free acids in water which acts upon the cement is quite destructive, in proportion to the concentration of the acid and to its strength or weakness as an acid. It is quite probable, however, that the liberation of colloidal silica by the action of acids would serve to a great extent as a protective influence against their further action. Sewage gases are generally effective by reason of the hydrogen sulphide which they contain. This gas is readily oxidized to sulphuric acid, and then its action is the production of soluble calcium and aluminum sul- phates, which are subsequently leached out from the mass. This action has been found greatest at the surface of the liquid (106). Hydrogen sulphide may also act by converting the iron of the cement into sulphide, and this becomes oxidized into ferrous sulphate and is leached out, or by its expansion causes disruption (28). The action of many other inorganic and organic solutions has been observed, but they do not come within the scope of this paper, since they are not met with in natural processes. PACINI, METAMORPHISM OF PORTLAND CEMENT 183 In general, the consideration is worthy of attention whether concrete structures which are under stress are not more lable to chemical disin- tegration than those which are in repose, or whether a single structure is not more liable to this action in its strained parts than in those not so affected. We have data to show that strained iron is more liable to corro- sion than unstrained, and it has been asserted that strained minerals are more acted upon by underground solutions (104). A number of protective measures against the action of saline waters upon concrete have been suggested and tried, but none has been so strik- ingly effective as to achieve universal recognition. The simplest remedy suggested is to make the concrete for such uses denser and more imper- vious by the employment of a greater proportion of cement, yet this may not always be practicable. When concrete is exposed to the gases result- ing from the decomposition of sewage, it is suggested that even such a proceeding may be of no avail (29). Previous air-hardening of the concrete before laying under sea water is acquiring more widespread use and is highly recommended (87). The cause of its protective action is attributed to the carbonation of the calcium hydroxide (48). Variations in the fineness of grinding and in the chemical composition of the cement used in concrete for sea-water construction have been pro-- posed. The French specifications for sea-water cements call for a finer grinding than that which is required for ordinary construction. Much has been claimed regarding the resistance to disintegration offered by the so-called “iron ore” cement, which contains a minimum of alumina, this being almost entirely replaced by iron. Having material in suspension.—The peculiar nature of the series of compounds forming and formed from cement, in that they are all of - relatively low solubility, tends, as has been before observed, to retard the reactions which may occur. Mechanical agitation, by promoting diffu- sion and by transporting the reacting materials to their possible spheres of action, will accelerate these reactions. The motion of water, per se, can and does produce this effect, and when the water is armed with sus- pended material, its activity in this direction is greatly enhanced. Where water has immediate access only to the outer surface of a mass of set cement and its pressure is low, the effect is a slow corrasion of the dense surface skin and ultimate removal thereof, rendering the interior gradually more accessible. Ordinarily, this process is a slow one, al- though under certain conditions, as in coast protection works where the velocity of the water is high and the suspended material coarse and plentiful, the destructive effects are more to be reckoned with. 184 ANNALS NEW YORK ACADEMY OF SCIENCES The effects from less spectacular processes are quite surprising. Where the pressure of the water is such that there is a marked motion of the water within the pores of the concrete, the erosion is internal and far more insidious. In this case, the suspended material is part of the struc- ture itself. Small particles of cement or, in the case of mortar, grains of sand which become detached from the parent mass are whirled around by the water stream and shortly enlarge the cavity in which they are rotating, until it merges with some adjacent cavity. Under favorable conditions this process may continue until the interior of the structure is greatly weakened. A factor which to some extent neutralizes the flow of water through concrete is the choking of the pores by sediment, coming from the water itself or furnished by the action of the water upon the concrete. If the flow is oscillatory, as in concrete exposed to the range of the tides, this protective effect will of course not be so marked (54). Diatoms and other microscopic marine organisms with siliceous or cal- careous tests undoubtedly play an extensive part in the preliminary- stages of this internal mechanical action, by choking the capillary spaces. At the same time, undoubtedly, the organic debris thus introduced may by its decomposition give rise to substances, carbon dioxide and hydrogen sulphide, for example, which have an accelerating action upon the proc- esses of solution, and the silting effect may thus be neutralized or even overbalanced. PART II EXPERIMENTAL INVESTIGATION In Part I, the ways in which water may influence the metamorphism of Portland cement were discussed qualitatively, and their possible effects upon the permanence of the structure of which cement forms the basis were pointed out. This question has now assumed economic and vital importance. eS In the following pages experimental data are offered, in elaboration of the outline laid down in the first portion of the paper. Points in the scheme which have been established beyond doubt by previous investi- gators are here omitted, and only such results are inserted as have been deemed necessary as additional evidence. The last division of the out- line, treating of the action of suspended material in water in effecting the erosion of concrete, has not been experimented upon, not having come within the scope of the writer’s activities, and therefore is omitted. PACINI, METAMORPHISM OF PORTLAND CEMENT 185 Other divisions have already been so thoroughly covered by previous in- vestigators that very little remains to be said about them. Emphasis has therefore been laid in this paper upon the little known fields. The problems which confront the user of concrete are of a high order of complexity. The generalizations of chemistry are not yet sufficiently developed to apply rigidly to systems of so many variables, and experi- mental work on a laboratory scale often fails almost entirely to reproduce the conditions of practice. The best guide to the truth, then, is the prag- matic sanction of experience—the investigator in this field can but point out probable directions for future experimentation. The theories which underlie past success are a safe guide, nevertheless, to future construc- tion, and the systematization thereof is a legitimate field of usefulness. While, strictly speaking, any aggregation of chemical compounds might be considered a rock, whether natural or artificial, a majority of the cases conceivable under such a classification would not present im- portant petrological problems in the study of their metamorphism. Such a problem as the action of water upon a mixture of sodium chloride and calcium sulphate can be partly solved in vitro, even though the action of sea water upon gypsum deposits is an interesting petrological investi- gation. The important components of Portland cement are everywhere about us in nature, and the reactions by which it is made artificially have been taking place for many geological ages without the intervention of man. Silica, alumina and lime are among the most important constituents of the earth’s crust; they are subjected in places to the same conditions that exist in the kiln, and are afterwards acted upon by water, under some of the same conditions under which man builds massive structures. The complex question of the history of rock magmas is not one to be solved by any one group of scientists, but by patient and concerted efforts of the chemist, the physicist and, above all, the petrologist. So the prob- lem of the constitution of Portland cement may be as yet somewhat inde- terminate; but an examination of the more general effects of metamor- phism may reveal some identity with conditions in natural rocks already studied and may direct us to the correct methods for investigation of the constitution of cement (67). Other important problems in the field of cement and concrete are re- ferred to in the following pages, and belong in great measure to the field of petrology. Not the least important of these is the suitability of vari- ous types of rocks for use as aggregates in concrete, and this work is claiming more widespread attention daily (19, 44, 111). ANNALS NEW YORK ACADEMY OF SCIENCES jt CO (op) TEMPERATURE OF THE WATER AT FIRST ADDED Two standard cements were gaged with the requisite quantity of mix- ing water for each at different temperatures. The effect upon the time of initial and final set was noted, as follows: TABLE 1 Effect of Temperature of Gaging Water on Time of Initial and Final Set Per cent by weight Temperature of Initial set, Final set, of mixing water mixing water hours hours A B A and B A B AM | B 29 21 70° F. 4.2% | 4.50 | 627) |leueae 22 2] 100° F. 1.50 4.00 4.00 7.00 22 Pall 150° F. 0.33 3.75 0.50 5.75 22 21 Alm e 1.00 2.79 2.75 | 6.00 The results seem to indicate that interference of hydrolytic decompo- sition with the setting appears between 150° F. and the boiling point of water. Below these limits, the effect of increase of temperature of the - mixing water, as is well known, is to increase the speed of setting (31). The setting time at these temperatures is a resultant of two opposed processes,—the formation of the water crystalline network, and the de- structive hydrolytic action of water upon the original constituents of the cement, resulting in a product which has no hydraulic qualities. Where the second process overbalances the first is the point at which the speed of setting ceases to increase and begins to diminish. This is true of course of the stage known technically as the final set (9). In the first few hours of setting, there is a period of relaxation, which McKenna has aptly termed reverse set, and which he has been able to detect with precision by means of an ingenious chronographie apparatus of his invention (60). ‘The phenomenon has been observed by the writer and his associates in the laboratory of the Board of Water Sup- ply, using the Vicat needle; but this apparatus does not lend itself to a scientific study of the finer differences in rigidity which occur during the setting period. McKenna’s apparatus should throw a great deal of light upon the initial metamorphism of cement. ; TEMPERATURE OF THE WATER THAT MAY SUBSEQUENTLY COME INTO CONTACT WITH THE SYSTEM High pressure steam—Wig (109) has recently presented an account of the excellent effects of high pressure steam when used in curing con- te _ PACINI, METAMORPHISM OF PORTLAND CEMENT 187 crete. He found that by using concrete that had attained its initial set and exposing it to steam at 80 pounds pressure the six months’ strength could be obtained in two days, a tremendous accelerating of the harden- ing process. This state of affairs is not very satisfactorily explained, if the harden- ing of cement is supposed to be due to the progressive crystallization of calcium hydroxide, since it is somewhat at variance with our knowledge of the conditions of crystallization to assert that continuous exposure to a high temperature, presumably constant, should accelerate crystalliza- tion ; particularly since in this case the amount of water present in the system remains the same. On the basis of the colloid theory, however, it is simply explained by supposing that adsorption of calcium hydroxide by the complex hydrogel is accelerated by higher temperatures. Cold storage.—A series of tests, embracing neat cements and mortars, was made upon tensile test specimens exposed, after the age of 24 hours, to low temperatures under diverse conditions. The following conditions were observed : 1. Chilling the briquettes at 24 hours’ age by filling the storage tank with water at the lowest winter temperature as it came from the tap. The water was then allowed to come slowly to normal winter temperature for the tank, about 60° F. 2. Chilling another set of specimens, otherwise normally treated, by filling the tank with cold water as before, 24 hours before breaking. 3. Storing another set in ice water for the entire period after remov- ing from the damp closet at 24 hours’ age. 4. Normal treatment. Two brands of well-known cement of high quality were run in parallel. The mortars were of proportions 1:3, Ottawa sand being used. The results obtained are summarized below: 188 ANNALS NEW YORK ACADEMY OF SCIENCES TABLE 2 Effect of Cold Storage on Strength ngt ounds % ature of Storage “per sauuare meen ree ae Number of Cement Mix Blorsee PaothiGd Species deg. F. 7 days | 28 days | 7 days | 28 days x Neat 60 Normal 730 739 0 0 10-12 43. 1 606 693 18 6 12-12 42 4 654 739 11 0 12-12 60 Normal 702 745 0 0 24-24 34 3 638 669 9 11 24-23 xX 23} 60 Normal 300 361 0 0 11-12 43 1 281 361 6 0 11-11 42 2 283 371 6 +3 11-11 60 Normal 313 408 0 0 24-24 34 3 262 312 16 24 22-24 nya Neat 60 Normal 628 843 0 0 12-12 43 ] 650 770 +3 9 12-12 44 2 649 872 43} +4 12-12 60 Normal. 628 697 0 0 24-24 34 3 530 622 16 11 22-24 Y 1:3 60 Normal 250 300 0 0 12-12 43 1 253 370 +1 +6" 12-11 44 2 287 327 +15 +7 12-12 60 Normal 228 317 ) 0 22-24 34 3 197 234 14 26 22-24 From these results, it is safe to conclude that, aside from the effects of frost, low temperatures are adverse to the development of the hardening process in cement, and that in general this effect is more pronounced in mortars than in neat cement. The adsorption of calcium hydroxide by the complex hydrogel may proceed at a lower rate at lower temperatures; or if this is not so, the primary hydration, of which this hydrogel is the product, may proceed more slowly, and thus less of the hydrogel be produced,—either of which processes will detract from the hydraulic activities of the mass. It would » seem from the experiments that the latter is the more satisfactory expla- nation, since the test specimens which were chilled at first and allowed to return to normal temperature show a tendency to return to normal strength at the longer periods, while the general tendency in the series kept constantly in cold water is to fall further off from the normal, indi- cating only a limited available amount of hydrogel to undergo the coagu- jating process. PACINI, METAMORPHISM OF PORTLAND CEMENT 189 The effect of sudden chilling at a period when a large proportion of the strength is already developed does not show any decided direction, both the positive and negative variations from the normal averaging the same. It may therefore be concluded that, for the temperatures studied, a chill- ing of this kind has no significant effect. An explanation according to the crystallization theory of hardening would fail to fit the facts so satisfactorily. In the specimens that were chilled at first and allowed to return to normal temperature, there should be under this hypothesis a more significant decrease of strength, owing to the formation of small, non-cohesive crystals from the rapid tempera- ture change. The return to normal conditions should not favor so nearly complete a recuperation as has been noted; unless a re-solution of the erystals and recrystallization were supposed, in which case it may be argued that such a process would require an abnormal solubility of small crystals when compared with large. In a normal specimen, re-solution and recrystallization are undoubtedly going on, strengthening the struc- ture, and the large crystals are growing at the expense of the small. If small crystals preponderate at seven days’ age, resulting in a weak mass, it is necessary to postulate a comparatively high solubility of the small erystals in order to arrive at a normal strength at 28 days. This, while by no means impossible, is not probable. Turning to the specimens kept continuously in cold water, it would seem that, although the first chilling should show severe effects, as it did, there should not be such a falling off in the rate of hardening, if the crystallization be progressive. It is quite possible, however, that crys- tallization at this temperature is not favored, and that the total number of binding crystals of calcium hydroxide is therefore less than at normal temperatures. QUANTITY OF WATER AT FIRST ADDED Size of cement particles——Other factors being equal, the amount of cement rendered inert by the action of water is proportional to the per- centage of fine particles. This is an absolute condition and presupposes free access of water to every particle. Needless to say, in practice this condition is seldom realized, except approximately in laying concrete under water, or in the careless use of an excess of water in mixing, or in protracted mixing. In the use of a very fine cement, then, if the proper proportion of water is added, the mixing time carefully regulated and proper precau- tions taken in depositing, the influence of texture upon the strength of the mass occasioned by the action of water is reduced to a small quantity, 190 ANNALS NEW YORK ACADEMY OF SCIENCES by virtue of the greater hydraulic activity of the fine particles, increasing the impermeability, as will be shown, and the confining therefore of the action of the excess water to a narrow zone. The bulk of the cement will be properly hydrated in spite of the fineness. The investigation of the effect of the size of particles due to the action of water thereon alone is not feasible, because no satisfactory measure of laitance formation, except the strength of the mass, has been devised. The measure of the strength would be unsatisfactory, since the propor- tion of fine particles affects the strength in other ways than through the formation of laitance, as has been pointed out in a previous communica- tion. From a study of the hydraulic properties of reground cement, Spack- mann and Lesley conclude (93) that only the very fine flour in cement, that portion not measured by the present tests using sieves, reacts when gaged with water and gives strength. It is difficult, of course, to draw a sharp dividing line between active and inactive material in cement, al- though it must be admitted that the greater part of the coarse material, even though it be of the same chemical composition as the fine, has little or no cementing value and serves mainly as a filler. Suitable fractional separation of the portion of cement passing the 200 sieve, by air-elutriation or other method, should with careful study be a valuable guide to the most efficient mechanical composition. Experi- ments upon the first method of separation are recorded by Peterson (71), and a scientific method of fractional elutriation using an inactive liquid has been worked out by Thompson (100). Much should be gained by the application and development of these methods. The influence of the size of particles of inert material added to the cement is also of great consequence, and a proper mechanical grading of the sand used in mor- tars is recognized as vital. The presence of clay in this sand, or the addition of clay alone to cement, come under this category, and have occasioned a great deal of discussion (8, 32, 33, 110). A comparison was made of the permeability of 1:4 mortar of Portland cement, when used in its ordinary condition, and when screened through a number 200 sieve. PACINI, METAMORPHISM OF PORTLAND CEMENT 191 TABLE 3 Permeability of 2-inch Cubes, Age 28 Days, Subjected to 80 lbs. Pressure Grams of water passing ‘Temperature per hour Nuniboase Cement of water. — testa Deg. Fahrenheit Unscreened Sereened EN Se Cee See eae RE 66 68 22 By) 5, 6 (SRA cere aoe 68 68 25 2 6, 6 Oc tte, ae eae 68 68 29 Trace 6, 6 Dre 68 68 331 81 5, 6 15r , Serge ey eae eee ee 64 64 rl Trace 5, 6 Ry las: 64 64 31 2 6, 6 CG ja ae 64 64 5 0 3, 6 Heys as. ote SES 68 72 6 Trace 5, 6 Me ye, 68 68 71 0 6, 6 ANY CLE s Goan OM TEC Oa een omer 61 13 The marked decrease in permeability resulting from the use of finer cement in mortar demonstrates that in impermeability, as in strength, the finest particles are the most active factors. Mechanical agitation when water 1s added.—Increased working should weaken a cement after a certain maximum point is passed. In order to establish this point, the effect of prolonged working was investigated. It was necessary to use a mix of fluid consistency, in which, for obvious reasons, the final set would not under normal conditions take place dur- ing the time over which the experiments were extended. Two grouts were employed: one in which cement was mixed with 50 per cent of its weight of water, and one in which an equal weight of water was used. The different tests were run respectively for periods from one minute to five hours, and they were mixed in a motor-driven stirring machine of the type common in chemical laboratories. After the stated period of stirring, the grouts were poured into glass tubes and kept in a damp closet for the twenty-eight-day period. Cylin- ders exactly two diameters high were cut from the specimens and crushed in the.compressing machine, two cylinders being crushed for each period, and the average of the compressive strengths being recorded. 192 ANNALS NEW YORK ACADEMY OF SCIENCES TABLE 4 Twenty-eight-day Tests of Grouts Mixed for Varying Lengths of Time Compressive strength, pounds per square inch, average Duration of mixing ss See ce EUs 50 perct grout /100 per ct.grout URSIMITNONC vest et or tess ses. cides ence ease 5240 3095 ND TNUTNUTTE Sh. ss o-< ccvsypato se eas eos 9045 4725 POMMMTUMULESe sels se nea gee See: 9710 4955 APO OUI eestis See ease ata mele tonne eerie 5879 4840 POIVOUNS a narcon Eeoiane ae ree ee eee 6075 4320 ENO UNE Sreatenciecveteec aisaeiceee Rare eae 4775 4538 The effect of mechanical agitation, when thus prolonged, is equivalent to that of the use of excess water—the strength of the cement is pro- gressively diminished as the working proceeds. It is noteworthy that the effect is only reached after a certain optimum period is passed. Be- fore this time, increased working increases the strength. We may con- clude that there occurs within this period a process which neutralizes the effect of hydrolysis; and this process is probably the formation of the network which constitutes the setting. As will be seen later, the effect of excess water is to reduce the ulti- mate strength. ‘The effect, then, of mechanical agitation must be to bring more cement into contact with water and, therefore, to increase hydrolysis. This is probably accomplished by stripping off the protective film of gelatinous material which envelops each cement particle when it comes into contact with water, which film regulates the hydration of cement and causes it to proceed in a regular manner. This film being stripped off, the cement is subject to the destructive action of hydrolysis. Where more water is originally present, the destructive action 1s sooner attained, as will be seen by comparing the 100 per cent grout with the 50 per cent. Evidently, the setting process proceeds best at high concen- trations, when the amount of water is low. This may be so regulated that the setting process will not take place at all, by using a large excess of water and much mechanical agitation, as has been repeatedly observed. by the writer. : Setting time of cement in laboratory ar and im damp closet.—The standard specifications for setting-time tests call for storing the specimen in the damp closet, whereas the tests as generally conducted in most laboratories are made in the open laboratory air. A series of experi- ments was made, for the purpose of noting the deviation from standard results caused by this departure from the rule. PACINI, METAMORPHISM OF PORTLAND CEMENT 193 TABLE 5 Setting Time in Laboratory Air and in Damp Closet Time of set in minutes Cement Laboratory air Damp closet Initial Final Initial Final RS ie oak See eet ang 255 37d 300 435 VEILS, See oid ates eas ane ae 120 360 300 450 EOE. ieee a as ae ca 300 420 360 480 “Soe atie eee eee 240 360 285 420 NA ssmita rene dco ntct ; 240 390 250 450 From these results, it will be seen that setting in a relatively dry atmosphere takes place in a shorter time than in a damp one; also that the setting time is more uniform under conditions of high atmospheric humidity. At the same temperature, evaporation takes place more rapidly in the former case; and allowing a cement mix to stand in such a position that evaporation of the mixing water may readily take place is practically equivalent to the use of an insufficient amount of mixing water. Effect of excess of mixing water on strength of concrete.—Concrete is often mixed so wet that, as it is filled into forms to a depth of several feet, the water rises above the concrete and throws out considerable lai- tance from the cement. The ease of mixing and placing very wet con- erete is the constant incentive for its use. This practice, however, is followed by a great deal of deterioration of the concrete in strength. The strength rapidly decreases with the increase in the quantity of _ water used in mixing. The visible effect of this weakening is the forma- tion of laitance, which has little or no setting power or strength, and which represents the loss of an active part of the cement, since, as is recognized, the finer parts are more hydraulically active. Tests were made by mixing concrete at normal consistency and shovel- ing one-half the batch into a tank containing three to four inches of water, the depth of concrete being about four inches. The water rose to about an equal depth above the concrete. In test No. 1, the concrete was allowed to settle in water four inches in depth for 30 minutes, when the excess of water was siphoned off and the remaining material poured into molds. In test No. 2, the depth of water in the tank was three inches, and the water was siphoned off immediately while in agitation. In test No. 3, the same process was repeated, except that the depth of the water 194 ANNALS NEW YORK ACADEMY OF SCIENCES was four inches, and the concrete used was somewhat leaner. Test No. 4 represents the direct qualitative effect of the addition of an excess quan- tity of mixing water without subsequent handling. All specimens were cylinders six inches in diameter and 12 inches high. The remainder of the batch of concrete in each case was poured directly into molds, and the specimens were broken at 28 days. The amount of cement lost was roughly ascertained where possible by filtering the siphoned water and weighing the amount retained on the filter. TABLE 6 Effect of Excess of Mixing Water on Strength of Concrete Strength at 28 No. of . Per cent Test No. = Proportions days, pounds specimens of water per sq. in. Per cent of cement lost 1240 760 8.2 8.2 8.2 8.2 8.2 8.2 8.2 0.3 »Specimens shoveled into water as described. Evidently, then, the mere presence of an excess of water is sufficient to produce the weakening effect, independently of any actual removal of - cement from the concrete. As may be seen from Nos. 1 to 3, the leaner mixes suffer the greater deterioration in strength. Effect of eacess of mixing water on permeability of concrete-—A par- allel series of tests upon the permeability of concrete treated with an excess of water was made, in which the correspondingly numbered speci- mens were treated in the same manner. The cylinders cast from these batches were eight inches in diameter and six inches in length, and were cased in the standard manner for permeability tests. Three specimens were made for each test, and at the age of 28 days were submitted first to 40 pounds pressure for one hour, then to 80 pounds for one hour, without interruption. The flow recorded is in grams passing during the last ten minutes of test. PAOCINI, METAMORPHISM OF PORTLAND CEMENT 195 TABLE 7 Effect of Excess of Mixing Water on Permeability of Concrete Gaps passing in last mn en minutes * Test No. Proportions REE cont of A aareoiine water 40 pounds 80 pounds Wend 24 8.2 67° F. 0 0 1 he 74 24 Sie 2ea ania e lercsperceors 479 456 B 2 24 8.2 58° F. 0 0 Be IZ 24 Sere Minne er alae 212 588 3 1: 2.33 25 8.2 56° F. 0 alt ole 1:2.33 :5. 8.2 60° F. 1814 Not tested 4 1 a jam aaa 8.2 67° F. 38 18 4 ieee 24 MOSSE aes [Utes temas carer 26 80 8 Specimens shoveled into water as described above. In the foregoing experiments, the decrease in strength and water- tightness may be referred to the deteriorating influence of excess water upon the cement (16). It may of course be argued that the more marked effects obtained in series 1, 2 and 3 than in series 4 are due to the method of making the tests; that is, that a considerable proportion of the active cement was actually removed from the body of the concrete by siphoning off the supernatant water with its laitance. Effect of excess of mixing water on the strength of neat cement.— With the idea in mind that the weakening effect was independent of the removal of cement (1), a further series of tests was instituted, using a neat cement of good quality. The cement was poured into a series of glass tubes in which increasing proportions of water had been put, the tests representing a series of grouts mixed respectively with 50, 75, 100, 150, 200 and 500 per cent by weight of cement of water. The tubes were shaken for one hour and then allowed to stand for 28 days. The cement settled into the bottom of the tubes in the order of its coarseness, the fine nebulous laitance settling last as a cheesy white layer of increasing thick- ness, as the percentage of water was higher. This layer was carefully trimmed off in preparing the test specimens. On breaking out the cylinders from the tubes at the end of the test period, it was decided to cut each cylinder into two, each exactly one diameter high, carefully noting the respective position of each in the tube. On submitting these to compression it was seen that the direction of difference between the upper and lower layers was not constant, nor 196 ANNALS NEW YORK ACADEMY OF SCIENCES — was the difference a significant one, so that it was considered wo to average the strengths. It will be seen by the table below that, even without actual removal of any cement, the formation of laitance has a weakening action ge cement. TABLE 8 Compressive Strength of Grouts Mixed with Varying Proportions of Water Crushing strength, Per cent of water eee pees ineh, Age, 28 days 50 6855 79 5900 100 4500 150 3430 200 2960 500 1810 The effect of excess of mixing water is therefore seen to result in decrease of strength as the water increases. Whether the effect is a permanent one was the next question that presented itself. 'To settle this point, a new series was undertaken, in which a larger number of differing percentages was introduced, and in which the resulting strength at two periods was determined. The cement was mixed with the stated percentage of water, and worked for two minutes, the drier mixes upon the table in the usual fashion, and the wetter mixes merely poured into the tubes and shaken. Paper mailing tubes were used, 2 inches by 48 inches, treated with molten paraffin and sealed with paraffined corks, so as to be absolutely tight. To obviate the effect of possible leakage, the whole series was stored in damp sand. Cylinders two diameters high were cut from the specimens at the stated periods, each cylinder being cut as nearly as possible the same dis- tance from the bottom, and care was taken to avoid including any of the soft cheesy top portion, the settled laitance. PACINI, METAMORPHISM OF PORTLAND CEMENT — 197 TABLE 9 Compressive Strength of Grouts Mixed with Varying Proportions of Water, Over Extended Period (Hach result is the average strength of three specimens. ) Compressive strength, pounds per : ; Percentage of Sy aera Per cent gain in water strength over 28 days 28 days 3 months 22 7076 7504 6 25 6174 9402 —13 30 4563 6030 32 38 3992 5059 27 50 2991 9312 77 79 2113 4078 93 100 1609 3044 120 150 1270 2379 87 200 1306 2579 97 500 399 1141 186 It is apparent from these figures that the effect of hydrolysis upon the strength of cement is a reversible one, at least to a certain extent, since the specimens in which an excess of water was used in mixing showed a greater recuperative ability at the longer period than the cement in which the normal amount of mixing water, in this case 22 per cent, was used. Upon inspection, it was observed that the three months’ specimens showed in each case much less laitance than the similar 28 days’ speci- mens had shown, and it was considered probable that the laitance, in standing, had adsorbed free lime from the remainder of the cement, through the activity of the water permeating the mass, and thus reverted to the original condition of the cement, or an approach thereto. An analysis was accordingly made of laitance scraped off from the top of one of the 500 per cent water specimens and thoroughly washed by decanta- tion. It probably represents a maximum condition in the hydrolysis of cement. TABLE 10 Analysis of Laitance from 500 per cent Specimen As obtained from | Treated with lime specimen water SiO ea See hon suet oat 15.28 15.91 © CRO Brae sere decease wvevaie a els, okt oneivove 2.28 2.42 PACE) raps cea, estat pcre ccdaiesie si 3.98 9. 82 (OPO) oR ahi yn al rt 26.96 - 86.67 LIVE Oe i cae ct crate ala este oy oie 2.86 1.28 SOM rics civscen vavsmicrs ie aikee chat 6.47 Ween COM TCO; Chile even sch whe os 42.17 35.18 198 ANNALS NEW YORK ACADEMY OF SCIENCES The normal ratio of silica to lime in unset cement may be considered 1 to 2.82. In this material we find the ratio 1 to 1.76. This indicates a great loss of lime; and it was thought possible, that, by adsorption of lime, this laitance might regain at least a part of its hydraulic proper- ties. Accordingly it was digested for several days with lime water at laboratory temperature, filtered off, carefully washed with distilled water and dried, as was the previous sample, at 100° C. An analysis showed the results tabulated in the second column. The ratio of Si0, to CaO had changed to 1: 2.30. Besides direct metathetical reactions between the components of ce- ment and the water solution which always surrounds a mass of hardening cement, adsorption of various materials from this solution is unques- tionably always going on. Were the fine particles of cement inert chem- ically, this would still take place, by virtue of the enormous total surface which they must present. Clay, it has been demonstrated, has the prop- erty of adsorbing ions of CO, from solutions of carbonates, and of Cl from solutions of chlorides (10). The laitance then may, by adsorption of calcium hydroxide given off from the cement adjacent to it, recover some of the lime lost by it. Whether the lime adsorbed restores the original status of constitution is of course mere speculation. The trend of the strength tests shows that this is probably not so, but that the adsorption is not entirely a reversion of the hydrolytic reaction; in other words, that “drowned” cement will probably never recover and attain to the strength it would have had with proper hydration. | Effect of the presence of clay and dissolved substances.—It is apparent that if the decreased strength be directly referable to the action of the excess water upon the cement, any means of preventing the access of excess water should prevent, if only to a degree, the destructive action. The colloidal nature of clay (6) has been utilized in the water-proofing of concrete, the principle of its action being the formation of continuous gelatinous films throughout the structure, which prevent the passage of water. Although the same problem is not presented in a grout that exists in finished concrete, it is probable that some blanketing action might occur upon the addition of clay to the mixed mass. The point was investigated. To correct for the effect of absorption of part of the mixing water by the admixed clay, a consistency test was made upon a sample of cement to which 10 per cent of clay had been added, and it was found to require 4 per cent more water than the same cement used neat. : The clay mixes were accordingly gaged with 4 per cent more water PACINI, METAMORPHISM OF PORTLAND CEMENT 199 than the corresponding neat cement mixes, and the following series of compressive tests was made: TABLE 11 Effect of Clay upon Destructive Action of ‘Eacess of Mixing Water (Average of two tests at 28 days) Neat cement Sere ace Compressive Compressive Water, ; Water, Beef ange ence ECE a gpereent (| ete ete poulndl 50 5782 54 1282 75 3134 79 1328 100 2273 104 2577 150 1896 154 2156 200 1381 204 1320 900 514 504 No strength developed If the action of saline solutions upon cement is to accelerate the hy- drolysis of the latter, it would appear that the destructive action of excess. water would be accelerated by the presence:therein of saline substances. in solution; also, it is legitimate to expect that the addition of clay restraining the hydrolysis due to excess water will in this case exert a similar influence. The following experiments, parallel to the foregoing ones, elaborate this point : TABLE 12 Effect of Clay upon accelerated destructive Action of Mixing Water Containing 5 per cent of Magnesium Sulphate (Average of two tests at 28 days) Cement, 10 per cent of which was Neaticemenp replaced by a fat clay (dried) 5 per cent solution Compressive 5 percent solution Compressive of magnesium strength, pounds of magnesium strength, pounds sulphate, per square sulphate, per square per cent inch per cent inch 50 2196 54 2774 75 548 79 1608 100 1512 104 No strength 150 556 154 a is 200 ' No strength 204 re es 500 No strength 504 os ee 200 ANNALS NEW YORK ACADEMY OF SCIENCES From these two series of experiments, it is qualitatively apparent that the presence of clay does prevent a certain amount of hydrolysis. From the first series, it is seen that this effect only begins to show itself as higher percentages of water are present, which would indicate that the clay may have taken up much more water than the constituency test revealed, and that, in the relatively drier mixes with clay, the cement suffered in strength because of insufficient water. On the other hand, experiments at this laboratory in which clay was used, replacing up to 10 per cent of cement in normally gaged material, showed that no signifi- cant decrease in strength was thereby obtained ; hence the loss in strength in the 54 and 79 per cent grouts cannot be due to this cause. _It is more probable that the colloidal nature of the added clay is brought into play more effectively at the concentrations in which in- creased strength is observed, and that the latter is due to the coagulation of the clay by electrolytes adsorbed at this optimum concentration. The same result would obtain where additional saline material has been added to the mixing water, as in the series where a 5 per cent solu- tion of magnesium sulphate was used. The clay here prevents the accel- eration of hydrolysis by the magnesium sulphate through adsorption of part thereof, and possibly by coagulating, forming an impenetrable bar- rier to the further action of water upon the remainder of the cement. QUANTITY OF WATER THAT MAY SUBSEQUENTLY COME INTO CONTACT WITH THE SYSTEM Permeability—The solvent effect of water coming into contact with cement structures is best studied by the permeability test. This consists in forcing water through a mortar or concrete at a known pressure and observing the amount of leakage through the specimen. In detail, the specimen is generally made up in the form of a cylinder, and this is cased with a thick coating of neat cement on all sides but the bottom. The water, under pressure, is applied on the full cross-section of the specimen and forced through, dripping from the bottom, whence it may be collected. ; With neat cement, of course, this method is inapplicable, because of the density of the material and the consequently enormous pressure nec- essary to force water through it, and moreover because of the mechanical difficulty in confining the water strictly to a passage through the speci- men. ‘The specimens tested, then, are lean mortars and concretes. Although this test is designed to ascertain the resistance which these materials offer to the flow of water, it is evident that this resistance is not a constant quantity in the case under consideration. PACINI, METAMORPHISM OF PORTLAND CEMENT 201 The temperature and pressure of the percolating water being constant, the flow is diminished by cementing and clogging, and increased by ero- sion and solution; the quantity of water flowing through the mortar or concrete therefore is a function of the balancing of these processes. . Cementing may result from deposition of material originally in solu- tion in the percolating water, or dissolved from one portion of the structure and deposited in another. Clogging, similarly, results from material originally in suspension in the percolating water, and deposited in the pores of the concrete, or from material eroded from one part of the mass, either mechanically or as a result of solution of the attacking portions, and deposited in another part. Erosion per se is a negligible factor; that is, the flow of pure water, carrying no suspended matter, will have very small mechanical effect upon an insoluble material. When the water is armed with suspended matter, however, its corrasive effects become proportionally magnified. Solution is the most important factor in the process of percolation. Following the order laid down by Van Hise for natural rocks (104, p. 536.), the basic materials removed are, firstly, the alkalies and, secondly, the alkaline earths, in the order calcium, magnesium. Since the alkalies exist In cement in the proportion of a little over one per cent and are not essential to the hydraulic properties or the strength, their solution is a matter of little consequence, except in that it may result in the for- mation of solutions which react upon the lime compounds and render their solution more easy of accomplishment. This reaction has been considered elsewhere. The removal of magnesium compounds proceeds at a lesser rate, although there is a greater percentage of them present ; and their removal, in the main, may be dismissed as insignificant. Since more than half the weight of fresh cement consists of lime, and since the strength of cement depends for the greater part upon calcium hydroxide, whether crystalline or adsorbed by colloids, the removal of calcium hydroxide from set cement is the factor of the greatest impor- tance. Considering its solubility in pure water, the reversion of the hydroxide to the crystalline form tends to diminish its solubility, or from the other standpoint, its adsorption by a colloid tends to remove it from the solvent action of water. Unfortunately, however, it must be borne in mind that without exception, cement structures are nowhere subject to the action of pure water alone. From rain water, with its appreciable burden of dissolved gases and atmospheric salts, to the water of the ocean and the more heavily laden rock and mine waters, concrete structures are everywhere in contact with saline solutions of varying concentrations. 202 _ ANNALS NEW YORK ACADEMY OF SCIENCES The effect of solution in percolation, then, is to a small degree de- pendent upon the solubility of the components in pure water. This effect diminishes as time goes on, because of the reversion of the soluble ma- . terial to a less soluble form and because of the protection afforded by the insoluble portions of the system decreasing the exposed area of soluble material. The washing away of these protecting films will of course neutralize the second factor. The increased solubility of the components of set cement in solutions of various electrolytes is the more important element in percolation. Even a very dilute solution may have tremen- dous total solvent power, when the time element is considered. In fact, it may be that the action of a dilute solution will on the whole exceed that of a concentrated solution, by reason of the greater cementing and choking action of the latter, tending to diminish the quantity of water that may come into contact with the soluble portions. A dilute solution, therefore, with its more insidious attack, is probably more to be feared in the end than the strong brine. Observation of the behavior of concretes and mortars during the per- meability tests gives a clue to the balancing of these processes, whether there is a preponderance of cementing and clogging on the one hand, or of solution and erosion on the other. Attempts were made, in the experi- ments noted below, to study chemically the reactions involved, by peri- odical analyses of the percolating water. To this end nearly four hun- dred complete analyses of the effluent water were made. Upon tabulation of these it was observed that any deductions based upon them would be inconclusive, as the chemical composition of the effluent water repre- sented one of a great number of variable factors that might occur at any point either within or without the concrete. The single qualitative generalization, that lime was removed from the cement at a diminishing rate, is the only permissible conclusion from the analytical data. The original purpose of these tests was to ascertain the suitability of various aggregates for use in concrete, with reference to their stability in the presence of percolating water. At the conclusion of the series, it was found that the effect of water upon the various aggregates was prac- tically negligible, during the period of observation, and that the action had been confined to the cement of the mortar. The aggregates had been protected from the action of water by the cement, it being probable, how- ever, that a continuation of the tests would have revealed the action of water upon these rocks, when the protective influence was removed. A series of sixteen aggregates was used, in as many concrete specimens. Since it is not the purpose of this report to discuss the relative suita- bility of these materials for concrete construction, but only to consider PACINI, METAMORPHISM OF PORTLAND CEMENT 203 the action of the water upon the cement, two cases alone will be con- sidered. The rock was crushed and screened for each experiment to the same average effective size, corresponding to the following mechanical analysis: TABLE 13 Mechanical Analysis of Aggregate used in Permeability Tests Sieve ee ve Per cent passing 1% 1.89 100 1% 1.58 94 1 1.02 a9 x .78 32 24, 59 21 7 .48 16 3 .30 6 4 .22 0 The sieve ratings are based on diameters of spheres of equivalent vol- ume to the largest sized stone particles that will pass. The fine aggregate was crushed quartz, the standard sand formerly used for cement testing, passing the No. 20 and retained on the No. 30 sieve. The cement used was a standard Portland of high quality. The specimens were made in the laboratory’s standard form for per- meability test, cylinders eight inches in diameter and six inches in length, the proportions used being 1:3.5:6, this being found the richest mix practicable to secure the porosity required for the test. They were cased in neat cement, and connected suitably for subjection to the pressure of the city’s water mains. Each specimen was protected from the direct flow of the water by a layer of one inch of clean coarse sand. The average pressure for the period of observation (52 weeks) was 22 pounds. The determinations of the rate of leakage were made weekly at first, and later every two weeks until the end of the test. The data appended below represent observations on the rate of percola- tion of water through two of the specimens which present the greatest interest from the standpoint of this paper, this flow being recorded in grams passing in ten minutes. The aggregate used in one specimen was a hardened neat cement, crushed to the size stated, and used in place of the rock generally employed in concrete. The parallel specimen selected for comparison was one in which the aggregate was a crushed granite, which showed a low solubility in hydrochloric acid (2.66 per cent dis- solved in one hour’s treatment with 1:1 HCl). 204 Temperature records of the percolating water were not kept, since these tests represent a part of a larger series in which this would have been impracticable. The other aggregates tested showed results from which it was quite difficult to draw any legitimate conclusions as to the relative suitability of different rocks in concrete subjected to these conditions. Concretes containing different aggregates.—A series of tests on con- cretes made up of different aggregates but with the same cements gave results which may be tabulated as follows: Flow in Grams of Water passing in 10 Minutes through Concrete Specimens TABLE 14 ANNALS NEW YORK ACADEMY OF SCIENCES subjected to continuous Water Pressure for 52 Weeks Pressure, Time pounds per square inch 24 hours 25 1 week 2 2 weeks 22 Ope 20 AT? 25 Satie 22 Ce =a 20 "/ ce 20) Sibi 25 LO: ae 25 Date 26 MA Sabte 24 Gy 24 alfeyen ee 20 20 17 2 aS 20 Pye ee 17 26ic0e- 22 DBR es 26 3007 = 22 Bede alr 20 Sy ie 19 ai 26 Sos 26 40 *“ 23 AO oe 25 44 * 22 Age 21 AG WIE 25 50 20 SDA Race 20 Month January.... February .. March..... August..... September. October.... November.. December. . January.... Grams passing in 10 minutes Conereve with Concrete with aggregate of erushed aggregate of crushed oat comonte| = Eee 2111 60 836 31 662 26 626 40 570 60 603 62 530 50 1295 38 1127 40 1310 45 870 25 997 36 985 28 973 32 639 20 792 40 731 36 802 43 800 49 781 46 763 50 © 115 Trace 105 2s 107 2} 110 3 93 oD, 75 9) 70 3 80 10 73 ee: 78 if PACINI, METAMORPHISM OF PORTLAND CEMENT 905 Comparison of these two sets of figures indicates that the cement of the concrete is more attacked than the aggregate. In fact, the flow obtained in this specimen was the highest but one of a series of sixteen, and the total lime content of the effluent water was also the highest but one. The visible effect upon examination of the interior of the specimens was a bleaching of the mortar, with evident solution of the cement. The original percentage of lime in the mortars was 12.8. Analysis of mortar from the granite specimen showed a content of 4.8 per cent, indicating that nearly two-thirds of the lime had been dissolved out. Further evi- dence of the loss of lime was found in the heavy white crust which formed on the exposed bottoms of the concrete specimens during the test. Small stalactites, quite soft to the touch, were abundant. The quantity of this deposit was not visibly different in the different tests. The calculated loss in lime of the mortar was greater than the loss computed from periodical chemical analyses of the effluent water, and this is due to the fact that much of the dissolved lime was deposited upon the bottoms of the specimens as the stalactitic growth above mentioned. There was no evidence that suspended impurities in the water had been carried into the interior of the concrete, and it is therefore supposed that the one-inch layer of sand by which the latter was screened from the direct flow of the water was an efficient filter for the purpose. The clogging action resulting from this source may therefore be dismissed as negligible. It may be concluded from these tests that concrete of this density tends to protect itself automatically from the action of percolating water, so that, for the period investigated at least, the flow tends to diminish to a minimum. The action of the water seems to be confined to the cement of the mortar, leaving the aggregate relatively unaffected. It is evident that, notwithstanding the utmost precaution in mixing concrete test specimens, wide differences in permeability may obtain in specimens mixed under the same conditions of handling and by the same workman, owing to structural differences in the resulting mass. How- ever, the results obtained are fairly comparable. _The most sensitive test for the internal changes which the concrete has undergone during percolation is the resulting strength of the concrete. Concretes containing different cements——A series of tests was under- taken in which the specimens were made up in the same proportions, 1: 2.5: 6, using in each specimen the same coarse aggregate, a crushed eranite, and the same fine aggregate, a standard quartz, but using differ- ent brands of cement. The specimens were stored in damp sand for a period of 28 days, then subjected to continuous water pressure of about 206 ANNALS NEW YORK ACADEMY OF SCIENCES 25 pounds for a period of 11 months. the specimens : : Percolation through Concrete Specimens Parallel specimens were stored in damp sand during this period and allowed to attain their full normal strength. The table following shows the leakage and final strength of TABLE 15 Months of percolation Brands of cement and grams of water passing in 10 minutes Compressive strength of specimens at the end of period Compressive strength of untreated specimens, pounds per square inch.. Loss of strength through percolation A B C D 1D) F 146 286 o 164 76 230 155 125 22 179 16 82 a6 70 90 167 11 85 37 47 52 161 11 82 U2 28 37 65 a 45 71 12 31 15 17 39 68 28 3 11 26 33 a7 46 14 6 16 21 40 43 atte 2 5) 11 18 oie 10 “5 a 14 8 13 10 ‘1 2 19 TABLE 16 Comparison of Strength before and after Permeability Test A B C D EK F avs ae cave eae ane 770"| 490 640 890 750 | 8590 1080 | 1210 | 1230 | 1125 1220 1090 29 60 A8 21 39 46 7 One specimen crushed. (DETZCemt hs eee are ee ener Other results are average of two specimens. Effect of the direction of flow through concrete.—Concrete seems to offer less resistance to the flow of water when the direction of the flow is parallel to the bed than when at right angles to it. A test covering this point was made with 8-inch cubes of concrete of the proportions 1: 4: 14, fine and coarse aggregate being a standard crushed bluestone. PACINI, METAMORPHISM OF PORTLAND CEMENT 207 TABLE 17 Rate of Flow in Gallons per Square Foot per Hour under 20-inch Head Age of specimens, 67 days. In specimen parallel to In specimen perpendicular Temperature of water, 64° F. bed to bed NSGe2emMIMUteS 2 sce se. oe 740.96 164.14 2d oes Ra ote ess ede ca" 585.28 159.54 3d TBE Teen A USI, 2 alee, 636.31 163.49 4th OM esr syt et Mer ca 539.53 158.3 | 5th USI Mea aa A Nt a 549.10 157.93 | Specimens immersed 24 hours, then retested : “US A ONO) Se Seee nes ee 665.38 182.46 2d CO AON th ied Ba We Oe Bee 642.77 177.54 3d AO TO tee EE eae Ne a 662.80 177.54 4th Dis SIT eer ame Reena 641.15 177.06 5th us eRe nner hs 659.57 173.67 In denser concretes, this effect was not found so marked. It will be noted that after storage following the first exposure to the effect of per- colating water, these specimens appear to offer less resistance to the flow of water. This may be due to the fact that in lean concretes the propor- tion of capillary and subcapillary voids is smaller and that of super- capillary voids greater, and that cementing and clogging actions, which have their greatest effect in capillary and subcapillary passages, are not so effective. The greater flow along the bedding planes has been observed in the case of rock, and is in all respects a phenomenon of the same nature. In the case of a stratified sandstone cited by King (51), the reason is ad- vanced that no more water can pass the more open layers, when advancing across the bedding planes, than was able to pass those of the closest tex- ture; whereas when the flow is along the bedding planes, each particular stratum carries water in proportion to the coarseness of its texture, uninfluenced by any other. In the case of water percolating into a concrete tunnel this would tend to emphasize lateral percolation, and in the case. of disintegration would exercise, in general, a localizing influence. It is not to be as- sumed that this is a rigid rule, inasmuch as a large number of factors, evidently, may neutralize this influence. From these considerations, it will be seen that the solvent effect of water upon set cement is of high importance in considering the perma- nence of concrete structures, and that this solvent effect tends to diminish as the set cement ages. This is not the only way, of course, that water 208 ' ANNALS NEW YORK ACADEMY OF SCIENCES may afterwards affect the metamorphism of cement. It has been pointed out by Goldbeck (43) and by White (108) that the expansion or con- traction of concrete depends upon whether the concrete remains wet or dry, and that the strains caused by alternate wetting and drying of con- crete may be a more fruitful cause of cracks than temperature changes. The presence of an optimum quantity of water is necessary, however, so that the proper reactions take place in the mass of setting cement, in order that the strength may increase normally. QUALITY OF WATER AT FIRST ADDED Compressive strengths of neat cements gaged with various solutions.— A normal Portland cement was mixed with the proper quantity of water (21 per cent by weight) in which was dissolved, in the different tests, varying concentrations of the salts indicated in the subjoined table. he cement was worked for one minute, and the plastic mass was tamped into glass cylinders approximately one inch in diameter, with the utmost precaution to avoid all air bubbles and at the same time to subject all specimens to the same pressure. TABLE 18 Compressive Strengths of Neat Cement Mixed with Solutions of Various Salts (Age of specimens, 28 days. Average of two determinations) Mikes Pounds per | Gain or loss, square ineh per cent il. Distilled qyater2 4-5 fo aerate ee eee eter 7330 2. 25% rock water® diluted with distilled water.... 6340 —14 3. 50% do. Ge 6495 —11 4. 75% do. ee, 6870 — 6 On AROck water alone: ...¢ ascertain 5605 —23 6. 2% sodium chioride.solution4.. 4.2. ee. eee ~ 6675 — 9 7. 4% GO. AL © a tgs Sag te Pees las eo ae ea 5815 —21- 8. 6% do. Fae ie Sar Aka eek Ree cite Ue Oma 5065 —31 9. 8% (a (Gea RNa Ca Ket Ree grees Fos eat pacchore es a 4215 —43 10. 10% Oe: Sepa linus Say ar outer eee Ce ee 5285 —29 1]. Saturated solution of calcium sulphate (+ 0.2%). 7025 —4 12. 0 2% solution of calcium chloride............... 6960 —5 13. 0.2% solution of magnesium sulphate............ 6680 — 9 14, 0.2% solution of magnesium chloride.. ane 59995. —23 15. Equal parts of 11 and 12 (CaSO, and CaCl, ae liad 6565 —10 16. Equal parts of 13 and 14 (MgSO, and MeCl, Nera 7355 +0.6 17. Equal parts of 12 and 13 (CaC), and MgSOQ,)..... 5810 —21 18. Equal parts of 11 and 14 (CaSO, and Mg(l,)..... 6200 —15 § This water contained: CaO, 1177 parts per million. MgO, 226 SO;, 408 Cl, 4360 PACINI, METAMORPHISM OF PORTLAND CEMENT 209 The glass cylinders containing the cement were then stored ina damp closet for 28 days, when the cylinders were broken out, and two speci- mens, each exactly one diameter high, cut from each cylinder. These were put into water for a few hours, so that they might be in the moist. state when crushed. The cylinders were kept in the damp closet instead of being stored under water, to avoid leaching out the salts contained in the mixing water, thus obtaining the maximum effect of the dissolved salts. It will be noted that there is a decided loss of strength in all but one ease (number 16). This particular case may be explained by the prob- able formation of an oxychloride, by the magnesium chloride and the magnesium hydroxide liberated by the action of the magnesium sulphate upon the calcium hydroxide of the cement. The oxychloride formed from these two materials has a tensile strength far superior to that of Portland cement itself, and its presence probably counteracted the de- structive action of the salts upon the cement. It is probable, however, that, at longer periods, this increase would disappear and become a de- crease. Otherwise, the presence of saline matter dissolved in the mixing water seems to have a decided deleterious effect upon the strength of cement. ‘This point is of marked importance in construction, inasmuch as the problem of mixing water is often solved by using the water nearest at hand, without inquiry into its qualities. Tt is the custom to specify that the water used in mixing concrete shall be free from oil, acid, strong alkalies or vegetable matter (77) ; but such _ a Specification does not cover the case in point, and the presence of large quantities of dissolved salts in water used for construction is easily over- looked. In concrete construction, it is of the utmost importance that the water which may be used in mixing be additionally subjected to such tests as will reveal either its mineral content or its action when mixed with cement and possible subsequent attack thereon. The action of sodium chloride appears to be nearly directly propor- tional to the amount employed. This salt is used in mixing water for construction carried on in cold weather, in order to prevent freezing of the deposited concrete. Its effect upon the strength of cement, if used in excessive quantities, is, as has been shown above, likely to become a seri- ous matter. Under the conditions of construction which generally pre- vail, however, much of the salt may be leached out of the mass. The results above represent a condition of maximum attack. Dieckmann (25) recommends the use of from 1 to 2.5 per cent of salt for concrete to be laid in cold weather, but states that percentages larger than this cause a marked decrease in the strength. ji) ANNALS NEW YORK ACADEMY OF SCIENCES. Effect of gaging with various solutions upon the strength of mortars afterward stored in water.—The above tests do not show, of course, a normal condition, since no water came into contact with the cement after it had set. Working with more porous material, a 1:3 mortar, so that in storage a heightened subsequent water action might take place, the following results were obtained : TABLE 19 Effect of various Salts dissolved in the Mixing Water, wpon the Strength of 1:3 Mortar (Sand, screened Cow Bay. Specimens stored in damp closet for 24 hours, then continuously in water for the rest of period) Compressive strength, pounds per square inch Number of specimens 7 days 28 days 3 months 815 1475 2600 1 1005 1185 1805 3 945 1310 2170 3 1010 1520 2420 3 885 1240 2100 i 910 1625 2145 2, 2 3 3 3 2 ~ ~ ~ ww we ewe wv ~ CD OD HE Od HI OD GD CO CO CD GO KD OO G9 9 OO bo bD OD 09 CO CD CO EI CD wv 865 1410 2115 935 1595 2710 ~ 2% (6 Co RCNA he. Va eM TEM os 1090 1580 2500 sie i ssolution Of BESO pa. s-eeseRe 930 1605 2670 135 2% OE 1 fo S20, Gee eee aioe 840 1480 2710 2.9. lZasolatonotNa@late.ce ere ener 1105 1385 2000 3,9, 2% GOs oA ee EE eee eae 1000 1035 1685 2,05 Tensile strength, pounds per square inch 7 days 28 days. | 3 months NUCH Ta" ores tie part pk Le ee ara ie Sin 179 272 326 5,6,6 1% solution of Al, (SO,),.......... 208 272 321 6,6,6 2% HOSS) 5S OSB Fa bea ete eae 193 262 340 4,6,5 1% solution of Na,SO,........ Gia 216 290 304 6,6,6 2% GOR oot adie are ee ene ears 205 300 343 4,6,6 1% solution of MgSO, ........2. 02 6: 194 260 317 3,6,6 2% CG LORIN Seca © pei gtaninn Tune 185 246 283 46,6 1% solution of ZnSO,.............. 126 263 315 2,0,6 2% QE! fact eee goer 205 272 319 3,6,6 1% solution of FeSO,.............- 201 266 314 5 ,6,6 2% (0 (On Celene Reena tc ARS 184 258 311 5,6,6 1% solution of NaCl............... 211 261 310 6,6,6 2% Os pra soe crease pee Weck dae 224 209 310 5,6,5 ~ ~ PACINI, METAMORPHISM OF PORTLAND CEMENT 911 The general conclusion that may be drawn from these values is that the effect of electrolytes in the mixing water, when the cement is after- wards subject to immersion in water, is to increase the strength at the early periods (7 and 28 days), but later to depress it (15). In general, the more concentrated solutions give a greater depression of strength. The early increase in strength is probably due, in the presence of an optimum quantity of water, to additional cementing or void-filling ma- terial precipitated in the pores of the mortar by reaction between the added electrolytes and the solutions resulting from the action of water upon cement. This deposited material may, in its later history, revert to a soluble form and be washed away, leaving abnormal voids, or else . in its growth may disrupt the cells it occupies, in either case reducing the strength. Effect of gaging grout with rock waters.—In grouting deep tunnels, the question has arisen as to the advisability of using the rock water at hand when fresh water was inaccessible. The water available in the instance in hand was an effluent from a shale bearing a small proportion of pyrites, and when it issued from the rock face it contained a quantity of dissolved hydrogen sulphide. As none of the water was immediately available for a laboratory test, an artificial mixture was made up, in which the quantities of dissolved salts and hydrogen sulphide occurring in the natural water was purposely exaggerated, to obtain accelerated effects. TABLE 20 Analysis of the Artificially Mineralized. Water Parts per million IETS iptrstevrate vatettostsna etait soderter eels titotes toys Rosh rateita tal Me ltite 891 AOI Gi ee trate ttta ater RR ORE I 1764 IM EXO) Sala pick o Oa OiGs 1050S COSMOS CRS areca sc 1461 S Olpreyseot oa cronnnesr dare atie dias cvcrinseisie ts auselaa 1948 (Ol aes saa encieites Sr itire ens ogee pe eRe eae aaa Ard 2920 A grout was made up according to specifications, using a normal Port- land cement, and Cow Bay sand with 100 per cent passing 10 sieve, 75 per cent passing 40 sieve; in the proportions 1: 1144 with 35 per cent of liquid. The wet mix was poured into glass cylinders, kept 24 hours in air until set had developed and immersed in water. Four sets of three specimens each were made, the first set mixed with 35 per cent of distilled water; the second, 35 per cent of the water above mentioned ; the third, 35 per cent of a 10 per cent dilution of this water, and the fourth, 35 per cent of a 1 per cent dilution. No discrepancy was observed in the setting time, as all the specimens 912 ANNALS NEW YORK ACADEMY OF SCIENCES developed a fair set within 24 hours. The grouts mixed with the undi- luted sulphide water turned a dark green, but otherwise no change was noticed in these or any other specimens. Three cylinders one diameter high were cut from each set of specimens, and, after storing 28 days in distilled water, were crushed. TABLE 21 Compressive Strength of Grout Mixed with Different Proportions of Water Containing Hydrogen Sulphide (Average of three specimens, age 28 days) Pounds per Mixes square inch DD USEMN OG: WiabOr 55.005 Soe wed vo, cen Sas Gr wale airatotlinte eilote cneiles She tatielte een ne 1424 Undiluted sulphide wateric)j.4.c. acc 0s Soe oes Se ae 1608 10 per cent of sulphide water, 99 per cent of distilled water.... 208 1 per cent of sulphide water, 99 per cent of distilled water...... 1110 Apparently, considering the average of the last three values, water of this composition will have no evil effect at 28 days upon the grout with which it is gaged. Three series of tests were undertaken, in which a 1:3 mortar of Ot- tawa sand and a cement of good quality was mixed with Croton water, and with two typical rock waters encountered in tunnel work. TABLE 22 Analyses of Rock Waters Parts per million E W COP T 0 RRR RTH ei ee eae ere ea) 85 943 INES Oto Si 2 spe SES ete rea eee cera eeeapeate 159 156 SOs Saar eve oi Sie antatoea os tag atl ncean cobayar ie ans Ree yeep 73 172 IER Ae Sasori pote ear ep oe sae Smee ee URNS 1380 3420 Total solids ..... Pe ral EA ce rae gM Do ge . 2978 7929 The normal amount of water was used to mix the mortars in each case, and the briquettes were stored in the damp closet over the stated periods. TABLE 23 Tensile Strength of 1:3 Mortars Mixed with Various Saline Waters Pounds per square inch Number of Mixed with— — | specimens in 7 days 28 days 3 months phe Crotomuwaters cn) aoe a eer 302 322 344 6, 5, 6 Water (HS ctu cre Mere Denia 297 343 363 6, 6, 6 Water sob ics tc). et yok Bese 296 335 383 6, 6,6 PAOINI, METAMORPHISM OF PORTLAND CEMENT 213 As was found in the case of the grouts last mentioned, waters of this general concentration do not appear to affect the strength of cement mortars with which they are gaged, and the probabilities are that no serious effects will result from this cause alone. QUALITY OF WATER THAT MAY SUBSEQUENTLY COME INTO CONTACT WITH THE SYSTEM Theoretical considerations —The action of dissolved salts in water that comes into contact with concrete, where such action is deleterious to the concrete, has been carefully studied by a large number of investi- gators (68, 81, 96, 112). Of the salts which have been found injurious, magnesium sulphate and magnesium chloride seem to have the greatest effects. What concentration of dissolved salts is necessary in order that disintegrating effects shall manifest themselves cannot be definitely stated. This is a field problem and is subject to wide variations under different conditions. A water containing relatively little dissolved material, acting under favorable conditions of porosity, pressure and wide temperature changes upon one concrete, may accomplish failure of the structure; while another water, of high saline content, meeting a dense, impervious con- erete, not forced through the mass by pressure and under conditions of small temperature change, may have practically no action. Manifestly, unless these varying conditions are taken into account, it is unscientific to draw any conclusions regarding the attack of different waters or the resistivity of different cements. It may be laid down as a basic principle, however, that the denser a concrete, other conditions being equal, the greater its resistance to the attack of saline waters (10, 41, 57). The alkali waters of the Western states have given a great deal of trouble in concrete construction. Most experimenters conclude that their action upon concrete is in the main mechanical and due to the disruptive force of crystallizing or efflorescing salts deposited in the pores by intermittent submergence and drying out (30, 38, 49, 56). ; Of course, as has been pointed out, action of this sort is not confined to concrete, and any material of construction possessing porosity is lable to a similar disintegration. The remedy, therefore, is to prevent the penetration of the saline solutions by the employment of courses of permanent, impenetrable materials, preferably asphaltic layers. Where the attack is not mechanical but chemical, this remedy is also applicable. Unfortunately, there are examples of construction which are exceptions, and, in these, some change in the chemical or mechanical 214 ANNALS NEW YORK ACADEMY OF SCIENCES constitution of the cement is the only way to prevent decomposition. In concrete block construction, where the blocks may be made long before they are actually put into the structure, it is found of great advantage to allow them to harden in air or in damp sand, and so permit to a great extent the carbonation of the lime compounds. Some investigators claim excellent results from this method (41, 55). As to the modifications in the constitution of the cement that will combat the action of saline solutions, there is a great disparity of opin- ion, which possibly is based upon lack of standardization of experimental conditions. It is generally conceded that high silica cements are best suited for the purpose (7). The use of puzzolan cements, or of addi- tions of puzzolan to the cement in use, is also well recommended (7, 37, 66); and the addition of clay, burnt or dehydrated, finds favor with some (7,75). As to the lime content of the cement, opinions are divided whether it should be high (5, 41) or low (92). Cement of greater density (57) and cement ground to a greater fine- ness than usual (72) are favorably commented upon. The subject, because of its great complexity and because of the questionable value of laboratory results, is at present in a chaotic state. The length of time that must elapse before judgment may be passed upon the permanence of a material under these conditions and the corresponding newness of the field of Portland cement render present conclusions largely a matter of speculation. Effect of storage in various saline solutions upon the strength of mortar.—In order to study the relative resistance to saline solutions offered by cements varying in chemical composition and in fineness of grinding, a series of 132 2-inch mortar cubes was made up, in the pro- pertion of 1:3, with standard Ottawa sand, the cements used being A. A high silica cement B. A low silica cement C. A cement of ordinary composition, sifted and remixed so that 98.8 per cent passed the 100 mesh sieve and 88.6 per cent passed the 200 mesh sieve D. The same cement as C sifted so that 92 per cent passed the 100 sieve and 75 per cent passed the 200 sieve ee PACINI, METAMORPHISM OF PORTLAND CEMENT 915 TABLE 24 Analyses of the Cements Used in Tests with Saline Solutions Per cent A B Cc SlOke casera ie ie ee as 23.50 19.74 22.99 GLO)! sereye tha let anes 2.36 2.75 2.42 I O)s eats ener asi tees 7.28 8.77 6.79 JEG) se apa aa ee ean Bae 62 18 60.86 60.84 Ce (De Ne ee ee 2.29 2.86 4.14 SOS REESE A aoe ne ean Meelelt 1.39 1.76 CO,H,O, alkalies...... 1.28 3.63 1.06 The cubes were stored 24 hours in the damp closet, and then trans- ferred to the solutions mentioned in the following table, three cubes to each liquid, and there stored for three months, then broken. n TABLE 25 Compressive Strength of Mortars Stored for Three Months in Various Saline Solutions (Hach value is the average of three determinations) Pounds per square inch Storage medium Easy é ' 5 High aa Low ae Finely wey Coarsely a silica Gani silica ont ground cent | ground een) Croton water:..| 2217 |...... PAGI. Naa < CARI: 3: ercrete 2066 Sodium 5%| 2267 2 2090 —1d 3266 53 2273 10 sulphate, 10%| 3264 47 2035 —18 2223 4 2262 9 Magnesium 5%| 3244 46 | 1787 | —28 | 2233 5 | 2759 33 sulphate, 10%| 2604 18 | 2646 7 | 3003 42 | 2489 20 Sodium 5%| 2365 7 1785 —28 2305 8 2695 30 chloride, 10%| 1778 —20 2019 —18 2968 40 2044 —1l Magnesium 5%| 2331 Sale 1827 26) | 273i 33 | 2305 12 chloride, 10%| 1757 —21 1769 —28 | 2570 20 2269 10 Calcium 5%| 2653 19 2516? 2 2219 4 1808 —13 chloride, 10%| 2224 0 1994 —19 2042 —4 2238 8 Average gain — sos, (percent) 2.4) --.- LOH ele eel eae CG Cie ae 12 9 Average of two determinations. 916 ANNALS NEW YORK ACADEMY OF SCIENCES The general deductions from these experiments for the period covered are that the high silica cement, notwithstanding its slower rate of har- dening, resists the action of these dissolved salts better than the low silica cement, and the finely ground cement better than the coarsely ground. Moreover, with the concentrations used, the stronger solutions in nearly every case had a more destructive effect upon the strength of the mortar than the weaker. | The strengths here obtained by storage in salt solutions are in general decidedly greater than those obtained by storage in fresh water. Hx- amination of the cubes, when removed from the solutions at the end of the test period, revealed under a lens that the exterior was being at- tacked, minute pittings being quite distinct. The strength attained by these specimens may be considered as a re- sultant of the balancing of two effects: the deposition of crystallized or precipitated material in the voids, which by packing the spaces with solids will increase the compressive strength; the creation of additional voids by direct solution or by the disruptive effect of metathetically pro- duced material. It is probable that the disintegrating effect for these concentrations is reached considerably beyond three months’ exposure. From the increases in the compressive strength, it is likely that at this period a great deal of crystallization or precipitation has proceeded, overbalancing in the main the disruptive effects. This is a general deduction, and single instances are notable in which the reverse holds good. In the case of the finely ground cement, the density of the mortar made therefrom has prevented the disruptive effect to a greater degree; and thus the deposition, while not necessarily as much as in the coarser cement mortars, has had a more marked effect in increasing the strength. Effect of storage in rock water upon the strength of lean cement mortars.—A series of briquettes of 1:4 Ottawa sand mortars was made up, using a normal Portland cement of high quality. The mix was made. lean purposely to accelerate whatever disintegrating effect might occur. Batches of the briquettes were stored in bottles in the laboratory for the 7-day and 28-day tests, and additional series were stored in the field, for the longer tests, at stations where the waters in question were encoun- tered. The field series were stored in running water, and the action upon these should be more severe than upon the laboratory specimens stored in still water. In each case a parallel test was made by storing a series in pure drinking water. | PAOINI, METAMORPHISM OF PORTLAND CEMENT 217. TABLE 26 Tensile Strength of 1:4 Mortars, stored in Rock Water Strength, pounds per square inch Water Stored in laboratory Stored in field Secure 7 days| Gain |28days| Gain ||3mos.| Gain |6mos.| Gain Drinking..| 211 |...... DOT | Meee 27) |e vee 324y ae 12,12,6,6 CO Oo 220 |+4% | 312| 5% || 323] 1% | 247 |—24%| 12,12,6,6 B? )..) 203 |-4% || 287 |=3z% || 318 |—2% | 303 |\— 5%) 12,12 6,6 Ce: OPQ mee 221 |14% | 288 |—3% || 340| 6% | 328| 1%] 12,12,6,6 TABLE 27 Analyses of Rock Waters in Previous Experiments Parts per million A B Cc EES the agit ealas.: 44 SHO), asian ee ee 20 15 4 Fe,0,+Al,0,.......... 7 5 4 CAO Re sae tues 284 399 87 GOR eer a ghee 124 118 38 See nies Ca eck 727 353 31 (OUR rca ted le the moa 826 046 270 CO,, Alkalies, ete...... 949 459 317 Total solids....... 3037 1895 751 The drinking water used to store the blanks contained in neither case more than 100 parts per million of total solids. The most consistent reduction of strength, although a slight one, is observed in the case of water B, a fairly typical sulphato-chloride water according to Clarke’s classification (18, p. 190). A strikingly high and sudden reduction occurs at six months in water A, a sulphate water charged with hydrogen sulphide, while water C, a chloride water, shows no marked reduction of the strength, which, however, may be due to a’ low salinity. The six-month briquettes stored in water A showed superficially much minute pitting, due to the removal of the sand grains, presumably by solution of the matrix of the cement. Two sections were cut from one of these briquettes, one transverse and one longitudinal, in the hope of discovering whether any replacement of the original material by sul- 218 | ANNALS NEW YORK ACADEMY OF SCIENCES phates or sulphides was going on. The microscopic examination did not reveal anything of the sort, the sections being in all respects similar to sections cut from the briquettes stored in drinking water. It was con- cluded therefore that the loss of strength was due to actual removal of material by solution rather than by replacement with material which would cause disintegration through a discrepancy in volume. The legitimate general deduction from these tests is that, over the period of experiment, the effect of these waters is greater in void filling by crystallization or precipitation than in disintegration by solution or disruption. — The void-filling material, if of a stable nature and not likely to return into solution, should be in a measure a protection against the further entrance of the saline solutions. It has been mentioned that this prop- erty has been suggested of magnesium hydroxide (70). Probably upon this possibility is based the reported effect of chemically inert fine ma- terials, added to the cement for protection against such destructive action. SUMMARY OF EXPERIMENTAL RESULTS 1. Increase of temperature of the water with which cement is mixed causes acceleration of the set up to a certain maximum perenne, then a retardation. 2. Storage in cold water, without freezing, retards the hardening of neat cement, and that of mortars more. 3. Increase in the proportion of fine particles in a cement decreases the permeability of mortar made therefrom. 4. Mechanical agitation increases the strength of cement up to a cer- tain maximum time; after which, if continued, it reduces it. 5. The setting of cement is accelerated by dryness of the atmosphere. 6. An excess of mixing water progressively reduces the strength of cement, This effect is partly reversive of itself, and the reversion may be increased by additional colloidal material in the original cement. 7. Water percolating through concrete dissolves the lime of the ce- ment chiefly, and this effect tends to neutralize itself by “healing.” 8. Percolation through concrete preferably follows the bedding planes. 9. Salts in solution in the mixing water tend to lower the strength of cement. This effect may be neutralized by precipitation in the pores. 10. Storage in saline water affects low silica cements more than it does high silica, and coarsely ground cements more than it does finely ground cements. PACINI, METAMORPHISM OF PORTLAND CEMENT 219 GENERAL CONCLUSIONS In general, the metamorphism of Portland cement represents on an accelerated scale the processes which occur in natural rocks. The accel- eration is of course due to the ease with which water has access to the finely comminuted particles in the initial stages of metamorphism. Many of the minerals found in natural rocks, when ground as finely as, or finer than Portland cement, undergo vastly accelerated reactions in the pres- _ ence of water; colloidal bodies are thereby produced, and the water is rendered alkaline (18). The end product of prolonged water action on Portland cement bears. a striking qualitative similarity to the end product in the kaolinization of feldspars. The same transformations evidently occur in both cases,— the alkalies and the lime are abstracted, and the water and alumina con- tents increased. The exceeding fineness and high adsorptive power of the resulting products are also similar. The action of water on nearly all silicate minerals is, in effect, a repetition of this process. The peculiar adsorptive properties of colloidal bodies render these hable to coagulation. As has been pointed out in preceding pages, much of the cementing material of conglomerates and sandstones, except where calcitic, may have its origin in a similar phenomenon. On a natural scale, the action of water is greatly retarded, because of the larger size of the bodies acted upon, and the consequent paucity of surface upon which water may exert its influence. When Portland ce- ment has properly undergone its initial metamorphism, the setting process being complete and the hardening process in great part so, it approaches the condition of a natural metamorphic rock, and activities towards its further change are katamorphic and vastly slower in their results than the initial changes. The component particles have now become consolidated and the surface offered to the action of water is minimized. Of course, this is truer of neat cement than mortar and truer of mortar than of concrete, these being in the order of increasing porosity. The hypothesis that crystal formation is responsible for the strength of hardened cement is not so complete and satisfactory as the colloidal hypothesis just referred to. In a compact mass, the growth of crystals can hardly be considered anything but an element of weakness. As has been shown by the foregoing results, the effects of varying some of the conditions of the action of water upon cement are best explained by considering the hardening a coagulative process rather than a process of crystallization. 999 | ANNALS NEW YORK ACADEMY OF SCIENCES BIBLIOGRAPHY 1. Abstract: Influence of the Proportion of Water on the Compressive Strength of Cement Mortar and Concrete. Conc. Eng., 3, 316. 1908. 2. AMBRONN, H.: Crystallization and Gel Formation in the Hardening of Cement. Tonind. Zig., 33, 270. 1909. 3. ANDERSON, A. O.: The Incrustation and Absorption of Concrete. Proce. Am. Soc. Test. Materials, July, 1911; reported in Cement Age, 13. 1911. 4, ANDERSON, G. G.: The Effect of Alkali on Concrete. Trans. Am. Soe. C, E., 67, 572. 1910. 5. ANON.: Tests on Concrete in Sea Water. Eng. News, 64, 483. 1910. 6. AsHLEY, H. E.: The Colloid Matter of Clay and its Measurement. Bull. 388, U. S. G. S. 1909. . BIED AND VIVIERS: Decomposition of Mortars. (Report.) Proc. Internat. Soe. Test. Materials. 1909. 8. BINCKLEY, G. 8.: Experiments on the Coarser Particles in Cements. Eng. Record, 61, 212. 1909. 9. BLount, B.: The Setting of Portland Cement. Eng. Digest, 3, 135. 1908. 10. Brown, F. W.: Action of Alkali Water on Cement. Cement, 9, 93. 11. BurcHARTZ, H.: Researches upon Mortar and Concrete Frozen and again Thawed Out. Mt. Kgl. Materialpruf, 28, 276. 1910. : Properties of German Portland Cements. Cement Age, 10, 250. 1908. 13. BURKE AND PINCKNEY: The Action of Alkali Salts upon Portland Cement. J. Ind. and Png. Chem, 3, 317. 1911. 14. CamMERON, F. K.: Application of the Theory of Solid Solutions to the Study of Soils. Report No. 64, Div. of Soils, Dept. of Agriculture. 15. CARPENTER, R. C.: Materials which Retard the Activity of Portland Cement. Cement, 8, 21. 16. CHAPMAN, C. M.: Waterproofing with Water. Eng. Record, 62, 767. 1910. 17. CLARKE, H. C.: Trans. Am. Soc. C. E., 14, 158. 1885. 18. CLARKE, F. W.: The Data of Geochemistry. (Second Edition.) Bull. 491, U. S. G. S. 1911. 19, CocHRAN, J.: Directions and Suggestions for the Inspection of Concrete Materials. Hngineering and Contracting, 37, 115. 1912. 20. Committee of Assn. R. R. Superintendents: The Action of Sea Water on Conerete. (Report.) Cone. Eng., 2, 139. 1907. 21. CuSHMAN, A, S.: The Effect of Water on Rock Powders. U. S. Dept. of Agriculture, Bureau of Chemistry, Bulletin 92. 1905. 22. D’ANS, J.: Do Clay and Cement Adsorb CO, Ions? Z. Chem. Ind. Kolloide, =I 6, 38. 1910. 23. DecKER, G.: The Hardening of Hydraulic Cements. Tonind. Ztg., 33, 1493. 1909. 24. DrescH, C. H.: The Chemistry and Testing of Cement. London, 1911. 25, DIECKMANN, G. P.: Use of Concrete in Freezing Weather. Rock Prod- ucts, 9, 52. 1909. 26. DITTER AND JESSER: Thermochemical Experiments upon Sintered Portland Cement Mixtures. Zentralbl. Hydr. Zement, 1, 71. 1910. 27. DorELTER, C.: Fusions of Calcium Metasilicate and Calcium Aluminum Silicate. Zentralbl. Hydr. Zement, 1, 104. 1910. PACINI, METAMORPHISM OF PORTLAND CEMENT 2) 28. DonatH, E.: Behavior of Certain Liquids towards Cement and Concrete. Chem. Ind., 34, 123. 1910. 29. DuNnN, W.: Effect of Sewage Gases on Concrete Work. Cem. Eng. News, 22, 484. 1910. 30. Hditorial: Effect of Salt Solutions on the Set and Durability of Concrete, Cone. Eng., 3, 318. 1908. 31 : Effect of Temperature on the Set of Concrete. Cement Record, i 242 Boe : What Percentage of Clay is it Safe to Permit in Sand for Cement Mortar? Eng. News, 57, 620. 1907. 33. ———: The Effect of Clay in Cement Mortar. Eng. Record, 55, 703. 1907. 34. : The Micrography and Constitution of Cement. Tonind. Ztg., 32, 1681. 1908. 35. : Retempered Mortar as a Bonding Material. Cement Age, 13, 97. 1911. 36. Fak, M. S.: Cements, Mortars and Concretes. New York, 1904. 48. . FeretT, M. R.: Resistance of Cements to Sea Water Increased by Admix- tures of Puzzuolan. Cem. Eng. News, 22, 11. 1910. . Fisk, A. J.: Effect of Alkali on Cement Mertars. Eng. News, 64, 168. 1910. . FULLER, W. B.: The Laws of Proportioning Concrete. Report to the Aqueduct Commissioners, New York. 1905. . Gand, W. L.: Composition of a Number of Proprietary Mixtures for Waterproofing Concrete. Cone. and Const. Eng. 3, 54. 1908. . GARY AND SCHNEIDER: Behavior of Hydraulic Cement in Sea Water. Mt. Kgl. Materialpruf., 27, 239. 1909. . GILLMORE, Gen. Q. A.: Cements, Limes and Plasters. . GoLpBECK, A. T.: Expansion and Contraction of Concrete While Harden- ing. Proc. Am. Soc. Test. Mat., 11, 563. 1911. . GREENMAN, R. S.: Practical Tests of Sand and Gravel Proposed for Use od in Concrete. Proc. Am. Soc. Test. Mat., 1, 575. 1911. . GRIMSLEY, ©. P.: The Gypsum of Michigan. Geol. Surv. Mich., 9, (2), 188. 1904. . HEADDEN, W. P.: The Destruction of Concrete by Alkali. Expt. Sta. Color. State Ag. Coll., Bull. 132. . Herser, A.: The Grinding Process and the Setting of Portland Cement. Tonind. Ztg., 34, 936. 1910. JANpDA, F.: Theories of the Setting and Hardening of Portland and Rom:n Cements. Oesterr. Z. Berg. Htittenw., 56, 432. 1908. . JEWETT, J. Y.: Cement and Conerete Work of the U. S. Reclamation Service, with Notes on the Action of Alkali Waters. Proc. Am. Soc. Test. Materials, 8, 480. 1908. . KARL: Effect of Sewage on Concrete. Eng. Record, 62, 28. 1910. . Kine, F. H.: Principles and Conditions of the Movement of Ground Waters. 19th Annual Rep., U. S. G. S., part 2. 1899. 2. LE CHATELIER, H.: Recherches sur les Mortiers Hydrauliques. Paris, 1904, : The Decomposition of Cements. Proc. 7th Int. Cong. Applied Chem., London, Sec. 2, 10, 5. 1909. 222 54. 5D. ANNALS NEW YORK ACADEMY OF SCIENCES MAcKENzIg£, W. B.: The Disintegration of Concrete Exposed to Sea Water between High and Low Tides. Hng. News, 58, 472. 1907. MatrHews, BE. R.: Action of Sea and Fresh Water on Cement and Con- erete. Eng. Digest, 5, 527. 1909. 56. MATTHEWS AND Watson: Effect of Freezing and Immersion upon Cement and Cement Mortar. Hng. Digest, 4, 405. 1908. . Maywnarp, E.: Iron Ore Cement. Cem. Eng. News, 22, 243, 1910. : Solution and Decomposition of Cements. Tonind. Ztg., 33, 1462. 1909. . McGee, W J: Outlines of Hydrology. Bull. Geol. Soc. Amer., 19, 193. 1907. . McKenna, C. F.: Hardness of Plasters and Cements, and a Simple Chronographic Apparatus for Recording Set. J. Ind. and Eng. Chem., 4, 110. 1912. . MgeAbDE, RIcHARD K.: Portland Cement. Easton, 1906. : The Influence of Fine Grinding upon the Physical Properties of Portland Cement. Cement Age, 7, 161. 1905. : The Ultimate Composition of Portland Cement. Chem. Eng., 10, 183. 1910. . MicHAELIS, W. A., SR., AND W. A., JR.: The Hardening Process of Hydrau- lic Cements. (Paper read before Assn. Ger. Port. Cement Mfrs., Ber- lin, 1907.) Cem. Hng. News, separate. : The Hardening of Cement Under Water. (Paper read before Assn. Ger. Port. Cement Mfrs., Berlin, 1909.) Cem. Hng. News, separate. . MicHae.tis, W. A., Sr.: Puzzuolan Mortars in Sea Water. Tonind. Ztg., 33, 1308. 1909. . MoNTEMARTINI, C.: The Hydration of Cements. L’Industria Chimica, 7, GOD 90S: . NEWBERRY, S. B.: Cement for Sea Water Construction. Cement Age, 9, 38, 1907. . NEWBERRY, S. B. and W. B.: Constitution of Hydraulic Cements. J. Soe. Chem. Ind., 16, 887. 1897. . O'Hara, J. M.: Action of Sea Water on Portland Cement. Cem. Eng. News, 22, 288. 1910. . Prererson, P. M.: Determination of the Finest Powder in Portland Cement. Tonind. Ztg., 33, 1687. 1909. . PIERCE, G.: Destructive Action of Alkali upon Cement. Mining Sci., 63, 130. 1911. . PLtums, R. A.: Waterproofing Concrete. Canadian Cement and Cone. Rev., 4, 91. . Porrson, L.: Studies of the Chemical Influence of Sea Water upon Port- land Cement. Ciment, Nos. 6, 7. 1910. . Porrer, C. J.: Chemical Changes in Portland Cement Concrete, and the Action of Sea Water thereon. Jour. Soc. Chem. Ind., 28, 6, 1909. . Poutsen, A.: Long Time Concrete Tests in Sea Water. Eng. News, 64, 3. 1910. . Progress Report of the Joint Committee on Concrete and Reinforced Con- crete of the American Society for Testing Materials.: Proceedings, 1909. PACINI, METAMORPHISM OF PORTLAND CEMENT 923 . ReaD, EH. J.: The Crystalline Products of the Hardening of Portland Cement. J. Soc. Chem. Ind., 29, 735. 1910. (Discussion by Blount.) . REIBLING AND R&yES: Physical and Chemical Properties of Portland Cement, III. Phillip. Jour. Sci., 6, 207. 1910. . RICHARDSON, CLIFrFoRD: Portland Cement from a Physico-chemical Stand- point. Atlantic City, June, 1904. Also, Eng. News, Aug. 11, 1904, Jan. 26, 1905, Cement 1904-1905, 5, 3, et seq. . Rowan, W. D.: The Effect of Sea Water, Alkali Water, and Sewage on Portland Cement. Eng. Cont., 24, 52. 1905. . ROHLAND, P.: The Weathering of Stones and Mortars. Zeits. Chem. Ind. Wolloide, 8, 48. 1911. : Influence of Electrolytes on the Setting Time of Cement. Zeits. Angew. Chemie, 16, 622, 1903. Stahl u. Hisen, 28, 1815 1908. : Causes for the Changes in Speed of Hydration in Cement. Zeits. Chem. Ind. WNolloide, 8, 251. 1911. . SABIN, L. C.: Cement and Concrete. McGraw, 1905. . SAGER AND CRAMER: The Hardening of Previously Set Cement. Tonind. Ztg., 32, 1746. 1908. . ScHort, O.: American Portland Cement Concrete in Sea Water. Chem. Ztg., 35, 256. 1911. . SHEPHERD AND RANKIN: Preliminary Report on the Ternary System CaO-A1,0.-SiO,. Jour. Ind. and Eng. Chem., 3, 211. 1911. 9. SLICHTER, C. S.: Theoretical Investigation of the Motion of Ground Water. 19th Ann. Report, U. S. Geol. Surv., pt. 2, 1899. . SMITH, A. HRSKINE: Some Old Cement Tests. Cone. Eng., 4, 124. 1909. . Society Scanp. PorTLAND CEMENT Mrrs.: Sea Water Tests of Portland Cements. Cement Age, 10, 184. 1908. . SPACKMANN, H. S.: Relation of Lime Content of Cement to Durability of Concrete. Concrete, 11, 52. 1911. . SPACKMANN AND LESLEY: The Hydraulic Properties of Reground Cement. Cement Age, 10, 1. 1908. . SPALDING, F. P.: Hydraulic Cement. Wiley, 1906. . STERN, E.: The Setting and Hardening of Cement. Mt. Kgl. Materialpruf, 28, 173. 1910. . STEWARDSON, C. W.: The Effect of Alkali on Cement. Brick, 33, 102. 1910. OT. . THATCHER, H.: The Effect of Sea Water upon Portland Cement, Mortar TAYLOR AND THOMPSON: Concrete, Plain and Reinforced. Wiley, 1905. and Conerete, and upon Steel Embedded Therein. Proc. Am. Soe. C. E., 34, (2). 1908. . THOMAS, E.: Colloidal Cement Waterproofing. Cement Age, 11, 158. 1909. . THOMPSON, G. W.: The Classification of Fine Particles According to Size. Proc. Am. Soe. Test. Mat., 10, 601. 1910. . TORNEBOHM (quoted in 80) : The Petrography of Portland Cement. IlKong. Int. Verb. Material. 1897. . TuTTLe, M. C.: Decomposition of Concrete in Sea Water. Eng. News, 60, 341. 1908. . UneER, C. (Tr. by MicHartis): Fused Portland Cement. Cement Eng. News, 22, 192. 1910. 109. 110. dail, 112. ANNALS NEW YORK ACADEMY OF SCIENCES . VAN Hiss, C. R.: A Treatise on Metamorphism. U. S. Geol. Sury., Mono- graph 47. 1904. . WALLING, W. A. B.: Some Notes on the Setting of Cement. Chem. News, 104, 54. 1911. . Weston, R. 8.: Effect of Sewage on Concrete. (Letter.) Eng. Record, 61, 714. 1909. . Wuitr, A. H.: Disintegration of Fresh Cement Floor Surfaces by the Action of Smoke Gases at Low Temperatures. Pr. Am. Soc. Test. Ma- terials, 9, 580. 1909. . WHits, A. H.: Destruction of Cement Mortars and Concrete through Ex- pansion and Contraction. Proc. Am. Soc. Test. Mat., 11, 531. 1911. Wie, R. J.: The Effect of High Pressure Steam on the Crushing Strength of Portland Cement Mortar and Concrete. Proc. Am. Soc. Test. Mat., 11, 580. 1911. f WILLIAMS, I. A.: Influence of Fine Grinding upon the Physical Properties of Portland Cement. Trans. Am. Cer. Soc., 10, 244. 1908. WitHey, M. O.: A Survey of the Concrete Aggregates of Wisconsin. Con- erete, 12, 55. 1912. Wormser, D.: Changes in the Strength of Portland Cement. Cement Age, 9, 254. 1907. of ee and similar matter. volume of the Annals coincides | in Sear. with the calendar year trations, and may ‘be learned upon sepicaton to the Librarian of ie .cademy. The author receives his peal: as soon as his paper has (2) The Memoik (quarto ey established in “1898, are issued at as intervals, It is apuended that each veg shall be JO to The - a pletion are sent free to Fellows and Active Members. THE LIBRARIAN, New York Academy of Sciences, = care of American Museum of Natural History, . New York, N. Y. - ANNALS OF THE NEW YORK ACADEMY OF SCIENCES Vol. XXII, pp. 225-258, Pll. XXINI-XXXIIl Editor, EpMuND Otis HOovEyY ' THE PHYSIOGRAPHY OF THE PERUVIAN ANDES py itt NOTES ON EARLY MINING IN PERU BY V. F. Marsters NEW YORK PUBLISHED BY THE ACADEMY 1% SEPTEMBER, 1912 THE NEW YORK ACADEMY OF SCIENCES (Lyceum or Naturau History, 1817-1876) : OFFICERS, 1912 President—HMERSON McMit.in, 40 Wall Street’ Vice-Presidents—J. EDMUND WoopMAN, FREDERIC A. Lucas CHARLES LANE Poor, R. 8. WooDwoRTH Corresponding Secretary—HENryY EH. CRAMPTON, American Museum ~ Recording Secretary—Epmunp Otis Hovey, American Museum Treasurer—HeEnry L. DoHeErty, 60 Wall Street Librarian—RatPH W. Towser, American Museum Editor—EpmunD Otis Hovrey, American Museum SHCTION OF GHOLOGY AND MINERALOGY Chairman—J. EH. WoopMan, N. Y. University Secretary—Cuar.es P. Berkey Columbia University SECTION OF BIOLOGY Chairman—FReEpERIc A. Lucas, American Museum Secretary—WiLL1aAmM K. Gregory, American Museum — SECTION OF ASTRONOMY, PHYSICS AND CHEMISTRY Chairman—CHARLES LANE Poor, Columbia University : Secretary—F. M. PEDERSEN, College of the City of New York SECTION OF ANTHROPOLOGY AND PSYCHOLOGY Chairman—R. 8. WoopwortH, Columbia University Secretary—FREDERIC LiyMAN WELLS, Columbia University The sessions of the Academy are held on Monday evenings at 8:15 o’clock from October to May, inclusive, at the American Museum of Natural History, 77th Street and Central Park, West. [ANNALS N. Y. ACADEMY OF SCIENCES, Vol. XXII," pp. 225-258, Pll. XXIII- XXNIII. 17 September, 1912] THE PHYSIOGRAPHY OF THE PERUVIAN ANDES WITH NOTES ON EARLY MINING IN PERU By V. F. Marsters (Presented in Abstract before the Academy, 5 February, 1912) CONTENTS : Page i TEMTTE DE TO MCULO ING! 5 Bs ceerchersie Ses Sle Gio ee a ae ORC AO LORE ene Ra naar ECR ee ea 225 PO pPOsiAap li GaPEOVINGCES OF REE. a5 Sb oo Sete cieiee ao Sion ie ele dele cele sa Ske dee eats 227 COBDS TAL TOIEMIMSES Se Sos sche co CIS NCR aE nN CECI CR CT a le 228 Orie ME iM aTye MCA rn ac tok situa c cise wie erst aloes w lscvesein sd sc etene ss (20D Locus of agriculture and its dependence upon physiographic [DLIVETIVOUINVETTS 653 os ate aro oro G GESTS Han TE RECT er erst eM Wiest slope and West Range of the Andes........-.....5...-.--+20-- 239 Hira o-Ovone sSeChlOMe sos sobs min cle sicko 's Gil ciate cis ceeds ero doce es coe LOO @cona-CoOrro, uno? SCChHON sc) kas sic see os a oe bs dls Seale Sees 240 Highland Plateau or Intercordilleran Belt........ PAT eeen or tol eae WETLOME CMR ASCOU SE CHHOMe i icic\ecrens siete vc es eS bcesele # co's ea ee bese a, ccs, LAD Casma-Huaraz-Huacaybamba section............. 0. ce ee eee ee 2AG Piuiraskiunan cabanas SCCHOMs Jac cco vise celiiseis cee cccleocecceescuwes 24T Hasthanzerand Hast -Slopesc.....2c5.ee2. cece ees ee at Veh ietena Stree 248 Topographic expression as related to the geology of the Peruvian Andes. . 249 Pal eMUM Sp cUIN ESOC IT OIMGS are sta crane ene el cue cca .c) eiisie a's couch aveve, ace we 14) ots teeta. 6 Steeles: 26 249 ZoOrritos-Wambeyagque Plait. . ose. 0 cote ss oe Sas cane cece cee eevee JAD Ghaimehra-O lino Ss Ane seer ares gins: uayens ate ste Sieae-ce ees Seow oie ieee =.) LOO @cona-Moquequa” S@ChHON. 2.4.0. cc cc. secs ces croc een es cer ees sees, ZO Huacho-Cerro de Pasco section................. tins ei areoeeeseaaiccoin ee 251 WMeona-COLeoe UN OMSECH OME dem elocra ces cies Hele A cele oe et BAS ee shee os OD NL CSROM CARLY aIMIMIN SINE POR IWs seals nies ee ois = Selelelo ses isle ebles viene ase) 2D Ee UIA ecgets, eWay nersno loca eres ets goeani's Gy Seah nits, Gata re aC mAela ws. 8 AAs ae atksicceniyere EOS INTRODUCTION So far as I am aware, very little has been written relative to the physiography either of the coast or of the Peruvian Cordillera. As preliminary to the presentation of a few observations on the above subject, I wish to summarize very briefly a list of the early pioneer geographers, geologists and naturalists who have visited the Peruvian shores and have published observations relative to the subject in hand. (225) 226 ANNALS NEW YORK ACADEMY OF SCIENCES Darwin was among the first to record observations relative to the coast of Peru. He noted the many old shore line beaches now standing at various elevations above the present sea level. The same facts were noted later by D’Orbigny. The occurrence of “kitchen middens” was recognized at various points. They appeared to be associated with the ancient beaches, thus suggesting that elevation had actually taken place since the accumulation of the shell heaps. The writer has seen between the mouth of the Rio Grande and Lomas as many as five well-defined old shore lines or beaches occurring in succession, the highest one being approximately some sixty feet above sea level. I find also that a geographical map of a portion of southern Pern, together with a portion of Bolivia and Chile, has been prepared by one Mr. Pentland, but to date I have been unable to locate it or any of his written contributions on the geography of the above section. Pissis and D’Orbigny were the first to contribute detailed geological and indirectly geographical information concerning southern Peru and the adjacent republics. Both these men constructed cross sections of the Peruvian and Bolivian Andes, the section starting at the coast near Tacna and ending on the east slope of the East Cordillera to the east of La Paz. Some years later Forbes went over the same ground and likewise con- structed a cross section. While the same type of relief is expressed in all these sections, the classification of the formations, based in part upon lithological and in part upon paleontological data, differs very widely in each case. The sections prepared by these men may be seen in the now rare publication of the Geographical Society of La Paz, Bolivia. Among the later pioneer naturalists who did much serious work in Peru was one Senor Raimondi, an Italian by birth. He came to Peru, bringing with him the training of an Italian institution and that pro- found interest in his field of investigation that is always sure to produce invaluable results. Raimondi was the Agassiz of Peru. Among his first efforts was the preparation of a topographic map of Peru; and to date it is the only map possessed by the republic and officially accepted by the government. The writings of Raimondi are likewise voluminous. While collecting his map data he likewise accumulated a mass of infor- mation relative to the geology, mineralogy, zodlogy and botany of the entire country. The results of his investigations were published in a series of volumes by the Peruvian government. Another source of information is to be found in the publications of the Cuerpo de Ingenieros de Minas, a department of the government in- stituted by Sr. José Balta for collecting information concerning the natural resources of the republic, such as mines and mine production, the . " a j i i A. Nt rn pi mines ody eee SoH Aipeme We ven wae pn Sed ee ete wy Sey ee eae Lo 3 ee nT a ET UREN Ca acer aT S ANNALS N. Y. Acap. Screncps r 80 CUAYRQUIL a iT NESS. & { { \ | Lobosise, Ete, 7, ESI Trujillo i ee CALLAO KE” SKETCH MAP SHOWING THE GEOGRAPHICAL PROVINCES OF PERU BY V. F. MARSTERS FS] Coastal Ridges Coastal Plains West Slope-W. Cordillera East and West Cordillera Intercordilleran Belt a, % TOP SBUGAIACES Of CEOCE YEH 0 : cee settee I : a erenepersabaeehir ote ee 7 ST MARSTERS, PHYSIOGRAPHY OF THE PERUVIAN ANDES 99% geology of mining districts, data concerning irrigation projects, under- ground waters, etc. More than two hundred bulletins have been pub- lished by this department. While much information exists in the series of publications, it is unfortunate that no attempt has ever been made to correlate it with a view to preparing atleast a preliminary geological map of the more important mining centers of the republic. So far as I am aware, the only attempt to correlate the formations in distinct parts of the republic was recently made by Dr. G. Steinman for the Cordillera an:] by the writer for the formations of the coastal plains as they appear in northern, central and southern Peru. Hence we may say that, while there are sources of information relative to the geology and geography of the Peruvian Cordillera, in no publication, so far as I am aware, has any attempt been made to give even a skeleton outline of the probable physic- graphic history of any section of the Andes. Let us now see if, from the information to be gleaned from the early writers and especially the map of Raimondi, combined with the observa- tions of the writer, we may be able to get a concise picture of the geography and at least a glimpse into the geology of a portion of the Peruvian Andes. The Republic of Peru occupies an area approaching 2,000,000 square kilometers ; it extends from 0° to 20° 8. latitude and from 64° to 84° W. longitude; its coast line is approximately 1,500 miles long, and it pos- sesses all the varieties of climate from typically tropical conditions in the north to cold temperate in the south and in the higher parts of the Andean Highlands. One need not go beyond the confines of the Inca Republic to find any of the variations between the extremes mentioned. One may leave Lima under a semi-tropical sun and, in the course of a few hours by rail, be riding over a snow-covered Puno. As soon as he reaches the montana, or the wooded part of the eastern slope of the east range, he passes immediately into warm temperate and tropical climates. TopoGRAPHIC ProviINcres oF PERU The distinct topographic and physiographic provinces of Peru are well defined. They may be summarized as follows: 1. Coastal Plains. 2. West Slope and West Range of the Andes. 3. Highland Plateau, or Intercordilleran Belt, and its associated sec- ondary ranges. 4, Hastern Range. 5. Kastern Slope and Lowlands. 298 ANNALS NEW YORK ACADEMY OF SCIENCES COASTAL PLAINS The distribution of the coastal plains may best be seen in the accom- panying sketch map. They occupy three sections, which may be designated as the northern, central and southern divisions. . The northern division begins in the region of Tumbes, near the bound- ary of Peru and Ecuador. It extends along the coast to the south, reach- - ing its maximum width in the cross section between Cerro del Yllesca and Salitral. From this point to the south, it grows narrow quite rapidly and finally ends at Salaverry. Going still farther south, we find but few remnants of what must have been the inner edge of the coastal plain tucked away in partly drowned valleys within the limits of the western slope. Between these valleys, the formations composing the foothills of the western range now occupy — the present shore line and continue to do so, until we reach the region of Canete. Here the coastal plain again makes its appearance. It can be traced to a point just to the south of the mouth of the River Yauca. Here again the foothills of the western slope reach the present shore line. If we were to travel along the shore line from Yauca to the south, we should be very much inclined to believe that, as far south as Ocona, the actual foot-hills of the Cordillera again formed the present shore line. This, however, we find not to be true. Should we pass up the Valley of the Atico, we should find a belt of country some fifteen to twenty miles wide occupied by Tertiary and post-Tertiary deposits, similar to those we have seen to the north but standing at a higher altitude. They are deeply cut by the streams coming from the west range of the Andes. It is not improbable that this belt may extend to and beyond the Valley of the Chaparra. Going to the south, we can trace the elevated coastal plain, where it is known under the various names of Paco Alto, Cuno-cuno, south of the Valley of Ocofa, Pampa de la Joya, south of the Valley of the Vitor, Pampa de Clemesi, south of the Tambo Valley, etc., and thus continues with decreasing altitude to the southern boundary of Peru and Chile. : One of the additional features of the coastal plain is the frequent occurrence of isolated hills and ridges near the present shore line. These are locally known as “morros.” They usually stand-at an appreciable elevation above the general level of the plain and hence form convenient landmarks for determining direction of travel to a given point. With the exception of the section of the coastal plain to the north of the Valley of the Ocofa, the principle transverse streams have succeeded in cutting only deep narrow valleys, with very little available floor for “MARSTERS, PHYSIOGRAPHY OF THE PERUVIAN ANDES 229 agricultural purposes. The interstream spaces are barely scarred by insig- nificant tributaries, so that we have huge expanses of desert plain extend- ing from valley to valley. As we approach Moquequa and Locumba, how- ever, the elevation has been less, so that the rivers have made broader valley floors on which has been developed considerable grape culture, while, in the Valley of Sama, sugar cane forms the chief agricultural product. Geologically considered, the coastal plains of Peru are composed of Tertiary and post-Tertiary sediments and lava flows, laid down on a post-Cretaceous surface, which at the beginning of Tertiary sedimenta- tion had not been worn down to grade. This is proved by the number of half-buried hills now standing above the general horizon of the coastal plain, the so-called “morros” already referred to. In many instances, these hills merge into ridges and form what is known in Peruvian geography as “Cadena de la Costa,” or coastal chain. Their distribution may be seen on the accompanying sketch map. As might be expected of a coastal plain bordering a mountain range of such tremendous proportions and geologically young, it has under- gone differential elevation to a great degree. In various localities, folding and faulting have taken place on a considerable scale. ~ In that section to the-north of Paita where I have made some detailed stratigraphical studies, on account of the development of the Zorritos, Lobitos and Nigritos oil fields, it is caleulated that not less than 3,009 feet of Tertiary sediments enter into the structure of the coastal plain. Some localities are rich in fossil gastropods and nautiloid forms. They are supposed to represent lower and middle Tertiary faunas. It is interesting to note the fact that the oil-bearing localities so far developed are associated with the sections of maximum disturbance of the formations. The oil field of Zorritos is located on the eastern flank of a folded section, much of which is located beyond the present shore line. It is also in this section that we find the largest amount and most minute type of surface dissection. The folding of the Zorritos section may be traced to the northeast for some distance. In the interior to the southeast, the folding gradually fades out until we reach the flanks of Amotape Mountain, where its contact with pre-Tertiary formations is encountered. Passing to the section of Lobitos, we find the same stratigraphical and structural relationships as noted in Zorritos. The area of maximum fold- ing is near the present shore line. About Lobitos, the original surface of the coastal plain has been completely eroded away. To the north of the Lobitos field, we find the original surface forming an extensive plain 930 ANNALS NEW YORK ACADEMY OF SCIENCES some 300 feet above sea level and with its shoreward edge somewhat cvt up by narrow, short valleys. Passing to the interior from Lobitos in the direction of the Amotape spur of the west range of the Andes, consid- ered in its broadest sense, the evidence of folding, faulting, etc., rapidly disappears, until we approach the inner edge of the plain, 2, we find that the Tertiary sediments are upturned. Following to the south towards Talara and Nigritos we find that while folding is much less accentuated than at Lobitos, it by no means disap- pears entirely. At Talara, however, we find another area of maximum disturbance, which extends to the southern end of the oil field now being operated by the London Pacific Petroleum Company. Following this section to the interior, we note, as in the preceding cases, a disappear- ance of the folding, until we reach a narrow belt adjacent to the Amotape Mountain, where the Tertiary strata are found to stand at a high angle and dip toward the northwest. I wish here to record another fact. The first oil springs noted in this region were located at the interior edge of the coastal plain at a point known as La Brea. A little later a residuum of petroleum was located on the playa near Nigritos. This was mined for some time by the Spaniards and prepared for painting the bottoms of their ships. In the Chira Valley, we find the Tertiary occupying an arm-like ex- tension or depression between the Amotape spur on the northwest and an edge of the foothills on the southeast. On the high plains to the rear of Paita, some 300 feet above sea level, there is a well-defined noncon- formity between red clays with a marked tilt and a series of sands and conglomerates containing numerous fossils. Many of these resemble very closely the forms now living on the present shore. At various points between Paita and Piura there may be seen on the surface shells of exactly the same variety as are now living on the Pacific shore. Just south of Paita we find a typical outlier of the western range stick- ing its higher points somewhat above the level of the adjacent plain. In the interior, the eastern boundary is located near Tambo Grande. It follows the general trend of the rivers Salitral and Serran (or R. Piura), or not far from the contact of the coastal plain deposits with the pre- Tertiary formations of the foothills. That is to say, the Salitral-Serran valley has been made by a longitudinal subsequent river, using the phrase- ology of Professor Davis. The maximum width of this division of the coastal plain is attained along a line from Cerro de Yllesca to Salitral. Following the inner edge, we pass to the south through Olmos, Motupe, Patope and Cayalti. A very narrow belt then passes along the present shore line to the Pacasmayo MARSTERS, PHYSIOGRAPHY OF THE PERUVIAN ANDES 931 Valley. The Cerro de Yllesca forms another of those outliers that serve to break the monotony of the sky line as seen from the interior. From the Tablazo de Paita, the elevation decreases somewhat in the direction of the Despoblado de Sechura and the Plain of Olmos, while a minimum uplift took place in the section through Morrupe, Chiclayo and Lambeyaque. As to the stratigraphy of the Sechura-Olmos area we ean say but little. This is one vast plain, a typical desert, strewn with wind-blown sands, with here and there small depressions occupied with “salinas” or salt deposits, but so slightly dissected that data giving a clue to the stratigraphy are wanting. Sections, however, between the mouth of the Sechura River and Cerro del Yllesca, as exposed on the shore line, lead me to mistrust that beneath the sheet of recent accumula- tions over this vast plain there must exist a thick series of Tertiary sedi- ments. Not a single morro of pre-Tertiary origin is to be found until we approach the region of Olmos. The accumulation of Tertiary sedi- ments in the central parts of the desert may thus be very considerable. As has been stated, the continuation of the coastal plain below Cayalti is represented by only a narrow rim, which, as we approach the Valley of Pacasmayo (the River Jequetepeque), again widens out to several kilo- meters. Here we find that the outhers of the foothills are quite numer- ous. An interpretation of the stratigraphy in this valley is difficult on account of the lack of good exposures. Sufficient evidence is at hand to suggest that the red clays corresponding with the lower part of the Paita section are present. Following to the south again, but a narrow band of the plain bordering the shore line connects with the Valley of the Chicama and Santa Cata- lina. ‘These valleys are probably filled in part with late Tertiary sedi- ments, but later were littered up with much post-Tertiary waste, a phase of coastal geography to be discussed later. The southern extremity of the northern division ends at Salaverry. The stretch of coast from Salaverry south to a point just south of Lima is largely occupied and confronted by the foothills of the Cordil- lera. Only the larger valleys have broad floors near the coast. Whether many of these contain remnants of Tertiary formations or not has not been determined by the writer. It is not improbable that during Tertiary time a fiord-like arm of the Tertiary Sea may have extended into the partially drowned valleys of the West Range. At all events, it is certain that a large amount of waste has very recently accumulated and spread over the lower stretches near the present shore line, presumably in post- Tertiary time. The valleys of Huacho and Chimbote are fair examples. Mention should also be made of the fact that, in the upper portions of 232 ANNALS NEW YORK AVADEMY OF SCIENCES many of these partially filled pre-Tertiary valleys the recent deposits of sand, gravel and clay have again been attacked by the present streams and redistributed at lower levels. The same type of physiographic history is repeated in the valleys of the Rimac and the Chillon. The lower part of the Rimac Valley and the broad area where it is confluent with the Chillon have been aggraded by the deposition of an enormous sheet of fluviatile material. Lima, the capital of the republic, stands on the edge of this fluviatile plain. Whether Tertiary formations exist beneath the sheet of waste or not has not yet been determined. Not until we reach the vicinity of Cerro Azul, Canete and Chincha does the typical coastal plain again make its appearance. In the section through Chincha it has a width of some three or four kilometers. ‘The transverse streams have incised themselves but slightly into the surface deposits. The formations thus far exposed here appear to be post-Ter- tiary waste, sands and conglomerates, undoubtedly deposited upon late Tertiary clays. The latter may be seen in sections at points between Chincha and Pisco to the south. These are regarded as the equivalents of the late Tertiary clays at the bottom of the Paita section. Folding and minute faulting may be seen at various points; the best exposures occur near Pisco. To the north of Pisco, the amount of Tertiary deposi- tion may have been very considerable, but to the south, there is reason to believe that no great proportions were reached. ‘The fact that a very large part of the shoreward area from Paracas to the mouth of the River Ica is occupied by outliers of the pre-Tertiary oldland, precludes, at least, the idea of Tertiary deposition on any great scale in that particular sec- tion. It was only between this broken line of hills or outliers and foot- hills of the west range that any great amount of Tertiary /leposits accu- mulated. In the upper part of the Ica Valley, within the limits of the coastal plain, a very thick sheet of post-Tertiary waste was spread out on the Ica plains as far south as Ocacaje. To the west of Ocacaje, the best exposures of the light-colored Tertiary clays containing fossil fish are to be seen. These rest upon the eastern flank of the outhers already referred to and probably pass under the fluviatile plains in the upper stretches of the Ica Valley and plains. From Ocacaje south we may trace the plain across the Pampa de Huayuri to Palpa, San José and Nasea. To the south of the Rio Grande, we reach the Pampas de Yunga Colo- rado and Bella Union on the northwest side of the Valley of Acari. The plain finally ends at a point a short distance south of the mouth of the River Yauca. The Pampas de Yunga are separated from the coast line by Cerro Yunga, composed of pre-Tertiary formations, another outlier of the West Range. Some distance to the south of Cerro Yunga, the coastal MARSTERS, PHYSIOGRAPHY OF THE PERUVIAN ANDES 233 deposits again form the shore line and continue with but little inter- ruption to the southern terminus of the central division. From the Pampas de Yunga, with an elevation of some 700 to 900 feet above sea level, the surface gradually descends, until, at Lomas, it passes beneath sea level. To the interior from Lomas, where the plain is known as Bella Union, the surface is some 300 or more feet above sea level. In the valley of the Rio Grande, sections of what appear to be late Tertiary may be ‘seen. On the coast near Santa Ana, similar sections may be found. On the southwest flanks of Cerro Yunga, the same sediments are exposed. In the interior, on the inner edge of the Pampa de Yunga, is the so-called Vallé de Carbonaria, where a considerable thickness of clays and sands, presumably late Tertiary in age, may be seen. To the south of the plains bordering the Valley of Yauca, the foot- hills once more form the present shore line. Nevertheless, in the mouth of the Valley of Atiquipa and again in the Chala may be found a tri- angular patch of what appear to be late Tertiary clays and sands: but beyond this point the foothills seem to present a solid front, until we reach the mouth of the Ocona Valley. This, however does not prove to be the case. Should we enter the mouth of the Atico Valley and journey in the direction of Caraveli, we should find that the supposed foothills of west range of the Andes prove to be outliers, separated from the main range by much-dissected plains made up of Tertiary and post-Tertiary formations, and that these extend to the actual foothills of the Andes in the vicinity of Caraveli. These plains probably pass to the northwest as far as the valley of the Chaparra and represent the northwestern limit of the south division of the coastal plains. The inner edge of this plain is not less than 3,000 feet above sea level and has a width of twenty-five miles or more. The pre-Tertiary coastal ridge as far as Oconia shows only its highest point at elevations exceeding the general level of the interior plain, while its shoreward flank has been nearly stripped of its post-Cre- taceous covering. It is in sections on the Caraveli and tributaries to the Atico that we find for the first time that mud flows of no mean propor- tions enter into the formation of the coastal plain. In Quebrada Chin- chin, we may see at least 300 feet of this flow exposed. Moreover, these flows extended far in the direction of the coastal ridge. From the region of the Atico section, we may trace the high plains to the southeast, being represented by the Pampas de Paca Alta, Bourbon, etc., to the valley of the Ocona; then we find the continuation in Pampas de Cuno-cuno, Majas and Vitor, where we reach the valleys of Siguas and Vitor. The inner edges of these pampas stand at least 5,000 feet above sea level. They are very deeply cut by the main streams coming from the West 934 ANNALS NEW YORK ACADEMY OF SCIENCES Range, in fact most of them have cut veritable canyons and, in many parts, are still continuing the process of incision. The inter-stream spaces are, as yet, barely scratched by tributaries to the master lines of drainage. The pampas thus far named are but high-lying plains, with their initial surfaces well preserved. They are veritable deserts. To the southeast of the Vitor Valley we find the pampas of La Joya and Chachendo, the two being partly separated by outliers of the West Range, while nearer the coast and on the northwest side of the valley of the Tambo we find the pampas of Islay and Tambo. Crossing the Tambo Valley, we meet the lower plains adjacent to the shore line and usually known as the La Punta Plains. Crossing the coastal ridge, however, we come upon a higher-lying plain of enormous dimensions and usually known under the name of Clemesi. It extends with little interruption to the valley of the Moquequa. It decreases very materially in elevation as we pass from the Tambo to the Moquequa Valley. On the Moquequa, the plains have a width of from twenty to twenty-five miles. Of all the pampas so far mentioned in the southern coastal plain, the lava flows already mentioned form no inconsiderable part of the deposits making the Siguas-Vitor La Joya sections. Excellent sections may be seen on the Siguas and Vitor valleys. Here can be seen two distinct flows, as indicated by the inter-stratification with sediments. The max- imum development of the flows seems to have taken place along a line or section from Arequipa to Quilea, the former town being located well within the west slope of the West Range of the Andes, and on the con- tact,of the flow with older formations. Just what may be the correct correlation of the sediments of these sections with those of the central and northern division of the coastal plain has not been determined with any degree of certainty; it is mis- trusted, however, that they represent rather late phases of Tertiary de- position. It is important to note, as well, what an enormous amount of elevation has taken place, since their deposition, simce we now find some of these at least 1,000 feet above sea level. There is no proof that the lava plains as a whole ever stood at sea level. The fact, however, that salt deposits are known to occur well up to the higher plains indicates a depression of such extent that a part at least of the mud-flow must have been near sea level. . Returning to our description of the areal extent of the southern divi- sion, we find that the Pampa de Clemisi has suffered but slight erosion. Near the inner edge of the Cadena de la Costa are some very fine salt deposits. The best sections are to be found in the slopes of the Moque- PLATE XXIV eae a OF PERU Heat WOM SION AULT ©) 4 “ i vsh ual al eta ic ee Went avi Rerentnes ey ” - a \ 4 aa isis, Siu Seve. 5 RR WEAN, IAARYD cyt enc aveeae . . ; i ret ¥; Mies af ako 40 alle, sift sooied jaf aenittt ate Ho en fhe’ f tener ite _ lisiq: Istadoo ot: _ehuises wien Dettiess “gro iat lon D npO NOME SMART aa te eer ne » foitooe with) aide uth t “ “ bs Ys : ya" : ¢ 4 > 7 > } t “R =a + : & . ata hi GR at Ri ee) ae ES ‘« ; 4 / m 4 % . : eget nye , ey v Us L~ ’ ned ih ute wade ties P } ¥ ¥ a oad r ‘ ‘ ae ( : , ‘ 4 mA S , x £. - Kae Lf 4 K ihe it saltaS 4 ‘fi Pana! Ota : | ity 2 rT ty . & By bs F 2 “ t 4, \ ' ‘ f : . \ z ’ ph Bey : Foes fe igh ; be - oy i - ut ‘ 3% i fA ” Ld ! ‘ aS, " rf t i 4 7 ¥. 7 as zy Fa Goes © poe it ba sean 4 7 ir % i resis a ; igh ‘i bedi Linaaay Sherpas OT: My Sty veROcTT het! Y Al 2 . Ay AIXX WGOVId ‘TIXX WwaAI0A SHONTIOG “GvOV “KX ‘N SIVNNY 4 OR rth iA ‘AY oa ppp aaa ; Nays Lez hath five nicl Istenon oat to toe! ai abicely) ode ehnoimibes ainiq [alanen os 8 obote Yo tte dT aqely eave ait “ 2 ‘ oe By 7 vi) i: AEN ‘ ; i 5 * ¥ k t } Lis i ay ; a } f ~ ' r KA x B i ; X ‘ nM ; t & i \ ‘ : ; BY ‘ J ra y = { 2 a ye x , , AXX GOV1d ‘TIXX aWwaT0A SMOINTIOS ‘dvoy ‘X ‘N SIVNNY ANNALS N. Y. ACAD. SCIENCES VoLUME XXII, Phare XXVI \ PLATE XXVI Gena ana ae a pe OELENDO , { Siw “sands IS ‘and bays of the coastal plain, eer ; W ee i ty, \ x ‘ Bay hs ‘ Aha ‘A t Fi IME XX tp XNXV ANNALS N. Y. ACAD. SCIENCES VOLUME XNII, Phare XXVII Riaime te Bde ue DOLE (01h) AUpMUOOIU. Wey bi ” i (, TH) anil dt osonda BOO SAD thor COLT sit). HP THOMR oat to weit. Moses) any Vue Mn i " Wy ‘orbist TlenoD odt iabeanog y bade a PLATE XXVII MOQUEQUA (ILO) RIDGE View of the Moquequa (the Ilo) near the coast where it has eut a deep gorge in the Coastal Ridge. MARSTERS, PHYSIOGRAPHY OF THE PERUVIAN ANDES 935 qua Valley and in the Cuesta de Mato de Caballo. Here is éxposed a series of somewhat coarse deposits on top, followed by a thick series of clays and fine sands. Folding and faulting have taken place to a con- siderable degree. The Pampa de Clemesi has an elevation near the Tambo Valley of some 4,000 feet on its inner edge, but it descends in the direction of Moquequa and Locumba. At the valley of Sama, which at present con- stitutes the dividing line between Chile and Peru, the elevation at the inner edge is not more than 800 to 1,000 feet. The same grouping of sediments may be seen on the Sama as in the Moquequa Valley. As regards the character and distribution of the outliers,.a word must be added. The pre-Tertiary coastal chain is persistently present from the valley of Atico to the mouth of the valley of Sama. Near the valley of the Tambo, the edge of the foothills practically joins the Cadena de la Costa. It is for this reason that we have a number of small pampas separated by a collection of outliers. In the stretch of coast from Moi- lendo to Quilea and Camana, the coastal ridge is most prominent in its topography and reaches a width of some kilometers. Their topographical strength, however, gradually fades away in the direction of Sama. To the south of this point, I am told, the ridge is replaced by a series of morros, of which the historic morro at Arica is a good example. Original Tertiary Area—At the present time we probably see but a small portion of the original area of deposition of Tertiary time. The Bay of Guayaquil is very shallow and may represent a slightly depressed part of the Zorritos-Tumbes plains. That Tertiary deposits extend some distance to sea in front of the Zorritos-Nigritos shore line is more than probable. This conclusion is borne out by the fact that, for many years past, mariners have repeatedly reported the occurrence of oil patches on the surface of the sea some miles out from the present shore line. Further, such reports are not confined to this part of the coastal shelf, for quite recently similar evidence has been noted in the region of the Lobos Islands, located in front of Pacasmayo. Jn fact, so marked was the evidence that parties of Lima have recently attempted to take up territory on the above-mentioned islands, with a view to developing a petroleum industry. So far as is known to the writer, petroleum-bearing formations of the coast are confined to the Tertiary horizons. Even in the interior of Peru, in the region of Lake Titicaca, there is every reason to believe that the petroleum deposits of the Pusi section are likewise confined to the Ter- tiary. 236 ANNALS NEW YORK ACADEMY OF SCIENCES Locus of Agriculture and tts Dependency upon Physiographic Phe- nomena.—tIn these days, much is said about physiographic features as determinative factors in the location of human industries. A word con- cerning the distribution of agriculture on the coastal plain is worth while. Mention has already been made of the differential elevation which the coastal plains of Peru have suffered. Where the elevation has reached its maximum, or where the plain has been aggraded to a much greater ele- vation, as by the addition of lava flows, the present streams coming from the flanks of the oldland (the west slope of the Andes) have cut out verit- able canyons.. Time enough has not yet elapsed to permit these streams to widen their floors to any appreciable extent. Here and there in some ot the canyons, the stream has reached the underlying oldland and its rate of erosion has been retarded. Above this point, the stream has thus had some opportunity to do some side cutting and hence has widened its floor ; only at these points do we see man availing himself of the agricultural opportunities offered him. In proportion as the elevation has been less so in a general way, we may say that transverse master streams have had a chance to make floor room for the use of man. Let us take a few ex- amples to illustrate the principal stages and their physiographic bearing on the topic in hand. Probably the maximum elevation of the coastal plains of Peru was attained in the Mollendo-Arequipa, or Vitor, section. While formations entering into the structure of this section are in part sedimentary in origin, it is not assumed that the present elevation of its surface above sea level represents the actual uplift above that plane. lt is very probable that this part of the plain has actually been aggraded by a lava flow, or flows, the present surface of which never stood at sea level, although the lower part of the earlier flows may have reached the edge of the Tertiary sea. At all events, the Chile-Vitor stream coming from the high plateaus to the interior have, within the limits of the coastal plain, sueceeded in cutting a very deep canyon. Further, it has developed its widest floor near the contact of the inner edge of the plain with the pre-Tertiary oldland, or the foothills of the West Range of the Andes. Fer that reason we have Quercotilla, an agricultural com- munity and a village, located at this point. Within the remainder of this valley to the sea, we do not find a single town until we reach the present shore line, where Quilca, a small village, is not placed on the valley floor, but to one side in a deep-water embayment of fiord-like character. This little village is here for the dispatch by water of agricultural products from the valleys of Camana and Ocofia, neither of which possesses a safe port of entrance. Mention has also been made of the special case of AND THE FOOT NGE OF THE ANDES 5 ie han SCALE ° yh pant A ORI Ale t ie trg { (it pe py kiey * PAT U0 he a tan oar YO. 21 HDOOW SHE ALA Pi TWIAXX Givi ‘TIXX ava10A REN RIGS sk Ge aT PLATE XXIX ae * CHILE ae NEAR es PO ie ‘ , ee ht) N ’ oth \ F 138 , E d 4 i ’ ant hy 4 io ’ Pell Alt’ } . ) ) _ a - \ wat | ‘ < wv t - , Nit ve } , ( vy J 5 ay « " Say | q'tatts 1 it s fi sh ; ’ Asan $ 4 | ‘ a we iit De A ae Ane} A » i; RRS ] aes \ ce is f x ey L 5 ey ( eres " a 1 Ao oy x\ " lint \ sap ctad ee iy WILE ALAT “Nine ge ee eae feu oh 0 cit YALIAT 7 4 t eh Fs ; XIXX GiVIid ‘IIXX DNOAIOA SMONHIOS “GVOV “X “N STVNNV MARSTERS, PHYSIOGRAPHY OF THE PERUVIAN ANDES 937 Arequipa. Arequipa is some miles within the interior of the West Range of the Andes. It stands on the slope and near the edge of the lava flow, or flows, playing so important a role in the lithological content of the coastal plain to the southwest. To the rear of the city, and forming a most picturesque background, the majestic cones of Misti, Pichu-pichu and other voleanic ridges rear their heads to 19,000 feet or more above sea level, and 9,000 to 10,000 feet above the light-colored lava plains of the Arequipa campo. The Chile River coming from the high plateau to the interior passes between two of these cones, where it has cut a most picturesque canyon, pushing its way across a part of the mud flow until it reaches the contact with pre-Tertiary formations of the foothills. The important fact to note here is that where the Chile succeeded in making a valley-floor, namely, at the moment it reached the contact referred to above, there was located the city of Arequipa. Here we find agricultural pursuits well developed. In all of the southern part of Peru there is no prettier bit of landscape to be found than the campo of Arequipa. It is the Switzerland of Peru. That is to say, physiographic features were the determinative factors in the location of this important trade center. In the Valley of Majas, we again find that a valley-floor has been developed near the contact of' the inner edge of the coastal plain with the edge of the foothills of the West Range. Here has developed a large grape culture and the manufacture of wine; but from this locality across the plain we find little chance for the farmer. Not until the Majas passes through a deep, narrow gorge cut in the coastal ridge, or cadena de la costa, do we find an additional opportunity for the development of agri- culture. As soon as the Majas passes beyond the west slope of the ridge, it has greatly widened its valley, having cleaned away the sediments on the west flank and built for itself a broad, semi-triangular fluviatile plain, which has pushed seaward sufficiently to be a menace to the mari- ner. On this plain, we find the agricultural town of Camana located and a campo alive with agricultural activities. The plain on which stands the city of Lima and its suburban towns is but a repetition of the same physiographic sequences, the difference being that the Lima plain was built at the edge of the cordillera or foothills proper of the West Range of the Andes, while the latter developed on the seaward edge of an outlier of the same physiographic province. In the Valley of Pacasmayo, we find that the river cut its way to the sea and at the same time widened its floor throughout its entire course; as a consequence, no inconsiderable part of the entire floor from the edge of the foothills to the present coast line is under cultivation, or has been at various periods. ‘The amount of cultivation is, in some cases, de- 238 ANNALS NEW YORK ACADEMY OF SCIENCES pendent upon the amount of available water. In the Pacasmayo case we find an additional feature worthy of mention. At the contact of the coastal plain with the foothills, the valley has not been deeply incised. On both sides of the valley are extensive plains, composed largely of the waste from the edge-of the foothills, plus the original surface of the coastal plain. So near are these plains to the level of the river, but a short distance within the foothills, that water has been diverted from its legitimate place on the valley-floor to irrigate large stretches of plain on either side of the valley. Success in this attempt has been foiled, in part, by the lack of water for the extent of territory taken up on the one hand and the strong tendency to salinity of large tracts of the plains on the other. It is only where the coastal plain has undergone the minimum amount of elevation that we find the last distinct stage to be described. This is fairly well illustrated by the Etén-Lambayeque-Motupe plains, where we have a group of small streams coming from the interior to a slightly ele- vated plain. Only along the inner contact of the coastal plam with the foothills have these streams slightly incised themselves, but as they ex- tended outward and over the plain they actually spread their waste over large areas. Under such physiographical relations, we have the condi- tions for the development of the most important rice industry in the entire Republic of Peru. The rice fields occupy the fluviatilly aggraded portions and such adjacent parts of the original plain as may be reached by the amount of water available. The growth of sugar cane has also become an important industry. Where the main streams have formed a well-defined valley-floor, the predominant culture is of maize, alfalfa and the staple vegetable products for the markets of the principal towns. In the valley of the Piura, we find a slightly incised valley, the floor of which is occupied by cotton culture, as the foremost industry, throughout its upper stretches. Notable lack of water has limited the territory under cultivation. Piura and Catacaos are located at these points. Again we find this portion of the valley-floor near the inner edge of the coastal plain. The lower portion of the Piura Valley has spread itself over the plain. Not enough water reaches this part to assure crops. The town of Sechura is located near the mouth of this sand-laden water-way. The people here can maintain themselves only in part by agricultural work. Not a small part of the inhabitants is engaged in transportation, fishing and salt-mining. ANNALS N. Y. ACAD. SCIENCES VOLUME XXII, Puarns XXX Wie. 2 ry ie ait OM seh eneenne “MOOT Rp so: Hei. $s sttogts wa ‘aso alt nd OM nt ak rae anche Pony a EE aa t A aa “ { Hebkibire’: heal uk Fs: ty twain Pana f ot, tho ecb ants low Dain Asn) auld, Ho) MG PLATH XXX WEST CORDILLERA NEAR CERRO DE PASCU Fic. 1.—Limestone summits in the West Cordillera as seen in the Oyon- Cerro de Pasco Section. WEST CORDILLERA NEAR UTOTO Vie. 2.—The West Cordillera in the region of Utoto, showing limestones on the left and voleanics on the right. Ore bodies exist near the contact. MARSTERS, PHYSIOGRAPHY OF THE PERUVIAN ANDES 229 WEST SLOPE AND WEST RANGE OF THE ANDES We now come to the geographical province which I have designated as the West Slope and West Range or Cordillera of the Andes. So far as I am aware, no one has attempted to outline the physiographic history of this section, nor is it my intention to try to solve this problem in all its details ; nevertheless, I wish to present a few of the larger physiographic features and their variations as observed in various sections on the West Slope, with the hope at least that these may lead in the direction of a correct and final analysis. As a basis upon which to formulate our views, I shall describe what to me appear to be two fairly typical cross sections of the West Slope in central and southern Peru. These IJ shall designate as the Huacho-Oyon section and the Ocofa-Cora Puno section. Huacho-Oyon Section.—The town of Huacho les about sixty miles to the northwest of Callao, on the coast, in the valley of the Huari. You may ascend from Huacho up to the Huari Valley and its important branches to Oyon. Oyon stands at the base of the more prominent peaks of the West Range, hence the choice of the two names to locate the section to be described. The distance from Huacho to the east side of the West Range is approximately 35 leagues, 175 kilometers, or 105 miles. Begin- ning at Huacho, we find that the valley of the Huari has a broad flat floor at the present shore line. It extends inland some eighteen to twenty miles. It is the seat of important agricultural industries. At the apex of the Huari floor we find the river running in a deeply incised valley. Its tributaries have likewise cut canyon-like side valleys. On either side of the Huari Valley, at the coast line, we find that the foothills of the West Range come practically to the present coast line. Should we ascend to a point in the foothills from which a long-distance view of the upland surface may be clearly seen, we shall at once note a moderately even but highly inclined surface descending in the direction of the sea and ascend- ing in the direction of the culminating points of the West Range. As we pass in the direction of the snow-capped prominences to the northeast, the inclined surface gives way to prominences standing out in clear relief above the upland surface. These finally culminate in the peaks of the West Range of the Andes. In the Lima-Oroya section, the above physi- ographic features can be discerned, but they are by no means so clearly defined. If we look into the details of topographic expression on the West Slope, we find that it has been most minutely cut by steep, narrow valleys leading to the master transverse lines of drainage from the in- terior. The formations entering into the structure of the West Slope are 240 ANNALS NEW YORK ACADEMY OF SCIENCES Cretaceous or older sediments and volcanics. The latest formations, at least, are not younger than Cretaceous. We also have good reason to believe that the oldest formations entering into the coastal plain are not older than early Tertiary. It therefore follows that the topography of the West Slope must have been developed in post-Cretaceous time, or more accurately, between the period of uplift of the Cretaceous sediments and their associated volcanics and that of the initiation of Tertiary sedi- mentation. It was during this interval that the present topographic detail in its large phases as now expressed on the West Slope was deline- ated. From the data at hand, the successive steps seem to have been as fol- lows: At the close of Cretaceous sedimentation and voleanic activity the West Range of the Andes was elevated; the west side of the uplifted sec- tion was subsequently worn down to a poorly graded surface, at least in the Huacho-Oyon section; this stage was followed by a strong uplift and subsequent deep dissection, or erosion. This was followed, in turn, by the depression of at least the shoreward edge of the post-Cretaceous land- surface and the initiation of Tertiary deposition. In other words, Ter- tiary deposition took place on the partly drowned edge of the West Slope. Ocona-Cora Puno Section.—This section has been chosen for the rea- son that the observed facts show considerable variation from the one already described. The town of Ocofia is situated on the coast some seventy miles to the northwest of Mollendo. The point known as Cora Puno is located near the western edge of the West Range. It is one of the highest collection of voleanic peaks on the west of the Cordillera. Starting at Ocona, we shall run our section across the immense plain or desert of Cuno cuno to the valley of the Chorunga, thence up the Andaray cuesta to Cora Puno and the West Range. The formational and topographical facts are roughly shown in the accompanying section. At Ocofa, we have the pre-Tertiary coastal ridge facing the present shore line. On its west slope may be seen patches of Tertiary and post-Tertiary still inviting the attack of the Pacific waves. Passing inland, however, we find that, as we reach the level of the pampas, pre-Tertiary formations may be seen at many places sticking through the thin sheet of late sediments at elevations of 3,000 or more feet above sea level. As we pass towards the north we soon realize the fact that the surface of the Cuno cuno plain has once been covered in part at least by a sheet of mud-like lava. Only in the deeper: side-valleys leading to the Ocoha canyon can we see the sediments as recognized near the coast. This light-colored sheet now persists until we reach the steep slope that passes into the Chorunga Valley. Here the Chorunga as well ANNALS N. Y. ACAD. SCIENCES VOLUME XNII, PLATE XXXI Iie. 1 0 adie HAVE: ‘dboqbial sy ont to ee ae pai oh ona (ft OH, ae (iit ae % | anol aia ont to nitonO- hs “Cit rou) ity 46! lua hae’ Wane t ra, mihi s it baie Dw et WOE Ol aH } ‘ V5 4a i ie : : ¢ w it alt ou rif ae PLATE XXXI INNER EDGE OF THE PAMPA CUNO CUNO Ite. 1.—A view of the inner edge (inface) of the Pampa Cuno cuno (coastal plain) as seen from the stripped upland of the west slope in the region of the Valley of the Choringa-Ocona-Cora Puno Section. CUESTA DE IQUIPI, NEAR VIEW Fre. 2.—A closer view of Cuesta de Iquipi, or inface of the coastal plain as seen from opposite the Rio Grande Valley-Ocona-Cora Puno Section. Corro Puno MARSTERS, PHYSIOGRAPHY OF THE PERUVIAN ANDES QA suse eo ete [<<*] Diorite Cretaceous RSs Sandstone and [&2%e"] Lavas and Volcanic Ash of Corro Puno BSS] Basalt *Syay] Porphyry Ti Mud Lava Tertiary =| Sands Shales Wie. 1.—G@ecological section from Ocona to Corro Puno as the Rio Grande and the Ocona have succeeded in cutting through, not only the soft lava referred to above and the underlying sediments, but also have deeply incised themselves in the underlying pre-Tertiary formations of the partly buried foot- hills, or, more accurately stated, have cut through the coastal plain formations until in the Valley of Chorunga the line of erosion adjusted itself to a pre-Tertiary topography. That is to say, the Chorunga simply cleaned out the mass of sedi- ments and sheet of lava which had been dumped into it during Tertiary time. This is evidenced by the patches of lava that still hang in the little recesses on the southwest slope at various points: between the Rio Grande and the Andaray cuesta. Ascending the Andaray cuesta, we go but a. short distance up the slope before we again see: remnants of the familiar lava sheet of the Cuno cuno plain, still nestling on the slopes and in the protected valleys of the uncovered foothills lead- ing up to the base of the Cora Puno domes. At Andaray and Yanaquiqua, we can see the same facts well illustrated. Andaray is located in a depression in the foothills, the bottom of which is still in part occupied by the light-colored lava. From Andaray, we pass up another steep slope, or series of slopes, composed in part of pre-Ter- tiary rocks, only to find that we are again on a somewhat dissected plain which extends away to the east in the direction of Chuquibamba and north in the direction of the picturesque Cora Puno. Lithologically, this is the same type of rock as that which we have already seen in the Chorunga Valley, the Andaray cuesta and the inner edge of the coastal plain. It is my belief that the plains at the base of Cora Puno group of domes are but parts of the same lava plain that we have already recognized in a portion of the Cuno cuno desert. Continuing our journey to the north, we come to the somewhat rugged, jagged edge of a still higher plain that is made 949 ANNALS NEW YORK ACADEMY OF SCLENCES up of a vesicular scoriaceous black to gray lava. Standing on top of this is the picturesque group of Cora Puno domes, four in number. So far as personal examination was made, it was found that lower slopes of the domes were composed of a succession of ropy lava flows alternating with voleanic ash and cinder-like layers. Nestling in the valleys between the domes are fine glaciers. On the southwest side of the group, there may be seen a very fine icefall, with an exposed edge of not less than 200 feet. At the time of visiting this locality, the domes were well covered with snow. So symmetrical and smooth did the domes appear, even to their summits, that the writer felt that with a specially constructed snowshoe an ascent might be made without serious difficulty. The base of Cora Puno is about 14,000 feet above sea level. The domes are not less than 20,000, and one of them may reach the 21,000-foot mark. Before interpreting the physiographic value of this section, let us’ look at another running from Ocofa to Caraveli. In this case we start at the coast with the nearly buried coastal ridge facing the Pacific. Going inland we pass over a succession of plains arranged in bench-like order. The natives have applied names to each of these. Near the coastal ridge, the plain is composed of sedimentary formations, but we do not go far to the interior before we recognize the fact that the light-colored lavas form the only visible lithological unit exposed in the deeper valleys of the plain. Should we go from Caraveli to Atico, we will again cross the lava plains, but in the lower part of the Atico group of valleys we will see that sands and conglomerates become important members of the coastal plain deposits. Furthermore, after we leave the outliers at the shore line of Ocona, we do not come again in contact with the pre-Tertiary oldland until we pass through the Quebrada de Chin Chin to the Caraveli Valley. From Caraveli inland, the foothills ascend rapidly to the elevation of the basal plain of Cora Puno. Above this elevation tower, here and there, occasional peaks, most of which are quite different in topography from those of Cora Puno. They represent elaborate groupings of spires and pinnacles, at places so steep that snow seems unable to cling to their sides. Should any one of the enthusiastic mountain climbers who have recently achieved noted success in mountain climbing in central Peru care to in- vestigate this region, he will be sure to find opportunities for testing his skill, of which he had never dreamed. We see, then, in the first-mentioned case, the oldland, or a portion of the West Slope, has scarcely been buried beneath the Tertiary deposits and no small part of these is composed of mud flows capping the entire series. In the section to the west, however, we have seen that sedimenta- + DOMES OF CORA PUNO eS pare eae £ beth Fi mt ¥ iy tRO SOLIMANA pat Hr. AP eee ae CE : +9 a x pes a acacia eE Og” eae i ee ak FON base RRC -A view of Cerro Solimana, a neighbor of Cora Puno. Note the dif ag ropourapmic corm: 6 Ye aI a I Ng be Fe ae fab “aalt 4 } ane a Yai oe LEE SOD) 40 Pd x Wein LOR, OFT) i it fess avemnicen hl Nie ERE: Ms i a ve ANNALS N. Y. ACAD. SCIENCES VOLUME XNII, PLare XXNII Fie. 1 Fig. 2 MARSTERS, PHYSIOGRAPHY OF THE PERUVIAN ANDES 943 tion was predominant near the edge of the coastal chain and also in the lower part of the Atico Valley; but it should be added, in the latter case, that much of the upper part of the sediments may be post-Tertiary and associated with glacial waste. At least, in the Atico-Caraveli section, we see nothing of the oldland upon which these late sediments rest, until we approach the vicinity of Caraveli. From these sections, it would appear that the West Slope in this locality was never reduced, or worn down to a grade, whereby even an approach to an even sky-line was attained, as noted in our northern section, but was occupied by valleys and hills of . large dimensions when Tertiary sedimentation and volcanic eruption be- came prominent phenomena. The pre-Tertiary foothills were only partly buried in sediments, but later they became almost obliterated by the ac- cumulation of mud flows of great thickness and areal extent. It is thus evident that the physiographic history of the West Slope is by no means simple. To get a clear insight into the possible variations and succession of events, sections should be studied in every department and province facing the Pacific coast. In the main, however, I believe that we have noted the chief succession of events and the principal factors involved, namely: (1) The development of an oldland surface upon Cretaceous and probably older formations and corresponding to the western slope of the West Cordillera. (2) Its elevation and deep incision of valleys. (3) The depression of the shoreward portion, which, during Tertiary time, was in part at least covered with Tertiary deposits, and portions of which now form the present coastal plains. (4) The differential elevation of the coastal plain deposits and the extension of the drainage of the oldland to the present coast line. In all probability, no small part of the original area of Tertiary deposition may now be beneath sea level. HIGHLAND PLATEAU OR INTERCORDILLERAN BELT To the belt of country lying between the eastern and_western ranges of the Andes, I have applied the terms Highland Plateau or Intercordilleran Belt. These terms are used in the broader sense and for the need of better ones. Although presenting a complex geologic history, it can be traced from central and northern Bolivia throughout the entire length of the Peruvian Republic and into Equador; its width, in general, varying from 100 to 200 kilometers. DA 4. ANNALS NEW YORK ACADEMY OF SCIENCES In the south of Peru, it is in part occupied by the most picturesque inland body of water of which the South American continent can boast, the well-known Lake Titicaca. To the southwest of Lake Titicaca, we have a somewhat dissected area drained by the Rivers Blanco, Canco- marca and Maure. To the southeast, we find the extension of the broad plains leading to La Paz, Bolivia. To the northeast, the Titicaca plains soon melt away in the foothills of the East Cordillera, whose long un- broken front, as seen from Lake Titicaca, presents a spectacle of moun- tain scenery that is not duplicated until we reach central Peru. To the northwest, we find that the Titicaca plains give place quite rapidly to a highly dissected belt. The streams coming to the Titicaca basin from the northwest are short, and their floors are very much ageraded, representing arm-like extensions of the enormous pampas to the north of the lake. As soon as we pass over the drainage divide at La Raya, we drop into an area that has been deeply cut by the drainage system of the Apurimac. It represents the headwaters of the Tambo and Ucayali, the latter join- ing the Marafion, and it, in turn, emptying into the Amazon. Of this stretch of country, the writer is personally acquainted with the Cuzco section, or rather, the belt etxending from La Paz, Bolivia, to a cross section through Cuzco. Let us search for a locality where we can get a clear sweeping view, both to the northeast and southwest. To do this, we must ascend to the upland surface into which the Urubamba and Apurimac systems of drain- age have deeply incised themselves. Various points of vantage may be found in the highlands overlooking the valley of the Urubamba be- tween Cuzco and the Pueblo Urubamba, or from high points overlooking the valley of the Apurimac. Selecting a high point among the latter, and looking to the northeast, we will see, in the foreground, as it were, reall two ranges looming up before us, and far above the level on which we are standing, one lying between the Apurimac and the Urubamba, and the other between the Urubamba and the head-waters of the Madre de Dios. While topographically these form parts of the East Cordillera, I am inclined to believe, from a geological point of view, that only the range lying northeast of the Urubamba-Villcanota valleys belongs in the Hast Cordillera proper. One could not wish for a more varied bit of mountain scenery than that from Sicuani to Cuzco. To the northeast, we have the pinnacle, spire and-knob-like topography that characterizes a region of great volcanic activity. To the southwest, however, we have a very dif- ferent expression, that which characterizes a belt occupied largely by sedimentary formations, here and there disturbed by local volcanic in- trusions. PLATE XXXIII INTERCORDILLERA NEAR CUZCO, LOOKING EAST Fie. 1.—A view of the sky-line of the intercordillera belt in the region of Cuzco, looking east. INTERCORDILLERA NEAR CUZCO, LOOKING SOUTH Fig. 2.—A view of the sky-line of the intercordillera belt in the region of Cuzco, looking south. u lik LY ATALT ; , a ; 4 Z . i ee 4 . C100 page QM TOOL COND AAR pac COMET 10 soiges off oi ila 4 rtallifrt0» tod ni andy to ouitl 7a, ont) | Yo wolr Be cea Se en A A ene ee iets dene eo ; Mee a iuk HTUO# OVLAOUA: 0 yD ATA Asien, aaogaurlvee : ibe thas wl b109" rotab odd to. ontit vale: “atts aul oy, \ a i ’ t ' ue te ny i } - / } ; ye ( i 1 A 5 th Aa he Oe ay ae F r ee) ts 5 J c fora ANNALS N. Y. ACAD... SCIENCES VOLUME XNII, Phare NXNIII ¢ ae SE Ree gin MARSTERS, PHYSIOGRAPHY OF THE PERUVIAN ANDES Q45 From the point of view selected, we see a billowy mass of ridges whose summits correspond roughly to the level of those on which we are stand- ing. This view passes across the Intercordilleran Belt, while the former, in part, belongs to the East Cordillera. From a physiographic point of view, we thus have, first, the Intercordilleran Belt, a highly dissected plateau, longitudinally drained by the Apurimac and its branches head- ing against the east flank of the West Cordillera; second, the East Cor- dillera, behind which run the Urubamba and Villcanota systems, also in longitudinal position ; but these finally succeed in cutting their way across the final range to join the Ucayali, which passes west of north through the Department of Loreto and drains for many miles the inner edge of the Amazonian plains. It is on this river that the important inland city of Iquitos is situated. Cerro de Pasco Section —Let us now examine the topographic charac- teristics to be observed in a cross section passing through the mining town of Cerro de Pasco and connecting with that which we have already de- scribed under the name of the Huacho-Oyon section. In this way, we can get a long-distance view of the principal features of these geographic provinces extending from the coast to the East Cordillera. As we have already stated, Oyon stands on the western slope of the West Range. Let us now go from Oyon to an advantageous point of ‘view in the Uchuc- chacua section, where we may obtain a clear view both to the east and to the west. Such a point may be found near the main trail from Oyon to Cerro de Pasco and not far from the “pass,” with an elevation between 16,000 and 17,000 feet above sea level. To the west and southwest will be seen the western slope of the West Cordillera, as already described. To the east, we will see, some 3,000 or 4,000 feet below us, a stretch of rolling undulating country extending to and beyond Cerro de Pasco. Occasional knobs and ridges may be seen at various points standing on the plain and serving as guides and mileposts to the explorer. From the same point of view, on a clear day, the East Range may be seen, away to the east and northeast of Cerro de Pasco. Here may be found the grandest and most magnificent group of ridges and peaks rearing their summits far above the line of permanent snow. There are peaks here that pass far above the 20,000-foot level. Into this Intercordilleran Belt the headwaters of the Huallaga have cut their way southward into the vicinity of Cerro de Pasco, having pushed through a part at least of the Hast Range. It is important to note here that the Huallaga did not succeed in cutting its way through the entire Hast Cordillera to join the Ucayali, but was forced to maintain a longitudinal position as far as Huanco. Then it broke through the East Cordillera to join the Marafion IAG ANNALS NEW YORK ACADEMY OF SCIENCES far to the north. It thus maintains a position approximately parallel to the Ucayali throughout a long distance. The Cerro de Pasco plain is quite clear as far south as Oroyo. In this part of the plateau we again find a pretty sheet of water, but very much smaller than the inland Titicaca. From this body of water, known as Junin, the drainage, instead of going to the Huallaga, now goes to the southeast, the main line of drainage being the so-called Mantaro. From Oroya to Huancayo, the Mantaro has made a most picturesque valley, on the floor of which are located many prosperous agricultural communities. Ascending the highland in the region of Huancayo, we can see to the east the picturesque Cordillera de Marca Valley, really a part of the Hast Cordillera, while to the west may be noted the West Range presenting an almost unbroken front. From Huancayo, the Mantaro continues its course to the region of Mayoc, where it breaks through the frontal East Range and deflects again to the northwest for some distance, but finally elbows its way across the last eastern barrier near Huaribamba and joins the Apurimac. We thus find here another illustration of a pattern of drainage which suggests an adjustment to structure, such as may be seen in many other regions the world over. Relative to the section between Mayoc, or the first elbow of the Man- taro and the Cuzco-Abancay region, I may add, from data obtained from explorers, that it 1s apparent that it is a broad highland belt into which the east tributaries of the Apurimac have incised themselves until they have cut the entire section into a veritable labyrinth of canyons. In other words, it is simply a continuation of what we noted to the southeast of Cuzco. Casma-Huaraz-Huacaybamba Section —Let us now look at a section through the northern part of the Department of Ancachs beginning with Casma on the coast and ending at Llata on the Marafion. At Casma, and far to the north and south, the foothills of the west slope form the coast line. The main transverse valleys have been aggraded near the present shore line, but we do not go far to the interior before the flat floors are replaced by steep-sided V-shaped valleys. Here again we have the rapidly descending surface of the West Slope; to the interior, this culminates in the West Range, which in turn overlooks the valley of the Huaraz. The highest points of the range in this section do not exceed 15,000 feet. From an examination of the Raimondi map, it will be seen that it is well defined throughout a large part of the Department of An- cachs. Note the fact, however, that in the place of the broad plain, as found in the Cerro de Pasco section, occupying the belt between the Hast MARSTERS, PHYSIOGRAPHY OF THE PERUVIAN ANDES 247 and the West Cordillera, we have a comparatively narrow valley running parallel with the West Range. It is drained by the River Huaraz, occu- pying a longitudinal position throughout some three provinces, and finally in the Province of Pallasca turning to the west to join the Santa, which empties into the. Pacific. Continuing our section to the east from Huaraz we find in the place of the deeply dissected intercordilleran plain another range of enormous proportions and running parallel with the one already described. It teaches elevations of more than 20,000 feet at various points. From its east slope, however, we now see the much-dissected Intercordilleran Belt extending to the valley of the Marafion, a distance of perhaps seventy miles. We thus have in this section two ranges on the west side of the Intercordilleran Belt, instead of one, as in other sections noted; further, instead of an intercordilleran lowland, as noted in the Cerro de Pasco sec- tion, we have a deeply dissected belt whose uppermost surface has a marked descent towards the valley of the Maranon. To the east of the Maranon, we again have a somewhat softened expression of a continuation of the East Cordillera. : Let us now consider the position of the Maranon drainage system. It heads to the south in the Department of Junin in a collection of lakes within the limits of the Intercordilleran Belt. Following the main line of drainage, we find that it, like many others, skirts the east edge of the intercordilleran highland for many miles to the north, and finally, in the Department of Amazonas, turns east to join the Amazon system, of which it is an important member. Note again the longitudinal position of the Marafion with reference to the geographical provinces as outlined. To the east of the valley of the Maranon, the Raimondi maps show a fairly well developed range separating the drainage of the Huallaga from its western neighbor. Just what this range is, geologically, is unknown to the writer, but it is believed to correspond to the East Cordillera. Piura-Huangabamba Section——To the east of the Piura-Salitral River the long western slope is absent, and we encounter immediately the con- tinuation of the West Range with a rather abrupt slope, though softened in its topographic expression, as compared with sections to the south. This is followed by the intercordilleran highland, already thoroughly cut to pieces by parts of two systems of drainage. The section of the Inter- cordilleran Belt between Huangabamba on the north and Chota on the south is drained by the rivers Huangabamba and Chotana. These be- come confluent in the river Chamaya, which empties into the Marafion. Note again the longitudinal position of the Huangabamba and Chotana rivers with reference to our geographical provinces, while the Chamaya is typically transverse in position. D48 ANNALS NEW YORK ACADEMY OF SCIENCES In the north part of the Department of Piura, we have some unex- pected topographic variations. By consulting the Raimondi map, it will be seen that the part corresponding to the West Range has projected a long spur to the southwest, known as Cerro da La Brea, or Amotape Mountain. It is composed of Cretaceous shales and dioritic intrusives. It is flanked on both sides by the coastal plain formations. Note also the arrangement of drainage between the spur and the headwaters of the Huangabamba. The rivers Chira and Catamayo represent the main line of transverse drainage, while its tributaries the Suipira, Quiros and Macara represent the longitudinal drainage of the somewhat depressed section of the intercordilleran plain. HAST RANGE AND EAST SLOPE Concerning the East Cordillera, we have already seen in our various views that it forms a prominent range of enormous altitude at various points and maintains its identity throughout the entire length of the Peruvian Republic. Its highest altitudes exceed those of the West Range. The most picturesque views may be had from the region of Lake Titicaca, where, on a clear day, it may be seen passing north from Bolivian terri- tory into Peru and presenting apparently an unbroken front, as far as the eye can reach, to the northeast. It maintains its identity throughout the departments of Puno and Cuzco, reaching enormous altitudes in the Department of Junin, where it is deeply cut into serrated forms. We then can follow it through the department of Huanuco and San Martin and finally into the Department of Amazonas, where it apparently sep- arates into two ranges and continues as such into Ecuador. While no great elevations are attained in northern Peru, they assume again in central Ecuador their old-time grandeur. Concerning the East Slope of the East Range, I regret very much that I had little opportunity to see enough of this geographic province to war- rant personal description in any detail. Only in southern Peru have I penetrated the East Cordillera to a point where a far-reaching view of the foothills and, beyond these, the great stretch of rolling and undulat- ing lowland, may be had. The foothills proper do not occupy a very wide belt. They quickly descend to an elevation of not more than 4,000 feet and probably less, where, from the long-distance view at least, one would Judge we should encounter the inland edge of the great Amazonian plains. These occupy the east portion of the Department of Puno, a very large part of the Department of Cuzco and nine-tenths of the Department of Loreto. It is on these eastern slopes and the huge plains below that we find uncivilized tribes of Indians, or the native “salvaje.” MARSTERS, PHYSIOGRAPHY OF THE PERUVIAN ANDES 24.9 TopoGRAPHIC EXPRESSION AS RELATED TO THE GEOLOGY OF THE PERUVIAN ANDES PLAINS AND SECTIONS Zorritos-Lambeyaque Plain.—The formations of this division of the coastal plain are, so far as known, Tertiary in age. They are entirely sedimentary and are composed in the main of clays and sands with occa- sional pebble and conglomerate beds. From a section prepared by me. between Fernandez and the shore line, something like 3,000 feet of sedi- ments can be calculated. More than 2,000 feet have been penetrated by the drill in search of new petroleum-bearing horizons both in Zorritos and Lobitos. I wish here to emphasize the point that the only localities where a topographic expression resulting from uplift and subsequent erosion may be found are confined to the three localities, Zorritos, Lobi- tos and Nigritos, where petroleum in large quantities has been obtained. Each one of these places is located on the east limb of a somewhat broken or locally faulted anticline. ‘The formations as seen in the Fernandez section are regarded as lower and middle Tertiary. In the Lobitos, it is believed that we have middle and upper Tertiary. In the Paita section, reference has already been made to the occurrence of a series of sands and conglomerates resting unconformably upon a mass of red clays. These deposits contain some fossils, apparently the same as those living at the shore line to-day. They are now 250 to 300 feet above sea level. As we pass in the direction of Piura, we can see little mesa-like elevations. ‘These were found to be composed of the red clays seen in the lower part of the Piata section. It is thus evident that the red-clay deposits were elevated and eroded before the deposition of the conglomerates referred to above; and, further, the erosion must have been largely confined to the outer half of the Paita section. It is important to note another fact at this point. At a number of places in the Paita-Piura Pampa or Despoblado, as the natives are accus- tomed to call these plains, you cannot fail to see small areas strewn with shells, all of which appear to be specifically the same as those living on the present shore line. This would suggest that, in very late geological time, the Paita-Piura Plain was beneath sea level, and that it was subse- quently elevated to its present position. As we pass to the south, the despoblado has been barely scratched by the Piura River and the surface subsequently littered up with wind-blown sands. That is to say, topographically, the plain is so young that it yet has the same expression as when it emerged from sea level. This con- 250 ANNALS NEW YORK ACADEMY OF SCIENCES dition is maintained throughout the Desert of Sechura and the Despo- blado de Olmos. From Chiclayo and Etén south, we have but a rim of the coastal plain left. It attains considerable proportions in the mouths of the valleys of Pacasmayo, Chicama and Santa Catalina. From a topo- graphical point of view, we have here but a narrow expressionless plain, with its seaward edge rapidly retreating under the attack of the Pacific waves. The two prominent points where topographic expression relieves the monotony of the plains are Cerro or Silla de Paita and Cerro de Yllesca. These are, in a word, half buried outhers of the foothills of the West Cordillera. Similar cerritas may be seen in the broad opening of the Pacasmayo Valley. All these were little islands in a Tertiary sea. They are composed of Cretaceous shales and sandstones which have under- gone metamorphism under the effect of intrusives. Chincha-Olmos Plain.—From Chincha to Pisco, we have the same monotonous plain. It is not until we reach Pisco that we find a little topographic relief and this time again associated with the uplift of a series of light gray to cream-colored clays. We need, however, only to go a little distance inland to see a continuation of the pampa in the direc- tion of Ica. Should we follow the coast, we should find that a consider- able area is occupied by the outliers of the foothills, but these are quite modest in topographic expression. These pre-Tertiary hills continue to the mouth of the river Ica and attain considerable width. ; I wish to turn aside here for a moment to refer to the Peninsula of Paracas, a short distance south of Pisco, since its geology is somewhat unique. Some time ago, coal was found in the cliffs of the peninsula, and a company was formed to exploit the deposits. Examination of the waste brought out of some of the prospects revealed the occurrence of true Carboniferous plants. It is the only locality known to the writer where undoubed Carboniferous and coal-bearing measures occur. So far as my observation goes, the Peninsula of Paracas is by far the most ancient “morro” on the entire coast of Peru. It is my belief that the formations entering into the remainder of the coastal chain, or Cadena de la Costa, are, geologically, much younger. Passing to the other side of the shore ridge of which Paracas is the northwestern extension, we find the lowland facing the actual foothil!s of the West Cordillera cut into by the Ica River. Here we are relieved to find another good section of the same light-colored clays as seen in the Pisco section. As soon as we reach the Pampa de Huayuri, we are again greeted by an enormous stretch of high plain, and its monotony is only relieved when we reach the modest canyon cut in it by the Rio Grande. MARSTERS, PHYSIOGRAPHY OF THE PERUVIAN ANDES O51 On the southeast side and close to the coast is the Cerro de Yungi, an- other pre-Tertiary outlier. From this point to the termination of the Chincha-Olmos section we see again just the young Tertiary plain wigs y eut into by the rivers de Acari and Yuaca. Ocona-Moquequa Section—In the southern section, which may be aptly termed the Ocona-Moquequa section, we have a repetition of the same topographic expression and the same formations involved, with one additional feature, and that is the role played by the mud flows. As already indicated, they form avery large part of the formations entering into the coastal plains extending from the Valley of Caraveli to the Pampa de Clemisi. The special feature of the south section is the per- sistence and strong relief of the coastal chain. It extends practically throughout the entire length of the coast line. I have indicated that the formations entering into the structure of the Cadena de la Costa are the same as appear in the north. While this may be true in some paris, I am inclined to think that, in the south, other formations than those of the foothills of the cordillera so far noted may be found in the Cadena. de la Costa. Huacho-Cerro de Pasco Section.—t\ have already called attention to the occurrence of an evident west slope facing the Pacific Ocezn and the coastal plains. It has a marked inclination towards the sea. It is thor- oughly cut to pieces by a network of transverse valleys. It is only when you ascend the uppermost edges of any of these valleys that this feature becomes apparent. Nevertheless, above the general sky line of the slope, we can see many elevations. The geology of the above section is as follows. Near the coast the formations are sandstones and shales into which have been intruded two: types of volcanics. These are then followed by a broad band of crystal- lines, probably diorites and related types. These are, in turn, followed. by shales and sandstones. Large coal deposits are to be found in this part of the section. The shales and sandstones extend to and beyond Oyon. In the vicinity of Ututo, the sandstones and shales are replaced by an enormous thickness of limestones. Into the limestones and the: inner edge of the sandstone and shale formations, enormous volcanic masses of at least two kinds have been thrust. In this section at least,. the volcanics form most of the crests of the West Cordillera. The sedi-. mentaries, especially in close proximity with the volcanics, have been folded and crushed on a very large scale. The eastern slope of the West Cordillera is remarkably well defined. Limestones are here turned up on edge, and adjacent to this horizon we- find the volcanics. Each formation presents its own type of topography. 952 ANNALS NEW YORK ACADEMY OF SCIENCES At some points along the eastern slope, the escarpment is of such a char- acter as to suggest faulting on a very large scale. Continuing across the Intercordilleran Belt, consisting of an undu- lating plain, some 3,000 to 4,000 feet below the summits of the West Cordillera, we find the formations involved in its structure to be sand- stones, shales and limestones, through which, at various points, knobs of voleanic rock have pushed their way and now form a part, at least, of the principal relief of the Cerro de Pasco lowland. The town of Cerro de Pasco is built on the slope of one of these knobs. Here we find a mass of voleanic rock in contact with a large body of limestone. It is on or near the contact that the famous ore body is located. Going northward over the intercordilleran lowland to Goyllarisquiseca we again encounter coal-bearing formations. Just what are the stratigraphical relations be- tween these andthe coal-bearing measures of Ututo and Cajatambo is not known. Beyond this point the writer has not penetrated the wilds of the East Cordillera. From data obtained from prospectors and en- gineers, I have reason to believe that the principal formations involved in its structure are very much older than any we have seen in the West Cordillera. Probably Devonian and Silurian and older terranes asso- ciated with a huge mass of intrusives make up the great part of the Hast Range. By way of comparison, I may add that the formations found in the Lima-Oroya section duplicate, in the main, the Huacho section, both in succession and kind. Oroya stands at the south end of the Cerro de Pasco lowland. Here we have the expected limestones and shales and associated intrusives noted to the north. Ocona-Cora Puno Section.—Starting at Ocofia, we meet first of all the outhers within the limits of the coastal plain, or the Cadena de la Costa. Ascending the shoreward escarpment, we pass over the ridge and on to the edge of the enormous pampa which we already know under the name of Cuno cuno and have recognized as an intergral part of the Coastal Plain province. Should we pass into the canyon of the Ocona, we should see the Tertiary sediments resting unconformably upon a series of sand- stones and shales. Inland along the line of our proposed section, the Tertiary sediments give way to the development of an enormous mud flow, which, on the inner edge, or escarpment overlooking the valleys of the Chorunga and Ocofia, is not far from 1,000 feet in thickness. In the canyon of the Ocofa, in the region of Piuca, the pre-Tertiary sedi- ments recognized nearer the coast have been largely replaced by various types of volcanics. There are at least three types to be found in the Oconha and Chorunga valleys, namely, a dioritic, a trachytic and a basaltic type. Their succession of volcanic activity was probably in the order MARSTERS, PHYSIOGRAPHY OF THE PERUVIAN ANDES 953 named. From the valley of Chorunga to Andaray, the dioritic forma- tions constitute the principal formation. Ascending the highland above Andaray, we again find a broad plain extending away to the southeast in the direction of Chuquibamba. The plain is composed of the same sort of mud flow as we have seen on the inner escarpment of the Coastal Plain overlooking the Chorunga Valley. Through this protrude various knobs of diorite, the same horizon as we saw in the trail from Chorunga to Andaray. Into this plain the west tributary of the Majas River chas cut a deep valley not only through the mud flow, but well into the under- lying crystallines. On this floor rests the city of Chuquibamba. Continuing our course to the north and northeast, we come finally to the edge of the lava flows forming a veritable platform on which were built the four confluent domes of the Cora Puno composed of lava, ash and scoria. A very large part of their slopes is above permanent snow lne. These domes undoubtedly surpass 20,000 feet in altitude. From the northeast slope of Cora Puno we have before us a gently ascending plain, with cerros appearing here and there above the general sky line. Among these are the majestic Solamana, Leon Wachang and others with unpronounceable Indian names. And here let me say, if any enthusiastic mountain climber wishes to test his real ability, he should not miss trying the spires of Solamana. The plains between the peaks are, in part, at least, made up of a light-colored mud flow, lithologically the same as that seen on the other side of Cora Puno and the inner edge of the Coastal Plain or the Iquipi cuesta. Occasional small knobs of limestone protrude through this sheet. Also, where some of the streams have cut to any great depth, we sometimes find limestone exposures. That is to say, the light-colored lavas, the dark basal lava platform and the superimposed cones of Cora Puno, probably rest upon a floor of limestone on the north- east and dioritic crystallines on the southwest of Cora Puno. Such of the smaller spire-like hills as were examined were found to be volcanic. In a word, then, comparing our sections, we shall find that we have a similar succession and order, with the exception that in the Huacho-Oyon case the Coastal Plain is absent. Physiographically, Cora Puno and the series of snow-covered domes to the southeast and northwest are situated well up the West Slope and not far from the West Cordillera. The im- portant mining camp, Caylloma, is located on the east slope of the West Cordillera, into which the headwaters of the Apurimac are now cutting their way. To the northeast from Caylloma to the Cuzco region we have the broad- ened Intercordilleran Belt, literally cut to pieces by the labyrinth of val- leys oceupied by the various tributaries of Apurimac and the Villcanote. O54. ANNALS NEW YORK ACADEMY OF SCIENCES To the southwest from Cora Puno, we have the West Slope extending to the edge of the Coastal Plain, then follows the Coastal Plain to the Cadena de la Costa, the latter now bordering the present shore line. — If time would permit, a section passing from Mollendo through Are- quipa to Puno could be shown to duplicate,.in the larger phenomena, the facts already brought out in the Ocofa-Cora Puno section. I wish, how- ever, to say just a word concerning the Titicaca region. he lake occu- pies a portion only of the Intercordilleran Belt. Within this basin, Ter- tiary sediments have been deposited. There is reason to believe that the Titicaca basin represents an area the depression of which was associated with down-faulting on a large scale. It probably extended from the north end of the Tertiary Titicaca well down to La Paz. Further, the Tertiary deposits rest upon limestones and shales. We are probably war- ranted in correlating the latter with the limestones and shales of the Cerro de Pasco lowland. To the west of the lake, we have the shales and limestones extending to the divide of Cerros de Toledo, where we again come in contact with voleanic intrusives, the true core of the West Cor- dillera. The belt of limestones and shales on the east slope of the West Cordillera has been traced to the Cuzco section. It physiographically belongs to the Intercordilleran Belt. The section of Forbes brings to light the same physiographic features, the Coastal Plain, the West Slope and West Cordillera, followed by the Intercordilleran Belt and finally the East Cordillera. NoTEs oN Harty MINING IN PERU To attempt to discuss the mineral resources of Peru in detail is not my intention at this time. I wish, however, to say one word relative to the early history of mining, its initiation by the Incas, its subsequent devel- opment during the period of Spanish rule, and finally to present a brief geographical and geological correlation and distribution throughout the West Cordillera and the Intercordilleran Belt. While the Incas as a race were decidedly agricultural and pastoral in their vocations, they were, nevertheless, not ignorant of the use of the precious metals. This is proved by the occurrence of gold and silver vessels discovered in their notable monuments, known under the name of “huacos.” The huacos are large quadrangular and pyramidal earth- works. They were probably used in connection with religious rites and ceremonies. While these constructions, or monuments, may be counted by the score near the coast, and usually are located on the floor of the broader valleys near a locality affording protection, the best preserved MARSTBERS, PHYSIOGRAPHY OF THE PERUVIAN ANDES 955 examples seen by the writer are to be found in the Valley of Santa Cata- lina (Salaverry). Very fine huacos may also be seen near Lima and within a day’s ride from the Capital. Quite recently a pair of vases, each apparently beaten out of a solid piece of gold, was found in a hhuaco near Lima. From the Valley of Nasca, I have seen a large col- lection of gold bands, undoubtedly used as wristlets and as ornaments for the head. All these were discovered in the’ interior of a huaco. Hence all must have been of Inca manufacture, and from the crude metal ob- tained from its original source by these people. With the invasion of Pizarro and his followers, and the subsequent establishment of Spanish rule over the Inca people, we have to note the introduction of a new regime. The invaders were primarily interested in the discovery and accumulation of the precious metals. On the other hand, the native was agricultural and pastoral in habit. As soon as the Spaniard had gained a knowledge of the gold-bearing possibilities of their newly acquired territory the labor problem became an important one. There could be but one outcome or solution on the part of the in- vaders. The natives were thus pressed into mining services and driven away from their chosen vocations. For a period of something like three centuries the native remained in a state of servitude to the Spanish rulers and people. If we can put any credence in various sources of informa- tion concerning the suffering of this inoffensive race, it seems most re- markable that they did not revolt and make at least one heroic effort to free themselves from the servitude into which they had fallen. It was ‘during the period of the Viceroys that the gold-bearing resources of Peru became legend. That Peru was rich in the yellow metal was evident to the invader Pizarro upon his first survey of Cajamarca. Here he dis- covered and took charge of gold and silver to the amount of some three millions of soles in actual value. To speak in general terms of the mining activities under Spanish rule, we may conveniently group the localities of maximum activity in the fol- lowing manner: (1) Cajamarca-Pataz Section. In this region, old Spanish prospects may be counted by the score in numerous valleys. (2) Huaraz-Cajatambo-Cerro de Pasco Section. While the old Spanish workings are not yet known to be as numerous in this section, it is never- theless certain that large amounts of both gold and silver were obtained, especially in the region of Cerro de Pasco. (3) Cotahuasi-Andaray Section. In this section and as far west as the. valley of the Chala, there are many abandoned prospects. In Andaray and the Cotahuasi vicinities, as well as in the Cerro de Pasco region, much work has been done since the establishment of the republic. 256 ANNALS NEW YORK ACADEMY OF SCIENCES (4) Cuzco-Cotabamba Region. The region of Cotacamba and adjacent valleys was the scene of great mining activity in the early Spanish days. Many of these prospects have likewise been worked in late times. (5) Poti-Sandia Section (East Cordillera). This is known to con- tain not only a large area of mineralized territory—gold-bearing quartz veins—but also an abundance of placer on the eastern slope of the Hast Cordillera. : (6) Huanaco Section (Hast Cordillera). Just a word as to the distribution of the principal mining localities and their relation to the geology of the Cordillera. After seeing a large number of ore-bearing sections in the south, center and north of Peru, the relation and association of zones of maximum mineralization with certain formations becomes very clear indeed. Let us return for a moment to our Huacho-Cerro de Pasco section. Near the coast, there are a few intrusives which have pushed their way through the sandstones and, shales. Apparently associated with these volcanics are gold-bearing veims which have been prospected from time to time. Nothing of very great value, however has been found here. The Spanish prospector did not find this little crop of veins very attractive. He was not slow to hunt more pleasing ground further to the interior. It is not until we pass to the zone of the West Cordillera, where there are enormous intrusive bodies bordered by limestones and shales, that we find ore-bodies of large dimensions. On the west side of the Cordillera, we have a group of silver-copper-gold veins, some of which can he traced for more than a kilometer, with widths approaching 20 meters. Should we pass over the divide to the East Slope, from which we see the Cerro de Pasco lowland, we will find both on the slope and in the valleys leading to the crests of the Cordillera another group of veins that are undoubtedly associated with the east contact of the intrusives with the limestones and shales. Throughout this region may be seen many old plants (arrastres) and the still more ancient quimbolete, where the ores were treated for the recovery of gold and silver. The amount of visible tailings show to what extent the early prospectors worked. Since the foundation of the repub- lic, the native has likewise continued to work in this region, but, of course, in the old-fashioned way. It is not at all uncommon to find Indians in possession of solid silver utensils hammered out of a single piece of silver. Should you visit the plazas of any of the villages in these sections, you would find the silversmith present with his little collection of silver utensils and ornaments of various kinds. Most of the metal he obtains from miners of the same locality. MARSTERS, PHYSIOGRAPHY OF THE PERUVIAN ANDES 257 Passing over the lowland of Cerro de Pasco, we come to the noted deposit of copper-silver ore now in part the property of the Cerro de Pasco Mining Company. The ore-body appears to be on or near the con- tact of eruptives with a thick series of limestones. According to late statistics, the amount of silver taken out of the surface portion of the Cerro de Pasco deposit between the time of its discovery (1630) and the end of the nineteenth century amounts to some 450,000,000 ounces. If we should return to the coast from Cerro de Pasco via the Cerro de Pasco Railway to Oroya, and thence via the Central Line to Lima and Callao, we should pass through other mining localities such as Rio Blanca and Morococha. Here again the respective ore-bodies are closely asso- ciated with contact phenomena such as have already been described. In the region of Matacana, a repetition of the same sort of occurrence may be seen. In the region of Lima, there are copper-bearing ores associated with eruptives in contact with limestones and shales. That is to say, the occurrences of ores in each section can be correlated both geographically and geologically. Let us look for a moment at a south section across the Cordillera, say from Mollendo to Puno and the east. At Mollendo, we find that outliers of the Cordillera, or Cadena de la Costa, are composed in part of gneissic and granitoid masses, which are probably intrusive in sandstones and shales. Within the gneissic zones occur small copper deposits. While an attempt has been made to develop some of these, they have never reached the productive stage. Passing into the edge of the foothills in the region of Carabaya, just below Arequipa, we again find a band of erystallines bordered on each side by sandstones and shales. Prospect- ing on a small scale has brought to light small bodies of copper-silver ore, mostly located in the crystallines, while the sandstones and shales are reported to carry coal beds. It is not until we reach Lagunillas that we again come to a region containing ore deposits. Here are silver- gold-copper ores in a belt extending from Santa Lucia to Marivillas. This is to be correlated with the Caylloma silver occurrence already re- ferred to. Here, as noted in other sections, we have eruptives associated with limestones and shales. These are probably to be correlated with the Morrococha Belt on the Lima section. In the Titicaca basin to the south of Puno, ore deposits carrying silver and lead have been prospected. I have not seen these, but I am informed that the formations involved are the expected limestones associated with eruptives. Geographically this should be correlated with the Cerro de Pasco occurrence, as it lies in the Intercordilleran Belt and possesses the same lithological relationships. 258 ANNALS NEW YORK ACADEMY OF SCIENCES Within the Titicaca basin, we have an additional occurrence which so far is not known to occur anywhere in the northwest extension of the Intercordilleran Belt. In this basin, there is a considerable series of sediments deposited upon lhmestones which are regarded as Cretaceous. The former are probably Tertiary. They contain petroleum, but devel- opment work has not been carried far enough to determine the areal extent of the oil sands. The only work of any moment has been done by the Titicaca Oil Company, backed by California people. Work has been suspended for the present. The gold-bearing veins of Poti as well as those of the Santo Domingo region to the east from Tiripata are associated with the eruptives and older sedimentaries of the East Cordillera. : It is also known that the region of Huanaco, to the north of Cerro de Pasco, is highly mineralized. From data at hand it would seem that the geological associations are similar to those of Santo Domingo. In the region of Huaraz and Recuay (valley of the Huaraz) there exists another mineralized belt containing silver, copper and gold ores, as well as lead ores. It is quite undeveloped. These are associated with intrusives of the West Cordillera. Iron, copper and silver ores also occur in the West Cordillera, to the northeast of Piura. They are associated with dioritic intrusions and bosses in the midst of a heavy series of shales and sandstones. ‘ RESUME In the Andes of Peru, we can easily recognize a series of parallel and well-defined physiographic provinces, which in the larger sense are defi- nitely related to the geological development of the Andean Range as a whole. . Whatever has been the succession of physiographic changes in the development of the Andes as a unit, there has evidently taken place, at least on broad lines, a marked adjustment of drainage to structure, thus affording a longitudinal and transverse arrangement, or pattern, such as may be easily recognized in many other continental mountain systems. A comparison of observed and recorded facts with reference to the occurrence of ore-bodies in Peru proves that they are generally associated with contact phenomena. ns of the » Acalemy consist of two series, viz. © lls ede in general wile the alee ee » mniform es a eee dollars oe volume. The articles ‘The ees of aoe upon ee length and ie shies of illus- ations, and may be learned upon application to the Librarian of the c The author receives is ee as soon as his Daper has ay tre a Fellows and ee a ‘The e & Tue LIBRARIAN, New York Academy of Sciences, care of Aree American Museum of Natural History, ; New: York, Noe NEW YORK PUBLISHED BY THE ACADEMY A to OctopER, O19: Ae ae aitias 0 FES 6 101 THE NEW YORK ACADEMY OF SCIENCES (Lyceum or Naturat History, 1817-1876) OFFicERs, 1912 President—EMERSON McMituin, 40 Wall Street Vice-Presidents—J. EDMUND WoopMAN, FREDERIc A. Lucas CHARLES LANE Poor, R. S. WoopwortH Corresponding Secretary—Hunry EH. Crampton, American Museum Recording Secretary—Epmunp Oris Hovey, American Museum Treasurer—HENrRY L. DoHERTY, 60 Wall Street Inbrarian—RaLPH W. Tower, American Museum Editor—EDMUND Otis Hovey, American Museum SECTION OF GHOLOGY AND MINERALOGY | Chairman—J. E. WoopMaN, N. Y. University Secretary—Cuag_Les P. Berkey Columbia University SECTION OF BIOLOGY Chairman—FREvDERIC A. Lucas, American Museum Secretary—WiILL1AM K. GreGory, American Museum SHCTION OF ASTRONOMY, PHYSICS AND CHEMISTRY Chairman—Cuares LANE Poor, Columbia University Secretary—F. M. Prprrsen, College of the City of New York SECTION OF ANTHROPOLOGY AND PSYCHOLOGY Chairman—R. 8. WoopwortH, Columbia University Secretary—F REDERIC Lyman WELLS, Columbia University The sessions of the Academy are held on Monday evenings at 8:15 o’clock from October to May, inclusive, at the American Museum of Natural History, 77th Street and Central Park, West. {ANNALS N. Y. AcAD, ScreNcES, Vol. XXII, pp 259-266. 15 October, 1912] NOTES ON THE STRUCTURE AND GLACIATION OF OVERLOOK MOUNTAIN By Nein EK. Stevens? (Read by title before the Academy 6 May, 1912) CONTENTS Page TUDETROG ITC ION 5:5 'S Wars e.b oars SPesG Onc Ggere) Cee are GTR NG SPCR te ene ent a ne nee 259 SUIST CHUN Cees cee tay ei ate icone al eae asa E Mey TU EL WALSaneoealie os, SoGccua a\dvs Goa es 260: ‘GUGHOM | abc's SAG Bea oes Sie eee eee Or or Oe TENE Hl ee ar et 262 WOTPD MDGS ees eh Sih Sie ates eit eaters ave ae ena 263 ID PRD BIRAS seo Sises Bla SG Se ae eee te none Ciel EEO RRS PU RHES AUDSCNE (ee AU rp 264 CLAD) AEs is.6 ebace Bua) wicca tee ea Cea aE re tT a are ae 266 INTRODUCTION Overlook Mountain is the southern terminus of the great central mountain chain which forms the backbone of the whole Catskill system. On the east, the mountain rises precipitously above the low ground of the Hudson Valley, the land at the base of the steep incline on this side being only about six hundred feet above sea level. This commanding position gave Overlook for many years the reputation of being the high- est peak in the Catskills, although it is actually more than a thousand feet lower than Slide Mountain (Slide Mt. 4204 ft.; Overlook Mt. 3150 ft.). The position of Overlook in the Catskill system makes it of par- ticular interest; and the present paper, though it by no means contains a complete account of the geology of the mountain, is offered in the hope that the notes contained therein may be of service to future investigators and may, perhaps, stimulate an interest in the geology of this region. The writer wishes to acknowledge his indebtedness to Professor H. H. Gregory of Yale University for generous criticism and suggestions. The summit of the mountain (Fig. 1) forms a triangle, from the apexes of which project three main ridges with smaller ridges between. The principal ridge stretches southwest for a distance of about four miles and ends in a series of three lower peaks, separated from the main peak by the Meads gap. The southern ridge is short and slopes rather sharply to the level ground of the valley. The northeastern ridge, on the other hand, is short and high, merging into the Plattekill Mountain 1 Introduced by James F. Kemp. (259) 260 ANNALS NEW YORK ACADEMY OF SCIENCES at an elevation of about 2500 feet. Aside from this narrow ridge, Overlook is separated from the mountains to the north by the valley of the Saw Kill. . STRUCTURE Unlike most mountains, the Catskills consist of a succession of piled up, nearly horizontal strata, showing that they are really but the remains of a great interior plateau now greatly eroded and dissected by water.? VA aS IL Fa w= — SS = PGE SE SFA Fie. 1.—Overlook Mountain and adjacent territory. A portion of Kaaterskill Quad- rangle, New York U. 8S. G. 8. Topographic Sheet. Reduced. Scale: 1/83332 Except for a layer of hard conglomerate which caps many of the higher peaks, the mountains consist almost entirely of alternate layers of soft red shale and harder sandstone. The sandstone is sometimes thinly laminated and frequently cross-bedded ; often however it shows a very even 2RALPH STOCKMAN TARR: The physical Geography of N. Y. State. New York, 1902, p. 41. STEVENS, GLACIATION OF OVERLOOK. MOUNTAIN 961 texture and is practically free from evi- dence of fine stratification. This latter variety of sandstone, generally designated as gray or blue flagstone, is extensively quarried for paving stone. This alternation of soft, easily eroded, shales with the more resistant sandstone gives rise to the abrupt ledges, flat moun- tain tops and “terraced” sides so character- istic of the Catskills. Wherever a layer of shale has been exposed, the surface has been quickly eroded down to the next layer of sandstone. Nowhere are these features better shown than on Overlook Mountain. Note in Fig. 1 the plateau-like tops of the lower peaks and the terraces of the south- ern ridges. Fig. 2 shows a section through the two lower peaks, between the Mead’s Gap and the Bear Clove. This section extends from where the strata first appear above the gla- cial soil of the Woodstock valley, altitude 660 feet, to their summits, altitude 2100 feet. It is altogether probable that some of the strata, represented here as of uni- form thickness, are really somewhat lense- shaped. ‘This however could not be deter- mined owing to the prevalence of glacial deposits. As will be seen from the figure, the cap of these peaks consists of nearly 600 feet of sandstone. The upper portion of this cap is hard and rather coarse, but the stone becomes softer and more finely lamimnated below. The lower layers of sandstone are characterized by a _ cross- bedded structure and contain numerous streaks of red shale too thin to indicate in the section. Besides the layers of sand- stone and red shale there are, as indicated by the figure, two layers of bluestone; the upper outcrops at an altitude of about 1500 On weet er ey Cate = O° 67 ONS . oO . away te: ON OOD Gyo ioya et Reh ls. 1° ° . es Qe onOh eats ————————————— 460° Shale = Bluestone WN Covered] _|Sandstone|:2=~*.] Fic. 2. —Diagrammatic section through the two peaks west of Meads Gap | 262 ANNALS NEW YORK ACADEMY OF SCIENCES feet, the lower, which outcrops on the southern slope of the mountain, is in this region covered with glacial drift. The strata as a whole dip gently west-northwest. GLACIATION Two glacier streams have swept over this region: the Continental Glacier, and later the more shallow Hudson Valley Glacier. The geo- logic structure of the region makes it somewhat difficult to trace satis- Fic. 3.—Map of Overlook region, showing direction of glacial strie (indicated by arrows) Contour interval 200 feet factorily the course of the glaciers; for the sandstones and soft shales of which the mountain is composed have retained glacial markings only in exceptional localities; while a considerable portion of the surface is, of course, covered with glacial drift. The map (Fig. 3) shows the direction of the striations left by the Hudson Valley glacier in the Overlook region. It affords an interesting 3A, A. JULIEN: “The Excavation of the bed of the Kaaterskill.”” Trans. N. Y. Acad. Sci., Vol. 1, p. 24-27. 1881. STEVENS, GLACIATION OF OVERLOOK MOUNTAIN 263 example of the extent to which the direction of the ice near the surface can be affected by the topography. A variation of more than ninety degrees in the direction of the motion of the ice is indicated by the strize on Overlook itself. Just northeast of Overlook, the course of the glacier was nearly south (8. 20° W.). A portion of it, however, flowed over the ridge between the Plattekill and Overlook Mountains and through what is now the upper valley of the Saw Kill in an almost westerly direction, moving nearly southward again through Meads Gap and at Shady. The ice moved almost directly west up the Woodstock Valley, its course be- coming gradually more southerly as it passed over the group of low hills, known as the Ohio Mountain,* and the region farther south (S. 60° W. near Glenford and 8. 40° W. at West Hurley). MORAINES While glacial action has probably not greatly altered the general out- lines of the mountain, the valleys have been more or less filled with gla- cial deposits. A moraine nearly a mile long fills much of what was once the much deeper valley south of Meads Gap; while the upper valley of the Saw Kill, between Overlook and the mountain ridge to the north, is: filled to a considerable depth with morainic material. Both of these are, in reality, parts of a large moraine which extends westward from Over-. look. As the moraines of the Catskills have been but little studied, a brief account of the material composing them may be of interest. About: eighty per cent of it is of local origin, consisting of the sandstone, shale and conglomerate found throughout the Catskills. Of these materials, conglomerate is the least common and forms less than ten per cent of the whole. About one-half the local material consists of bowlders of various sizes, with which pebbles and gravel are mixed with no sign of stratifica- tion. From this it is apparent that water has played no part in the deposition. The foreign material consists largely of quartz and several kinds of granite, with occasional pieces of water-worn, stratified rock and some sandstone containing brachiopodous shells. Some shells picked up in the bed of the Saw Kill, about two miles from its source, have been iden- tified as Spirifer arrectus,® a species characteristic of the Oriskany sand- 4This mountain is called ‘“‘Tontshi Mt.” on the U. S. G. S. topographic sheet. It seems, however, that this must be an error, as this elevation is locally known only as “Ohio Mt.”: while the name “Tontshi” is applied to the much higher peak, left un- named on the government map, just east of Ticetonik. 5The writer is indebted to Professor Charles Schuchert of Yale University for the identification of this specimen. 264 ANNALS NEW YORK ACADEMY OF SCIENCES stone, a thin layer of which outcrops in the Little Catskills near the Hudson River due east from Overlook Mountain. It also appears in the Helderbergs to the northeast. One rather unusual variety of metamorphic rock, several pieces of which were found, has been kindly identified by Professor J. F. Kemp of Columbia University, who writes: “It is a type of rock fairly well known in the Adirondacks. It has obviously been pretty well crushed and granulated, but it is a member of the anorthosite series, which when unchanged has large rectangular crystals of labradorite in a mass of small crystals of augite. . . . The rock outcrops at the very head- waters of the Schroon River in North Hudson township and also in the Keene Valley. I think it probably occurs in many other places, where it has not yet been specially observed. . . . In Bulletin No. 138 of the New York State Museum, on page 43, under the name of “The New Pond Locality,’ you will find a brief description of the rock.” That both the specimens just referred to are characteristic of the Adi- rondacks, together with the fact that all the metamorphic rocks found are common in those mountains, indicates that most of the foreign ma- terial in these moraines is of Adirondack origin.® DRAINAGE Except for a small portion of its eastern slope, Overlook is drained entirely by a single stream, the Saw Kull, which forms a loop, some ten miles or more in length, extending nearly around the mountain (Fig. 1). From its source in Echo Lake at the base of the northeastern ridge, it flows directly southwest through Shady, then east through the Wood- stock Valley to a point directly south of its source. In the course of this loop, the Saw Kill receives the smaller streams which flow from both sides of the Overlook ridge. Owing to the small size of their watersheds, many of these smaller streams are dry during a part of the year. The rainfall is much greater in the spring than at any other season, and this gives rise to floods which make the erosive power of these streams much greater than it would otherwise be. The floods at this time are greatly increased by the melting of the winter’s snow, and there is added the erosive force of the ice as it breaks up. 6 These moraines yielded more foreign bowlders than did the ones on the north side of the district, in the valley of Schoharie Creek, described by J. L. Rich in the Jouwr- nal of Geology, vol. 14, p. 113, 1906, especially p. 120. Mr. Rich found but one bowlder of gneiss. It may be that later glacial action, radiating from local centers, had con- cealed earlier bowlders, brought in from the north. Mr. Rich’s paper has also some general comments on the movement of the continental glacier, and, at the outset, upon the present conditions of rainfall. STEVENS, GLACIATION OF OVERLOOK MOUNTAIN 265 Although no data as to the actual stream-flow on Overlook are at hand, a comparison of the flow of neighboring watersheds can not fail to be of interest in this connection. Fig. 4 gives the comparative discharge of the Schoharie, Esopus and Catskill Creeks for the different months over a period of years.‘ It shows that more than one-third of the total run-off of these streams occurs during two months, March and April. As the curves represent an average of several years, they give but little idea of the size of some of the floods. In the Schoharie, for instance, the maximum daily discharge in November, 1907, was 13,100 cubic feet per Cyonio = SOPs: -——— SCHOHARIE =~ Wie. 4.—Average discharge of the Catskill, Esopus and Schoharie Creeks for each month over a period of years Expressed in second-feet per square mile run-off. Taken as follows: Catskill Creek, South Cairo, N. Y., for period 1901 to 1905; Esopus Creek, Kingston, N. Y., for period 1901 to 1906; Schoharie Creek, Prattsville, N. Y., for period 1905 to 1908. Data com- piled from Summary of the Climatological Data for the United States by sections. Sec- tion 104 (Weather Bureau) ; Water Supply and Irrigation Papers Nos. 166 and 202; and the Report of the State Hngineer and Surveyor, State of New York, 1907 and 1908. second, while for the same month, in 1908, the maximum was onlv 268 cubic feet per second. The valley of the Saw Kill is primarily of erosive origin. Like many 7 Records of rainfall, kept at Reservoir No. 1 of the Kingston City Water Works, in- dicate that the rainfall near Overlook is much like that of the other watersheds men- tioned. 4 266 ANNALS NEW YORK ACADEMY OF SCIENCES other valleys in the Catskills,* however, it was partly filled with debris by the continental glacier, so that now the Saw Kill, for the first five miles of its course, has cut its way through a moraine of varying height, 70 ft. at Shady, and 40 ft. half a mile below its source. This fact is clearly shown by the character of its bed, which is strewn with huge bowlders of bluestone, sandstone and conglomerate, together with smaller ones of granite and quartz, the harder ones still showing the marks of ~ glacial action. Ecuo LAKE® Echo Lake, the only considerable body of water near Overlook, is clearly of glacial origin. It is a shallow pond, about three hundred yards long by two hundred wide, and about eighteen feet deep in the deepest part, situated in the angle formed by the Plattekill and Overlook Mountains, just at the base of the high ridge connecting them. Across this deep valley, the moraine forms a huge natural dam which holds back the water of Echo Lake. The lake is thus bordered on three sides _ by high wooded ridges; while on the west, extending out over the mo- raine, is a swamp larger than the lake itself. As is to be expected from its situation, Echo Lake is apparently de- creasing in size rather rapidly. The swamp on its westeru side is slowly invading its waters. On this side, too, the lake is being narrowed by the action of its outlet, the Saw Kill, in cutting back through the Glacial drift which forms its bed. The pitch of the Saw Kill, which falls 1200 feet in the first four miles of its course, together with the floods men- tioned above, makes this erosion relatively rapid. In addition to this cutting away on its lower side, the lake is being rapidly filled in from above. The silt and leaf mold washed from the steep mountain ridges above the lake are deposited in the still water, and the amount of this material is very considerable. On the north and east this deposit forms a bed extending into the lake for more than two hun- dred feet and reaching a depth of four or five feet. Here the deposit is, to a considerable extent, protected from further action of the water by a dense growth of the yellow pond lily, Nymphea advena, for which the fine silt and leaf mold furnish a favorable substratum. The combined effect of these agencies in reducing the size of the lake is so great as to make it probable that, at no very distant date, Echo Lake will be oblit- erated. WasuHineron, D.C. — 8 Joun C. Smock: “On the Surface Limit or Thickness of the Continental Glacier in New Jersey and Adjacent States.” Am. Jour. Sci., vol. 25, pp. 339-350. 1883. ® Also known as Sheu’s Lake. ons and oe of researches, ae ‘with the Tec- s and similar matter. ‘volivne are ected. separately, each in its own cover, and ed i in goalies: on an Saves of three per year. The pa of y- “The me receives ee Se eues as soon as his paper has aos ae te of se arma ove the te each Woes ; ris devoted fo ee Menicins Volume ITI to Zodlogical encase e. The price is. one dollar per part as issued. AL | publications are sent free to’ Fellows and Active Members. The THE LIBRARIAN, New York Academy of Sciences, care of agaeiiean Museum of Natural History, New Yorks oN. PEDAL LOCOMOTION : “AND OF THE OF THE LIMBS IN HOOFED ANIMALS | inn1aM K. Gregory THE NEW YORK ACADEMY OF SCIENCES (Lyceum or Natura History, 1817-1876) OFFICERS, 1912 President—EmeErson McMiuuin, 40 Wall Street Vice-Presidents—J. EDMUND WoopMAN, F'REDERIC A. Lucas CHARLES LANE Poor, R. 8. WooDwoRTH Corresponding Secretary—Henry HK. Crampton, American Museum _ Recording Secretary—Epmunp Oris Hovey, American Museum Treasurer—HeEnRY L. DoHzERTY, 60 Wall Street Iibrarian—RauPu W. Towser, American Museum Editor—EpmunpD Otis Hovny, American Museum SHCTION OF GEOLOGY AND MINERALOGY Chairman—J. E. WoopMAN, N. Y. University Secretary—Cuartes P. Berkey Columbia University SECTION OF BIOLOGY Chairman—FREpDERIC A. Lucas, American Museum Secretary—WiLuL1amM K. Grecory, American Museum SECTION OF ASTRONOMY, PHYSICS AND CHEMISTRY Chairman—Cuar_Es LANE Poor, Columbia University Secretary—F. M. PEDERSEN, College of the City of New York SECTION OF ANTHROPOLOGY AND PSYCHOLOGY Chairman—R. S. WoopwortH, Columbia University Secretary—FREDERIC LiyMAN WELLS, Columbia University The sessions of the Academy are held on Monday evenings at 8:15 o’clock from October to May, inclusive, at the American Museum of Natural History, 77th Street and Central Park, West. [ANNALS N. Y. Acad. ScrENCES, Vol. XXII, pp. 267-294, Pl. XXXIV. 18 October, 1912] NOTES ON THE PRINCIPLES OF QUADRUPEDAL LOCOMO- TION AND ON THE MECHANISM OF THE LIMBS IN HOOFED ANIMALS By WILLIAM K. GREGORY CONTENTS Page NATTA CECE [N@ Inmet ameen ter Tepe Aca apes screen, cls meg eivariclrs lar WleGieca! ai bud) Auoie Water aly srisice Boag leks 268 PUMP SeEESALCeMa se COMMPOWMIG NEVES! tn cece cede ais celsleelc weiss veces cee 269 Hvolutionary stages of quadrupedal locomotion..................+.++---- 270 Factors of long distance travelling power in ungulates.................. 270 TENG NTO ORS Si Seen oe eee ee BS 6 ot Os RC RI eae 271 Adaptations for minimizing waste of energy..............+-+-.eee0: PAT IEGMTRETITITT. & Soci etree ae RISO Bee cee ee ee ar a ee ea 271 Rropulsionyand= thexcenter Of Gravity. . ss. 0s5o-s - scr s+ eee ke oe eos 272 Sinuous movement of the body in running........................ 272 Spiral configuration of limb bones and articular surfaces.......... 273 AO Seamer Oil OMe eee rete ees aot A sere cates ayo icra Sieud Sie jes suswaieod! » aUolsterwheu a ate ae 273 Renaiihummrerionuota the Mim See = ice sides cad as cee ea aee wo oie alte sin Sere 273 SYOESGle coreicer. do G24 ca luteic sen calc ARERR SkG MCMC cl asset ok cr ares ike eo pee a 273 SGN Oe titel CL Caeewayey en eya actrees caer eoevede cue soe. ale fe oie ecu eb eis se rue gue e le Hee Siere 274. eT SsalaO fannie raee oa remit an reetn Ala eos lie Cale ee ba ie cles ae 274 ANOGUE Ol -SUIBIGIO SSG Scie oo GcOte SRaraNe GOREN, ee een Rone a Pe OTE Pi ee 274. Acceleration increment of stride due to ballistic power of limbs.... 274 HFECHeA TON Mit yan Clgn SUT G] Comers ieee evo sia cielerts elalei's eaeieis cece; ay aveney ) cneusevcis suse erates eye epaue 276. Mechanical and physiological relations of “power” and “speed” in the loco- SOTVOUEL WE, OVO ICTUIS 5 ceo ele ee Bec ye ous SG ser UE en OS aoa ea eT! Conitrachle toreevor locomotive muscles: <2 5.46 fadeansc. ce os cess eee ces 276 Augmented by “hold and let go” arrangements................... DT PAM OUT sy Olen OMG AGT O Mars tercnstar ys doll MAN Face SSE udtlagc, wage ravens Rake a eva ele PST The “angle of insertion” and the principle of the parallelogram of forces. 277 Relations of speed of movement and angle of insertion............ 279 Variableness of the rotation component......................i... 219 “Moment of Resistance” and “Diminishing Load”..................... 280 Summary of mechanical and psychic factors in power and speed........ 287 Application of the foregoing principles to the study of the limbs of ungu- VAS Seer teaerer ew cree epee Laie rata aye co au Ne iraafeurer aya been chase Chonan ay ene aaolo eon aconaeica gee oer 282 Functional significance of the angulation of the limbs................. 282 Mechanics of the foot in graviportal and cursorial types.>............. 283 MerAarSO-KEMVOLal Tm CAlOS ace oieie trie se erelc wtciete aia clejeed sie alae snes 284 Otper-adaptuve contrasts! im the feet. .3......5 625.052.2502. c i ee eee 287 Graviportal and cursorial types of tibia. ............... cee eee eee es 288 Cravaportalvand cursoniall types of femur. 2.25... seece. cee tee alee be 290 Graviportalandscursorial bypes Of Pelvis... 5)... . .ctslcs eee cee betes 292, Graviportal and cursorial types of fore limb......................-.--- 293 Diagnostic vs. convergent evolution values of limb ratios.............. 294 Comparative: table of limb ratios. 22% 6... .6. 2b ote lee nee Facing page 294 (267) 268 ANNALS NEW YORK ACADEMY OF SOIENCES INTRODUCTION The movements and locomotive mechanisms of animals were the subject of a classic work by Borelli in 1680, entitled “De Motu Ani- malium.. .” The chief pioneers of modern research were the brothers Weber (Eduard and Wilhelm), authors of “Die Mechanik der mensch- lichen Gehwerkzeuge,” Gottingen, 1836. Marey, the author of the hand- book on “Animal Mechanism”! (1874), invented elaborate apparatus for analyzing and graphically recording the movements of the limbs of men and animals, while more recently the mathematics of human locomotion have been developed by O. Fischer and others. General reviews of animal mechanics and of human locomotion are given by Marey,” Haycraft? and Luciani. A very able analysis of the mechanism of locomotion in the horse is given in “The Horse in Motion,” by J. D. Stillman,® while the extensive series of photographs by Muybridge® record the actual positions of the limbs and body assumed in motion by ungulates and other animals. None of the above mentioned works considers the subject from the evo- lutionary point of view. Ryder,” and especially Cope,® pointed out cer- tain adaptations in the feet of ungulates, such as the reduction of the digits in the “Diplarthra” and the so-called displacement of the metacar- pals, and Cope used these observations in his argument for the hypothesis of the transmission of acquired characters. He also made the following very important observations :° “In animals which leap, the distal segments of the limbs are elongated; in those which do not leap, but which merely run or walk, it is the proximal seg- ments of the limbs which are elongated. “Animals which run by leaping are divided into those which run and leap with all fours, as Diplarthra, and those which run and leap with the posterior limbs only, as the jerboas and kangaroos. In both types, the distal segments of the hind limbs are elongated, and in the Diplarthra, those of the fore limb also. ale “Animals which do not leap in progression (elephants, Quadrumana, bears) are always plantigrade and have very short feet but elongate thighs and mostly tibias.” 112mo. London, 1874. 2 Op. cit. 3H. A. SCHAFFER: Text book of Physiology. Edinburgh and London, pp. 228-273, 1900. 4Physiologie des Menschen . . . Ins Deutsche tibertragen and bearbeitet von Dr. Silvestro Baglioni und Dr. Hans Winterstein . . . Siebente Lieferung. Jena, 1906. 5 Hxecuted and published under the auspices of Leland Stanford, 4to, Boston, 1882. 6 Animals in Motion . . . Third Impression 4to, London, 1907. 7Am. Nat., vol. 11, p. 607. 1877. 8“The Mechanical Causes of the Development of the Hard Parts of the Mammalia.” Journ. of Morphology, vol. 3, pp. 137-290. 1889. 8 Tbid., p. 151. GREGORY, QUADRUPEDAL LOCOMOTION 269 Cope neglected to follow up the important mechanical and adaptive corollaries of these facts. He merely drew the very questionable infer- ence that “those elements which receive the principal impact in progres- sion are those which increase in length [in phylogeny ].”?° In his paper on “The Angulation of the Limbs of Proboscidia, Dino- cerata, and other Quadrupeds, in Adaptation to Weight,” Osborn? con- cluded that “The straightening of the limb [in the Dinocerata, Proboscidia, etc.] is an adaptation designed to transmit the increasing weight through a vertical shaft. Correlated with it are the shifting of the facets into the direct line of pressure and the alteration of their planes from an oblique to a right or horizontal angle with relation to the vertical shaft.” Gaudry,’? in describing the limbs of extinct South American ungu- lates, endeavored to show how the pose of these animals could be inferred from a study of the facets——an idea which had been previously advanced by Osborn.** Gaudry also designated as “rectigrade” the pose of ele- phants and similar heavy forms which stand with straightened limbs and toes, resting the weight chiefly on the pad. The marked contrasts in the limbs and musculature between the slow- moving heavy-bodied ungulates and the slender swift-footed or cursorial types 1n various phyla, constitute a subject which will be discussed in considerable detail in the monograph on the Titanotheres by Professor Osborn. At his suggestion, the following notes, forming a part of the present writer’s studies on this subject, are now published, together with some of the drawings which have been made by Mr. Erwin 8. Christman, under the direction of the writer, for the monograph above mentioned. The writer is also indebted to Dr. W. D. Matthew for valuable criticisms and suggestions. LIMBS REGARDED AS CoMPOUND LEVERS The simple principle that the limbs of quadrupeds are compound levers and that the relative lengths of the upper, middle and lower segments © are adapted to specific loads, muscular powers and speeds, although well understood by students of human and equine locomotion, has apparently not hitherto been applied to elucidate the adaptive contrasts between eursorial and “graviportal”’** ungulates. 10 Loc cit. 11 Am. Nat., vol. xxxiv, pp. 89-94. 1900. 12 ““Hossiles de Patagonie. Les Attitudes de quelques Animaux.’’ Ann. de Paléon- tologie, tomei. 2 ‘‘Patriofelis and Oxyena restudied as Terrestrial Creodonts.” Bull. Amer. Mus. Nat. Hist., Vol. 13, pp. 270, 271. 1900. : 144This word has been invented by Professor Osborn to describe the conditions in heavy-bodied animals with long proximal and short distal limb segments. Q"0 ANNALS NEW YORK ACADEMY OF SCIENCES EVOLUTIONARY STAGES OF QUADRUPEDAL LOCOMOTION In the Stegocephalian stage of quadrupedal locomotion, the short limbs were held widely outward from the body, the humerus and femur were very short and the feet were spreading and flat. Crawling was effected in part by a sharp downward pull of a proximal segment (humerus or femur), thus tilting the body upward on the same side and throwing the weight on the opposite foot. The long axis of the body was meanwhile thrown into alternate lateral curves, the advancing fore limb being on a convexity, the advancing hind limb on a concavity. In the late reptilian or early mammalian stage, the feet were brought around partly under the body, the elbow and knee began to be drawn in, the scapula was rotated backward as the coracoid lost its connection with the sternum, and the body became well raised off the ground. According to a hypothesis advanced elsewhere by the writer,*® this process was asso- ciated with the acquisition of climbing, or semi-arboreal, habits, struc- tural vestiges of which remain in the partly divergent first digit and many other characters of early Eocene mammals."® The Lower Eocene ancestors of the various orders of peedirtes had probably all long since passed through these earlier stages of quadrupedal terrestrial locomotion, and at that time many of them had perhaps become more or less digitigrade. The primitive “Protungulates,” Meniscothervum, Periptychus, Pantolambda, may give us some idea of what the several ancestors of the subungulate series (Hyracoidea, Embrithopoda, Pro-. boscidea, Amblypoda) may have been like. They also preserve apparent traces of arboreal ancestry in the relatively short, spreading hands and feet, long limb bones, the humerus with large entocondyle and entepicon- dylar foramen, the undiminished power of pronation and supination of the forearm and many other characters. The Basal Hocene Huproto- gonia, the ancestor of the Condylarth Phenacodus, with slender subun- guligrade feet, represents a more advanced stage of evolution, in the direction of the Perissodactyls. FACTORS OF LONG-DISTANCE TRAVELLING POWER IN UNGULATES The primitive ungulates of the Lower Hocene were doubtless sur- . rounded by environmental conditions which set the premium of survival upon improvements in long-distance travelling power and in speed. These improvements have been attained in various ways and in the most 15 ‘The Orders of Mammals,” Bull. Amer. Mus. Nat. Hist., Vol. 27, p. 226, 1910. 16 MATTHEW: Arboreal Ancestty of the Mammalia, Amer. Nat., Vol. 38, 1904, pp. 811-818. GREGORY, QUADRUPEDAL LOCOMOTION Oral diverse lines of evolution,—in the elephant no less than in the antelope. The general factors of long-distance travelling power may be grouped broadly under the headings: (A) Endurance and (B) Speed. A. ENDURANCE Endurance may be measured either by (1) the length of time an ungu- late can keep in motion without rest or refreshment, or by (2) the. amount of reserve strength left after a stated expenditure of energy, or by (8) the relative quickness of recuperation. Endurance increases with practice. As metabolism increases, the muscles, lungs, heart and other organs of the thorax become stronger and larger. More rapid metabolism requires more food and larger digestive ap- paratus. Although the Lamarckian hypothesis in its crude form is very probably untenable, it is a fact that herbivorous animals have longer and heavier digestive tracts than carnivores. Moreover, as fast as the denti- tion has become adapted for the harder, less nutritious kinds of food, the digestive apparatus must have become more complex and wuch heavier. While the enlarging thorax and abdomen have made availal.‘e a great increase in energy, they have caused an even more rapid increase in total weight. With increasing total weight, the internal and external resistance to be overcome in locomotion also rises, and hence the margin between the total energy available and the energy required in progressing a given dis- tance-is lessened. In other words, endurance, the first great factor of long-distance travelling power, is directly proportional to the efficiency of the adaptations for minimizing the waste of energy. ADAPTATIONS FOR MINIMIZING WASTE OF ENERGY Higher efficiency in locomotion has doubtless been attained, first, by advantageous modifications of the organs of propulsion (such as are de- scribed below), secondly, by improvements in the supporting frame- work, thirdly, by improvements in the methods (a) of conserving the inertia of forward motion, (0) of taking up shock, (¢) of preventing dis- location and (d) of minimizing lost motion. Momentum.—The shocks and strains to which the locomotive apparatus is subject vary with the momentum of the body in motion. Hence as momentum is the product of mass by velocity, the shocks and strains experienced by heavy animals in rapid motion are very great, and devices for lessening them become conspicuous. Pee) ANNALS NEW YORK ACADEMY OF SCIENCES Propulsion and the center of gravity—Perfect quadrupedal locomo- tion,” says Stillman,*’ “requires uniform support to the center of gravity (of the whole animal) and continuous propulsion by each extremity in turn.” In walking and running, by the straightening or extension of the limbs, the center of gravity is raised and thrown in advance of the centers of support. The body thus falls forward, the center of gravity describing a curve of greater or less convexity, the forward motion being accelerated by the thrust of the propelling limb. ‘The first shock of the downward fall in the running horse is taken up by the forwardly stretched and slightly bent hind limb (Fig. 1) placed beneath or in advance of the center of gravity; the gradually stiffening muscles of the thigh and back checking the downward momentum (Still- man, p. 91). The rearing muscles thus come into play and serve to let the fore part of the body down gently. Dislocation of the fully extended forelimb in landing is prevented partly by (a) the crutch-lke action of the limb itself (which is slung from the converging fibers of the serratus magnus attached to the top of the scapula) and by (0) the contraction of certain muscles of the shoul- der, neck and back (Stillman, pp. 61, 62, e¢ seq.). ‘To these arrange- ments and conditions, Stillman attributes the absence of the clavicles in the horse. The center of gravity in the smoothly trotting horse describes a rela- tively flat trajectory, whereas in the “bounding” movement, or gallop, the center of gravity ricochets and the trajectory consists of a series of cycloids of marked convexity. This mode of locomotion, while very rapid for short distances, is too wasteful for heavy-bodied animals, which require a relatively flat trajectory and a maximum saving of inertia. Sinuous movement of the body in running.—By the bending backward of the pelvis, first on one side and then on the other, the thrusts of the femora are brought more nearly into line with the anteroposterior axis, while wrenching of the pelvis is prevented by the contraction of the longissimus dorsi of the opposite side (Stillman, p. 36). By this means also, the length of the stride is directly increased. The same sinuous motion of the body is associated with the “figure-of-8” movement of the limbs noticed by Pettigrew.*® . In this connection may be noted also the devices for avoiding “interfer- ence” of the limbs (e. g., “stifle action” of the iliacus, preventing the knee from striking the abdomen; oblique trochlea of the astragalus carrying the advancing foot around its fellow of the opposite side). Dislocation 17 Op. cit., p. 87. 18 Animal Locomotion, p. 39. 12mo. New York, 1874. GREGORY, QUADRUPEDAL LOCOMOTION 273 is provided against not only by the ligaments, but also by the metapodial keels, by the grooved trochlea of the astragalus, by the cnemial crest of the tibia, ete. Spiral configuration of limb bones and of articular surfaces—Good- sir, Pettigrew*® and others have shown that the articular surfaces of the elbow, ankle and calcaneo-astragalar joints are spirally warped surfaces which act after the manner of screws. The limbs as a whole also are twisted levers with the ridges and muscles arranged spirally. “This arrangement,” says Pettigrew, “enables the higher animals to apply their traveling surfaces to the media on which they are destined to operate at any degree of obliquity so as to obtain a maximum of support or propul- sion with a minimum of slip. If the traveling surfaces of animals did not form screws structurally and functionally, they could neither seize nor let go the fulera on which they act with the requisite rapidity to secure speed, particularly in water and air.” Lost motion—Lost motion through backward slipping of the foot upon the ground is provided against in the horse by the form and details of the hoof, and in the elephant by the plantar pads. Pendulum action of the limbs.—The brothers Weber held that in rapid locomotion the limbs swing freely as pendula, but Marey and later in- vestigators, according to Luciani,” hold that the natural swing of the leg is very largely damped and controlled by the flexor muscles. In favor of the view that there is some measure of analogy to the pendulum, we ob-- serve that in the horse, the center of gravity of the limb, corresponding: to the “bob” of a pendulum, is relatively proximal in position, and this: is associated with rapid oscillation of the limb, whereas in the elephant, the center of gravity of the limb is farther down the shaft, and here we have a slower oscillation of the limb. It will be observed that while the body is moving forward, the propelling limb is moving backward, and its own backward momentum, due to weight alone and to the pull of the ex- tensors, must be overcome by the forward pull of the flexor muscles and by the forward pull of the body as a whole. Hence the heavier the limb, the greater the force expended in overcoming and reversing the mo- mentum of each limb at the end of each stride. SPEED The speed of a quadruped or biped in motion is measured by the product of the length of the stride into the rapidity of the stride. 12 Animal Locomotion, pp. 23-24, 28, 29. 1874. 20 Op. cit., p. 126. 274 ANNALS NEW YORK ACADEMY OF SCIENCES The diverse adaptations in the limbs, considered as compound levers, are related to either or both of the factors, “length of stride” and “rapid- ity of stride.” LENGTH OF STRIDE Length of limb.—Length of limb is the first factor of length of stride. Tt is generally proportional to height at the shoulders and hips. Length of limb has been attained in cursorial animals by lengthening the lower and middle segments of the limb, in graviportal animals by lengthening especially the proximal segments of the limb. Angle of stride—Angle of stride is the second factor. It is measured by the are described by the lower end of the femur or humerus in swing- ing from the position of extreme extension to that of extreme flexion. A wide angle of stride not only lengthens the stride, but also enables each limb, first, to be placed in turn below or beyond the center of gravity in order to secure more continuous support for the center of gravity. and, secondly, it enables the propelling limb to exert its propulsive effort for a relatively long period. ; Acceleration increment of stride due to ballistic power of limbs.—Those portions of the stride that are due simply to the length of the limb and to the angle of the stride might, if determined, be illustrated by moving the inert limbs of a dead animal suspended in the air. In the slow walk of a biped, the successive positions of the legs might for our purposes be rep- resented by a series of inverted V’s (/\/\/\/\/\/\ ) with the lower ends touching. Each /\ represents a single step and two successive /\’s repre- sent a stride. In the rapidly moving animal, however, the stride receives a very considerable increment, due to the impetus imparted by the pro- pelling limb and to the forward motion of the body as a whole, which carries the forwardly moving foot to a position far in advance of its own unaided reach. This “acceleration increment,” as it may be called, in- creases with the velocity of the movement and is proportional to what we may designate the ballistic power of the limb. This ballistic power may be defined as excess propulsive power over and above that which is necessary to move the limbs as stilts and to support the weight of the body; it is expended in lengthening the stride. Ballistic power and the acceleration increment of the stride are measured by the length of time at least three of the feet are off the ground together during a single stride of a quadruped running at full speed. In Fig. 1, representing an elephant in rapid motion (ambling), it will be observed that three of the feet are never off the ground at the same instant; whereas Stillman’s Figs. 2-10 show a galloping horse in which at least three of the feet are GREGORY, QUADRUPEDAL LOCOMOTION 275 off the ground at once in seven out of nine phases of one stride. The same figures show that the elephant has three feet on the ground together during five-sixths of one stride, and during the remaining sixth the body is supported by one fore foot and the opposite hind foot, whereas the horse in question never has three feet on the ground together during the stride there pictured. Thus it will be seen that the “acceleration incre- Fic. 1.—Graviportal and cursorial modes of locomotion contrasted in the amble of the elephant (I) and gallop of the horse (II) I after Muybridge; II after Stillman ment” of the stride and the ballistic power of the limbs are at a maximum in cursorial animals and at a minimum in graviportal animals. The “acceleration increment” will no doubt increase also with the potency of the psychic motive and of the neural stimulus (cf., p. 282). In brief, the factors of length of stride are (1) length of limb, (2) angle of stride, (3) acceleration increment. 2"6 ANNALS NEW YORK ACADEMY OF SCIENCES RAPIDITY OF STRIDE Rapidity of stride, the second major factor of speed, is determined by the rate of oscillation of the limbs, especially of the proximal segments. The conditions determining rapidity of stride are discussed below (p. 281). MECHANICAL AND PHYSIOLOGICAL RELATIONS OF POWER AND SPEED IN THE LOCOMOTIVE APPARATUS CONTRACTILE FORCE OF LOCOMOTIVE MUSCLES The contractile force of a muscle (1. ¢., its ability to overcome inertia at a given instant) is proportional to the number of its contractile fibers, when these fibers are parallel to the direction of contraction. 'The force of such a muscle is therefore proportional to the sectional area of the muscle.*t In the case of weights lifted vertically by isolated muscles, the work (W) performed is measured by the product of the muscular force (F), multiplied by the distance (D) through which the load is hifted?? (W=F*X D). This distance is proportional to the length of the muscle,* for the “shortening” of a muscle is proportional to its length. Hence the total work performed will be proportional both to the length of the muscle and to its sectional area, and hence to the mass or the weight of the muscle.?* The work performed by a long muscle is greater than that of a shorter one of the same sectional area.2> Long and slender muscles such as the sterno-mastoid and the sartorius of man exert a small power over a long range; short and thick muscles such as the pectoralis major, the gluteeus maximus or the temporalis develop a relatively great power multiplied by a short range.2® The contractile force of muscle per unit of sectional area is much less in cold-blooded than in warm-blooded animals. It is lessened by disuse and extreme fatigue and is increased by exercise, and hence is dependent upon the nervous system and general systemic condi- tions. The contractile force is inversely proportional to the number of con- nective tissue fibers mingled with the striped muscle fibers. Hence muscles grade into tendons and ligaments. When a muscle is stretched, 21 HAYCRAFT in Schifer’s Text hook of Physiology, p. 242. 22 Toid., p. 245. 23 Toid., p. 244. 24 Tbid., p. 246; also Marey, p. 62. 1874. 2 FAAYCRAFT, p. 246. 26 MARRY, p. 62. 27 HLAYCRAFT, p. 243. GREGORY, QUADRUPEDAL LOCOMOTION any it serves partly as a ligament. All muscles in situ are stretched to a certain degree, and thus act as ligaments.** “Over extension” of the muscle is prevented by the inextensible connective tissue fibers.*° Ac- cording to Stillman,*° the length of the muscles cannot be increased by exercise, otherwise the tension necessary to prompt action would be lost. The contractile force is highest when a muscle is stretched to its full “physiological length” (that is the greatest length it ever assumes during life). As shortening takes place, the contractile force becomes less and less (Haycraft) .** Contractile force and speed of movement augmented by “hold and let go” arrangements.—Fick and Helmholtz showed** that the greatest force and velocity of contraction are developed when the movement of the muscle is checked during the initial stages and when the resistance is suddenly diminished. Amount of contraction.—The shortening of individual muscles is in general proportional to their length when in repose, but different in- vestigators give somewhat different estimates. “While Weber described a muscle as shortening 70 per cent. of its length, when unweighted, more recent observers incline to put the shortening at 20 to 30 per cent. of its length” (Haycraft).** Marey** estimates “the mean shortening of a muscle while contracting, when it is not detached from the animal,” as “about a third of its length when in repose.” Bishop’s estimate is one- fourth (Stillman, p. 31). “When the fibers are not parallel, but obliquely set, as in the gastrocnemius, we have a greatly extended trans- verse area of muscular fibers, which act therefore very powerfully, though, on account of their short length, they can exercise their pull but a comparatively short distance” (Haycraft) .*° THE “ANGLE OF INSERTION” AND THE PRINCIPLE OF THE PARALLELOGRAM OF FORCES In Fig. 2 (1), let AC represent a rod free to rotate around the point A in the direction CC’; let BD represent a contractile spring fastened at D, inserted on AC at B and forming the angle ABD (a). Assume that the length of BD is proportional to its contractile force; then from 28 Tbid., p. 245. 23 Toid., p. 242. 30 Op. cit., p. 32. 31 Op. cit., p. 242. 34 HAYCRAFT, op. cit., p. 248. 33 Op. cit., p. 244. 34 Op. cit., p. 62. 33 Op. cit., p. 242. 278 ANNALS NEW YORK ACADEMY OF SCIENCES the principle of the parallelogram of forces, we may resolve BD into two forces, the first AB acting in the direction of the rod AC and tending to press AC against its fulerum A, the second component BR acting at right angles to the first and tangent to the are of rotation BB’. The first component AB may be called the “centripetal component,” the sec- ond BR may be called the “rotation component.” In Fig. 2 (IL), the contractile spring bd is of the same length as before, but the angle of SS “i Fic. 2.—Diagram illustrating the direct relation of the angle of insertion (a,a’) to the “rotation component” (BR, br) and the inverse relation of the angle of inser- tion (a, a’) to the “centripetal component’ (AB, ab) and to the speed of the insertion point (proportional to BB’, bb’). insertion abd (a’) is increased; then the centripetal component ab will be less than AB, but the “rotation component” br will be greater than BR. Accordingly as the angle of insertion increases, the pull across the shaft becomes more direct, while the pull along the shaft decreases; in other words, the rotation component varies directly, the centripetal com- ponent inversely, with the angle of insertion. GREGORY, QUADRUPEDAL LOCOMOTION 349 Applying this to Fig. 6, Il, we see that in the horse, the muscles fig- ured are inserted at more open angles of insertion (a, B, y) than in the mastodon or elephant and that their rotation components are therefore relatively greater, the centripetal components relatively less. The foregoing principles were worked out independently by the writer, but Luciani*® gives similar principles for human locomotion. He states that not all of the muscular force is available for the movement of the skeleton, that this is only the case when the insertion of the muscle on the bone reaches almost a right angle, as in the case of the masseters, which can exert their whole strength in pressing the lower jaw against the upper jaw. He says that generally, owing to the conditioning form of the skeleton, the muscle is attached more or less obliquely, so that the direction of its fibers makes a more or less acute angle with the long axis of the bone. In all these cases, a great part of the total force of the muscle is lost as regards movement. In every case, however, whatever the form of the muscle or the size of the angle of insertion may be, by resolving the total pull into its components, in accordance with the law of the parallelogram of forces, one can determine how much of the total pull is expended in the movement of the bone, assuming the other bones to be stationary. Liuciani*’ also shows that the more acute the angle of insertion is, the smaller will be the component of rotation, and the nearer the angle of insertion approaches a aga angle, the greater will be the ee of rotation. : Fig. 2, we see that if DB contracts to DB’, the point of insertion will move from B to B’. If now the angle of insertion be increased to a’ and db (equal to DB) contracts to db’ (equal to DB’), then the point of in- sertion moves only through bb’, which is less than BB’. If the contrac- tion time as well as the distance be constant, then B will move faster than b; that is, when the rate of contraction and length of muscle are constant the speed of the insertion point varies inversely with the angle of insertion. It is also evident that if the angle of insertion and other factors re- main constant the speed of the distal end of a long bone will increase as the point of insertion is moved toward the head of the bone. (Because BC will be larger.) Variableness of the rotation component—From Fig. 2, it will be seen that the angle ab’d is somewhat greater than abd, that is, both the angle of insertion and the rotation component increase as the muscle contracts. 36 Physiologie des Menschen, Siebente Lieferung, p. 115. 31 Tbid., p 116. 280 ANNALS NEW YORK ACADEMY OF SCIENCES “MOMENT OF RESISTANCE” AND “DIMINISHING LOAD” In every lever, whether of the first, second or third order, the “power” and the “resistance,” acting along parallel lines, but in opposite direc- tions, are in equilibrium when the power multiplied by its effective dis- tance from the fulcrum is equal to the resistance multiplied by its effect- ive distance from the fulcrum. The “effective distance” is measured by a line passing through the fulerum and perpendicular to the line of direc- tion of the force. The product of a force multiplied by its effective distance from the fulcrum is called ds “moment.” m M W Wie. 3.—l. Hindfoot of an extremely cursorial type (Neohipparion) showing at the in- stant of greatest extension of the foot a low moment of power of the calf muscles (M X BB’) and a very high moment of resistance (W X B’ A’ v of the pressure of the tibia upon the ankle. II. Hindfoot of an extremely graviportal type (Mastodon) showing at the in- stant of greatest extension of the foot a much higher moment of power of the calf muscles (m X bb’) and a relatively lower moment of resistance (w xX b’ a’) of the pressure of the tibia upon the ankle. In Fig. 3, I, it would appear natural to assume that the point A, on the ground, is the fulerum, and that the “resistance” is the pressure of the tibia upon the ankle joint at B’, while the “power” is the contractile force of the muscles of the calf, applied at B. Similarly, Eduard Weber®® described the human foot as a lever of the second order and gave for the relations of the forces and movements of the foot in raising the 33 Cf. HAYCRAFT, op. cit., p. 251. GREGORY, QUADRUPEDAL LOCOMOTION 281 weight of the body a formula which translated into the terms of our Fig. 3 would be as follows: M X BA’=W X B’A’ But Knorz, Henke, Ewald and others, as quoted by Haycraft (Joc. cit.), showed that the effective distance of the muscular force M is not BA’, but BB’, and that we should rather conceive the foot as a lever of the first order with the pivot at B’, the “power” at B and the “resistance” (offered by the reaction of the ground upon the foot) at A. In that ease, the moments around B’ are ag ollows : M X BB’=W X B’A’ Hence, other things being equal, the longer the foot, the greater will be the moment of resistance to be overcome by the muscles of the calf. If the angle B’AD’ be increased, as when the foot assumes a more vertical position, the effective distance B’A’ decreases; that is, the mo- ment of resistance decreases as the foot becomes more vertical. In other words, the “load” diminishes as the calf muscles contract. It has been shown by Fick and others (quoted by Haycraft, loc. cit., p. 246) that when the force of- muscular contraction is opposed to a diminishing mo- ment of resistance, the muscle is capable of performing more total work (force X distance) than when the resistance is constant. Consequently, the diminishing resistance, conditioned by the raising of the foot, ena- bles the calf muscles to perform their work under the most favorable conditions. SUMMARY OF MECHANICAL AND PSYCHIC FACTORS IN POWER AND SPEED The speed of the distal end of a “long bone” of the limbs will depend upon (1) the nearness of the point of insertion of the principal muscles to the joint or axle, (2) the smallness of the angle of insertion of the muscles, (3) the position of the muscle fibers with reference to the long axis of the muscle, and (4) the speed of contraction of the muscle itself. If a long muscle and a short muscle were isolated for experiment, it might prove that the short muscle would contract faster than the long one, but, in nature, a single movement of a long bone is produced by the simultaneous and coordinated action of muscles of varying length. Thus, in the act of extending the whole arm from the fully flexed position, the relatively short, broad thoraco-scapular and scapulo-humeral muscles contract in the same time as do the relatively long extensors of the fore- arm, irrespective of their lengths. The speed and force of contraction naturally depend partly upon the strength of the stimulus and partly on 989 ANNALS NEW YORK ACADEMY OF SCIENCES the resistance to be overcome.*® The end result, as it were, determines the rate of contraction of the codrdinated muscles. The effective regu- lation and correlation of muscular action is obviously an extremely com- plex function of the peripheral and central nervous systems. The force and speed of contraction of a given set of muscles in a living animal at a given moment are determined not only by many mechanical factors, of which a few have been mentioned above, but also by the whole psychic constitution of the animal and by the psychic effectiveness of the ex- citing “motive.” . Fic. 4.—‘Graviportal”’ adaptations for the walk and amble in the Mastodon 2 APPLICATION OF THE FOREGOING PRINCIPLES TO THE STUDY OF THE Limps oF UNGULATES ~FUNCTIONAL SIGNIFICANCE OF THE ANGULATION OF THE LIMBS As noted above (p. 269), the straightness of the limbs in the Probos- cidea and similar heavy-bodied animals was interpreted by Osborn in 1900 as “an adaptation designed to transmit the increasing weight through a vertical shaft.” While this is no doubt an incidental advan- tage of the straightness of the limbs, it is probably not the chief teleo- logical “object.” From a consideration of the mechanical principles governing the use of the limbs as compound levers (see pp. 278-281) and 39 J. BuRDON SANDERSON, in Schiifer’s Text Book of Physiology, Vol. 2, p. 363, 1900. GREGORY, QUADRUPEDAL LOCOMOTION 983 from a comparison of the photographs (Fig. 1) of an ambling ele- phant and of a galloping horse, it seems probable that the straightness of the limbs in graviportal animals has been evolved pari passu with the short rectigrade feet and with an ambling even gait, in combination with a long stride of minimal acceleration increment (p. 274). Conversely, the bent or angulate character of the limbs in the horse and other cur- sorial animals is correlated in part with the very long, slender unguli- grade feet and with a bounding galloping or trotting gait, in combination with a long, very rapid stride of maximal acceleration increment. Fic. 5.—“Cursorial” adaptations for the run, gallop, etc., in the Neohipparion In other words, the use and structure of the feet have been the teleo- logical dominants which have determined the diverse modifications in the musculature, proportions and angulation of the proximal segments of the lhmbs, just as in early stages of aquatic adaptation in reptiles (e. g., Thalattosuchia, Nothosauria, etc.) the aquatic habits are reflected more clearly in the feet or distal segments rather than in the proximal limb segments. MECHANICS OF THE FOOT IN GRAVIPORTAL AND CURSORIAL FORMS Comparing the structure and function of the graviportal and cursorial types of feet, we see (Figs. 3 and 4) that, in the elephant, the very mas- sive gastrocnemius and soleus muscles are attached at a wide angle to the massive calcaneum, while the foot itself is very short. In the position shown in Fig. 3 (11) and as compared with conditions in the horse, this 984. ANNALS NEW YORK ACADEMY OF SCIENCES gives a relatively high moment of power (proportional to 6b’) to the calf muscles and a relatively low “moment of resistance” (proportional to b’a’) to the great pressure of the shaft of the limb upon the astrag- alus. A considerable part of this “moment of resistance” is also sub- tracted by the supporting effect of the great pad of elastic tissue under- neath the foot. No doubt the specimen from which Fig. 3, II, was drawn should have been mounted with the feet more nearly vertical ; this would greatly shorten 6’a’ and further increase the advantage of m X bb’. As the plantar pad is raised from the ground, more weight is thrown on the toes, but, at the same time, they are brought further back almost beneath the astragalus, thus reducing 6’a’ to a minimum, so that the load decreases as the muscles contract. This arrangement not only compensates for the fact that the greatest absolute force of a muscle is developed when it is stretched to its full psysiological length, but it also permits the muscle to perform a greater total quantity of work than would be the case if the load were increasing instead of diminishing (p. 280). Similarly, in the horse, the greatest “moment of resistance” is when the foot is fully flexed forward (which is at the instant the foot touches the ground) ; the action of the extensor muscles is thus suddenly checked ; this conditions physiologically a corresponding and sudden increase in the available energy (p. 277). By the raising of the heel the moment of resistance (W X BA’), as in the case of the elephant, also decreases, a. @., the load diminishes; but in the elephant, the motion of the foot is relatively slow and the acceleration increment of the stride (p. 275) is therefore slight, whereas in the horse the motion of the foot is very rapid, and the acceleration increment (through the high velocity im- parted to the relatively light body by the spring-like extension or open- ing of the angles at the stifle, hock, fetlock and pastern) is very great, so much so that at least three of the feet are off the ground during a great portion of the time. In brief, short feet (as in the Proboscidea) slightly bending at the ankle, raise a heavy load through a short distance with a minimal acceleration increment of the stride; long feet (as in the horse), sharply bending at the ankle, throw a smaller load a long distance, with a maximal acceleration increment. Metatarso-femoral ratios.*°—That in cursorial animals the hind foot is long as compared with the femur, while in graviportal animals the 40 The investigation of limb- and arch-form and proportions, and especially the estab- lishment and significance of definite ratios between the limb segments, were suggested by Professor Osborn and taken up conjointly by him and the writer in the Hocene sec- tion of the Titanothere Monograph; the following observations on limb ratios are in part a preliminary publication of the joint results attained. GREGORY, QUADRUPEDAL LOCOMOTION 285 reverse is the case, is shown by the table of ratios, Plate XXXIV, and especially in the following examples: Length of metatarsal III Pane: Length of femur Graviportal Mediportal Subcursorial Cursorial Coryphodon... .14 Rhinoceros .37 Eohippus... .50 Higquusisesaes 18 Uintatherium. .10 Paleeosyops .21 Tragulus.... .56 Antilope. ... 1.00 Mastodon..... met Mesohippus. .57 Odocoileus .. 1.00 Wlephas.... ....+. 13 Brontotherium .20 oxod om =). ....': Sly These ratios definitely prove the connection between the mode of loco- motion and the length of the middle metatarsal as compared with the femur. The wide differences in the metatarso-femoral ratio, ranging from .10 in extreme graviportal forms to 1.00 and upward in cursorial forms, are partly bridged over in the mediportal and subcursorial types, and even more completely in a fifth group including certain primitive fee Mitse el: ungulates, the Condylarths, in OS ae ranges from .43 to .31. Perhaps the most important facts to keep in mind in comparing these and similar ratios (below) are that the ancestral Placentals probably had relatively short hands and feet and long limb bones (p. 270), but that there was doubtless a considerable range of variation in this respect even as far back as the Upper Cretaceous epoch. We are unfortunately unable to follow the ratios through approximate phyletic series except in a few cases (especially Titanotheres, Equid, Rhinocerotide), but, in every case, we can feel sure that the precise ratios attained in the end- forms are conditioned largely by the nature of the ratios in the stem- forms of each family. Thus, the exceptional shortness of the feet in the Amblypoda is con- ditioned by the fact that this group, as represented by Pantolambda, had comparatively short feet before gigantism was developed. Hyrax is another example of a small form with very short feet, and from some such forms the Proboscidea probably arose. Besides those phyla which had short feet in the ancestral forms and which merely emphasized this feature, there are many phyla which started from animals with feet of moderate length and later shortened up the feet to a considerable extent. Thus in the Titanotheres, the oldest and most primitive form (Hotitanops) has a metatarso-femoral ratio of about .34, which is not far from that of other early Perissodactyls, but 236 ANNALS NEW YORK ACADEMY OF SCIENCES in the collateral descendants of Hotitanops, we observe a relative short- ening of the digits, correlated with increasing body size and straighten- : Mts. III. ing of the knees, so that ce from .34 through .31 and .30 in the Middle Eocene genera to .28, .26 and even .20 in the gigantic ' Brontotherium. Again, in the Rhinoceroses, the oldest, smallest and most ay Mts. IIT. primitive forms have a foot of moderate length aod = 43-42), but by progressive relative shortening and broadening, the ratio drops to .24 in Metamynodon. In the Hippopotami, which are probably descended from animals proportioned about as in Oreodon (with an index of .38). gigantism and aquatic habits have brought about a reduction of the index to .26. Similarly in the Toxodonts, the smaller and more primitive forms had relative long, slender feet, while in the gigantic Toxodon fallSto cai: In those groups in which the most primitive known members had already attained a slender foot, with reduced side toes, gigantism is unable to effect a complete approximation to the graviportal type. Thus, in the bisons, which are undoubtedly descended from slender-footed forms having a metatarso-femoral ratio not less perhaps than .75, the sudden increase in size causes the ratio to fall but slightly (to .65). In the. gigantic Irish elk, whose ancestors probably had a very high metatarso- femoral ratio (perhaps 1.00 or more), this ratio falls to .71. The evolution of cursorial forms is also indicated by marked changes in the ratio under consideration. Thus, in the Equide, it rises from .53 in Hohippus, through .68 in Mesohippus to .99 in Hypohippus, reach- ing the extreme of 1.16 in the slender-limbed Upper Miocene horse, Neohipparion. In the relatively small and slender kiang, the ratio is still 1.00, but in the relatively heavy-bodied Equus scotti of the Pleisto- cene, the ratio is only .84, while in modern horses, we observe even in race horses a falling off of the index to .78, and in the stocky-limbed Hippidion, it drops to .72. Correlated with this fall in the length of the metatarsal III, we observe a straightening of the knee. These fig- ures possibly may mean that the modern Hquus caballus, on account of its great size, 1s somewhat less adapted to extreme cursorial locomotion than was the slender Neohipparion which closely paralleled the deer Odocoileus.** On the other hand, race horses seem to have compara- tively long femora (cf. Stillman, p. 80) and cheetahs, hounds and other “J. W. GipLrey : Bull. Amer. Mus. Nat. Hist., Vol. 19, pp. 474-476. 1903. Mts. ITI. - GREGORY, QUADRUPEDAL LOCOMOTION 987 a forms that progress by bounding have quite long femora. In this con- nection, it must be remembered (p. 278) that a long femur, implying small angles of insertion of the principal extensors, gives relatively high speed of rotation of the insertion points, but low power for the rotation components. ‘The long femur of graviportal forms has a wholly different meaning (p. 289). Progressive reduction of the side toes in the Artiodactyla as in the Perissodactyla is accompanied by the elongation of the cannon bone (here represented by two coalesced digits, metatarsals III. and IV.) and by other cursorial adaptations. The primitive Artiodactyl foot with four complete side toes is represented in Oreodon, which has a ratio of .38. In Sus the ratio is .84, in the primitive four-toed camel Hotylopus reedi it rises to .52, in Tragulus to .66, whence it rises rapidly to 1.00 and more in the deer and antelopes, culminating in 1.35, an extreme figure, in the giraffe. In the giraffe, the development of great body size has not brought about any reduction in the length of the metacar- pals, but, in correlation with the long neck, has even lengthened them. Other adaptive contrasts in the feet.—There are other adaptive con- trasts in the feet of graviportal and cursorial animals, as follows: In graviportal forms, the astragalus is flattened down, for, when the weight of the body is raised, it is easier to force the tibia up a gentle slope than a steep one. No lateral keels are needed on the trochlear surface to prevent dislocation because of the breadth and spreading character of the tarsus and of the large size of the fibular malleolus. In eursorial animals, on the contrary, the curvature of the astragalar troch- lear is steep and the range of movement wide; the trochlea keels help to keep the narrow tarsus in place. With regard to the phalanges, in cursorial animals, the ae bending at the fetlock and pasterns and the sudden straightening out of these joints under the pull of the powerful flexors of the foot greatly assists in projecting the body into the air (Stillman, loc. cit., p. 89). In graviportal animals, on the other hand, the terminal phalanges are reduced ; the massive flexors of the foot raise the weight slowly and assist the animal in rolling from one foot to the other. 288 ANNALS NEW YORK ACADEMY OF SCIENCES GRAVIPORTAL AND CURSORIAL TYPES OF TIBIA Some of the tibio-femoral ratios (=) given on Plate XXXIV may here be grouped as follows: F Graviportal Mediportal intathertunie see eee 53 Palzeosyopshia yen... oe eee Teh Cory phodom. cert sen eee cee 61 Rhinoceros indicus............. 79 LEV PONE MUI. Sy peoaueone dear sac 56 Tapirus.c) cs.6 22s. 2 eee .80 Mastodon rey an eee eee ner .69 Pantolambdas:.... 3,-2.5 .76 Hiephashndicusiy.eee eee .60 Brontotherium............. ... 54 Metamynodon.................. 08 Cir gorten Peleoceras. eee ae cae .O7 : Eohippus \22.25 oes8 ao eee -. 1.00 Subcursorial or Primitive Mesohippuss).). 225m s4. 4-eeeeee 1.08 Phenacodus primeevus.......... 84 INeohipparion sss.4)- 2. eee 1.17 te WOMNOOAMN soon0420c 97 Equus caballusi i. aq = Jl ox = & an ~ = i 2 an = a Fr Species. = ta =| ah P z a3 = @ I x 82 =) = & o ss = 5 | = ss al & aot g ap ia 3 o 5 oS 2 a) | 5 oC 2 = ene 2 a | oe) a | 8 = a | 3 CoNDYLARTHRA MM. mum. mm, mm mm atl Euprotogonia puercensis ....-...... -| 105 45 43 107 ; er SLEDS Os enorCor esonccas|! Sakurce|| accocn Phenacodus wortmani...........-..-. 134 51 .38 { ne r| 1.00 \ 107 36 65 89 83 Phenacodus primzevus ...........+--- 234 74 31 198 -84 167 70 42 146 .87 Meniscotherium terreerubrie........-- 100 2 -29 91 ht 82 22 .27 58 .70 AMBLYPODA Pantolambda bathmodon............. 149 36 .24 114 -76 124 31 .25 82 .66 Coryphodon lobatus ........... ....- 423 62 14 260 -61 363 70 19 240 .66 Uintatherium (Dinoceras) mirabile...| 692 70 -10 360 53 540 106 at) 380 -70 PYROTHERIA Pyrotherium (figd. by Gaudry).......| 622 | ......]...... 351 -56 Cb | ceAnina [oso vor 238 52 PROBOSCIDEA Mastodon americanus................ 1,020 e 117 Al 705 69 885 165 18 670 .75 Boab ites bth eS arasarnaccues OC 1,020 138 13 618 .60 810 183 -22 685 -80 KE. (Loxodonta) africanus............ 1,050 144 13 755 7i | 1,000 205 20 870 87 HyRacorDEA PSOE ANE Oran lita lotatetercSalst= sinlvisirieiaialeiatst 71 19 .26 69 Ne/ 69 16 -23 46 .66 ToxODONTIA ALOR OGOURADemerietateislcisthicter feleinicnterarereratele 577 101 Abt 325 56 387 147 .33 298 77 EDENTATA GRAVIGRADA lab ferilo)s):01:\ os see once dpe OOnDueOUmE ae 150 18 12 108 :72 13 22 16 106 .80 estodon armatus,......2........-0.06 640 78 12 330 51 530 94 17 325 -60 PeRISssODACTYLA TAPIROIDEA Heptodon calciculus ............... at) ude 75 e 43 175 1.00 115 67 -58 114 -99 Tapirus americanus - | 262 108 -41 208 ‘79 205 106 -50 177 .86 AA OMIE Eh UNC A gooadconoanounonor 320 120 -37 258 -80 250 120 48 228 91 PrrissopacryLa RaiNoo: Hyrachyus agrarius.... 254 110 43 243 95 197 93 47 197 1.00 Hyracodon nebrascensis c 267 114 42 220 -82 202 114 56 210 1.03 Rhinoceros indicus...............+-+ 495 180 .37 395 -79 385 186 48 385 1.00 Teleoceras fossiger........0:..sce00 408 105 .25 233 57 305 e 114 -37 238 -78 Metamynodon planifrons............- 480 118 :24 280 .58 393 103 39 320 81 PerissopAcTYLA TITANOTHEROIDEA Hotitanops borealis .........-........ PANG |) eae easub |p enuees 203 tf} a ieee 1 eal erie ns A [oer cr ene Paleeosyops major..............2sse0es 433 137 31 832 ED revste eayeel | petmeetoees teal (catetarereta || eaarertene Palseosyops leidyi.........--...--..-. 370 110 .30 290 -78 325 2113 235 IFAP Nose ose Ae loan ola blnkep PsP eigen ll asnediellesetine \jesrnnoaa lPoaceen 2340 106 237 Limnohyops sp. (A. M. N. H. No. 11689} 35 e 111 31 285 -79 293 109 228 ae st fs ** No. 11690 387 123 31 283 BA: Se eeavesinel becca: |jeonece Crd sme. Gl Manteoceras manteoceras............. MGR Vitsantas secon s 272 Bok Wes eee Aen ese mcicatie eocsier. on feat os e Mesatirhinus petersoni, No. 11659. 308 118 33 283 A Mela S Seis apeane.||RRaonode i neemaa Dolichorhinus hyognathus, No. 13164.| 386 119e OO eS fctevateie) || Beyaratese LONE) |Meat 284 Titanotherium trigonoceros........... 770 220 e -28 430 55 620 240 520 Brontops/robustus?.. ses. 010s emesis 812 212e .26 448 As 608 230 504 Brontotherium gigas 9 (518 A. M.)...] 780 200 -20 427 54 528 214 78 PERISSODACTYLA HiPPOIDEA OHIDPUSIAD saeraneien sees eck cser 162 82e -90 162 1.00 121 64 53 110 -90 MGR OpodiY OWI eh moagascangsucHesonuod 178 121 .68 193 1.08 136 92 .68 136e} 1.00 Hypohippus osborni.................. 278 218 -78 277 1.00 205 203 -99 260 1.27 Neohipparion whitneyi ... ats 249 252 1.01 293 1.17 187 218 1.16 244 1.30 IQUE KAN Over a eee tore area 313 277 .88 310 +99 237 238 1.60 302 1.27 UGUIMISISCOLEL eye sternisrmtaicterabs enisrereds sterner 370 263 71 330 .88 289 243 84 342 1.18 Equus caballus (race horse).........-. 392 288 -73 363 92 305 240 -78 363 1.19 Hippidion neogeeum................- 340 214 -62 305 .89 273 198 :72 287 1.05 ARTIODACTYLA Oreodontidxe = aoredon Cul bextSOnie sents 161 62 .38 142 .88 138 57 41 113 81 wide SUSI RCROL ese rces cre viele Sistelere Aes 248 86 -34 216 .86 208 77 .37 168 -80 Hippopotamids Hippotamus amphibius.......... 498 130 -26 332 -67 395 152 -38 270 .68 Camelidee Kotylopus reedi.................. 148 78 52 142 :96 118 68 57 100 .84 Camelus arabicus................ 470 325 .60 400 -80 363 330 -90 455 1.25 Tragulidx AUEMUE GAN eS docseonussaaanane 94 62 -66 103 1.09 74 42 -56 62 83 Cervidee Odocoileus hemionus............. 253 255 1.00 295 1.16 198 208 1.05 223 1.12 i Genie MMOPACBLOS <1 crane ereicinter cine 430 350 -71 454 1.05 334 342 1.02 358 1.07 ovidse Gazella dorcas juv. .............. 140 132 81 176 1.25 93 134 1.44 118 1.26 Antilope cervicapra.............. 183 183 1.00 223 1.21 133 180 1.35 168 1.26 IBINON) DISON. incisive oe eye aveutere 369 243 -65 355 -96 290 198 -60 293 1.01 Antilocapride Antilocapra americana........... 210 218 1.03 260 1.23 164 213 1.30 202 1.23 Giraffide CPE CT hls ee geScorraneeece 2.) 466 630 1.35 550 1.18 4385 618 1.42 698 1.60 e—Estimated. ~ PUBLICATIONS OF THE pane. of the tals soutleides in general with the calendar year old at the uniform price of three dollars per volume. The articles the volume are printed separately, each in its own cover, and ‘ibuted i in bundles on an average of three per year. The price of a) The Manele ‘(quarto series), established in 1895, are issued at rregular intervals. It is intended that each volume shall be devoted to All publications are sent free to Fellows and Active Members. The Annals are sent to Honorary and Corresponding Members desiring them. ee viene. and inquiries concerning current and back numbers of THE LIBRARIAN, © New York Academy of Sciences, care of American Museum of Natural History New York, N. Y. | Editor, EpMUND Oris Hovey ie AGE OF THE BEDFORD SHALE OF OHIO BY sy Grorce H. Girry Bo EW YORK PUBLISHED BY fea ACADEMY Orricers, 1912 _ Pecstiee Ie unio McMILu1n, 40 Wall Street neta oh | Vice-Presidents—J. EDMUND Woopman, -FREDERIC. she Lu . CHartes LANE Poor, R. 8. Woopworta 5 ae Corresponding ee E. CRAMPTON, / American - f Be , Treasurer—HENRY L. Douerry, 60 Wall Sivect ty - Librarian—Rawen W. TOWER, American ‘Nate | | Bditor—Eowowp Oris ONE American Museum | SECTION OF GBOLOGY AND MINTRALOGY : o e nd E. Woopman, N.Y. University . ! i" Cd aa P. BERKEY Columbia University | { SECTI ON OF BI OLOGY am rie : iN - vet Chairman—Freperic A. Lucas, American ATuseurn ‘Secretary WILLIAM K. Grecory, American Museum "SECTION OF Gian MOS PHYSICS AND CHa Chairman—CHaRLes Lane Poor, Columbia University . ee M. PEDERSEN, College. of the City of new York o ral SECTION OF ANTHROPOLOGY AND PSYCHOLO '] Chairman—R. S. WoopworrH, Columbia University Secretary—FREDERIC LYMAN WELLs, Columbia University. _[Anwnats N. Y. Acap. Sciences, Vol. XXII, pp. 295-319. 13 November, 1912] GEOLOGIC AGE OF THE BEDFORD SHALE OF OHIO? By Grorce H. Girty (Presented in abstract before the Academy, 7 October, 1912) The Bedford shale takes its name from Bedford, in Cuyahoga County, Ohio. The typical locality is the gorge of Tinkers Creek below the falls, where the formation comprises about 75 feet of bluish clay shale lying between the Berea sandstone above and the black slaty Cleveland shale below. To the west, the lower portion of the Bedford develops a sand- stone member which is quarried as the “Cleveland bluestone,” while the upper portion undergoes a change of color to a strong red. In fact, the Bedford is better known as a red than a blue formation. The fossils of the Bedford shale are largely confined to the basal por- tion, though a few species are represented by rare individuals at higher horizons. At Bedford, fossils are abundant in immediate contact with the Cleveland shale, where they are more or less crushed, and also a few feet higher in large calcareous concretions, where the surface characters are apt to be obscured. Some of them are broken and rounded as if by wave or current action, but this is not the general character of their occurrence. The Bedford shale lies close to what has been considered the boundary between the Devonian and Carboniferous systems, and it is the purpose of the present paper to present such evidence as I have bearing on the geologic age of the formation. The question then is whether the Bedford shale shall be included in the Devonian or the Carboniferous system. I shall treat this largely as a paleontologic question, and my fossil evi- dence is derived from typical sections at Bedford and other points in Cuyahoga County. Tt will not be out of place to consider some of the principles controll- ing such an attempt as I have taken in hand. Theoretically, the great geologic systems were defined by movements cre- ating extraordinary changes in the conditions of land and water, always undergoing changes more or less gradual, and these conditions entailed corresponding changes in the character of the plants and animals which had in them their habitat. As expressed lithologically, the rocks of the several systems, perhaps generally in the typical region and not infre- 1 Published by permission of the Director of the U. S. Geological Survey. (295) 296 ANNALS NEW YORK ACADEMY OF SCIENCES quently in regions far removed, show a basal sandstone or conglomerate with more or less pronounced evidence of preceding erosion or uplift. Such events though often far reaching must in the broadest sense have been only local, but we may fairly expect the more subtile changes of environment, such as depth, temperature, currents, which govern the character and distribution of animal life, to have extended far beyond the area of great disturbance, and we may look for evidence of such disturb- ance in the fossil floras and faunas, when none is to be found in the rocks themselves or their relation to one another. In fact, such paleontologic evidence may have been all but universal. Conglomerates and uncon- formities and faunal changes occur at other horizons than the division lines between systems, but such evidence will be of great importance in deciding the point in question. Yn addition to the kinds of evidence already mentioned as often accom- panying the transition from one geologic system to another, that is, an interval of erosion, a basal sandstone or conglomerate and a well-marked faunal change, there are also certain other considerations of a more ad- ventitious or incidental nature. For practical purposes, it would be unfortunate if this line (that between two systems), which of all lines it is desirable to represent on a geologic map, were taken where it could with difficulty be recognized in the field, as in the middle of a uniform lithologic interval, or where the evidence would often be concealed, as would be the case in some regions if it were assumed to he between two formations of soft and easily disintegrated material. Furthermore, the importance of convention also enters the consideration. Other things being equal, it is clearly preferable to take for the boundary which is sought the same horizon at which it has been drawn elsewhere, if that horizon can be determined. The boundary between the Berea sandstone and the Bedford shale in some degree satisfies all these requirements. It is easily recognized and easily traced; it also appears to be the locus of an unconformity.” The Berea answers to the basal conglomerate of theory. While the Bedford seems properly to form part of the great shale series which preceded it, the Berea marks the change to another and different type of sedimenta- tion. The passage from Bedford to Berea is also marked by an abrupt and strong faunal change. Not only are the two faunas widely different, but the Bedford has a preponderant Devonian facies and the Berea a pre- ponderant Carboniferous facies. To some extent, the faunal change at ; 2 This erosional unconformity was clearly stated by Newberry as early as 1874 (Ohio Geol. Survey, Geol., vol. 2, p. 91). More recently it has been mentioned by J. E. Hyde (Jour. Geol., vol. 19, 1911, p. 257) and described by W. G. Burroughs (idem, p. 655). GIRTY, GEOLOGIC AGE OF THE BEDFORD SHALE 997 the top of the Bedford is offset by one equally marked in the opposite direction, between the Bedford shale and the Chagrin shale. In support of this statement I shall not feel called upon to submit evidence, because in so far as it is untrue, the Devonian affinities of the Bedford faunas are by so much stronger than they are here represented. Finally, the deter- mination of the boundary between the Devonian and Carboniferous at the base of the Berea satisfies to some extent the canon of convention. This is true, however, only from one point of view, for if the Berea (“Corry”) sandstone correlates with the Kinderhook beds of the Missis- sippi Valley which represent the base of the Carboniferous system in its type section, and if some horizon below (?) the Chagrin correlates with the Chemung group which represents the top of the Devonian in its type _ section,*® the canon of convention or usage would be ambiguous in its bearing upon these intermediate beds whether they should be classed with one system or the other, because by one approach they would be found above the recognized top of the Devonian, just as by the other they would be found below the recognized base of the Carboniferous. This too is only partly true, however, because Hall* classed as Che- mung, or more often “Upper Chemung,” the intermediate group of strata here under consideration, so that, although their stratigraphic position is above the Chemung proper and therefore theoretically above the top of the Devonian, they have in practice been included all along in that system. Incidentally, this circumstance seems to show Hall’s opin- ion of the affinities of the “Bradfordian” faunas,° though it must be re- membered that he correlated with the Chemung the Waverly group lying still above. Now the statements which it is proposed to present the evidence for and to discuss are the following: the faunal change incident to the pas- sage from the Bedford shale to the Berea sandstone; the Carboniferous aspect of the Berea fauna, and its probable correlation with the Kinder- hook of the Mississippi Valley; the Devonian aspect of the Bedford fauna, and the relation of the Cuyahoga, Berea and Bedford faunas to those of the typical Mississippian sections of Missouri, Illinois and Iowa. | %The typical sections of the Devonian and Carboniferous in the United States are meant. The original sections are of course in Hurope. 4This statement is based upon the fact that Hall described and included in the Che- mung fauna numerous pelecypods from many of the well-known “Bradfordian” locali- ties of Pennsylvania. 5 Here and elsewhere in this paper, ‘‘Bradfordian’”’ is employed in the sense in which I defined it in 1904 (Science, nN. s., vol. 19, p. 24)—-for a series of strata about 500 feet thick, lying between the typica] Chemung and the Waverly group (as limited below by the Berea sandstone), and comprising in the typical sections near Bradford, Pa., and Olean, N. Y., the formations described under the names of Cattaraugus, Oswayo and Knapp. : 998 ANNALS NEW YORK ACADEMY OF SCIENCES Some diversity of opinion exists as to the principles which should govern the interpretation of paleontological evidence in determining the boundaries of geologic systems. JI propose to consider these different views and to decide which is to be employed in the present investigation. Thus, to take a put case, it is sometimes said that the line between the Devonian and the Carboniferous should be placed at the first introduc- tion of Carboniferous species. This principle seems by implication or otherwise to be adopted by Glenn and Butts and Clarke in their discus- sion, of the geology of the Olean quadrangle and other areas in western New York.® It is true that, in adopting this principle in the Olean rock section, the authors were influenced by the belief that the interval below the Olean conglomerate member of the Pottsville formation in the Olean quadrangle corresponded to the interval similarly underlying the Olean conglomerate in northwestern Pennsylvania and northeastern Ohio, so that several hundred feet in the one section would have to correspond to the authentic Waverly group in the other.*. Now that it is known, or at least seems highly probable, that, owing to erosion which preceded the Olean conglomerate, that member rests on lower and lower strata as it is followed eastward, so that at Warren, Pa., most of the Waverly rocks are missing and they are not known to appear-in the sections farther east,—with this condition of affairs granted, it is possible that the au- thors would not have adopted the principle of first appearance. It is necessary, however, to examine this principle to see whether it is gener- ally applicable or applicable to the present case. The meaning of the principle as stated above is clearer at first sight than when it is examined more closely. The implication largely turns on, the meaning which is given to the terms “Devonian” and “Carbon- iferous” species, and, at the risk of drifting into something like the Greek dialectic, it will not be unprofitable to consider this question. By Carboniferous species may be meant (a) any species which has been found in rocks of Carboniferous age; (b) a species especially abundant or widely spread and persistent in rocks of that age though occurring elsewhere, or (c) a species known only from the Carboniferous, of at least known there but not in the other geologic system with which com- parison is made. It is evident that, in the present instance, the words “Carboniferous 6 New York State Mus. Rept., vol. 56, pt. 2, 1903, pp. 985, 991 and 999. Clarke gave a somewhat more extended discussion the year previous (idem, vol. 55, 1902, p. 524) in which he showed that a marked faunal change took place at the base of the “‘Brad- fordian,”’ but not that the ‘“Bradfordian’’ fauna had a preponderating Carboniferous aspect. " Loc. cit., p. 991. GIRTY, GEOLOGIC AGH OF THE BEDFORD SHALE 299 species” cannot be taken with the first meaning (a), because some types found in the Carboniferous have a very long range and originated at much lower horizons. Leptena rhomboidalis is an example.® If we assume the second meaning (b) for the expression “Carbon- iferous species,” the principle would mean frankly a redetermination of the Devonian-Carboniferous boundary, not only in other areas but in the typical area as well, a redetermination, moreover, which would never cease, because the Devonian beds which would thus be added to the Car- boniferous would carry over other species having a sporadic appearance at lower horizons. This would entail a new adjustment and so on appar- ently until the bottom of the stratigraphic column was reached. Fur- thermore, is it one such species, or two, or a score that the application of this principle involves? Reasonably but one. The decision might easily then come to depend on the identification of one or two specimens re- sembling several closely related species. If we now take the last meaning (c) for the expression “Carboniferous species,” 7. e., species characteristic of the Carboniferous, those not known to occur at any horizon outside of the Carboniferous, the statement under consideration becomes hardly more than a truism, but a truism which assumes that we have complete knowledge of the range of species, that the species in question are not only not, known to range outside of the Carboniferous but that they cannot so range, and this is an assumption which everyone knows is quite inadmissible, since the range of species as recognized at any time is continually being changed by new data. Other considerations might be brought forward, but it is already plain that this is not a workable principle, no matter how understood, for determining the boundaries between systems, or for classifying the Bedford shale. It might be held, on the other hand, that the proper way to fix the line between the Devonian and Carboniferous systems is by the disap- pearance of the last Devonian forms. This principle is the antithesis of that just considered. It is open to the same objections and is equally untenable. | The only practicable method of interpreting paleontologic data in most cases of this sort is evidently by taking the balance of evidence. By this method the decision hangs not upon one or two forms but upon the entire number known, and, although the evidence of each form may be impaired by poor material and close relationship between species, it be- comes cumulative. Here the expressions “Carboniferous” and “Devo- nian” species signify species especially common in their respective sys- 8 De Koninck, however, distinguishes the Mississippian Leptzena as a distinct species under the name Leptena analoga. 300 ANNALS NEW YORK ACADEMY OF SCIENCES tems. If the species are absolutely unknown outside of their respective systems, the evidence is so much the stronger, but the difference is one of degree only, and no assumption is made that a species at present known in the Carboniferous may not subsequently be found in the Devonian or vice versa. 'That a species has not yet been found in the other system, however, rightly gives it exceptional weight especially in regions where the available data on range are considerable. This principle is of course unsatisfactory. It is impossible to estimate or to state the results mathematically. Different species have different values, and the same species may have different values in different re- gions. While these values are not expressible in numerical terms and indeed must vary somewhat with each observer according to the character and extent of his experience, nevertheless, in practice, the evidence is seldom so nicely balanced or the antecedent experience of those who judge it so diverse that reasonably satisfactory and unanimous conclusions are impossible. In determining the relationship of geologic formations, which includes also the determination of their geologic age, some species have, as already pointed out, greater importance than others. This is partly because we know more about some than others. The determination of such rela- tionships as I have mentioned rests very largely on our knowledge of the range in geologic time of different types of fossils and involves one of the most fallible of all processes of inductive reasoning. Because a fossil has not yet been found above or below a certain horizon it does not follow that it never will be so found, yet that is virtually the inference on which all correlations and age determinations are based. At best, this fur- nishes conclusions which are fairly safe, and at worst it furnishes con- clusions which are highly unsafe. At its best, the conclusion depends upon the concurrence of a large number of species and upon species whose range has been ascertained by a large number of observations. For the same reason, it is clear that common species are more significant than rarer ones, because our knowledge about their range is more trust- worthy, and, in some cases, fairly sound inferences can be drawn from a single species. It is, however, not only the trustworthiness of our knowledge which lends greater significance to some forms than to others, but also the length of range in geologic time, which differs with almost every form; for, obviously the presence of a form which had an established range of 100 feet would be much more significant in correlation than that of one whose range was 1000 feet. Here enters also the consideration of groups larger and smaller than species, the range of which is, generally speak- GIRTY, GHOLOGIC AGE OF THE BEDFORD SHALE 301 ing, in proportion to their size,—long as the biologic rank of the eroulp increases, short as the biologic rank of the group decreases. Our knowledge of the range of fossils is conditioned by other peti than the amount of data as to their occurrence, since it manifestly in- volves the identification of the fossil as a species and the identification of its geologic horizon. Another factor then which makes some forms more significant than others is that their generic and specific relations can be determined with greater certainty. Variation in this particular is in some cases intrinsic, in others extrinsic; often it depends upon both factors. Manifestly, in groups where variation is restricted in degree, or is restricted to a few characters, or is marked by complete intergradation, discrimination of species and even of genera is more difficult than others. ‘Types which vary in shape alone have in my ex- perience proved especially unsatisfactory, while those which are sculp- tured or which possess other features of relief give more reliable results, especially when, as is almost always the case, this is combined with varia- tion in configuration also. Preservation, which in fossils has always destroyed the soft parts and the coloration, often obscures other charac- ters too, and this deterioration, owing to peculiarities inherent in whole groups of shells, is more liable to befall some types than others. As is well known, owing to their physical rather than their chemical structure, the shells of pelecypods and cephalopods are apt to be removed by solu- tion so that in Paleozoic rocks they are as a rule reduced to the condi- tion of molds, while the shells of brachiopods retain their original compo- sition. Also, partly because of the solution of the shell, which tends to obscure both sculptural or specific and structural or generic characters, partly because of being marked by growth lines alone and having the generic characters largely developed along the hinge where they are at best difficult of observation, many types of pelecypods show differences only in shape and configuration (which are peculiarly liable to be altered by compression), while at the same time quite widely different genera are in general aspect, which is about all that can be determined, very similar,—for these reasons pelecypods often prove an unsatisfactory eroup for stratigraphic paleontology, since even the generic position of specimens as they ordinarily occur in the Paleozoic is often undeter- minable except on characters which are not of themselves strictly generic, although they are, or appear to be, correlated with generic characters. It is obvious that, while the larger zodlogic groups are as a rule of less value because of their longer range, they are at the same time usually of greater value because of the precision and certainty with which they can be distinguished, for it hardly needs to be stated that a specimen 302 ANNALS NEW YORK ACADEMY OF SCIENCES can often be referred to a genus with certainty while its specific position is a matter of doubt. On the other hand, a provisional specific identifi- cation is sometimes only possible on the assumption of a generic one. Though because of their more extended range (in part compensated by the certainty of delimitation), the larger groups are less serviceable in correlating different sections, they are more instead of less valuable in estimating the importance of faunal changes in the same section, since they indicate a greater degree of change and possess the added advantage of increased certainty of discrimination. Of equal importance with identification of a form in its biologic rela- tions is the identification of its geologic horizon and this is frequently unsatisfactory. It seems to be true, and it is natural that it should be so, that in geologic time, as at the present day, the progress of sedimenta- tion and the course of biologic development varied in different areas or provinces, and that deposits may be, so far as one can tell, essentially contemporaneous, and yet very different in lithologic character and in the character of their fauna and flora. The limits of geologic provinces are not clearly defined, if indeed it is not ultimately shown that they are without definite limits but are continuous with one another; and still less is it known what were the factors which produced their differentia- tion. For my own part, I have very little faith in the theory of barriers (in the sense of land barriers) as a panacea for all the ills of strati- graphic geology. On the contrary, I believe that during geologic time, as today, the conditions controlling the character and distribution of faunas are depth, temperature, food supply, current action, salinity, bot- tom and so forth. At all events, as between different provinces, most correlations are at present more or less provisional, so that while the paleontologist must not disregard the data from any area, his deductions concerning one province should be largely guided by the data from that province. Since the data of range and distribution of species are in large meas- ure not on record even when they have been ascertained, and since the records are very scattered, each investigator must approach a problem with a different store of facts on which to base his inference as to geo- logic age and correlation, but it by no means need follow that such partial or even one-sided knowledge must lead in different cases to different conclusions. A number of years ago when much engaged with the investigation of these “Bradfordian” beds, from which I have since been temporarily diverted, I began and all but completed a descriptive study of the fauna of the “Corry” sandstone (since correlated with the Berea sandstone). GIRTY, GHOLOGIC AGE OF THE BEDFORD SHALE ‘This fauna I will now list in the terms in which it was then prepared as follows: Crania levis Keyes Rhipidomella n. sp. Schuchertella desiderata Hall Clarke Producitella n. sp. Productus vn. sp. idem, n. var. idem, n. var. n. sp. arcuatus Hall ? levicosta n. var. Strophalosia-like form, n. gen. n. sp. Spirifer marionensis Shumard disjunctus Sowerby ? Cyrtina triplicata Simpson Syringothyris angulata Simpson extenuata Hall and 303 2 Camarotechia metallica White ? Paraphorhynchus striatum Simpson mediale Simpson Pterinopecten alternatus Simpson Aviculipecten equalatus Simpson ? patulus Hall ? cancellatus Hall ? Paleoneilo sp. Sphenotus sp. Sanguinolites senilis Herrick Spathella ? sp. Cypricardinia sp. Mytilarea sp. Edmondia ? sp. Straparollus roberti White ? Platyceras varians Simpson dorsale Simpson Athyris lamellosa L’tveillée Cliothyridina squamosa n. var. Camarotechia heteropsis Winchell Tropidodiscus crytolites Hall Conularia byblis White ? If this list is compared with the one which I shall give farther on, it will be seen how very differeht the “Corry” fauna is from the fauna of the Bedford shale. Thus the statement that a pronounced faunal change marks the transition from Bedford to Berea time seems amply justified. ‘The second point which I wish to make in this connection is that, for the first time in this region, in the ascending series, we have a fauna of dis- tinctly Carboniferous type. The “Corry” fauna contains much that is new, but the development of species of the Productus rather than the Productella group (though on this I do not lay much stress because of the difficulty of adequately determining one group from the other), and especially of a Productus of the cora type, an abundant Spirifer of the marionensis type, Athyris lamellosa, a species of Cliothyridina, two spe- cies of the Kinderhook genus Paraphorhynchus and a few other forms, identify this horizon as Carboniferous and probably Kinderhook. There is, to be sure, some evidence pointing the other way, as for instance a Spirifer doubtfully identified with S. disjunctus and the two Aviculipec- tens, also doubtfully referred to Devonian species, but no one will ques- tion on which side the evidence is stronger. The fauna of the Bedford shale has never been described in full. In- complete faunal lists have been given in two or three instances. A few species identified or figured from this formation may be found scattered 304 ANNALS NEW YORK ACADEMY OF SCIENCES among Hall’s paleontological monographs (as Macrodon hamiltonie) and an occasional species has been described by other writers (as Paleo- nelo bedfordensis by Meek). MHerrick® has published a plate of figures drawn from specimens obtained at Central College in the central part of the State, and Foerste’® has described and figured a very limited devel- opment of the fauna as represented in eastern Kentucky. It may be said, however, that the Bedford fauna is very imperfectly known. In my work on the “Bradfordian,” the fauna of the Bedford shale was collected and in part described. As represented in my collections, the Bedford fauna in the typical localities of Cuyahoga County, Ohio, com- prises about 50 species, which may be listed as follows: Lingula n. sp. Pterinopecten ?n. sp. irvinensis Foerste ? Macrodon hamiltonie Hall Lingulidiscina n. sp. Edmondia aff. subovata Hall and el- newberryi Hall ? lipsis Hall Pholidops nu. sp. Cypricardella aff. gregaria Hall and Schuchertella herricki Foerste tenuistriata Hall Chonetes n. sp. Sphenotus aff. cuneatus Conrad and Productella pyxidata n. var. contractus Hall Strophalosia sp. Pholadella newberryi Hall ? Rhipidomella n. sp. sp. Cranena ? aff. subelliptica Hall and Ptychodesma ? sp. Clarke Bellerophon aff. pelops Hall, mera Cryptonella ? sp. Hall and jeffersonensis Weller Camarotechia sappho Hall Tropidodiscus aff. acutilira Hall, bre- Delthyris n. sp., aff. sculptilis Hall vilineatus Conrad and cyrtolites and mvissowriensis Weller Hall Spirifer aff. marionensis Shumard ? Pleuwrotomaria aff. sulcimarginata Syringothyris carteri Hall Conrad Nucleospira ? sp. Platyceras sp. Camarospira ? sp. Loxonema ? sp. Athyris aff. hannibalensis Swallow Conularia aff. newberryi Winchell and fultonensis Swallow Hyolithes sp. Paleoneilo bedfordensis Meek. Orthoceras sp. Leda diversa Hall Goniatites sp. Solenopsis ? sp. Proetus ? sp. In addition, there is a doubtful species of Rhombopora, several species of conodonts, which are rare, and abundant though ill-preserved ostracods suggesting the genera Primitia, Cytherella, Beyrichia (2 species) and Paraparchites. The identifications and comparisons given above are subject to re- vision, but, in spite of such possible changes, the list will serve to show ® Sci. Lab. Denison Univ., Bull., vol. 4, pl. IX. 10 Ohio Nat., vol. 9, p. 515 et seq: 1909. GIRTY, GEOLOGIC AGE OF THE BEDFORD SHALE 205 the general character of the typical Bedford fauna as represented in very complete collections. Before commenting on the characters of this fauna, as shown by my list, it will be desirable to consider some species which have been recorded from the Bedford and which I have not identified there. Newberry has cited the following species from the Bedford shale: Syringothyris typa Win. ; Hemipronites crenistria Phil. Orthis michelini Lev. Chonetes logani Hall Spiriferina solidirostris White Lingula cuyahoga Hall Macrodon hamiltonie Hall Rhynchonelia sagerana Win. upon which he comments in these words: “In this list there are several which have peculiar interest and significance, Syringothyris typa and NSpiriferina solidirostris, for example, from the fact that they are characteristic of the Lower Carboniferous rocks of other States, while Orthis michelini is common to the Carboniferous formation all over our country and in Europe.” Herrick,’ referring I doubt not to this passage, says: “Dr. Newberry has decided that the Bedford shale is Carboniferous on the basis of such fossils as Syringothyris typa, Hemipronites crenistria, Chonetes logani, Orthis michelina and Spiriferina solidirostris and a few more. Having searched in the same localities without finding these forms in the typical Bed- ford as it appears in southern Ohio and on the other hand finding the species above mentioned [in a preceding list] we feel some hesitation as to the occa- sion of the confusion. These species may indeed occur below the Berea, but in flags and greyish shales not in the blue or red Bedford shale!” As to the closing remark I may say that though Newberry did not de- scribe or figure the species which he named, there is, owing to the consti- tution of the Bedford fauna, no reasonable doubt as to what types he wished to indicate in each case, and Herrick is quite in error in supposing that these species did not come from the true Bedford shale. Now, as to the species mentioned by Herrick, Orthis michelina and Hemipronites crenistria are the species which I have listed as Schucher- tella herricki and Rhipidomella n. sp. The difficulty of discriminating species among the Rhipidomellas and Schuchertellas is such that these types are of minor importance in correlating faunas. The Waverly Schuchertellas are so closely allied to S. chemungensis that it would de- mand considerable temerity to say that a given suite of fossils belonged to a species of the one fauna rather than to a species of the other and indicated either Carboniferous or Devonian age. Much the same is true 11 Op. cit. p. 109. 306 ANNALS NEW YORK ACADEMY OF SCIENCES of the Rhipidomellas, but if reliance may be placed on the size and shape of the muscle scars, which are usually regarded as good specific characters in this group, I may say positively that the Bedford form is not Rhipi- domella burlingtonensis (which was described as a variety of michelini and is the most probable species indicated by that name which was orig- inally applied to a European form). Chonetes logani is the form which I called Chonetes n. sp., and which was beyond question wrongly identi- fied with Norwood and Pratten’s species. For Syringothyris typa, I have adopted Schuchert’s identification, S. carteri. When Newberry wrote, and when Herrick wrote for that matter, the genus Syringothyris was regarded as a diagnostic Carboniferous type and very justly, so far as the facts were then known, but it has subsequently been found that the genus occurs abundantly in direct association with Spirifer disjune- tus in the “Bradfordian” rocks of northwestern Pennsylvania. Since S. disjunctus has always been regarded as being as emphatic a marker of the Devonian as Syringothyris was of the Carboniferous, it is clear that the evidence of either type is disqualified for deciding the question at issue. Schuchert?? has even described a species of Syringothyris from the middle Devonian of Missouri, and furthermore a tendency to develop the syrinx seems to be manifested in several Devonian species of Spirtfer, so that it would seem as if the evidence of S. disjunctus should be es- teemed of greater weight in favor of the Devonian than that of Syringo- thyris in favor of the Carboniferous age of the “Bradfordian” strata. Newberry’s Spiriferina solidirostris is the Deltthyris n. sp. of my list. It is absolutely certain that this form is not the Kinderhook species S. solidirostris and almost equally certain that it is not a Spiriferina at all. It has, it is true, the general expression and the median septum which are found in Spiriferina and which are also found in the group of Spirifers to which the title Delthyris has been applied, but it does not possess the punctate shell structure which is an indispensable character of Spiriferina. The form in question is not rare in the Bedford shale and I have been able to examine a considerable number of specimens. This I have done both with a hand lens and with a compound microscope without success in finding the punctate structure which is usually a feat- ure easily detected in species really belonging to the genus. Thus, I am forced to conclude that the form is not a Spiriferina, which is a typical Carboniferous genus, but that it is a Delthyris, which is an almost equally typical Devonian one. I have thus traversed all the forms thought by Newberry or by Her- rick to indicate a Carboniferous age for the Bedford fauna, and their 142Am, Jour. Sci., vol. 30, p. 223. 1910. GIRTY, GEOLOGIC AGE OF THE BEDFORD SHALE 307 supposed significance has for one reason or another quite disappeared under impartial criticism. Consideration seems to be demanded at this point of a small list of “Carboniferous” species cited by Mr. Butts from the “Bradfordian” of the Olean quadrangle. These species are not known in the Bedford shale, but the Bedford interval is probably represented in the Olean sec- tion, though not distinctly recognizable there. At all events, if the Knapp and Oswayo formations of the Olean section are Carboniferous, it is clear that the Bedford shale must be Carboniferous, whatever its fauna, since if it does not represent some horizon in those formations, it must represent one above, rather than below them. It is therefore ger- mane to this discussion to scrutinize the evidence for calling the Knapp and Oswayo formations Carboniferous. I suspect that the authors of the work in which these species are cited would have proceeded differently if they had not assumed as a postulate the general equivalence of the Olean rock section with that of northwestern Pennsylvania and northeastern Ohio, so that the question which they considered was not, “Is the Car- boniferous actually represented in the Olean section?” but, “Since the Carboniferous is represented in the Olean section, where should the line best be drawn between it and the Devonian ?”'* The “Carboniferous” types cited by Mr. Butts make up a total of but seven out of a list of 59 species. All the rest are Devonian forms, most of which, and possibly all, have never been found in rocks of Carboniferous age, so that were we to consider the question whether the faunas show a predominating Devonian or Carboniferous facies, there could be but one answer. It is only by adhering to the rule of “first appearance” that these formations can with any justification be called Carboniferous. Let us, however, consider the Carboniferous character of the seven spe- cies on which this age determination depends. In addition to two fishes referred to the Carboniferous genera Ctenodus and Gyracanthus, the list includes five invertebrates. These are Oehlertella pleurites, Orthothetes crenistria, Glossites (Sanguinolites) amygdalinus ?, Sphenotus eolus ?, Crenipecten winchelli. I have already expressed the opinion that but little reliance can be placed upon the Schuchertellas in matters of correlation because of the difficulties of drawing any satisfactory lines between species or supposed species in the genus. The significance in the present instance is still 13This reference of the Knapp and Oswayo formations to the Carboniferous has re- cently been reaffirmed by Hartnagel (New York State Museum, Handbook 19, p. 87 et seq., April 1912), without the discussion that would seem to be demanded by the subsequently known fact that these formations occur below the Waverly group of Ohio (at least if the Berea sandstone is taken as the base). 308 ANNALS NEW YORK ACADEMY OF SCIENCES farther reduced by the fact that Schuchertella (Orthothetes) crenistria is a strictly European species, though the name has been applied to sev- eral forms in the Waverly which might be regarded as a single species or split up on rather small differences into several, according to the dispo- sition of the investigator, and of which some are doubtfully distinct even on trivial characters from the common Chemung form Schuchertella che- mungensis. From my own experience if any brachiopods are less satisfactory for identification and therefore for correlation than the Schuchertellas, they are the Discinoids, to which Oehlertella pleurites belongs. On the whole, however, this species, which is a rather common “Bradfordian” type, must be regarded as Carboniferous rather than Devonian in its bearing. The Pectinoids are greatly diversified in shape and especially in exter- nal ornament, and a number of fairly distinct groups can be made on superficial characters, some groups small and peculiar, others more com- man and generalized. To the commonest and most general of these, Crenipecten winchelli belongs. Now species superficially very similar to this are found in other genera, such as Pecten, Aviculipecten, Deltopec- ten, so that, as the hinge characters are very seldom to be observed, it is usually impossible to determine with certainty the generic group to which these commonplace Pectens belong, and this fact naturally brings into doubt the specific identification even when the superficial resemblance is close. The identification in this case is made without a query, however, and Crenipecten winchelli must be regarded as a distinctly Carboniferous type. Glossites amygdalinus and Sphenotus colus are also distinctly Carbon- iferous species, but in their case the identification is admittedly doubtful. I have not consciously made little of the evidence presented by Mr. Butts, yet two of the species in his list are obviously identified with doubt and it seems to me that of the remaining five the three invertebrates and possibly also the two fishes belong to types in which the discrimination of species is difficult and unsatisfactory even with very good material. Furthermore, in interpreting the evidence I would think it wiser to ex- tend the range downward of seven species rather than extend the range upward of 52. In seeking to determine whether the Bedford shale should be classed as Devonian or Carboniferous, the problem is not perhaps whether it carries a Chemung or a Kinderhook fauna; it is not a mere matter of correla- tion, though correlation is involved. The fact seems to be that the Bed- ford shale represents part of an interval between the base of the Missis-— sippian and the top of the Devonian as those systems have usually been GIRTY, GHOLOGIO AGH OF THE BEDFORD SHALE 309 defined and the real question is whether as such, from all the evidence at hand, it belongs more properly with the one system or with the other. The base of the Carboniferous system in this country, as usually recog- nized, is the Kinderhook group of the Mississippi Valley; similarly the top of the Devonian system is the Chemung group of New York. Now, there is substantial evidence for believing that the Berea sandstone repre- sents about the horizon of the Kinderhook and that it occurs several hundred feet above the top of the true Chemung. The evidence for this may be briefly summarized as follows: The representative of the Berea in Crawford and Erie counties in northwestern Pennsylvania appears to be the “Corry” sandstone. The “Corry” is more fossiliferous than the Berea and contains a varied and characteristic fauna. The “Corry” horizon carrying this fauna can be traced eastward to Cobhams Hill just east of Warren, where it comes in immediately above what has been called the “sub-Olean conglomerate” (Knapp formation), in the short interval which separates that formation and the Olean conglomerate. Beyond this the “Corry” horizon cannot be recognized, but it seems to be a mat- ter of common agreement™ that the “sub-Olean conglomerate” at Warren and the beds beneath represent the formations which in the Olean quad- rangle come in below the Olean conglomerate, where an interval of about 500 feet, comprising the Oswayo and Cattaraugus formations, occurs above the typical Chemung. Similar facts are indicated by I. C. White’s work in Crawford and Erie counties,’* since he recognizes the Venango oil sand group (which he calls Upper Chemung), with a thickness of 310 feet, and the Riceville shale with a thickness of 80 feet as intervening between the Chemung proper and the formations for which he used the names Corry and Cussewago. However many errors in detail there may be in these tracings and correlations, it seems safe to conclude that. an interval of 400 or 500 feet does intervene between the top of the true Chemung and the “Corry” (Berea) sandstone in this area, which is prob- ably represented in Ohio by the Bedford, Cleveland and Chagrin forma- tions. The first point to be considered in the paleontologic aspect of the problem is the affinity of the Bedford fauna, its predominant Devonian or Carboniferous facies interpreted on the facts of the general region in which the Bedford shale and the Bedford fauna were developed. Many genera and a few species, after a greater or less development in the Devonian, pass upward into the Carboniferous, ranging to various horizons in the Mississippian or even above. In most cases, there is no -14 See the report by Glenn, Butts and Clarke already cited. 158 Sec. Geol. Sury. Pennsylvania, Rept. Q. 4, 1881. 310 ANNALS NEW YORK ACADEMY OF SCIENCES general character or characters by which the Carboniferous species as a whole differ from the Devonian species. In some instances, however, certain general types within a genus appear to be restricted to one sys- tem of rocks or the other. Thus, the punctatus group of Producti is a distinctly Carboniferous development of the genus so far as known. Now, the great bulk of the Bedford fauna, as will be seen from an ex- amination of the table given on another page, belongs to types not char- acteristic of either system. Most of them would not appear out of place in either a distinctly Devonian or a distinctly Carboniferous fauna. In such an association, one might say “This is a new species in this fauna” but not “This is a Devonian species” or “a Carboniferous species,” as the case might be. Thus most of the Bedford species, considered in their broader relations, are ambiguous in deciding the Devonian or Carbon- iferous affinities of an intermediate fauna. One might indeed take up the Bedford fauna species by species and draw an inference from the number of Devonian, of Carboniferous and of new species as to whether the fauna should be grouped with the Devonian below or with the Car- boniferous above. Such a careful canvass of the relationship of the dif- ferent Bedford species would require more time than it has been possible for me to give and would almost need be accompanied by a discussion of each species, such as would be out of place in a paper of the present scope. Besides this, as between closely related species in the Devonian and the Carboniferous the conclusion reached in the identification would many times be a matter of personal opinion. Comparisons sufficiently ample have been made, however, to show that many of the Bedford spe- cies are new and that the Carboniferous alliances are at least not more numerous than the Devonian. J propose, on the other hand, to point out a few instances of larger groups than species, about the identifica— tion of which there can be less room for personal differences of opinion and which, because they do represent larger groups, carry more weight than species themselves, for I take it that the horizon which marked the extinction of the genus Spirifer would be more noteworthy than that which marked the extinction of some one species of Spirifer, such as S. keokuk. Of such peculiarly pre-Carboniferous types, the first in my list is the genus Pholidops, which has never the world over I believe been found at horizons recognized as Carboniferous. The next on the list is Delthyris, which has usually been identified in the Bedford fauna as Spiriferina and which I have already discussed at some length. This is a distinctly Devonian type of Spirifer and with one or two exceptions, to which reference will be made later, has never been cited from Carbon- iferous rocks. Next come the types which I have called Nucleospira ? sp. and Camarospira ? sp. GIRTY, GEOLOGIC AGE OF THE BEDFORD SHALE ant These two forms present serious difficulties of exact identification, being complicated with each other and with the two terebratuloids which I have called Cranena aff. subelliptica and Cryptonella ?sp. These are among the rarer forms of the Bedford fauna. When preserved in the shale they are apt to be badly crushed, but they often retain the shell, so that its structure can be determined. When preserved in the calcareous nodules, the shell is not retained (or its structure is obscured), but the proportions are not seriously altered. Thus, among these poorly charac- terized, generally ovate forms there are clearly two types, one with a punctate and one with a fibrous shell and of each type there appear to be two species, distinguished more or less strongly by size and configuration. Where the specimens have their real characters obscured by crushing or in other ways (and this is true of many of them) they cannot be satis- factorily placed in this scheme. The shells with fibrous structure have the general appearance of Athyroids, and for such they might casually be mistaken, but the ventral valve (and in one type both valves) is fur- nished with a well-developed median septum. ‘This character is not only alien to the Athyroids, such as Composita which the configuration sug- gests, but I do not know of any Carboniferous genus which has at once this shape and this structure. The larger of the two species suggests Camarospira more than any other genus with which I am acquainted, and the smaller more transverse one, which has a dorsal as well as a ven- tral septum, is certainly very suggestive of Nucleosmra. I have even observed what appear to be traces of fine sete on external molds. It cannot be positively asserted that these forms belong to the genera named, but it is true so far as I am aware that no genera having the character of these Bedford shells are known in any Carboniferous rocks of the Appalachian region. A few occurrences of Delthyris and Nucleo- spira have been noted in the Kinderhook group of the Mississippi Valley, but aside from this the Pholidops, the Delthyris, the Nucleospira ? and the Camarospira ? are peculiarly Devonian types and are not found in the Carboniferous. On the other or Carboniferous side must be mentioned the Syringo- thyris, which can, however, no longer be regarded as distinctly Carbon- iferous in its generic range. The Bedford form is, however, identified with a Carboniferous species. Again, I have a single very poor Spirifer which seems to belong to the marionensis group (a Carboniferous type), but which may be a somewhat abnormal S. disjunctus (a Devonian type). Lastly there is a species of Pholadella which is more nearly allied to the Carboniferous P. newberryi than to the Devonian P. radiata. These Carboniferous affinities, it will be noted, are specific, while the Devonian - ones are generic. 3479) ANNALS NEW YORK ACADEMY OF SCIENCES To summarize the matter so far as considered, the Bedford fauna is in many respects unique. It can be traced southward into Kentucky, but it cannot be traced eastward into Pennsylvania. Its place in the “Brad- fordian” of Pennsylvania has not been determined. It is distinct from the “Bradfordian” fauna. It is distinct from the Chemung fauna. It is quite distinct from the Chagrin fauna, which underlies it in the same section, and which, while differing in important particulars from the typical Chemung fauna, has nevertheless more of a Chemung aspect. It is equally distinct from the overlying Berea (“Corry”) fauna, which has more of a Mississippian aspect. It has a Devonian, or, as has sometimes been said, a Hamilton facies, because, while it consists mostly of genera which range into the Carboniferous and of species many of which have Carboniferous affinities, it is nearly lacking in the strictly Carboniferous types which abundantly accompany the latter at higher horizons and proclaim the geologic age, and because it contains a few Devonian types which very rarely and in the region under consideration never, so far as. known, range up into the Carboniferous. The Bedford and Cleveland formations may be lacking in northwestern Pennsylvania owing to pre-Berea erosion, or to some other cause, but I hardly believe this to be the case. If the Bedford does represent part of the typical “Bradfordian” section, and if its fauna is a peculiar and local development of the “Bradfordian fauna,” then we must enlarge the dis- cussion to include the “Bradfordian” (“Upper Chemung’) faunas, whose Devonian facies is conspicuous.*® In the one case (if properly lying above the “Bradfordian”’ but removed by erosion in the typical! sec- tion) the stratigraphic evidence, and in the other the paleontologic evi- dence, is stronger for classifying the Bedford shale as Devonian. To this must also be added the fact of a conspicuous faunal break between the Bedford and the Berea, and the fact already noted that in the Berea (“Corry”) we have, for the first time in this region, a fauna with a pre- dominating Carboniferous aspect, one which shows many new features when compared with the typical Mississippian, but which is distinguished by the absence of most of the Devonian types of lower horizons and the presence of many characteristic Carboniferous ones. ' Thus far it seems that the evidence has been strongly favorable to classifying the Bedford with the Devonian. If we broaden the discus- sion so as to include a larger field, that of the typical Mississippian area, which apparently represents a conspicuously different province, or at 16 See the “Upper Chemung” species of Hall’s New York reports and Butts’s lists of the faunas of the Knapp and Oswayo formations. These show a fauna with well marked differences from the typical Chemung, yet with, in my opinion, a distinctly De- vonian aspect. GIRTY, GEOLOGIC AGE OF THE BEDFORD SHALE 313 least represents a conspicuously different faunal development, the ques- tion becomes more confusing and the conclusion somewhat less satis- factory. A search among the American faunas now known for one which is comparable with that of the Bedford shale probably reveals none so simi- lar in a general way as a certain phase of the Kinderhook developed at Hamburg, Illinois, and the closely related one developed in the Glen Park limestone member of the Kinderhook of Missouri. The constitu- tion of these faunas, generically considered, is surprisingly close and many of the species seem to be related. These may be arranged in parallel columns, as follows: Limestones of the Kinderhook at Glen Sealine, Salle Park, Louisiana or Hamburg Schuchertella herricki Schuchertella chemungensis Productella pyxidata var. bedfordensis Producitella pyxidata Rhipidomella n. sp. Rhipidomella missouriensis Cryptonella ? sp. Cryptonella sp. Delthyris n. sp. Delthyris missouriensis Syringothyris carteri Syringothyris carteri Spirifer aff. marionensis ? Spirifer marionensis Athyris aff. hannibalensis Athyris hannibalensis Macrodon hamiltonie Macrodon sulcatus Leda diversa Leda diversoides Bellerophon aft. jeffersonensis Bellerophon jeffersonensis Platyceras sp. Platyceras erectoides Tropidodiscus cyrtolites ? Tropidodiscus cyrtolites There is also a Chonetes in both faunas, though of not very close rela- tionship; a Camarotechia, though the Bedford form is large and the Hamburg form small; and a Nucleospira, though also specifically dis- tinct. In fact, while many of the same genera are present in both areas, there are very few species which are really identical, and those for the most part belong to genera in which the identification of species is diffi- cult and a satisfactory identification impossible. Syringothyris carteri is an example. Professor Weller’” has pointed out the close relationship which exists between the Kinderhook fauna at Glen Park, Missouri, and that from the odlitic beds at Hamburg, Illinois, and he has also called attention to the conspicuous Devonian facies which these faunas present (p. 463), a resemblance which (like that of the Bedford) would seem to ally them with the Hamilton rather than with the later Devonian faunas. Pro- fessor Weller finds that 12 out of the 31 species at present known from 17 Acad. Sci. St. Louis, Trans., vol. 16, p. 462 et seq. 1906. 314 ANNALS NEW YORK ACADEMY OF, SCIENCES Glen Park can be paralleled in the Hamilton formation of New York, while only six species have parallel forms in the Chouteau limestone of Missouri. Although the Chouteau fauna is so near at hand, he is able to find specific identity in only two species and those are Schuchertella chemungensis and Tropidodiscus cyrtolites. When we consider this fact and that the fauna comprises such Devonian genera as Hunella, Atrypa, Nucleospira and Delthyris, not to mention the fish Ptyctodus eastmani, it would seem that the Devonian proclivities of the fauna far outweigh the Carboniferous ones, even with due consideration for the two doubt- - fully identified crinoid genera. This evidence is largely neutralized, however, when other factors are taken into consideration. The Kinderhook faunas of the upper Mississippi Valley show local ' facies to an almost unprecedented degree. To some extent, this differ- entiation may have a zonal explanation, but it is also probably local and environmental, since the lithologic character of the beds is also extremely variable. Professor Weller recognizes a northern and southern type of Kinderhook fauna which were contemporaneous, but almost entirely dif- ferent. The Chouteau limestone exemplifies the southern fauna and with this, as just noted, the Glen Park fauna has only two species in common, although Professor Weller apparently regards them as occupy- ing the same horizon. Both the northern and the southern faunas are also highly diversified. Beneath the fauna of the odlite at ghasiues, referred to above, which so closely resembles that of the Glen Park limestone, there is another having a considerably different facies. The latter Professor Weller cor- relates with the well-known fauna of the Louisiana limestone and this in turn with the typical Kinderhook of that ilk, which corresponds to the lower and larger portion of the Kinderhook section at Burlington, On the other hand, he correlates the fauna of the odlite at Hamburg with the Glen Park fauna and the Glen Park fauna with the Chouteau fauna, and with the upper part of the Kinderhook section at Burlington, if I understand him aright. Several very different facies are presented by these faunas. That of the Louisiana limestone (at Louisiana, Missouri, and Hamburg, Illinois) is distinctly more Carboniferous than that of the odlitic limestone at Hamburg and Glen Park, which, as already noted, are rather conspic- uously Devonian, though they occur above the other in stratigraphic position. The faunas of the Chouteau limestone and the topmost Kin- derhook at Burlington, with which the faunas of the odlitic limestone at Glen Park and Hamburg appear to correlate, are still more conspicuously Carboniferous, and they have so been recognized for a long time. GIRTY, GEOLOGIC AGE OF THE BEDFORD SHALE 315 As I have just shown, the fauna of the Bedford shale and the fauna of the odlitic limestone at Glen Park and at Hamburg are in some respects strikingly alike, but, though the resemblances are undoubted, there are also numerous and important differences. The resemblances consist of the presence in both faunas of identical genera and of related species. Identical species, however, are few and not of the first importance. The Delthyris, the Nucleospira, the Macrodon, etc., of the Bedford shale are not the same species as the Delthyris, the Nucleospira, the Macrodon, etc., of the odlite at Glen Park and at Hamburg. Correlation by similar species is certainly much more hazardous and less satisfactory than corre- lation by identical species. Indeed, although we of course know that all these faunas are more recent than the Hamilton, the table compiled by Professor Weller would indicate that the Glen Park fauna is almost as closely related to the Hamilton faunas as it is to that of the Bedford shale and much more closely related to the Hamilton than to the contem- poraneous Chouteau fauna. Restricted to their own showing, therefore, I believe that a correlation of the Bedford and Glen Park faunas would not be justified, except in a very provisional and tentative manner. However that may be, if, instead of considering the two faunas as iso- lated occurrences, we include, as we are forced to do, the faunas asso- ciated or correlated with them—the typical Kinderhook, the Chouteau and the Louisiana faunas of the Mississippi Valley in the one case, and the “Bradfordian” faunas of Pennsylvania in the other—it seems clear that we have two entirely distinct faunas, the one showing a strongly Carboniferous and the other a strongly Devonian facies, and we cannot conclude that they are contemporaneous sees of the same faunal zone on any evidence now known. The Bedford and Cleveland shales cannot be definitely identified in the “Bradfordian” rocks of northwestern Pennsylvania, either litho- logically or paleontologically, but there is an interval between the Berea (“Corry”) sandstone and the Venango oil sand group which seems to correspond in a general way to that represented in Ohio by these forma- tions, and I personally but little doubt that Bedford and Cleveland do correspond to strata in the “Bradfordian.” Even, however, if they do not, and the “Bradfordian” with its strongly Devonian fauna does en- tirely underlie the Bedford, I believe that the correlation of the Bedford shale with the odlites at Hamburg and Glen Park would not be justified at present. The small number of identical species and the almost complete ab- sence of all those characteristic Carboniferous types which by Professor Weller’s correlations occur at the same horizon as the odlitic limestones 316 ANNALS NEW YORK ACADEMY OF SCIENCES at Glen Park and Hamburg or below, though not yet found associated with them, would, in the absence of substantial evidence to the contrary, indicate that the Bedford shale was not really a contemporaneous forma- tion. The correlation of other horizons in the same sections is intimately connected with that of the Bedford shale. Professor Weller?® recognizes in the famous goniatite-bearing limestone at Rockford, Ind., a repre- sentative of the Chouteau limestone or southern Kinderhook. It is perhaps impossible to tell in fact, as it is certainly impossible to tell from his discussion, whether the limestone at Rockford corresponds to that part of the Chouteau which correlates with the northern Kinderhook or to that part which he recognizes in the Burlington section as occurring above the northern Kinderhook. If the Bedford represents the Glen Park horizon, it would apparently on the one hypothesis correlate with the goniatite-bearing bed at Rockford (and it does contain some goniatites of scarcely determinable genera), while on the other hypothesis it would come in above it. The position of the Bedford shale above the black Cleveland shale is at first suggestive of the position of the goniatite- bearing bed at Rockford above the Devonian black shale of Indiana, but there is no assurance whatsoever that the two black shales represent the same horizon and, even if such were shown to be true, it would not neces- sarily follow that the succeeding formation in the one case corresponded to the succeeding formation in the other. Indeed, until quite recently it has been the general consensus of opinion that the goniatite-bearing bed and the black shale beneath were quite separate and distinct forma- tions divided by a long-time interval, the one of Carbonifrous, the other of Devonian age, and no satisfactory evidence has yet been produced for believing otherwise. On the other hand, such facts as I am acquainted with both of stratigraphy and paleontology go to show that the Bedford and Cleveland shales are related in the closest manner and must be classed together wherever they are classed. The early Mississippian sections of Ohio and of the Mississippi Valley show great differences of development, both in the sediments which accumulated there and in the animal life which those sediments helped to condition. They probably constitute distinct provinces. There are no faunas in Ohio closely allied to the typical Burlington and Keokuk faunas,—nothing to correspond to the rich development of crinoid life which is found in those faunas and which doubtless did much to deter- mine the character of the associated life, unless still different influences determined both. The 18 species of crinoids known from the Cuyahoga 18 Loc. cit., p. 469. GIRTY, GEOLOGIC AGE OF THE BEDFORD SHALE 317 shale, though belonging to genera well represented in the early Missis- sippian of the Mississippi Valley, do not occur outside the State. Any- one, however, who will compare the fauna of the Chouteau limestone with that of the Cuyahoga shale, as found at such points as Medina, Richfield, Lodi and Royalton, cannot fail to find great similarity and not a few identical species. I am not prepared to state the exact extent of this resemblance, but my studies would indicate strongly that, if the Cuyahoga fauna is to be found anywhere in the Mississippi Valley, it is to be found in the Chouteau hmestone. The Waverly localities which I have mentioned are all, I believe, in the upper Cuyahoga. By definition, the Chouteau limestone is part of the Kinderhook group and therefore in stratigraphic position inferior to the Burlington limestone, but I am much disposed to think that the Chouteau limestone really correlates with the lower Burlington, the fauna which we know as the Burlington fauna being developed in and largely confined to the upper Burlington. I have already given a list of the fossils found in the Berea (“Corry”) sandstone which underlies the Cuyahoga shale. This fauna contrasts _ strongly with both the Bedford fauna below and the Cuyahoga fauna above. It has a much more marked Carboniferous aspect than the Bed- ford fauna, even if we exclude the faunas apparently contemporaneous with the Bedford having a more distinctly Devonian facies. Though, of course, showing great individuality, the “Corry” fauna is not only distinctly Carboniferous, but in some of its elements it is distinctly Kinderhook, as for instance in the genus Paraphorhynchus, a type which Professor Weller regards of special importance and which is said to be characteristic of the northern Kinderhook. _ It is interesting to find three faunas in the Ohio section showing re- semblances, more or less illusory perhaps, to these three aspects of the Kinderhook faunas of the Mississippi Valley, and it is also interesting to compare the stratigraphic relations of these faunas in the two areas on the assumption that the Cuyahoga shale correlates with the Chouteau limestone, the Berea sandstone with the northern Kinderhook (the Chonopectus fauna of the Kinderhook sections at Burlington), and the Bedford shale with the odlite at Glen Park or Hamburg, as is to some extent suggested by faunal similarities. According to their strati- graphic relationship in typical sections in the Mississippi Valley, the Bedford shale should not lie below the Berea sandstone, but above it. It should in fact even be contemporaneous with part, if not with all of the Cuyahoga shale. Tf, however, the Kinderhook relationship of the Bed- ford be eliminated, as I believe it can be eliminated owing to its probable relationship to other “Bradfordian” faunas, this contradiction largely 318 ANNALS NEW YORK ACADEMY OF SCIENCES disappears. The Cuyahoga and Berea together represent the Kinder- hook of the Burlington section (in which, if I understand him aright, Professor Weller thinks that the upper 15 feet corresponds to the upper part of the Chouteau, both stratigraphially and faunally, while the lower part of the two sections corresponds stratigraphically but not faunally), and they have the same relative position in both sections. If my hypoth- esis of the equivalence of the Chouteau with the lower Burlington is correct, then of course the Berea alone represents the entire Kinderhook section at Burlington and presumably its correlates at Louisiana and at Hamburg with their varying faunas. Professor Weller, as already described, recognizes two types of Kin- derhook faunas, the northern one, the typical Kinderhook, not being found at all southward in southern Missouri and Arkansas; the other, the Chouteau, occurring in Arkansas and Missouri and represented by a few feet of rocks above the northern Kinderhook in the section at Bur- lington. Professor Weller’s interpretation of these facts is that the two faunas were developed contemporaneously in disconnected basins, to the more northern of which the southern fauna gained access near the close of Kinderhook time. Tentatively, I would prefer to explain these rela- tions by supposing that the southern Kinderhook was entirely later than the northern and was represented in the Burlington section not by the topmost Kinderhook alone, but by the lower Burlington also. However that may be, the disappearance southward of the northern Kinderhook fauna is somewhat suggestive of the southward thinning of the Bedford and the Berea formations, as recently described by W. C. Morse and A. F. Foerste.*® The careful stratigraphic work of these writers, combined with that of Professor Prosser, indicates that the Bedford and Berea gradually thin to a feather edge and presumably disappear as recognizable formations in east-central Kentucky. To some extent, they also lose their distinct- ive lithologic characters, so that Morse denominates as “Bedford-Berea” the interval of shale and sandstone which they fill between the black Ohio and Sunbury shales. The tracing by stratigraphy is corroborated by the occurrence of more or less characteristic Bedford fossils at the base of this interval in Kentucky, and Dr. Foerste himself very justly raises the question whether the final appearance of these sediments, which con- sist of shale alone, should not be referred solely to the Bedford forma- tion, and the preceding occurrences in which the shales predominate below and the sandstone above should not be divided into Bedford and 79 Jour. Geol., vol. 17, p. 164 ét seq. 1909. GIRTY, GEOLOGIC AGE OF THE BEDFORD SHALE 319 Berea, respectively. In view of the fact that the Berea possesses a char- acteristic fauna which has not been found in the sections under consid- _ eration, whereas the Bedford fauna has been found there, and in a shale, and at the base of the interval, it seems to me that that portion of the interval would better be identified as Bedford alone, whatever is done with the upper part, though the evidence would suggest the advisability of calling the upper sandy beds Berea where they are present, and the whole interval Bedford where they are not. The bearing which this aspect of the Bedford-Berea stratigraphy has on the question of the geologic age of the Bedford shale is not entirely clear. From one point of view, one might say that it did not affect the classification of the beds at all, except insofar as it made them difficult to distinguish in the field and to delineate on a map. On the other hand, it might be urged with some force that since by the expansion and differentiation of the Chattanooga shale in a northward direction, that formation seems to cover an interval including the lower Cuyahoga. the Berea, the Bedford, the Chagrin and probably the Huron formations, and in a manner to bind them together into one group of sediments, they ought all to be classed as Devonian or all as Carboniferous. This, however, does not at all agree with the facts, where these formations are differentiated and developed in an unequivocal manner, and I believe that it should not prejudice such a classification of the rocks as is indi- cated by the facts ascertainable under those conditions. Therefore, while the weight of the evidence is not entirely cast on one side of the question, I believe that so far as the facts are known they indicate the line at the base of the Berea sandstone as the proper position of the Devonian-Carboniferous boundary in northern Ohio. This is because that boundary is marked by an unconformity, by the presence of a basal sandstone and by a pronounced faunal change, such that while the fauna of the Berea (“Corry”) sandstone has a distinctly Carboniferous facies and is probably to be correlated with the Kinder- hook group of the Mississippi Valley, that of the Bedford shale, though its stratigraphic position is above the typical Chemung, has, in connec- tion with the other “Bradfordian” faunas, a distinctly Devonian facies. “PUBLICATIONS OF THE h EW YORK ACADEMY OF SCIENCES (Lyceum or N ATURAL Huisvory, 1817-1876) The publications of the Academy consist of two series, viz. : 5 (1). The Annals (octavo series), established in 1823, contain the | _ scientific contributions a ous of researches, together with the rec- A seks. of the Annals read in general with the calendar year nd i is sold at the uniform price of three dollars per volume. The articles omposing the volume are printed separately, each in its own cover, and the separate articles depends upon their length and the number of illus- i - trations, and may be learned upon application to the Librarian of the "Academy. The author receives his separates as soon as his paper has ; ase printed, the date of issue appearing above the title of each paper. (2) The Memoirs (quarto series), established i in 1895, are issued at : ee calue intervals. It is intended that each yolume shall be devoted to monographs relating to some particular department of Science. Volume _ ete. The price is one dollar per part as issued. _ All publications are sent free to Fellows and Active Members. The Annals are sent to Honorary and Corresponding Members desiring them. i Subscriptions and inquiries concerning current and back numbers of : ny fe the pnecations of the Academy should be Ee to THE LIBRARIAN, New York Academy of Sciences, care of American Museum of Natural History, New York, N. Y. re distributed i in bundles on an average of three per year. The price of Tis devoted to Astronomical Memoirs, Volume II to Zoélogical Memoirs, “ANNALS OF THE NEW YORK ACADEMY OF SCIENCES Vel. XXII, pp. 321-326: ; ¢ Editor, EpmunpD Orts Hovey ra CHANGES IN THE BEHAVIOR OF THE EEL DURING TRANSFORMATION BY BasHFORD DEAN : NEW YORK : PUBLISHED BY THE ACADEMY 5 DrcemBer, 1912 THE NEW YORK ACADEMY OF SCIENCES (Lyceum oF Narurau History, 1817-1876) . — OFFICERS, 1912 President—HMrrson McMIttin, 40 Wall Street Vice-Presidents—J. E>pMUND WoopMAN, FREDERIC A. Lucas CHARLES LANE Poor, R. 8S. WoopwortH Corresponding Secretary—Hunry H. Crampton, American Museum - Recording Secretary—Epmunp Otis Hovey, American Museum Treasurer—HeEnry L. DoHERTY, 60 Wall Street Inbrarian—RauPxH W. Tower, American Museum Editor—HKpmunp Ot1s Hovey, American Museum SECTION OF GEOLOGY AND MINERALOGY Chairman—J. E. Woopman, N. Y. University Secretary—Cuartes P. Berkey Columbia University SHCTION OF BIOLOGY Chairman—Frepveric A. Lucas, American Museum Secretary—WiI.Lu1amM K. Grecory, American Museum . SECTION OF ASTRONOMY, PHYSICS AND CHEMISTRY Chairman—Cuar.eEs LANE Poor, Columbia University Secreiary—F. M. PuprrseEn, College of the City of New York SECTION OF ANTHROPOLOGY. AND PSYCHOLOGY Chairman—R. 8. Woopwortx, Columbia University Secretary—F REDERIC LYMAN WELLS, Columbia University The sessions of the Academy are held on Monday evenings at 8:15 o’clock from October to May, inclusive, at the American Museum of Natural History, 7th Street and Central Park, West. [Annats N. Y. Acap. Scr, Vol XXII, pp. 321-326, 5 December, 1912] CHANGES IN THE BEHAVIOR OF THE EEL DURING TRANSFORMATION By BasHrorp DEAN (Read in abstract before the Academy 13 May, 1912) The literature of animal behavior gives as yet little attention to the changes which occur in animals during the period of metamorphosis. This is a gap in our knowledge not remarkable perhaps when we con- sider how little is yet recorded of the behavior of many types of adult animals, even of common forms. None the less, it is precisely during the period of transformation that one may expect to find clues as to interesting conditions in mind-mechanism, for during this short period adjustments are completed which change, as it were, one functional “species” into another; for an animal may remain for years in its larval form almost unaltered, and it may subside again into a changeless form after a kaleidoscopic transformation. In fact, the more sudden the change in transformation, the more interesting it should be from the point of view of connecting habits with structures, for it would here bring into sharpest relief morphological changes and make them the more easily linked with changes in behavior. In the larval history of fishes, observations in this field have rarely been recorded. The teleosts, where conspicuous larval stages occur, are little studied, even in the case of those members of the group which have the most complete metamorphosis. The form-changes of eels have been described by a number of authors (Grassi, Calandruccio, Cunningham, Eigenmann and others)., and the changes are so marked that we can readily predict from them striking changes in behavior. That the latter actually occur, and in marked degree, was clearly brought home to the writer when an opportunity came to him in Japan (Misaki) to observe the transformation of a Leptocephalus into a Conger,—possibly Conger (Leptocephalus) mala- baricus (Day).* His notes, especially upon its behavior, are perhaps worthy to be recorded on account of the interesting nature of the “larva” and from the fact that this form is not apt to be observed. In point of fact Leptocephalus seems rarely to have been kept living in an aquarium tWRANCIS DAy: “The Fishes of Malabar.” Pl. xix. 1865, (321) 322 ANNALS NEW YORK ACADEMY OF SCIENCES more than a few hours. The specimen in question, I may mention, was in perfect condition when taken. It was noticed in the bag of a seine,— by accident rather than by design,—for had the fish not been actively moving at a particular moment, it would have escaped unnoticed on account of its glassy transparency. It proved to be hardy and lived in an aquarium for over three weeks, during this time undereot its metamorphosis. September 13. Larva (Fig. 1) almost colorless, even in light of dif- ferent intensities; it is rarely at rest; it is apt to swim rapidly and with Fic. 1.—#el larva. September 13. About natural size a kind of lurching movement. When advancing slowly, its height, which insures contact with a large surface of water, allows it to move with precision,—in the sense that a pencil held vertically in the first in-bent curve of the fish’s body will not be touched by the fish as it advances,— in other words, that the fish does not show lost or slipping movements. When resting, the young fish arranges itself in irregular vertical coils, thus probably keeping its balance. When disturbed (snout touched with a pencil point), the fish retreats tail foremost, the head remain- ing passive: if disturbed again, the head will be quietly drawn back, the motion starting as before with a withdrawal of the tail and hinder trunk. If disturbed repeat- Fic. 2.-—Hel larva. Retracted edly, however, the fish will either swim meee about actively or draw itself into a close coil (Fig. 2). This position, however, it will sometimes assume without artificial stimulus, e. g., after it has become “tired” swimming around the wall of a circular jar. Leptocephalus from time to time secretes con- siderable mucous: this remains attached but is finally “brushed off’ (in a mass) at the tail end of the body. Such a bit of slime will occasionally be touched by the young eel when swimming about; it is evidently dis- tasteful, for the young fish speedily frees itself, shaking its head in a curiously energetic way. DEAN, BEHAVIOR OF THE EEL DURING TRANSFORMATION 3923 September 15. The fish is now more easily seen. There is a slight clouding of its transparent sides, especially near the lateral line (Fig. 3). Its eyes are conspicuous and show numerous movements. It is more active than in the earlier stage, sometimes swimming with broad undu- lations (Fig. 4) different in type from earlier movements. A patch of ys WH ddddiee Me j DN \\ ° \\ Fic. 3.—£el larva. September 15 color, brownish orange, appears on the ventral body wall, just behind the gill opening (jugular villi). . September 18. The larval length and breadth are rapidly becoming reduced (Fig. 5). The caudal fin and dorsal ridge appear. The colored jugular patch has now de- veloped into a velvety mass bd of lighter color, and possibly serves as a larval adhesive organ, which hangs freely in the water. The intestine can be outlined. Pigmenta- tion is noted, especially along the lateral line and on the head-roof, and the en- tire fish has a faint purplish , \ tone. It remains more often at the surface than before, \\y eee here occasionally floating and swimming on its side, now and then thrusting its head out of water. It remains longer in one position than heretofore. If disturbed (head touched), it will wriggle its head backward,—and does not initiate the backward movement from the tail as in the earliest stage. September 19. Changes progress rapidly (Fig. 6). The coloration is distinctly purple, with whitish spots near the tail, and pigment patches on the ventral wall of the head.and within the neural axis. Vertebre lie. 4.—Hel larva. Position in swimming 394 ANNALS NEW YORK ACADEMY OF SCIENCES Fie. 5.—Eel larva. September 18 Fic. 6.—LHel larva. September 19 Fic. 7.—Hel larva. September 22 Fic. 8.—#eil larva. September 28 Fic. 9.—Eel larva. October 2 Figures about natural size DEAN, BEHAVIOR OF THE EEL DURING TRANSFORMATION 325 are now seen. The patch of jugular villi is reduced in size. The be- havior of the young fish is eel-like. It remains motionless for longer periods, occasionally lifting and turning its head, and there are pro- nounced movements of its opercula. September 22. Larval coloration is apparent in the white spots above and below the tail. In general the advances are clearly in the direction of the mature eel. The vertebre are conspicuous; the visceral wall be- comes opaque; the jugular larval organ is represented by a clump of scattered filaments; the gill region is more conspicuous (Fig. 7); the gill arches show the red lamelle, expand broadly and contract; the mouth opens wide; there is no movement in the neck region; the pec- toral fins function; and the swimming is snake-like, with more effort than propulsion, 1. ¢., slipping, unlike the precise movements of earlier stages. September 26. From now onward, the changes are less noticeable. Larval coloration is retained, e. g., in the light colored spots. The trunk is opaque, even in the gill region. September 28. At this stage, the last trace of the larval jugular organ was noticed. The body is thickened; the white spots have disappeared (Fig. 8). October 2. The last stage recorded (Fig. 9). Transformation is practically complete. Measurements of this contrasted with the earliest stage show a surprising shrinkage in the length and height of the young eel,—more exaggerated even, than in the cases described by Grassi. In the present instance, the young fish is about one-half the length of the earliest stage, and one-third of its height, after a growth period of about three weeks. It is another example of the paradox that development may be accompanied by considerable diminution in size. Especially interesting in the foregoing transformation is the rapidity with which the behavior of the young eel changes. This is not brought out in adequate detail in the present note, but it may be said that the observer could not but feel that the larva behaved like an animal suite of a different species from the one of the days before, or of the days following. This state of affairs predicates, obviously, kaleidoscopic changes in elements of the central nervous system, and astoundingly delicate and rapid adjustments; but whether these can be actually de- termined, 7. e., in the physical characters of the cells of brain and cord, must yet remain an open question. It can be solved only when an abundant material of Leptocephalus falls into the hands of a specialist who can bring to his aid the latest neurological technique. 326 ANNALS NEW YORK ACADEMY OF SCIENCES From a phyletic point of view, on the other hand, the origin of the rapidly changing behavior correlated with morphogenetic changes is less difficult to understand. We have, first of all, numerous grounds for concluding that the larval stages of teleosts are secondary, and that eccentric forms,—i. e, those with extraordinary fins, colors, outlines,— are derived from “larve” in which such extreme structures did not exist. ‘In the case of the eels, therefore, we can reasonably picture a progressive form of development, such, for example, as occurs in many of the older groups of teleosts. We next suggest that within this progressive series of closely similar stages, one stage should become especially important as adapting the young to a particular environment. The young eel then would tend to remain longer unchanged in the special environment favorable to its feeding, movements, lack of pigment, temperature-re- quirements, etc., and this phase in its life-history would come to sup- plant the adjacent steps in the progressive series. In other words, if we erant that the development of a young eel (the montée) might be accom- plished in the space of fifty weeks, and that at the end of this period it completed the fiftieth of its intergrading stages, we could also admit that with the same total period of larval growth certain of the stages might have expanded while others contracted. Thus, to take an ex- ample, stage twenty, which earlier may have been passed through in a week, might become successively protracted to two weeks, three weeks, or months. And the stages of the montéc intervening between twenty and fifty would be correspondingly reduced. The interval was at first thirty weeks, and included thirty stages; it was next, say, fifteen.weeks in which to represent thirty stages, and finally in the case of our Lepto- cephalus, it was reduced to the astonishingly short term of three weeks, in which to represent the many stages. In such an instance there can be little question that the marked changes in behavior are correlated with abbreviated phases of development. e Ae (okies ties. established in 1823, contain the contributions | and reports of researches, together with the rec- meetings and similar matter. * : ae a a oS coincides in ae with the calendar year) Th Wie receives his separates as soon as his paper iis ‘the date of i issue epeeette above the title of each PADER: vat ions of the Academy should be addressed to THE LIBRARIAN, - New York Academy of Sciences, care of a American Museum of Natural History, New York, N. Y. BoD. 0. JoHNson: acts THE NEW YORK ACADEMY OF SCIENCES (Lyceum or Natura History, 1817-1876) OFFICERS, 1912 President—BKMERSON McMILtin, 40 Wall Street Vice-Presidents—J. EDMUND WoopMAN, FREDERIC A. Lucas CHARLES LANE Poor, R. S. WooDwortH Corresponding Secretary—Henry H. Crampton, American Museum Recording Secretary—EpmMuND Otis Hovey, American Museum Treasurer—HENRY L. DoueErty, 60 Wall Street JInbrarian—Raupx W. Tower, American Museum Editor—Epmunp Oris Hovey, American Museum SEOTION OF GEOLOGY AND MINERALOGY Chairman—J. EK. Woopman, N. Y. University Secretary—CuHAgLes P. Berkey Columbia University SECTION OF BIOLOGY Chairman—Freperic A. Lucas, American Museum - Secretary—Wi.LLiamM K. Grecory, American Museum SECTION OF ASTRONOMY, PHYSICS AND CHEMISTRY — Chairman—CHARLES LANE Poor, Columbia University Secretary—F. M. Peprrsen, College of the City of New York SHCTION OF ANTHROPOLOGY AND PSYCHOLOGY Chairman—R. S. WoopwortH, Columbia University Secretary—F REDERIC LyMAN WELLS, Columbia University The sessions of the Academy are held on Monday evenings at 8:15. o'clock from October to May, inclusive, at the American Museum of Natnral History, 77th Street and Central Park, West. [Annats N. Y. Acap. Sct., Vol. XXII, pp. 327-333. 20 December, 1912] NOTES ON THE HABITS OF A CLIMBING CATFISH (ARGES MARMORATUS) FROM THE REPUBLIC OF COLOMBIA By R. D. O. Jounson (Read in abstract before the Academy 13 May, 1912) Introductory remarks, offered by BASHFORD DEAN at the meeting—The group of catfishes (Siluroids) holds a puzzling place among fishes. That it represents one of the ancient groups of bony fishes, there can be no doubt, but whether Siluroids are descended from some special line of ganoids or whether they have been derived through a long series of specializations from some ancestor essentially carp-like remains ever an open question. The trend of later work, certainly, tends to ally them more closely with plectospondylous forms, but many of their most important structural characters have never been explained on such a basis. Thus, the limb-girdles of some of the catfishes, with their accompanying muscles, have appeared to be primitive, and there has, as far as I am aware, been adduced no evidence to show that these structures were derived from highly specialized conditions of such living plectospondyls, for example, as characinids. The habits of Siluroids, which would help to explain the signifi- cance of these abdominal structures, have not been known to be remarkable, and there is no suggestion, therefore, that the characters in question might but be interpreted as highly modified rather than primitive. Accordingly, the present paper of Mr. R. D. O. Johnson merits, I believe, the attention of the Academy, for he shows that under conditions of stress, the ventral structures of the catfish Arges have an especial value to the fish in enabling it to creep against the strongest currents and to climb with great rapidity and skill. The conclusion, therefore, is evident that we may now reasonably interpret the puzzling fin-structure of Siluroids as developed in relatively recent times, and as having little significance in terms of more ancient groups. Mr. Johnson, it may be mentioned, spent several years in the highlands of the Republic of Colombia, and although the region he studied has been visited by but few naturalists, it is nevertheless hardly to the credit of our “cloth” that these observations on fishes should first be made by a mining engineer. The creeks and rivers of the Andes Mountains in the Republic of Colombia, South America, are torrential in character. The great major- ity of them are but a succession of falls, cascades, pot-holes and short “rifles.” The rainfall in the mountains is heavy and the rock under- lying the stream beds is schistose in character and comparatively soft. whe rate of erosion is exceedingly rapid, yet the grade lines of these (327) 598 ANNALS NEW YORK ACADEMY OF SCIENCES streams stand at high angles. This unstable condition seems to be due to the elevation of the Andes during a late geological period. The heavy rainfall, at times amounting to four or five inches within a few hours, produces floods of immense volume. These go charging down the can- yons with fearful fury, and at times it would appear that nothing could withstand their sweeping energy. Yet these turbulent waters are the habitat of fishes so wonderfully adapted to their surroundings that they are able to grow and to multiply in great numbers. In external appearance they resemble the catfish or horned pout of the north. The skin is smooth and scaleless. The color is a dark mottled gray shading into a slightly yellowish tint on the posterior parts. They rarely attain a length greater than twelve inches. As an article of food Fic. 1.—Arges marmoratus Regan ; side view they are esteemed by the natives and are well known by the local name “Capitan.” They have lately been described by C. Tate Regan as Arges marmoratus.* Under usual conditions they are clumsy and awkward swimmers, wrig- gling through the water like tadpoles, but as creepers and climbers they are without rival in the fish family. The mouth is small, but is sur- rounded by a broad, soft, rubber-like flap, very thin and flexible at the edges (Fig. 2). It is a sucker mouth and the entire mechanism is so perfectly adapted to the needs of the fish that it finds no difficulty in firmly attaching itself to any convenient object. It is this ability to make a quick anchorage that enables the fish to stay at home when nature seems bent upon sweeping the canyons and water-courses clear of every- thing movable. If, however, these fish were able only to keep themselves from being washed out in flood times, they would be insufficiently equipped to main- tain an existence in these mountain streams. If they depended upon their imperfect swimming alone as a means of locomotion, whatever migratory movement they attempted would inevitably have to be made 1 Trans. Zodlogical Society of London, XVII, p. 314. 1904. JOHNSON, HABITS OF A CLIMBING CATFISH 399 m a down-stream direction. The final result would be the same as though they were unprovided with a means of anchoring themselves at will. But they are equipped with another and very efficient apparatus for locomotion, The flat sucker mouth is half of the mechanism; the other half is located on the belly. Under the skin of the ventral side, just behind a line joining the pectoral fins, there is a triangular bony plate to which are attached the ventral fins (Fig. 2). The main anterior ITuscles Fig. 2.—Arges marmoratus Regan; ventral view ribs of these fins are broad and flattened, and the flat. surfaces are thickly studded with small, sharp teeth pointing backwards. The triangular plate and its attached fins are free to move in a longitudinal direction through a distance equal to about one-sixth of the length of the fish. This movement is accomplished by means of four muscles in two pairs attached to the plate; the anterior pair extending from their attachments on each side of the plate forward to the middle point on the bony arch just below the gill openings; the posterior pair extending from an 330 ANNALS NEW YORK ACADEMY OF SCIENCES attachment at the center of the posterior edge of the plate to the anal fin. It is evident that the fish is able to create a suction pressure in the re- gion of the plate, though how this is accomplished is not apparent from the structure. By means of the alternate action of the mouth and of this curious apparatus, the fish is able to creep against a current that would baffle its efforts entirely, if it relied alone upon its fins and tail. When it Orifices for the Inflow of water to Gills —--- rr lig. 3.---Arges marmoratus Regan; dorsal view is engaged in creeping or in sticking fast to some object, the sucker mouth necessarily is closed. It is evident that the gills must be sup- plied with the life-maintaining flow of water through some other avenue. At the upper extremity of each gill slit there is an orifice provided with a valve opening inward (Figs. 1 and 3). During the diastole of the gill covers, the water flows inward through the orifices and is expelled through the gill slits during the systole. | JOHNSON, HABITS OF A CLIMBING CATFISH 331 On clear sunshiny days, these fish may be seen in the depths of the clear water hitching themselves along over the surfaces of rocks, occa- sionally swimming short distances in the more quiescent places, but seeming to depend for locomotion primarily upon their creeping mechan- cays --as “i BS = ENeN SANS SS a \ SS =o SS =: ! ! i Al iit \ ee i Ni \ZAI e | ye — a aa ) : sy Wy ay ; maT = = | AWS es il ia yj PAZANMNY, \“ MMC ZZ LEE Fic. 4.—Section of a pot-hole, twenty-two feet deep, in Santa Rita Creek, Colombia, showing ‘‘capitanes’’ ascending its rocky walls ism. They are to be found in all parts of these mountain streams, from the most slender tributaries to the foot of the mountains. It is evident from this fact that they are able to travel up stream. They are too 539 ANNALS NEW. YORK ACADEMY OF SCIENCES ‘sluggish in movement and are provided with a swimming apparatus -altogether too inefficient to enable them to dash up the high and frothy falls. At one time, the writer had occasion to divert the water of a small mountain stream so that access could be gained to a deep pot-hole from which the water, rock and gravel were subsequently removed. This pot- hole was twenty-two feet deep, nearly circular in horizontal cross-section and it varied in diameter from six to ten feet. Generally, the sides ap- proached the vertical and in some parts inclined inwards. When the water had been lowered to within four feet of the bottom, the remaining water was seen to contain a large number of “capitanes.” ‘They were greatly excited and distressed and were swimming and creeping about in all directions. A small stream of water in a thin film ran down one side of the pot-hole from a leak in the dam above. Several fish, after nosing around the edge of the water, discovered this small inflowing stream and started to creep up in it, but becoming frightened by the movements of the working men near, dropped back. When work was stopped for the noon hour, four of the smaller fish started up, following the thin stream of water. The water ran over their noses, down their backs and trickled off their tails in small streams. They would hitch themselves up rapidly for the distance of a foot or so and remain quiet for a minute or two; then another foot and another rest. In half an hour, the four had reached the water in the pool at the foot of the dam above. In making the ascent, they were obliged to pass a part of the wall, about two feet in length, that inclined inward at an angle of about 30° from the vertical. When they reached this overhanging part, in no observable manner did they change their tactics, but they ascended it as rapidly and safely, and apparently with no more effort than the other portion of the wall. During the afternoon, several more of the fish climbed out. A large number were in the water at the bottom of the hole when work was suspended for the evening. In the morning not a fish remained. For the greater part, the path followed by the fish in making their ascent lay over smooth, water-worn surfaces free from any coating of vegetable matter. The upper part, however, was covered by a thin film of an alga-like growth that may have served for the engagement of the ‘sharp-pointed teeth on the movable ventral fins. The total vertical dis- tance through which the fish climbed measured eighteen feet. When ‘undisturbed, they covered the distance without a slip or fall. The water, diverted around this pot-hole, flowed through a large pipe and fell from the end upon the steeply inclined water-worn rock at the side of the JOHNSON, HABITS OF A CLIMBING CATFISH 333 channel below. A day or two after the water had thus been diverted, a dozen or more of these fish were observed to be clinging to the rock at the foot of the fall at the end of the pipe. They were evidently on their way up stream, but had encountered an artificial condition that inter- rupted their further progress. They were nosing about in search of a small stream or film of water sufficient to keep their gills wet and to lead them to the main body of water above. As there was no such stream, their further progress was prevented. ‘They made no observed attempt to swim up the fall, but confined their efforts to making short excursions up the rock above the water. Failing to find any leading stream, they crept back. They deposit their eggs in the deepest pot-holes and attach them indi- vidually to the under sides of large rocks. BLICATIONS or THE Ww YORK ACADEMY OF SCIENCES (Lyceum OF N ATURAL HIsTory, 1817-1876) are ublications of the Academy consist of two series, viz.: n tings and similar matter. % ume of the Annals coincides in general with the calendar year . old at the uniform price of three dollars per volume. The articles “buted | in bundles on an average of three per year. The Bee of rate articles depends upon their length and:the number of illus- ns, and may be learned upon application to the Librarian of the The author receives his separates as soon as his paper has > i hed date of issue ae above the title of each pee ‘a are sent free to Fellows and Active Members. The THs LIBRARIAN, New York Academy of Sciences, care of American Museum of Natural History, New York, N.Y. THE NEW YORK ACADEMY OF SCIENCES |, (Lycrum or NaturaL History, 1817-1876) OFFICERS, 1912 President—EMeErson McMitui1n, 40 Wall Street y Vice-Presidents—J. EDMUND WoopMAN, F'REDERIC A. Lucas CHARLES LANE Poor, R. 8. WoopworrH Corresponding Secretary—Hernry W. Crampton, American Museum Recording Secretary—Epmunp Ot1s Hovey, American Museum” Treasurer—HeEnry L. DoHERTY, 60 Wall Street | Librarian—Ratreu W. Tower, American Museum Editor—Epmunp Otis Hovey, American Museum ~ SECTION OF GEOLOGY AND MINERALOGY Chairman—J . E. Woopman, N. Y. University Secretary—Cuartes P. Berkey Columbia University SECTION OF BIOLOGY Chairman—FReEDERIC A. Lucas, American Museum Secretary—Witu1am K. Gregory, American Museum SECTION OF ASTRONOMY, PHYSICS AND CHEMISTRY Chairman—Cuar.Es LANE Poor, Columbia University Secretary—F. M. Peprrsen, College of the City of New York SECTION OF ANTHROPOLOGY AND PSYCHOLOGY Chairman—R. S. WoopwortH, Columbia University Secretary—FREDERIC LYMAN WELLS, Columbia University The sessions of the Academy are held on Monday evenings at 8:15 o’clock from October to May, inclusive, at the American Museum of Natural History, 77th Street and Central Park, West. — [ANNALS N. Y. Acap. Sct., Vol. XXII, pp. 335-337, Pll. XXXV-XXXIX. 27 December, 1912] THE KINGSTON, N. M., SIDERITE! By Epmunp Otis Hovey (Read before the Academy 4 November, 1912) In the year 1891, a prospector was ransacking the region along the North Branch of Percha Creek, near the Solitary Mine, about four miles north of Kingston, Sierra Co., New Mexico, in a search for horn silver (cerargyrite), when he stumbled upon what he supposed to be a solid mass of the ore for which he was hunting. ‘This was lying upon a ledge of granite and, according to the account of the finder, had been brought to view by erosion. The approximate position of the locality is latitude 32° 58’ North, longitude 107° 50’ West. A small fragment was broken off and sent to an assayer, who informed the finder as to the real nature of his specimen. Another small piece was sawed from the other end of the mass. These two pieces would weigh together about 5 ozs. (142 gm.), while the mass as delivered to the Foote Mineral Co. weighed 28 lbs., 6 ozs. (12,870 gm.) ; hence the total original weight was about 28 lbs., 11 ozs., or 18,012 gm. Mr. Warren M. Foote of Philadelphia very kindly sent the specimen to me for description and gave me the foregoing information regarding its discovery and history. The iron, which is a holosiderite, has been named Kingston from the post office nearest to the place of discovery. As found, it was lenticular in shape, and its dimensions were 204x167x70 mm. No sign of the original crust remains, and the “thumb marks’ have been much ob- scured by oxidation. The exterior, however, presents several broad, shal- low depressions, which are clearly shown in Plates XXXV and XXXVI, but nothing to indicate which was the briistseite. Crevices along which oxidation has taken place to a considerable degree penetrate the mass along the plates and are particularly noticeable in Plate XXXV, fig. 1 and Plate XXXVI, fig. 3. A combination of these cleavage crevices per- mitted the breaking out of the fragment that was first submitted to assay. The surface is thickly indented with oxidation pits 1-3 mm. across and shallow in proportion to their diameter. The material furnished me for investigation consisted first of the whole mass and then of the two end pieces and seven slices into which 1 Published by permission of the Director, American Museum of Natural History. (335) 336 ANNALS NEW YORK ACADEMY OF SCIENCES the mass was first cut by the Foote Co., two fragments for specific grav- ity determination and three for chemical study. The last, aggregating about 50 gm. in weight, were sent to Booth, Garrett & Blair of Phila- delphia for analysis, with the following result: Per cent He Cby: difference)! ihc sekck-t si sernes pains Bn cee 92.376 ING oie ertbene osc a ea EGET OC a an 6.980 LO Mth ea eam n teal ote ete aad val ey aise eins mc. DED 0.505 Os) Rete p ere ner Pies Sere TCA ACIRE ny RUN te oo 0.018 SSO ER ney Cater Orie pnts kits aan iaitn eo neuen aI es can Ree 0.014 Peieitas se ioeiey satis ta hava icaenctirey eee See eee Us OE Uae SC 0.099 Siig iat Sct Sule curs Bees CUD Stes rh elostea Sata UD te Ce 0.008 The analysts report the apparent presence of a trace of some other element, probably of the iron-platinum group of metals, but state that they were unable to isolate and determine it in the amount of material available and that it must be present in variable quantities. The specific gravity as determined at the American Museum on a fragment weighing about 12 gm. is 7.63. This is a low value, even when the small percentage of Ni + Co present is taken into consideration. End piece No. 1 contains the angular hole left by breaking out the fragment for assay. As submitted to me, this piece was about 40 mm. thick and weighed 1174 gm. The Widmanstiitten lines (Pl. XX XVII) were traceable over the whole polished and etched surface, but they were obscured by a flecky granulation extending likewise over the whole surface. The flecks, whose appearance reminds one of the particles making up a flocculent chemical precipitate, are irregular in outline and are from 0.4 to 0.5 mm. across, seldom reaching the latter dimensions. They have no particular orientation and are so strongly developed in places as to suggest an approach to ataxitic structure. ‘This portion of the mass closely resembles Tazewell in appearance. 'The etched surface of slice No. 1 was about 13 mm. distant from that of the end piece. It showed the flocculent granulation over about three fourths of its surface (Pl. XXXVIII), the remainder being occupied by a subcentral oval area about 90 x 40 mm. in size in which the kamacite was practically free from granulation. In the succeeding slices, the oval clear area ex- panded until in slice No. 6 there was no granulated area (Pl. XXXIX). End piece No. 2, however, showed a bright fleck here and there. The granulated portion of the mass, therefore, originally formed a cap over a roughly cone-shaped development of iron that was practically free from the flecks. | Turning now to the other features of the meteorite we may say that the Widmanstiatten lines are well developed, forming kamacite bands HOVEY, THE KINGSTON, N. M., SIDERITE 3317 from 0.5 to 1.5 mm. broad, but the usual width is 0.75 mm., where they are not doubled or trebled. Some were measured that were practically continuous for a length of from 75 to 85 mm. Neumann lines are abun- dant and distinct, though occasionally obscured by a minute network of curved lines. Teenite is practically absent and plessite is extremely subordinate in development. ‘Thin, short lines of schreibersite may be seen here and there, some of which are associated with little nodules of troilite and some with the bands of kamacite. About fifty small nodules of troilite were noted, varying in size from 1 to 8 mm. in diameter. Lawrencite exuded rather freely from the crevices in the slices during the dampness of summer. The iron is octahedral in structure, and the breadth of the lamelle throw it into the medium octahedrites (OM) of Brezina’s classification. A circle whose radius is 70 miles would pass through or close to the places of origin of the following New Mexico siderites: Kingston, Luis Lopez (Magdalena), Oscuro Mts., El Capitan and Sacramento Mts. The irons, however, seem to be independent falls. ha? sete oe eo (pe Fay hia 2 ape 7 7 At: S em ae eh Se feata ston meteorite. k view of the original mass. 1% natural size Ga lee aes : Sige, A tS <7 on ta ar as oe ace 7 iia. #ebot leaigito edt te wale abed bas iaovd 94 ANNALS N. Y. ACAD. SCIENCES JOLUME NNIT, Puatp XXNXV ee rant ¢ =O ft \ & l[eustea i OS IAXXX Givid ‘IIXX GNAIOA SMONGIOS “AVOV ‘X ‘N STIVNNY i PVAKX coe “watts: it ov ee, its 30 sot ttt 1 Pat Podatnt % sa < i ti} = i id = \ x ANNALS N. Y. ACAD. SCIENCES VOLUME XNII, PLare XXXVII ae x ns etched surface showing 2 which the granu- fata) {% 7 x Ae ae : ee ‘ S oF ca X Sali us cane Resco: t ’ 1 s ¥ salsye Baottods - Pets. ‘oat ANNALS N. Y. ACAD. SCIENCES VOLUME XXII, PLhaty XXXVIII shed and etched surface from which — tural size. ANNALS N. Y. ACAD. SCIENCES VOLUME XNII, PLaty XXNIN PUBLICATIONS Spe 7 OF THE _NEW YORK ACADEMY OF SCIENCES (Lyceum or Natura History, 1817-1876) : 1) The Annals (octavo series), established in 1823, contain the ientific contributions and reports of researches, together with the rec- ords of meetings and similar matter. A volume of the Annals coincides in general with the calendar. year and is sold at the uniform price of three dollars per volume. The articles composing the volume are printed separately, each in its own cover, and are distributed in bundles on an average of three per year. The price of the separate articles depends upon their length and the number of illus- trations, and may be learned upon application to the Librarian of the Academy. The author receives his separates as soon as his paper has been printed, the date of issue appearing above the title of each paper. Ww . (2) The Memoirs (quarto series), established in 1895, are issued at ‘irregular’ intervals. It is intended that each volume shall be devoted to “monographs relating to some particular department of Science. Volume is devoted to Astronomical Memoirs, Volume II to Zodlogical Memoirs, ate.” The price is one dollar per part as issued. All publications are sent free to Fellows and Active- Members. The ae are sent to Honorary and Corresponding Members desiring them. fA Subscriptions and inquiries concerning current and back numbers of any of the publications of the Academy should be addressed to THE LIBRARIAN, New York Academy of Sciences, care of ‘ American Museum of Natural History, New York, N. Y. Editor, Epaunp Or1s Hovey \ fogre) H RTER, ‘CONSTITUTION AND. MEMBER- — SHIP AN 1912 OF THE a 2 nw YORK” | PUBLISHED BY THE ACADEMY 20 a, 1913 THE NEW YORK ACADEMY OF SCIENCES (Lyceum or Naturat History, 1817-1876) Ns OFFICERS, 1912 President—Emerrson McMiuur, 40 Wall Street Vice-Presidents—J. EDMUND WoopM4N, FREpDERIC A. Lucas” _ CHARLES LANE Poor, R. S. WoopwortH Corresponding Secretary—HENry H. Crampron, American Museum Recording Secretary—EpmunpD Otis Hovey, American Museum Treasurer—HEnNRY L. DoHERTY, 60 Wall Street Inbrarian—RawtpPuH W. Tower, American Museum Editor—EHKpMUND OTIS Hovey, American Museum SECTION OF GEOLOGY AND MINERALOGY Channa . HE. WoopM4n, N.Y. University . Secretary—CuarLEs P. Berkey Columbia University ~SHCTION OF BIOLOGY Chairman—FREpERIc A. Lucas, American Museum Secretary—WiLL1AM K. Grecory, American Museum SECTION OF ASTRONOMY, PHYSICS AND CHEMISTRY Chairman—CHar.es LANE Poor, Columbia University Secretary—F. M. PEDERSEN, College of the City of New York — SECTION OF ANTHROPOLOGY AND PSYCHOLOGY Chairman—R. 8S. WoopwortH, Columbia University Secretary—F REDERIC LiyMAN WELLS, Columbia University The sessions of the Academy are held on Monday evenings at 8:15. _o’elock from October to May, inclusive, at the American poe we of Natural History, 7 7th Street and Central Park, West. [ANNALS NEw YorK ACADEMY OF SCIENCES, Vol. XXII, pp. 339-414. 20 April, 1913.] RECORDS OF MEETINGS OF THE NEW YORK ACADEMY OF SCIENCES. January to December, 1912. By Epmunp Oris Hovey, Recording Secretary. LECTURE. 3 JANUARY, 1912. Edward E. Barnard : THE PLANET Mars. BUSINESS MEETING, 8 JANUARY, 1912. The Academy met at 8:23 Pp. M. at the American Museum of Natural History, Vice-President Woodman presiding. ° The minutes of the last business meeting were read and approved. The following candidates for membership in the Academy, recom- mended by Council, were duly elected: ActTIvE MEMBERSHIP. A. B. Pacini, Board of Water Supply, New York. Henry Arctowski, New York Public Library. ASSOCIATE MEMBERSHIP. Charles R. Fettke, Livingston Hall, Columbia University. The Recording Secretary announced that a special fund had beer provided so that during 1912 each of the four sections of the Academy might have $100 at the disposal of the sectional officers for expenses or honoraria connected with an effort to make the meetings more interesting (339) 340 ANNALS NEW YORK ACADEMY OF SCIENCES to the general public and extend the eee of the Academy. A vote of thanks was extended to the unnamed donor. The Academy then adjourned. EpmMuNpD Otis Hovey, Recording Secretary. SECTION OF GEOLOGY AND MINERALOGY. 8 JANUARY, 1912. Section met at 8:28 Pp. m., Vice-President J. E. Woodman presiding. Thirty-three members and visitors were present. The minutes of the last meeting of the Section were read and approved. In the absence of Dr. Charles P. Berkey, Secretary of the Section, Dr. E. O. Hovey was elected Secretary pro tem. The Secretary pro tem presented an application in the name of Mr. George Borup for a grant of $500 from the Esther Herrman Research Fund, as a contribution to the Crocker Land Expedition which he and Mr. Donald B. MacMillan were organizing under the auspices of the American Museum of Natural History and the American Geographical Society. On motion, the application was referred to the Committee on Grants from Research Funds with power. The programme for the evening was then taken up as follows: A. B. Pacini, THz MeTAMoRPHISM OF PoRTLAND CEMENT. I. Remarks were made by Professors Kemp, Woodman and Grabau. The paper has been published as pages 161-224 of this volume. Dr. EK. O. Hovey gave a brief summary account of the Washington meeting of the Geological Society of America and a few of the paper presented there. Professor A. W. Grabau gave a similar account of the Washington meeting of the Paleontological Society. The Section then adjourned. EpmMuND OTIs Hovey, Secretary pro tem. SECTION OF BIOLOGY. 15 January, 1912. Section met at 8:15 Pp. m., Vice-President Frederic A. Lucas pre- siding. RECORDS OF MEETINGS 341 The minutes of the last meeting of the Section were read and approved. ; The following programme was then offered: Henry Fairfield Osborn, PHYLOGENY AND ONTOGENY OF THE HoRNS oF MAMMALS. Henry Fairfield Osborn, SkuLL MrasurmpMENTS IN MAN AND TH Hoorep MAamMMALs. Frederic A. Lucas, WHALING IN THE OLDEN TIME. SUMMARY OF PAPERS. Professor Osborn said in abstract: The recent discovery of the modes of origin of the horns in the titanotheres, a perissodactyl group remotely related to horses, tapirs and rhinoceroses, permits of a comparison of phylogenesis with the ontogenesis of the horns in bovine mammals. The latter is based upon an osteological series recently prepared by Mr. S. H. Chubb, the former is based on the rich phylogenic series of Hocene titanotheres in the American Museum of Natural History. The conclu- sion is that ontogeny closely recapitulates phylogeny, that the genesis is gradual or continuous, that the horns arise definitely and deter- minately. In the bovine series it seems, in accord with the conclusions of Dist, that the horn first appears as a circular thickening of the skin, accompanied by accelerated growth of the hair preparatory to the forma- tion of the keratin of the horny substance, at a period considerably prior to any sign of the horn in the bony structure of the frontals. This raises the problem, which will form the subject of a special paper in the Annals of the Academy, as to what element first arises in connection with horn evolution, namely: (1) the psychic, or desire to use the horn; (2) the epidermal callous or keratin protection of the bony horn center, or (3) the bony or osseous horn itself. It would appear that the psychic tendency must precede the epidermal and that the latter precedes the osseous, but this disputed point requires further investigation. The paper was illustrated with lantern slides, drawings and specimens. Professor Osborn, in his second paper, said in abstract: Comparative anatomists and zodlogists have been slow to introduce into mammalogy systems of measurement by indices and ratios, which have proved of such universal value in anthropology. It is found among the hoofed mammals, from studies undertaken by the author with the co-operation of Dr. W. K. Gregory, that cephalic indices and limb ratios between different segments of the skeleton are far more significant than systems of direct measurement. These cephalic indices of the gradual changes of proportion between different regions of the skull have the value of 849 ANNALS NEW YORK ACADEMY OF SCIENCES specific characters and sharply distinguish members of different phyla. For example, in the cross between the horse (f. caballus) and the ass (EZ. asinus), it is found that the cephalic indices are transmitted as pure non-blending characters. Among the most significant indices are the following: a) ¢ the cephalic, which is obtained by dividing the total or basilar length of the skull by the zygomatic breadth; (2) the cranial, which is obtained by divid- ing the basilar length by the postorbital length of the skull; (3) the facial, obtained by dividing the basilar length by the preorbital length of the skull, ete. The horses show proopic dolichocephaly, or elongation of the face, and a static condition of the cranium, while the titanotheres, in contrast, show opisthopic dolichocephaly, or elongation of the cranium, and abbreviation of the face. Like the phyletic differences of proportion between the horse and the ass, these differences are most exactly ex- pressed by the method of indices. The application of the ratio method to the limbs of the hoofed mam- mals has again produced most surprising results. It is found that mam- mals of different phyla adapted either to “weight” or to “speed” con- verge respectively toward typical “weight” or “speed” ratios, which are obtained by dividing the length of the lower segments, tibia and radius, respectively, by the upper segments, femur and humerus, metacarpus and metatarsus, respectively. These “weight ratios” and “speed ratios” are far more significant as regards function and phyletic change than the actual or direct measurements. Dr. Lucas exhibited lantern slides illustrating some interesting pic- tures from old works on whaling and showing the methods practiced by the early Japanese, European and American whalers. The Section then adjourned. WILLIAM K. GREGORY, Secretary. SECTION OF ASTRONOMY, PHYSICS AND CHEMISTRY.. 22 JANUARY, 1912. Section met at 8:15 Pp. m., Vice-President Campbell presiding. The minutes of the last meeting were read and approved. Dr. F. M. Pedersen of the College of the City of New York was then elected Secretary of the Section for the year 1912. The following programme was offered : C. C. Trowbridge, Recent DiscoveriEs CONCERNING A CHEMICALLY Active Moprrication or NITROGEN. William Campbell, Some Notrs on Iron AND STEEL. The Section then adjourned. F. M. PEDERSEN, Secretary. RECORDS OF MEETINGS 343 SECTION OF ANTHROPOLOGY AND PSYCHOLOGY. 29 JANUARY; 1912. Section met in conjunction with the American Ethnological Society at 8:15 p. M., Gen. James Grant Wilson presiding. The minutes of the last meeting were read and approved. The following programme was then offered: Pliny E. Goddard, Notrs oN THE JICARILLA APACHE. SUMMARY OF PAPER. Dr. Goddard in his paper said: The Jicarilla Apache are, from the point of view of material culture, a buffalo-hunting Plains people dwell- ing in skin-covered tipis. Their social organization differs from that of the Navaho and neighboring Pueblo tribes in lacking exogamous clans, there being two geographical divisions with ceremonial and political, but not marriage-regulating, functions. Among the ceremonies the speaker mentioned an annual feast celebrated on the 15th of September and probably connected with the corresponding celebration at Taos, the conspicuous feature of both consisting in a relay race. A ceremony resembling the Bear Dance of the Southern Ute is performed in cases of illness and is characterized, among other things, by .sleight-of-hand performances of masked dancers. The girls’ puberty celebration is very. prominent; a distinctive feature of the Jicarilla form of this ceremony seems to be the association of a young man with the adolescent girl. Among the myths of the Jicarilla that of the twin heroes is prominent. In the course of the discussion Dr. Goddard stated that he had been: unable to discover myths definitely connecting the mythology of the Jicarilla with that of their linguistic congeners in California and the Far North. In reply to another query he expressed his belief that, owing to the linguistic differentiation of the Apache, this tribe must have occupied its southwestern habitat a considerable period before the first historical notice of it. ( The Section then adjourned. F. Lyman WELLs, Secretary. BUSINESS MEETING. 5 FEpruary, 1912. The Academy met at 8:21 Pp. M. at the American Museum of Natural History, Vice-President Woodman presiding. The minutes of the last business meeting were read and approved. 344. ANNALS NEW YORK ACADEMY OF SCIENCES The following candidates for active membership in the Academy, rec- ommended by Council, were duly elected: Charles HE. Sleight, Ramsay, New Jersey. R. B. Earle, New York University. On motion, the following minute was unanimously adopted and or- dered to be engrossed and transmitted to the family of the late Mr. Charles F. Cox: The Academy suffers irreparable loss through the death, on 24 January, 1912, of Mr. Charles Finney Cox. For thirty-six years an Active Member and Fellow of the organization, his influence has been felt ffom the first in all progressive movements of the Academy. He served the Academy diligently as Curator, 1884, 1885; Councilor, 1891, 1892; Treasurer, 1893-1907; President, 1908, 1909. At the time of his death he was again acting as Treasurer. When President he was active in the organization of the Academy’s Darwin Cen- tennial celebration, and he delivered a masterly address on Darwin at the close of each of his two years of incumbency. Always the friend of investigation, he was one of the founders of the Scien- tific Alliance of New York, the first association of the scattered organizations that were striving independently to advance the interests of science in the city. Some five years ago he was again active in establishing the closer affiliation which now obtains among them. Mr. Cox’s consuming interest outside of his daily duties in the railways of the New York Central system was the study of the life and writings of Charles Darwin. In its pursuit, he became a keen and devoted collector of Darwiniana, and the portraits, first editions, manuscripts and other priceless memorials which he brought together constitute a remarkably complete exhibit of Dar- win’s scientific work and influence upon the thought of the last fifty years. Another of his avocations was microscopy, in which he was active for many years, while his interest in botany was evidenced by his participation in the founding of the New York Botanical Garden and in its management up to the time of his decease. ; In character, Mr. Cox was a man of great simplicity and natural refinement. He attracted. and held his friends with bonds of attachment that were altogether exceptional in their strength. While he will be missed and mourned by all who knew him, the sense of loss is peculiarly deep in the circle of the New York Academy of Sciences. To his family the Academy extends its profound sympathy. The Academy then adjourned. EpMunpD Otis Hovey, Recording Secretary. SECTION OF GEOLOGY AND MINERALOGY. 5 FEBRUARY, 1912. Section met at 8:31 Pp. M., Vice-President Woodman presiding. RECORDS OF MEETINGS | 345. In the absence of the Secretary, Dr. Hovey was elected Secretary pro tem. The minutes of the last meeting of the Section were read and approved. The following programme was then offered: E. O. Hovey, THE Amaia Farm METEORITE. Vernon F. Marsters, A SKETCH OF THE PHYSIOGRAPHY AND EARLY Mining DEVELOPMENTS OF PERU. Henry 8. Washington, RELATIONS OF THE FELDSPARS, LENADS AND VEOLITES. : George H. Girty, On Some FossILs OF THE LyKINS FORMATION. (Read by Title.) SUMMARY OF PAPERS. Dr. Hovey exhibited a polished and etched slice of the iron meteorite from the Amalia Farm near Gibeon, Africa, and called attention to the interesting curvature of the Widmanstatten lines in certain portions of the slice, apparently due to the softening of the neighboring surface of the mass as it passed through the air; also a line of discordance be- tween the lamella apparently due to welding by impact of two masses or two fragments of the same mass before the meteorite reached the earth. Mr. Marsters in his paper described the coastal plains and cordilleras of Peru and gave sections at several points from the sea to the summit of the eastern range of the Cordilleras, the petroleum deposits along the coast and the great deposits of coal, Lake Titicaca and the mines vf gold, silver, copper and vanadium along the contacts of the eruptive rocks with the sandstones and the shales of the middle and eastern Cordilleras. Dr. Washington in his paper gave an ingenious regrouping of the molecules in the standard analyses of the feldspars and related minerals, bringing out the isomorphism of the groups more clearly than is done by other methods of writing the formulas, provided one can admit that silicon acts as a base as well as an acid. The Section then adjourned. EpmMuND Otis Hovey, Secretary pro tem. 346 ANNALS NEW YORK ACADEMY OF SCIENCES SECTION OF BIOLOGY. 12 Fesruary, 1912. ‘Section met at 8:15 Pp. M., Professor Bashford Dean presiding. The minutes of the last meeting of the Section were read and approved. The following programme was then offered : John D. Haseman, Some Factors OF GEOGRAPHICAL DISTRIBUTION © IN SoutTH AMERICA. This paper has been published as pages 9-112 of this volume. The paper, which was illustrated with maps and diagrams, was dis- eussed by Professor Dean, Dr. W. D. Matthew and others. The Section then adjourned. WILLIAM K. GREGORY, Secretary. ‘SECTION OF ANTHROPOLOGY AND PSYCHOLOGY. 26 FEBRUARY, 1912. Section met at 8:15 P. M. in conjunction with the New York Branch ‘of the American Psychological Association, R. S. Woodworth serving as ‘chairman. The minutes of the last meeting of the Section were read and ‘approved. The following programme was then offered: 3. E. Hickman, THE INFLUENCE oF NARCOTICS ON PHYSICAL AND MENTAL TRAITS OF OFFSPRING. A. KE. Chrislip, AUDITORY AND VISUAL MEmory. Henry H. Goddard, THe Herrepity or Mentat Tratrts. SUMMARY OF PAPERS. Dr. Hickman said in abstract: The purpose of the study was to learn if the use’ of narcostimulants (tea, coffee, tobacco and alcohol) had any ‘effect on the offspring. The research extended over a period of four -years. It included 306 families with 2,560 children; 620 of this num- — ‘ber were students of Murdoch Academy, Utah. These were carefully ‘measured by medical experts and teachers to get their physical and imental status. The measurements and examinations included height, RECORDS OF MEETINGS B47 weight, eyes, ears, nose, throat, teeth, heart, lungs, stomach, spleen, liver, kidneys and nervous condition. A record of the death-rate in the families was obtained as well as a record of the student’s intellectual standing. The students were divided into eight claasses, according to the kinds and quality of stimulants used by the parents. The examination showed: first, that there was on an average a very decided difference between the offspring of abstainers and those of users, even where tea or coffee was used by only one parent, for the offspring of the abstainers were superior in size, intellect and bodily condition to those of the caffein parents; secondly, as the use of caffein was increased by the parents, from once to three and four times a day, a gradual de- erease in height, weight, bodily condition, etc., of the offspring was manifest ; thirdly, in families where not only tea and coffee were used, but also tobacco, the children were still more inferior mentally and physically, increasingly so with the increase of caffein drinks in connec- tion with tobacco; fourthly, where alcohol was used with the above narcostimulants the lowering of the physical and mental status was very marked.” : Comparing all the offspring of the narcostimulant parents with those from abstaining parents, the latter were found to be better in all the 22 measurements than the former. Some of the differences were very great, especially in weight, height, eyes, ears, physical health and rate of mortality. There are over 100 per cent more eye, ear and physical de- fects in the offspring of narco-parents; 72 per cent more children died in this than in the abstaining class; 79 per cent of the narcostimulant families had lost one or more children, while only 49 per cent of the abstaining class had lost any children. It was also shown that the death-rate of the parents in this latter class was 41 per cent higher than in the former. The research also brought out the fact that it took the ofispring of the narcostimulant parent eight tenths of a year longer to graduate from the grades. In the Academy they were on an average a year and seven months older than the students from the abstaining class. Mr. Chrislip said in abstract: Experiments have been carried on in the psychological laboratory of Columbia University and elsewhere for the purpose of comparing visual and auditory memory. The points in- vestigated in the first experiment were to determine: the number of repetitions required by each sense to reproduce in a certain order cer- tain total series of like construction; the average number of characters of a series recalled in their proper order for each repetition of series of like construction for each sense; and to determine, if possible, the best material for testing the two senses. 348 ANNALS NEW YORK ACADEMY OF SCIENCES The material used consisted of numerals, nonsense syllables and words. Series composed of 12 and 16 characters of each material were used in testing both senses. The result shows that when series of 12 numerals similarly con- structed were presented to the two senses, that out of 26 cases 20 are visual, 8 auditory and 8 show no difference. In the case of the series of 16 numerals, 19 visual, 4 auditory and 13 show no difference. With 12 nonsense syllables there are 15 visual and 15 auditory, the rest show- ing no difference, but for 16 nonsense syllables, 25 visual, 7 auditory and 4 show no difference. With the 12 words there are 14 visual, 10 auditory and 12 no difference; with 16 numerals, 22 visual, 9 auditory and 5 show no difference. For each repetition of each series the result shows that in the mem- ory tests for visual reproduction the greater average number is repro- duced. The nonsense syllables were the best material, as they offered few combinations or devices for memorizing them. Experiments, in which stories of 100 words each have been used to test the two senses, have been carried on for some time. The two senses have been tested for both immediate and delayed recall. In both the immediate and the delayed reproductions the visual has been better than the auditory. There is an experiment now in operation in which the method is somewhat different from that in the former experiments con- ducted with logical material. While the results are not all deter- mined the indications are that the auditory may surpass the visual. Dr. Goddard said in abstract: It is not the purpose at the present time to present any results, but rather to make some suggestions and point out possible lines of research in the hereditary transmission of mental traits which may be of interest to psychologists. In connection with our studies of the cause of mental deficiency at the training school at Vineland, much material has been accumulated show- ing the hereditary transmission of deficiency. In connection with these data many facts have come to hand which make it clear that not only deficiency, but many positive traits are directly transmitted. It is fur- ther suggested that psychology would gain valuable data and contribu- tions to many of its problems from a study of this question of heredity. Indeed, it seems quite possible that many problems which are now so complex as to elude our powers of analogy would be easily analyzed if we were able to study the heredity problem and thus eliminate the hereditary factor. For example, if the goodness of memory depends, as Professor James said, upon the natural retentiveness of the brain tissue plus the logical association that the individual establishes, then we may RECORDS OF MEETINGS 049 reasonably expect that the condition of the brain tissue may be a quality that is transmitted and could be eliminated through the study of mode of transmission ; or, in other words, we could determine to what extent the differences in memory are due to acquired factors. It would seem equally possible that sensory conditions may be traced through families, just as peculiar eyes or eyesight, peculiar hearing, kinesthetic sensations, taste, or smell may be dependent upon organic conditions which may be found to be directly transmitted. The inborn habits or instincts are so bound up with acquired habits that it makes a very complex problem. It seems quite possible that a study of the in- stinctive activities of members of different generations might reveal to us a good deal about the nature of instinct and its transmission which would have very important bearings upon many of our problems of in- stinct and emotions. Even the study of such a. complex problem as the inheritance of mental deficiency may possibly yield us some most im- portant results. . It seems hardly likely that mental deficiency is due to the absence of any one characteristic, but of several, and that it may be pictured more as though normal mentality is the result of a hundred factors of which a person must have, say, seventy-five in order to have what is called nor- mal mentality. Now the twenty-five that are lacking may be any twenty-five, perhaps, in the whole list and a tracing of the hereditary traits might lead us eventually to determine some things about the re- sulting mentality when the missing factors belong to different groups. We shall work on these problems at Vineland as rapidly as possible, but they should be studied in normal people as well. It is perhaps true that it would not be possible to go back farther than the living genera- tions ; but even so, if careful studies and tests were made of the mental traits in living persons, it would be possible to get the records of two and sometimes three generations, and these records could then be kept and supplemented as the years go by and the newer generations come on. There would thus be laid the basis for most valuable studies later on. The family histories, that we have secured in connection with our children at Vineland, suggest two or three interesting questions. For instance, there are several families in which alcoholism is strong in sev- eral generations. It is possible that we have in these families an un- usual appetite for alcohol, which appetite has been transmitted. It looks as though it would not be impossible to eliminate to quite an ex- tent the environmental factor, and so be able to determine whether this was hereditary or not. The same is true of the sexual life. A great a ae oar 350 ANNALS NEW YORK ACADEMY OF SCIENCES many charts show very much sexual immorality: and possibly here we may have, in some cases at least, an unusual development of the sex in- stinct which has broken over all bounds of conventionality and has shown in different generations. It appears that all of these problems are not only worthy of study, but might yield most important results. The speaker showed graphic charts illustrating the family histories of a number of families. ‘These charts showed the strong inheritance of feeble-mindedness and also illustrated the points made in regard to alcohol and sexuality. Considerable discussion followed. The Section then adjourned. f F. Lyman WELLS, Secretary. BUSINESS MEETING. 4 Marcu, 1912. The Academy met at 8:16 p. m. at the American Museum of Natural History, Vice-President Woodman presiding. The minutes of the last business meeting were read and approved. The Recording Secretary reported that Mr. Henry L. Doherty had been elected Treasurer to fill the pnexpired term of Mr. Charles F. Cox, - deceased. The Academy then adjourned. EpmunpD Otis Hovey, Recording Secretary. SECTION OF GEOLOGY AND MINERALOGY. 4 Mancn, 1912. | The Section met at 8:22 Pp. M., Vice-President Woodman presiding. Fifty members and visitors were present. The minutes of the last meeting a the Section were read, corrected and approved. The following programme was then offered: J. E. Woodman, ForELANDS oF THE Bras p’OR LAKkEs, CAPE BRE- TON IstAND, Nova Scorta. Charles P. Berkey, Is THERE FauLT CoNTROL oF THE Hupson RIVER CouURSE? V. F. Marsters, DiIsTRIBUTION OF PETROLEUM DeEPosITs IN PERU. RECORDS OF MEETINGS ©— 351 SUMMARY OF PAPERS. Professor Woodman illustrated his paper with many photographs showing exceptionally fine views of the topography of the region and the cliffs, bars and spits of the shores. 'T’wo peneplain surfaces are to be seen—the higher, characterized by crystalline rocks, and the lower by Carboniferous strata. The hooked spits and loops and bars of re- _ cent wave and current work are developed on a scale that is seldom equaled. Professor Berkey discussed the reasons for the usual belief that the straight course of the river is due to fault structure and illustrated this view by the use of Professor Hobbs’s map of Atlantic border hneaments. Tt was then shown that a detailed study of the structural geology of the lower Hudson region shows many strong faults crossing the river obliquely N. HE. to 8S. W. and some much less important running N. W. and S. E., but none of any real consequence running N. and SB. or par- allel to the Hudson River. It would be considered most éxtraordinary if a great river like the Hudson should be controlled by small, insignifi- cant faults and pay so little attention to the large fault zones. Remarks were made by Professor Woodman. . Mr. Marsters showed maps and drew sketches of structure and topog- raphy. ‘Two belts of oil-bearing strata were described, both Tertiary— a coastal region and an interior region. The Section then adjourned. CHARLES P. BERKEY, Secretary. SECTION OF BIOLOGY. 11 Maron, 1912. By invitation of Professor C.-E. A. Winslow and his colleagues, the Section met at 8:15 P. m. in the Department of Zodlogy of the College of the City of New York, Vice-President Lucas presiding. The minutes of the last meeting of the Section: were read and ap- proved. — The following programme was then offered : C.-E. A. Winslow and I. S. Kligler, THz Number anp Kinps oF Bac- TERIA IN Crry Dust. C. V. Chapin, THe ARIAL TRANSMISSION OF DISEASE, 852 ANNALS NEW YORK ACADEMY OF SCIENCES SUMMARY OF PAPERS. C.-E. A. Winslow and I. 8. Kligler in their paper presented the re- sults of the examination of about 170 samples of dust from streets, schools, houses and public buildings in New York. The total numbers of bacteria found varied from 150,000 per gram to 145,000,000, aver- aging from 3,000,000 to 5,000,000 from the indoor dusts and 49,000,000 from the street dust. Spores made up usually less than one-tenth of the total. The count obtained at body temperature was about one-half that at room temperature, averaging from 2,000,000 to 3,000,000 per gram in the indoor dusts and 22,000,000 in the street dusts. B. coli was usually present; in the street dust an average of 51,000 per gram was found and in two samples over 100,000, while none showed less than 100. The indoor dust, on the other hand, showed an average of between 1,000 and 2,000. Acid-forming streptococci, such as are char- acteristic of the mouth, were present to the extent of over 1,000 per gram in three-fourths of the street samples and one-half of the indoor samples. The average for the street samples was about 40,000 per gram; for the indoor samples about 20,000 per gram. The large pro- portion of these organisms, particularly in the indoor dusts, appears to be significant of buccal pollution. The paper, which was illustrated by charts and diagrams, was dis- cussed by Dr. Lucas. Dr. Chapin said in abstract: The diffusion of contagion through the room or out-of-doors only was considered, not droplet infection, which does not take place beyond a meter. Bacteriological evidence was not discussed, though the quantitative work of Winslow on sewer air and spray infection was referred to, a work which he is now extending to dust. Epidemiological study and experiment have been rapidly nar- rowing the list of alleged air-borne diseases. We now know that yellow fever and malaria are never air-borne; experiments have shown that bubonic plague and Mediterranean fever are not. There is no evidence that cholera and typhoid fever are ever air-borne and much that they are not. The spread of influenza out-of-doors does not take place, and perhaps not indoors. The alleged evidence that smallpox virus is air- borne around hospitals is very weak. Careful observation in hospitals has shown that typhus fever, cerebro-spinal meningitis and poliomye- letis do not pass from patient to patient in the same ward. ‘The same is true for uncomplicated scarlet fever and for diphtheria except by con- tact or close droplet infection. Probably measles and whooping cough, rubella, mumps, chickenpox and smallpox are not air-borne, even in the RECORDS OF MEETINGS 353 same room, but further observation may show that such infection may rarely take place. The paper was discussed by Professors Winslow and Bristol. The Section then adjourned, the members visiting and examining the laboratories and lecture rooms of the Department of Zoology, College of the City of New York. Witiiam K. Gregory, Secretary. LECTURE. 15 Marou, 1912. F. S. Archenhold: Astronomy, EpucaTION AND CULTURE. SECTION OF ASTRONOMY, PHYSICS AND CHEMISTRY. 18 Marcu, 1912. Section met at 8:15 Pp. M., Vice-President Poor presiding. The minutes of the last meeting of the Section were read and ap- proved. The following programme was then offered: Louis H. Friedburg, Propucrs or CHEMICAL ART. SUMMARY OF PAPER. Dr. Friedburg said in abstract: There are three lines along which synthetic chemistry is to-day advancing. First, the production of things found in nature; for example—wintergreen oil, vaniline and camphor. Second, ennobling one substance into another; for exam- ple—transformation of cellulose into artificial silk. Third, prepara- tion of substances which are similar to natural substances, but which are not found in nature; for example—celluloid and bakelite. There are some important discoveries which have been made by acci- dent, such as that of glass 2,600 years ago. The chemist must be alert — enough to recognize the value of such accidental discoveries. The speaker described in a very interesting and entertaining manner the manufacture of parchment paper, mercerized cotton, gun cotton, collodion and artificial silk. He showed on the screen many beautiful examples of the Lumiére colored photographs, and the glass caterpillar or spinneret which is used for making the artificial silk fibres. Reau- mur in 1784 first suggested the possibility of artificial silk. Celluloid bo4+ ANNALS NEW YORK ACADEMY OF SCIENCES -and «heckerboard screens for photo-color printing, and, lastly, the making of bakelite or artificial amber (so-called) were explained. Many fine specimens of all substances mentioned were shown. The Section then adjourned. ‘ F. M. PEDERSEN, Secretary. SECTION OF ANTHROPOLOGY AND PSYCHOLOGY. 25 Marouw, 1912. Section met in conjunction with the American Ethnological Society at 8:15 Pp. m., General James Grant Wilson presiding. The minutes of the last meeting of the Section were read and ap- proved. The following programme was then offered : Robert H. Lowie, Dr. RapbosavLJEVIcH’s CRITIQUE OF PROFESSOR — Boas. SUMMARY OF PAPER. Dr. Lowie stated that Dr. Radosavljevich had misrepresented Professor Boas on a number of important points. He had entirely misrepresented page 32 of Professor Boas’s preliminary report on “Changes in Bodily Form of Descendants of Immigrants.” The “economic” theory scoffed at by Dr. Radosavljevich is a figment of his imagination. What Professor Boas says is that the arrivals dur- ing the period following the financial panic of 1893 were under-devel- oped in every direction. When Dr. Radosavljevich reproaches Professor Boas for not studying the effect of American soil and financial panics. on the same individuals during a period of time representing the age of his subjects, he shows that he has not the faintest notion of what Boas‘is — discussing in connection with financial panics. Secondly, Radosavljevich’s contention that Boas’s own observations do not support his theory of a change of type is a most naive instance of conceptual realism. The conventional classificatory divisions of head forms have for Radosavljevich an absolute biological value, and unless the head forms of descendants of immigrants fall outside the conven- tional class of their parents he refuses to admit a change in head form. _ Dr. Lowie explained what statisticians and biometricians actually mean by a real difference between two series. In the discussion which followed, several visitors asked for informa- RECORDS OF MEETINGS 355 tion relating to the nature of Professor Boas’s methods, which were ex- plained by Dr. Goldenweiser, who had taken part in the investigation. The Section then adjourned. F. Lyman WELLS, Secretary. BUSINESS MEETING. 1 Aprin, 1912. The Academy met at 8:25 Pp. M. at the American Museum of Natural History, Vice-President Woodman presiding. The minutes of the last business meeting were read and approved. The following candidates for membership in the Academy: recom- mended by Council, were duly elected: ACTIVE MEMBERSHIP. Mrs. Henry W. Hardon, 315 West 71st Street. ASSOCIATE MEMBERSHIP. F. F. Hahn, Columbia University. The Recording Secretary then reported the following deaths: Edward Russ, an Active Member for 5 years. John B. Smith, an Active Member for 5 years. A letter was read from Dr. Hermann Credner thanking the Academy for the honor bestowed upon him by his election to Honorary Member- ship in the Academy; also an invitation from W. W. Gilchrist, Jr., artist, to the friends of the late Mr. Charles Cox, to view a portrait of the former Treasurer of the Academy at the Folsom Galleries. The Academy then adjourned. EpmunpD Otis Hovey, Recording Secretary. SECTION OF GEOLOGY AND MINERALOGY. 1 Apri, 1912. : Section met at 8:25 p. M., Vice-President Woodman presiding. The minutes of the last meeting of the Section were read and ap- proved. The Secretary announced that Mr. Alfred H. Brooks, of the United States Geological Survey, had been secured for the May meeting, at 356 ANNALS NEW YORK ACADEMY OF SCIENCES which time he would give a public lecture on the “Geology and Mineral Resources of Alaska.” Dr. Hovey announced that the.seismograph which had been presented to the Academy by Mr. Emerson McMillin had reached New York and had been passed by.the Custom House. The following programme was then offered: Wallace Goold Levison, IntusTRATIONS OF MINERAL ASSOCIATIONS BY MrAns oF CoLoR PLATE AND OTHER PHOTOGRAPHS OF OPAQUE SPECIMENS. A. B. Pacini, THE METAMORPHISM OF PORTLAND CEMENT. dale Charles T. Kirk, ALTERATIONS IN THE SNAKE River BASALTs. (Read by title.) SUMMARY OF PAPERS. Mr. Levison said in abstract: On a previous occasion the writer pre- sented to the Academy a “Note on Photographs of Minerals for illus- trating Books, Papers and Lectures” (Annals N. Y. Acad. Sciences, Vol. XII, pp. 661 and 663). The examples consisted of lantern slides and prints of light-colored mineral specimens of cabinet size mounted on standard size blocks with standard size labels and of microscopic mounts in Rakestraws. (See “Report of Committee on Standard Sizes,” N. Y. Acad. of Sciences, 1894.) The writer found it difficult to obtain satisfactory photographs of highly colored minerals in or on colored matrices, as such associations usually afford poor contrasts on ordinary plates. To produce representations of colored. minerals at that time hand- painted or colored lantern slides or prints were the chief expedient. Some time later the M. A. Seed Dry Plate Co. introduced its G. B. P. R. (green, brown, purple and red) plates, which served better than ordinary plates for lantern slides of certain colored minerals. Thus malachite and prehnite looked well on the green, native copper on the red and rusty or yellow-colored minerals such as the stilbite and cal- cite from Upper Mt. Clair, N. J., on the brown plates. The entire plate was usually of a tint much similar to that of the specimen, but different parts of the specimen usually developed in tints intermediate between the four possible colors above mentioned, so that pictures on these ‘plates made desirable lantern slides. This method was not applicable to paper prints. The recently introduced color plates of Lumiére, Jougla and Dufay RECORDS OF MEETINGS SOK now afford remarkably satisfactory lantern slides of colored minerals of either cabinet or microscopic size. The methods of production of all these plates were explained, but in the writer’s experience the Jougla and Dufay plates seemed to afford slides preferable in transparency and resistance to the heat of the lantern. Photomicrographs in color of thin sections of rocks and minerals by transmitted polarized light were made by Francois Frank on Lumiére plates as early as the year 1907 (Ch. A. Francois Frank, “La Micro- photographie en couleur avec les plaques autocrome de M. M. A. et L.. Lumiére,’ Comptes Rendus 1** Semestre, T. CXLIV, No. 24, p. 134L, 17 June, 1907). The first attempt to make photomicrographs, on autocolor plates, of microscopic colored minerals by ordinary reflected light was made by Mr. Frank La Manna, of the Borough of Brooklyn, N. Y., about Feb- ruary, 1911. Mr. La Manna thus photographed on Jougla and Dufay plates he brought from Paris several specimens of microscopic colored minerals mounted in Rakestraws by the writer. (F. La Manna, Ex- hibit at the Annual Reception of the Department of Microscopy of the Brooklyn Institute, 11 March, 1911.) The deep black interior of the Rakestraws served as a superior black background. Through the courtesy of Mr. La Manna, the writer received some of these plates upon which he photographed other similar specimens. These photomicrographs, jointly with those of Mr. La Manna, were ex- hibited before the New York Mineralogical Club in April, 1911, and again at the reception of the Brooklyn Institute’s Department of Micros- copy, March 9, 1912, and with additions in illustration of this paper. ~ (W. G. Levison, Exhibit at the Annual Reception of the Department of Microscopy of the Brooklyn Institute, March 9, 1912.) In making these photomicrographs the writer used a lens with a small stop giving a desirable depth of focus, a long bellows, a suitable color screen, long exposure, reflecting screens to soften the shadows and a very rigid adjustment of the apparatus and its supports. These autocolor plates have likewise afforded the-writer very satis- factory lantern slides of colored cabinet specimens. Each picture ob- tained is a direct positive. Such positives may afford approximately similar copies by the camera or other color plates, but duplicates made directly from the specimen are preferable. They may also, like any colored transparencies, be copied by contact on approximately similar colors, on a paper called T'to color paper by Dr. J. H. Smith, recently introduced from Paris. 358 ANNALS NEW YORK ACADEMY OF SCIENCES Mr. Pacini’s paper has been published as pages 161-224 of this volume. Remarks were made by Mr. Johnson and Mr. Price and Mr. Gaines of the Board of Water Supply testing laboratory. Several questions were asked by Professor Arnold of New York University. Remarks were also made by Professor Arnold. The Section then adjourned. CHARLES P. BERKEY, Secretary. SECTION OF BIOLOGY. 8 APRIL, 1912. Section met at 8:15 Pp. m., Vice-President Lucas presiding. The minutes of the last meeting of the Section were read and approved. . The following programme was then offered: Thomas H. Morgan, Sex-Linkep INHERITANCE IN POULTRY. Louis Hussakof, THE Spawnine Hasits oF THE SEA LAMPREY, Petromyzon marinus. i John T. Nichols, Notes oN CuBAN MARINE FISHES. SUMMARY OF PAPERS. Professor Morgan’s paper has been published as pages 113-133 of this volume. Dr. Hussakof said in abstract: The observations were made on the Nissequoque River at Smithtown, Long Island, June 1 and 2, 1911, while collecting material for an exhibition group of Petromyzon for the American Museum. The nests are depressions in the gravel of the river bottom, two or three feet in diameter, and six inches deep at the center. ‘The method of their construction and the general behavior of the specimens of the nest are very similar to those of the Brook Lamprey. But, owing to the large size of this species, all its movements can be minutely observed. The speaker exhibited a small model of the Lamprey group now under construction in the American Museum of Natural History and also life- size models of adult lampreys. The paper was also illustrated by lantern slides. Mr. Nichols dealt with the results of a brief collecting trip to Cuba and exhibited various specimens. He passed in review some of the Scombriform fishes. The king fish, Scomberomorus cavalla, is highly esteemed, but another species, S. regalis, is said to be occasionally RECORDS OF MEETINGS 359 poisonous. S. maculatus, the Spanish mackerel, was not seen. While regalis and maculatus occupy more or less distinct areas, cavalla is abundant both in Florida, with maculatus, and in Cuba, with regalis; in the speaker’s opinion these two last-named species, which are still closely related, have recently become separated through the competition on cavalla. Two very widely separated forms, Arbaciosa rupestris and Gobius soporator, were found inhabiting adjacent rock pools; both were concealingly colored and could have been confused until their distinctive color patterns were noticed. The paper was illustrated by means of lantern slides. The Section then adjourned. WILLIAM K. GREGORY, Secretary. SECTION OF ANTHROPOLOGY AND PSYCHOLOGY. 22 APRIL, 1912. Section met at 8:15 P. M., in conjunction with the New York Branch of the American Psychological Association, R. S. Woodworth acting as Chairman. The afternoon session was held at the Psychological Labo- ratory, Columbia University, and the evening session was held at the American Museum of Natural History. The following programme was offered : Gertrude M. Kuper, InpivinuaL DIFFERENCES IN THE INTERESTS OF CHILDREN. T. H. Kirby, PRACTICE IN THE CASE OF CHILDREN OF SCHOOL AGE. C. D. Mead, THE AGE OF WALKING AND TALKING IN RELATION TO GENERAL PRACTICE. G. C. Myers, SEX DIFFERENCES IN INCIDENTAL Memory. A. J. Culler, RELATION OF -INTERFERENCE TO ADAPTABILITY. E. 8S. Reynolds, J. T. Gyger and L. L. Winslow, EXPERIMENT IN THE CaTcCHING or PENNIES. D. O. Lyon, THE OPTIMAL DISTRIBUTION OF TIME AND THE RELATION OF LENGTH OF MATERIAL TO TIME TAKEN FOR LEARNING. SUMMARY OF PAPERS. Miss Kuper said in abstract: That interest plays a very important dynamic role in the educational field is only too evident from such 360 ANNALS NEW YORK ACADEMY OF SCIENCES treatises as Dr. Dewey’s article, “Interest as Related to Will” and Dr. Montessori’s “Pedagogia Scientifica.” But interest is a general term and can not have an absolutely universal value for every individual or every subject of thought or desire. Individual interests are as important in the social world as are individual capacities. They should, therefore, be a fruitful field for scientific investigation. 'The experimental work done with advertisements has brought to light group differences in the preferences of men and women for various appeals. The investigation to be reported was of a like nature, except that it dealt with children. The formal experiment consisted in asking an individual child to arrange nine pictures in the order in which he liked them best. The nine pictures were chosen to represent nine specific appeals: landscape, children, animals, religion, pathos, sentiment, patriotism, heroism, and action. (They were Cosmos prints and therefore of uniform size and finish.) In all, there were three series of these pictures, each parallel so far as possible with the other two in their appeals. The children numbered over 200, 10 girls and 10 boys for each year’s age from 6.5 to 16.5. They were almost entirely attendants of the public schools of New. York City and came from quite varied sections of the city. The results were tabulated according to age differences, broad social distinctions, and nationality. In the last-named case the number of subjects was so limited (10 girls and 10 boys to each of the following nationalities: Irish, French, German, and Italian, and only 9 girls and 8 boys to the Spanish) that the results are not held as significant. The positive data showed a sex difference in the order of preference for these several appeals. The girls’ order was: (1) Religion, (2) patriotism, (3) children, (4) pathos, (5) animals, (6) sentiment, (7) landscape, (8) the heroic, (9) action. The last two were decidedly lowest in the scale and the first three were quite clearly highest for all ages; but the picture representing these nine curves was one of bewildering intersections as the values changed from year to year. The boys’ order was: (1) Religion, (2) patriotism, (3) action, (4) the heroic, (5) pathos, (6) animals, (7) sentiment, (8) landscape, (9) children. The boys’ chart representing the curves for these appeals showed greater — agreement from year to year. Religion and patriotism, the heroic and action, and landscape and children kept rather parallel courses all along the age scale, and no very decided tendencies appeared with progressive age differences. Girls seemed to lose interest somewhat in pictures of children and animals and to take greater interest in the heroic and action pictures. The latter change is explained by the fact that, as the girls increased in school knowledge, they read an historical background into these more or less warlike scenes. RECORDS OF MEETINGS 361 A great sex difference was found in the variability measures, as cal- culated for the various ages, appeals, social classes, and nationalities. In every case but two, the girls exceeded the boys in their P.E.; and in these two exceptions the boys’ P.H. was once greater than the girls’ by only 5 per cent., and another time exactly equal to the girls’ P.H. The amount of sex difference was, as a rule, anywhere between 12 per eent. and 57 per cent. This held true in every scale, whether according to age, appeals, social class, or nationality. The girls’ average P.E. was 1.66; that for the boys was 1.36. Both girls and boys were least variable about the subjects they liked best, 7. e., religion and patriotism; but apart from these appeals there was no correlation of variability with relative likes or dislikes. It is a noteworthy fact that in range of variability the boys far ex- ceeded the girls. The limits for the boys’ P.E. were .82 (patriotism) and 1.60 (landscape), giving a range of difference of 78 per cent.; the limits for the girls were 1.47 (religion) and 1.95 (animals), showing a range of only 48 per cent. In this particular experiment this in- dicates that boys are very much more agreed about some likes than are girls, and yet quite as varied about others. In other experiments such a range of variability may point to greater individuality of the male sex among themselves while as a group they are relatively homogeneous. Another sex difference noted was the number of positive dislikes expressed by each sex. The girls gave 161, or 6 per cent. dislikes as against the boys’ 65, or 2.4 per cent. Boys seemed to entertain relative indifference toward the appeals at the bottom of the list. The things the girls disliked most were (1) scenes of action suggesting death and (2) pictures showing angry attitudes. The reasons given by the boys for their dislikes were (1) gloomy, indistinct scenes, (2) sentimental pictures, (3) costumes worn by men which were feminine in style or left the figure partly nude, and (4) pictures suggesting illness. A certain age difference revealed itself in the remarks made by the children about the pictures. The seven and eight year olds showed limited powers of observation. Some detail, and, in landscape scenes, al- ways the human detail, no matter how small, was made the focus of attention to the complete overlooking of the larger subject. Unfa- miliar details when pointed out to them received as many different in- terpretations as there were children. As the children grew older their remarks of both girls and boys. Emotional attitudes, actions and even of the pictures and they drew upon all their known sources for filling in their perceptions. At the ages between 11 and 13 the critical spirit made its first appearance among the girls. Only at fourteen did it oceur in the boys’ comments. At these ages the emotions prompted the 362 ANNALS NEW YORK ACADEMY OF SCIENCES remarks of both girls and boys. Emotional attitudes, actions and even words were ascribed to the pictorial persons. At 15, the remarks be- came more laconic, but what was said was significant and definite as to the persons, place and action of the picture. This age marked the first signs of hesitation in speaking of the pictures of sentiment. Up to the age of nine the remarks had been very naive; after that the pictures were dismissed with the phrase, “they’re lovers” or “a love pic- ture”; often the characters were named Romeo and Juliet, Paul and Virginia, etc. In all their comments the girls were far more personal than the boys. The personal pronoun and references to their individual experiences were the usual preface to their statements. With the boys it was quite. otherwise; they discussed the picture as an objective thing, independent of their conscious existence. Boys tended to locate scenes in definite historical time and specific geographic places. The effect of uncertainty about a picture, crudely averaged, was a displacement of about five places toward the lower end of the scale. Dr. Kirby said in abstract: This experiment was conducted to get some information concerning (1) the value of the practise experiment as a method for school work and (2) the value of practise periods of different lengths. 339 fourth year children belonging to 10 different classes took part in the practise, which consisted of adding columns, each of 10 numbers, 0’s and 1’s not included, as rapidly as was consistent with accuracy, each child competing with his own past record. Seven different sheets of columns of equal difficulty were used (Thorndike’s Addition Sheets). In every case there was one hour of practise, but for different classes this hour. was broken into 2214-, 15-, and 6-minute periods, an initial 15-minute period and a final 15-minute period being given to form the basis for determining the gain per cent. The hour’s practise for the 339 children taken as one group resulted in an average gain of 55 per cent.; median gain of 48 per cent. In a similar test with 19 university students, Professor Thorndike found an average gain of 29 per cent., median 33 per cent. from about 53 minutes of practise, and said: “The amount of improvement in this experiment may also add to our confidence that the method of the ‘practise experiments wherein one works at one’s limit and competes with one’s past record may well be made a regular feature in many school drills. Even if the same length of time produced in children a percentile improvement, only half as great as here, the gain would still probably be far greater than the gain by any of the customary forms of drill.” or RECORDS OF MEETINGS 363 For the classes which took the hour’s practise in 2214-minute periods, there was an average gain of 61 per cent., median 49 per cent.; in 15- minute periods, average gain 55 per cent., median 43 per cent.; in 6-minute periods, average gain 54 per cent., median 44 per cent. Dr. Mead said in abstract: 1. Data.—50 “normal” children (25 boys and 25 girls), averaging less than six years of age, of graduate students of Teachers College and Columbia College. Ages were thrown to the nearest month. Walking means: “To take a step unassisted.” Talking means: “To use a word intelligently, 7. e., to associate the idea with the object.” Results——The median “normal” child begins to walk at 13.5 months, with a probable error of 1.06 months. The chances are 999 to 1 that the true median will not differ from the median obtained by more than .66 month. The extreme range is from 11 to 30 months. .90 per cent. of the cases fall between 11 and 17 months. The median “normal” child begins to talk at 15.7 months, with a probable error of 2.83 months. The chances are 999 to 1 that the true median will not differ from the median obtained by more than 1.96 months. The extreme range is from 9 to 25 months. 90 per cent of the cases fall between 10 and 21 months, with 18 months as the mode. Il. Data.—145 “schoolable” children (boys and girls) of the Indiana School for Feeble-minded Youth, in reply to the question on the personal descriptive entrance blanks: “At what age did the child commence to walk?” and 92 in reply to the question: “At what age did the child commence to talk?” _ Results—The median feeble-minded child begins to walk at 21.8 months, with a probable error of 7.56 months. ‘The chances are 999 to 1 that the true median will not differ from the median obtained by more than 3 months. The extreme range is from 12 to 72 months. 90 per cent. of the cases fall between 13 and 50 months. The median feeble-minded child begins to talk at 34.2 months, with a probable error of 12.6 months. The chances are 999 to 1 that the true median will not differ from the median obtained by more than 6.5 months. The extreme range is from 12 to 156 months (only one case going above 108 months). 90 per cent. of the cases fall between 14 and 84 months. Dr. Myers said in abstract: A test was desired wherein the thing to be remembered should be merely incidental and where the focus of the subject’s attention should be directed away from the facts to be called for after the exposure of the stimuli, but where these facts would have to enter, wholly or in part, into the experience of the subject. To this end a 864 ANNALS NEW YORK ACADEMY OF SCIENCES list of six simple words were used as stimuli. The subject was told that he would be given a spelling test and he was led to believe that it would be a real test in speed and accuracy of spelling. A practise test with digits was given for three successive times before the real test began, to delude the subject as to the purpose of the ex- periment. A dozen or more digits were pronounced at random so rapidly that the subject could scarcely keep up in writing them. In the midst of this series of digits the experimenter, without any warning, gave the signal for the subject to turn the page upon which he was writ- ing, and continued to pronounce digits at the same speed. The subject was told that the words would be given in the same manner, but not quite so rapidly. The following words were then pronounced: angel, pickle, dirt, busy, onion, women. ‘The last word was pronounced in such a manner that another word was expected by the subject, but the signal, “turn,” was given instead, and the subject was told to write as many of these words as he could remember, to place them in the order in which they had been given, and to indicate by a line the place for each omitted word. The time each individual required to reproduce the words was. recorded by a stop-watch. After testing over 100 individuals the writer applied the test to groups of college, normal-school and public-school subjects. Aside from imme- diate reproduction, records were secured after various intervals, ranging from 14 hour to 3 months. In all such cases a practice test of rapid fold- ing of papers was added. After the words were pronounced the papers were promptly collected and the experimenter left the room. The sub- jects thought the work was ended, but at various times the experimenter reappeared and asked for the reproduction. The time for all group repro- duction was limited to 1144 minutes.. The best results were secured immediately after presenting the stimuli. Practically the same efficiency was shown for the reproduction after 6 hours as for that after 14 hour. But there was a decided fall after 7 days. and a still greater fall after 3 months. No appreciable difference was shown in efficiency between the lower grades and the college students for immediate reproduction; but after: various intervals there was a gradual decrease in efficiency with age. Of the 1,515 subjects, 757 females and 758 males, only 29 of the former and 18 of the latter reproduced the six words in exact order. In all grades the females were markedly superior to the males, both for the number of words remembered and for order. They had a higher central tendency and were more variable than the males in the 5th, 6th, 7th, and 8th grades, while for the other groups the males were more: variable. RECORDS OF MEETINGS 365 108 other subjects were tested with 10 letters and digits. Here the girls answered more, but the boys were better for order. Mr. Culler said in abstract: The purpose of this experiment was two- fold: to determine the effect of differently distributed practise series upon learning given material; and to make observations upon the learn- ing process in general. _ The material to be learned was the path from the beginning to the end of the Hampton Court maze. The paper (8 by 6 inches) on which the maze was printed, was affixed to a board. Over it was placed a large circular piece of cardboard, easily movable, having in the center a small opening (5% to 11/16 inch) through which extended a pencil to mark the course of the subject?s movement. At no time could the subject see more of the maze than the part visible through the opening. At the beginning of the experiment the subject was thus instructed: Pencil is now at the entrance tothe maze; keep on moving until you reach the end. Never cross a line; always keep to an open path. Mazes are all the same and will be placed in the same position. At each trial the time was recorded and number of errors was counted and recorded. ‘To each subject were given 12 trials. Subjects were di- vided into 6 groups as follows: 12 trials at one time, 6 on 2 successive days, 4 on 3 days, 3 on 4 days, 2 on 6 days and 1 on 12 days. There were 5 men in each group except the last, in which were 3. With regard to time of day, subjects were divided into two groups: one group each ‘day for the required number of days, after lunch (1-2 P. m.): the second group each day after dinner (7-8 p. M.). In comparing men of the two groups no account was taken of this slight difference, as it was considered practically negligible. Good light was uniformly provided. The interval between successive trials of a subject at the same sitting was 30-40 seconds. Subjects were all graduate students, age frou 22 to 28. Three classes of errors appeared: Wrong choice between alternative courses, retracing when on right course, and (accidentally) crossing a line. The first kind are major errors (value 1) and the other two kinds minor (value 144). These are arbitrary values for computing results. The major errors were counted as follows: There are 6 (or 7, depending upon the course taken) places where choice must be made between alternative paths of which only one is right. Each time the subject moved from one of these places in a wrong path, 1%. e., away from the goal, it was counted one error. Errors of retracing when on the right path were usually small and due to defective attention or eyesight— subject either thought he had accidentally passed an opening and moved 366 ANNALS NEW YORK ACADEMY OF SCIENCES back to see, or on coming to a turn failed to notice the opening and thought he had run into a blind alley. The results are as follows: I. TABLE OF ABSOLUTE TIME AND HRROR VALUES ATTAINED IN HacH GROUP (The different groups are indicated thus: One—12, etc.; the word indicates the number of trials each day, the figure the number of successive days. The two columns show the average of number of seconds consumed and number of errors made in the last three trials in each group; thus showing the relative standing of groups at end of practice period. ‘The figures in parentheses show relative position. ) Time, Errors, ‘ Per Cent. Per Cent. Dy sree La NDR cre ORE RI Vata une N agian ts 50 (3) 4.8 (4) WO == Gie iece) rte sot ccs vost mravo to nerwaa eles eieteinual Siero nett 61 (5) 5.2 (5) TATOO ih evans iis se Gide sus Seco ea paoseie, eee eT 59 (4) 3.2 (8) QU OU Re cog Maan tot UW AR hea ea aA 39 (1) 9 (1) NSD. aaa tas Ant tye | EERE ae iran es aM oeates Ree oe ae 75 (6) 5.5 (6) Ce Ly ahs os UCL gO Be Ala eA Oa 48 (2) 3.0 (2) II. TABLE OF PERCENTAGE GAINS (In each ease the percentage represents the ratio between the average of first three trials and last three trials in the same group. This table is intended to show improvement of each group irrespective of absolute values attained.) Time, Errors, Per Cent. Per Cent. Or aaa eta ante eae bas ogeeer cence pene 210.0 (4) 147.9 (5) ID O=—=G wis, ol spel cnsicets eleicus encuecersetsencisianctens 253.0 (3) 161.5 (4) PAT C GA 5 aviaisira te a. 'or0f sila staves cle ete ce esiees 195.0 (6) 802.0 (1) MOUTH Gh aed abs cyselereieiey sven ered ones 341.0 (2) 218.5 (3) SS TiO icin Crete SU aestatnaeiere anette ra Noalisveviens 206.6 (5) 125.3 (6) MW OLV Oa 0 sii 28, sole eof evedtseley see! svelte 368.7 (1) 236.6 (2) (It must be said that the results of Six—2 were vitiated by the professed indifference of one subject, because of which both time and errors for the last few trials in that group are abnormally high.) The results seem to point to the following conclusions: In general, outside the Six—2 group, the One—12 and Two—6 groups made the lowest absolute records and also least improvement; this apparently in- dicates that the learning period was too prolonged, with insufficient practise at any one time. On the other hand, the ''welve—1 and Four— 3 groups show in general the highest absolute records and greatest im- provement. Here the practise was more thorough each time and not so prolonged. ‘The curve of greatest regularity is the Four—3 curve. The three groups, then, in which practise periods were longer and confined to a few days show better results than the three in which practise periods are shorter and prolonged over 4-12 days. The applica- RECORDS OF MEETINGS 367 tion to learning any material would seem to be that better results are secured by a few more prolonged or persistent periods of study repeated perhaps for several days than shorter periods prolonged over a greater number of days. Some observations were made on individual methods of learning which can not be included here. Messrs. Reynolds, Gyger and Winslow. The authors said in abstract: The experiment had two aims: (1) To investigate the learning process. (2) To find what transfer, from the right hand to the left, if any, would be shown. Three subjects took part in the experiment which follows. It was carried on in two series: (1) That in which the subjects caught the pennies, two at a toss, palm of the hand down. (2) That in which they caught three. The first series was of 7 days’ duration; the second, 10 days’. The time for tossing was from 1 P. M. to 2 P. M. on Mondays and Wednesdays. Conditions were as nearly constant as possible, the same room being used throughout the experiment. In the case of the two-penny series, the subjects caught for 10 trials and then rested for 10. In the three-penny series two subjects caught at the same time, the third subject resting. In the first case, score was kept by the two unemployed subjects in turn; in the second case, by the one unemployed subject. Certain conditions influencing accuracy were noted, among which are the following: Some parts of the room were more conducive to accurate catching than others, that nearest the window being the most favorable. The pennies could be caught with most accuracy if no ob- jects were in front of the subject to distract his attention. The tossing, when carried on before a blank, light-colored wall, was most successful. An increase in confidence and in accuracy resulted when a window was opened to admit new air. An interruption, as that caused by another person entering the room, was followed by a corresponding fall in score. The subject, by counting to himself his successful tosses, was stimulated to a better score. The nervous feeling of haste as well as nervousness caused by outside matters of importance to the subjects (such as press- ure of work) tended rather to increase than to diminish their scores. Each subject discovered and followed his own methods of tossing. After finishing the two series, the subject who had followed the method of throwing his pennies high into the air was able to catch an additional penny (making four in all) with very little effort. The other subjects tried this continually and failed, their hands striking the floor before the fourth penny was reached. The quick shutting of the hand was an important factor. One subject was materially helped by thinking of 368 ANNALS NEW YORK ACADEMY OF SCIENCES the word “grab” previous to each trial. In some instances, the second penny would be caught and lost, the first and third being retained. Although occasionally a subject would catch all three successfully with- out knowing it, yet the tossing can not be said to have become automatic. The progress in learning was unsteady. Yet in each case there was a gradual advance, noticeable particularly in the beginning. A warming-- up period was universally experienced by each subject at the beginning of each day’s practise. In the second series, a transfer test was tried with the left hand before and after the practise series. This showed a considerable increase in ability to catch with the left hand. AMOUNT OF TRANSFER CATCHES Subject. Before Test. After Test. Per Cent Gain. x NaS cecoiese cape prcee eareauentts seuievradiay euetrameul ae taste 3 14 4662 Dive aietasatera aides So hclng Smee gaa eae ee 11 32 290+ SD ta la asl aves cheers sal ot antebel aanare tele che ia edaae arenes 1 29 2900 Total gain......... eenoeenno 15 és) 500 Mr. Sax said in abstract: Although art and science are widely sepa- rated, they may co-operate in art education. Prevailing methods are in- direct, depending upon a never certain transfer of training. During the three years the average student spends at art school, his course is as follows: Casts and still life in charcoal; still life in color; anatomy and perspective as formal subjects; the figure in charcoal; some composition, and, finally, painting the head and figure in oils. | Results show little transfer; for example, compositions show little knowledge of anatomy or perspective. Charcoal and oils have few identical elements in substance or procedure; in. fact, specific habits formed in mastering charcoal often act preclusively when the student attempts to paint. Students who can draw, but not paint; construct, but not compose, or are draughtsmen, but not colorists, and their oppo- sites are in the overwhelming majority. Experiments now under way on the learning process as applied to painting seem to show that (a) preparation in charcoal and still life*is unnecessary in painting figures; (b) efficiency depends largely upon correct analysis; (¢) muscular coordination plays a minor part; (d) a direct method and generalized idea of procedure are essential and (e) the control of attitude is most important. Dr. Lyon said in abstract: This paper was divided into two parts, it being in reality a discussion of two distinct questions: (1) “The Distri- bution of Time in Relation to Economy in Learning and Retention” ; RECORDS OF MEETINGS 369 and (2) “The Relation of Length of Material to Time Taken for Learn- ing.” Concerning the first of these, it was shown that in estimating economy, not only must we consider the time spent, but the degree of retention as well. It was shown that individuals differ greatly, and that where one could learn a set of ten stanzas in less time by the continuous - method (i. ¢., doing the work in “one sitting”), another individual could lower his total time by dividing the time spent into several periods, é. g., by spending 5 minutes per day. With but three exceptions re- tentiveness was decidedly better by the divided-time-method. This was notably the case with nonsense-syllables and poetry. The most general statement that can be made, taking all materials and methods of presen- tation into consideration, is that the most economical method is to dis- tribute the readings over a rather lengthy period—the intervals between the readings being in arithmetical proportion. For example, with one individual in memorizing a poem of twenty stanzas the highest retentive- ness was obtained by distributing the readings as follows: 2 hours, 8 hours, 1 day, 2 days, 8 days, 16 days, 32 days, ete. The practical bearing of the results obtained on education in general was then considered. The above individual found that the most economical method for keeping material once memorized from disappearing was to review the material whenever it started to “fade.” Here also the intervals were found to be, roughly speaking in arithmetical proportion. For similar reasons the student is advised to review his “lecture-notes” shortly after taking. them, and if possible, to review them again the evening of the same day. Then the lapse of a week or two does not make nearly so much difference. When once he has forgotten so much that the various asso- ciations originally made have vanished, a considerable portion of the material is irretrievably lost. 2. The Relation of the Length of Material to Time Taken for Learning.—Tables were presented to show that the relation depended almost wholly upon the division of the time spent in learning, 1. e., the distribution of the time intervals. In other words, the relation, or ratio, depends upon the method used in memorizing. Only three methods were considered: The “continuous” or “mass” method; the once-per-day method; and the once-per-week method. Up to a certain point, with some individuals, when digits were used as material, the time varied directly as the square of the number of digits, when the continuous method was used. By the once-per-day method, however, the time varied, roughly speaking, directly as the length of the material. It was shown that in order to get the best results the same subject should take all the various lengths of material used, and that it would be unfair to dis- 370 ANNALS NEW YORK ACADEMY OF SCIENCES tribute the varying lengths among different subjects. As only one _ method can be tried at a time, an experiment of this nature must needs extend over a period of several years. In the case of prose, by the once- per-day method, 500 words were memorized in as few days as the 95- word passage. The time may therefore be said to vary directly as the length of the passage. The same holds true for digits and nonsense- syllables, but not to so great a degree; for the number of days needed for 200 nonsense-syllables was considerably greater than that needed for 20. By the “continuous” method, however, we observe that where the 100-word passage was memorized in 9 minutes, the 500-word passage took 52 minutes—nearly 6 times as much time being required, although the passage is only 5 times as long. This is much more strikingly shown when we examine the curve obtained for the digits. Here we see that although it took only 5 minutes to learn 24 digits, it took 2 hours and 34 minutes to learn 200—more than 31 times as long instead of 8. In short, it is obvious that the once-per-day method is—to say nothing of giving a far superior retention—far more economical than the “con- tinuous” method. This is especially so for material memorized by motor associations such as nonsense-syllables or digits. } The Section then adjourned. F. Lyman WELLS, Secretary. BUSINESS MEETING. 6 May, 1912. The Academy met at 8:15 p. mM. at the American Museum of Natural History, President McMillin presiding. The minutes of the last business meeting were read and approved. The Recording Secretary reported the deaths of the following. members : | Col. John Jacob Astor, an Active Member for 18 years, lost with the Titanic. Mr. Isidor Straus, an Active Member for 6 years, lost with the Titanic. Col. John Weir, a Life Member for 5 years, lost with the Titanic. Mr. George Borup, an Active Member for 4 months. The Recording Secretary spoke of the great loss to the Academy, the Museum and the scientific world at large entailed by the death of Mr. Borup, who was to have been the leader of the Crocker Land Ex- pedition organized under the auspices of the American Museum of RECORDS OF MEETINGS Syl Natural History and the American Geographical Society for the purpose of arctic exploration during the years 1912-1914. The Academy then adjourned. EpmMuNpD Otis Hovey, Recording Secretary. SECTION OF GEOLOGY AND MINERALOGY. 6 May, 1912. Section met at 8:30 Pp. m., Vice-President Woodman presiding, about 150 members and visitors being present. The following papers were read by title: Neil E. Stevens, Nores on THE STRUCTURE AND GLACIATION OF OvER: LOOK MouNrAIN. George F. Kunz, THE GmM-BEARING PEGMATITES OF LOWER CALI- FORNIA. ie The meeting was then given over to the following public lecture: Alfred H. Brooks, GrEoLogy anpD MINERAL ReEsouRCES OF ALASKA. Mr. Brooks has been in charge of Alaskan exploration for the United State Geological Survey for the last ten years or more. No one is more intimately acquainted with the geography, geology and resources of that country. The lecture was illustrated with lantern slides, maps and charts. The Section then adjourned. CHARLES P. BERKEY, Secretary. LECTURE. ry WW UNe ISLE James F. B. Bowles: SANITATION OF THE PANAMA CANAL. SECTION OF BIOLOGY. 13 May, 1912. Section met at 8:15 Pp. m., Vice-President Lucas presiding. The minutes of the last meeting of the Section were read and approved. j 372 ANNALS NEW YORK ACADEMY OF SCIENCES The following programme was then offered: R. D. O. Johnson, Notr ON THE HABITS OF THE CLIMBING CAT- FISH (Arges marmoratus) FROM THE UNITED StTaTEs OF CoLomBi4A. (Read in abstract by the Secretary.) Bashford Dean, ON THE CHANGES IN THE BEHAVIOR OF THE EEL (Conger malabaricus) DURING ITs TRANS- FORMATION. Bashford Dean, Do DrvELoPING EmBryos GivE REAL CLUES AS To Lines oF DESCENT? William K. Gregory, Norrs on CERTAIN PRINCIPLES OF QUADRUPEDAL LocoMOTION AND ON THE MECHANISM OF THE Limss oF Hoorep ANIMALS. F. F. Hahn, On THE DictyonEMA Fauna oF Navy ISLAND, 3 New Brunswick. (Read by title.) SUMMARY OF PAPERS. Mr. Johnson’s paper is published as pages 327-333 of this volume. Professor Dean said in abstract: When at Misaki, Japan, ihe speaker: had made observations upon the structure and behavior of a living leptocephalus larva which was kept alive in an aquarium for over three weeks, during this time undergoing its metamorphosis. Especially in- teresting is the rapidity with which the behavior of the young eel changes from day to day in its methods of swimming and resting, response to stimuli, ete. The speaker suggested that these marked differences in be- havior in successive stages were correlated with kaleidoscopic changes in elements of the central nervous system ; that when more fully known this would probably afford a suggestive case of parallelism between psychic reactions and neurological conditions. The paper was illustrated by drawings and diagrams. Professor Dean in his second paper said in abstract: After reviewing the history of the question and touching upon the modern reaction against the extreme views of Haeckel the speaker endeavored to show that a comprehensive study of the anatomy and embryology of ganoid and teleost fishes in the light of paleontological data gave strong evidence in the affirmative. The Secretary gave an abstract of a communication from Dr. P. Bachmetjew, of Sofia, relating to the physiology of “Vesperugo pipi- strellus’ and “Muiniopterus schreibersu.’ In some cases these bats RECORDS OF MEETINGS 373 have been thawed out and the heart action had resumed even after the body had been cooled to —7° Cent. below the body temperature. The Section then adjourned. Wititam K. GREGoRY, Secretary. SECTION OF ASTRONOMY, PHYSICS AND CHEMISTRY. 20 May, 1912. Section met at 8:15 P. m., in the Doremus Lecture Room of the Chemistry Building of the College of the City of New York, Professor Poor presiding. The minutes of the last meeting of the Section were read and approved. The following programme was then offered: Charles Baskerville, TunGsTEN. SUMMARY OF PAPER. Professor Baskerville pointed out that tungsten at one time was hardly mentioned in text books, but that now it is a substance of con- siderable importance. It was discovered in tin-ore by Carl W. Scheele in the year 1781. In 1848, tungsten salts were used for fixing colors in cotton, and, in 1857, the fireproofing of draperies by tungsten was pro- posed. Tungsten is used in making bronze and steel. Tungsten steel retains its temper even when red hot and is better than the best carbon steel known. The rims of car wheels are made of tungsten. steel. The speaker then gave an interesting account of the invention and development of tungsten lamps. He spoke at some length of the very great practical difficulties that had to be overcome owing to the fact that the tungsten filaments were brittle. Finally, however, this was overcome so that now the tungsten incandescent lamp is the best one on the market. The Section then adjourned. F. M. PEDERSEN, Secretary. BUSINESS MEETING. % OcToBER, 1912. The Academy met at 8:25 Pp. m. at the American Museum of Natural History, President McMillin presiding. 374 ANNALS NEW YORK ACADEMY OF SCIENCES The minutes of the last business meeting were read and approved. The Recording Secretary reported the following deaths: Ferdinand Zirkel, Honorary Member for 8 years. Jules Henri Poincaré, Honorary Member for 12 years. The Academy then adjonrned. EpmMuND Otis Hovey, Recording Secretary. SECTION OF GEOLOGY AND MINERALOGY. 7 OcToBER, 1912. Section met at 8:30 p. M., Vice-President Woodman pres 35 members and visitors betta present. The minutes of the last meeting of the Section were read and approved. Before the announced papers were called for the chairman asked Professor J. F. Kemp to give some account of his recent trip to Panama. A first-hand general account of his experiences and geologic observa- tions was given. Questions were asked by members of the Section. The following programme was then offered: D. W. Johnson, THe Wesrwarp TRIP OF THE TRANSCONTINENTAL EXCURSION OF THE AMERICAN GEOGRAPHICAL SOcIETY. k George H. Girty, Grotocic AGE oF THE BEDFORD SHALE OF OHIO. SUMMARY OF PAPERS. Dr. Johnson gave an interesting account of the make-up of the party, the method of travel, the places of greater interest and some of the special features to which most attention was given. Dr. Girty’s paper was read in part by Professor Grabau and its bear- ings were commented upon. It has been published as pages 295-319 of this volume. | The Section then adjourned. CHARLES P. BERKEY, Secretary. RECORDS OF MEETINGS Bid) SECTION OF BIOLOGY. 14 OctoBER, 1912. ~ Section met at 8:15 p. m., Vice-President Lucas presiding. The minutes of the last meeting of the Section were read and approved. The following programme was then offered: Roy W. Miner, TypicaAL Marine INVERTEBRATE ASSOCIATION FROM Woops Hotz To Casco Bay. Roy C. Andrews, AN EXPLORATION OF NORTHEASTERN Korwa. SUMMARY OF PAPERS. Mr. Miner described and illustrated a series of typical invertebrate faunal complexes or associations of the eastern Atlantic coast. He gave views of many beautiful models and faunal groups which had been mode for the American Museum, illustrating ecological relations and the dominance of certain groups in particular localities. Mr. Andrews gave an account of an exploration made for the Ameri- can Museum of Natural History in a territory not hitherto studied by zoologists. The Section then adjourned. WILLiAM K. GREGORY, Becrey, - SECTION OF ASTRONOMY, PHYSICS AND CHEMISTRY. 21 OcToBER, 1912. Section.met at 8:30 Pp. M., Vice-President Poor presiding. The minutes of the last meeting of the Section were read and oparoret. . Professor Charles Lane Poor was sonianted for Teen of the Academy and Chairman of.the Section for 1913. Professor F. M. Pedersen was elected Secretary. The Committee on the future of the Section, consisting of Professors Poor, Trowbridge and Pedersen then made its report, which was referred to the Council for consideration. F. M. PEDERSEN, Secretary. 376 ANNALS NEW YORK ACADEMY OF SCIENCES SECTION OF ANTHROPOLOGY AND PSYCHOLOGY. 28 OcroBER, 1912. Section met in conjunction with the American Ethnological Associa- tion at 8:15 Pp. m., General James Grant Wilson presiding, about 128 members and visitors being present. The minutes of the last meeting of the Section were read and approved. The following programme was then offered : Franz Boas, A YEAR IN MEXIco. SUMMARY OF PAPER. Professor Boas gave an outline of the work of archeological excavation and linguistic research conducted by him during his stay in Mexico. His lecture was illustrated by numerous stereopticon slides showing especially pottery found in different layers in the Valley of Mexico. The Section then adjourned. F. LyMan WELLS, Secretary. - BUSINESS MEETING. 4 NOVEMBER, 1912. The Academy met at 8:30 Pp. Mm. at the American Museum of Natural History, Vice-President Woodman presiding. The minutes of the last business meeting were read and approved. The Recording Secretary reported the following death: Morris Loeb, an Active Member and Fellow for 20 years. The preparation of a suitable minute for the records of the Academy was referred to Professor J. J. Stevenson and Dr. HE. O. Hovey. Announcement was made from the Council of the engagement of Dr. Alexis Carrel, to give an address before the Academy on 11 No- vember, regarding his recent experimental work in physiology and of Professor Hugo de Vries for an address upon experimental evolution to be delivered 6 December. The Academy then adjourned. Epmunp Oris Hovey, Recording Secretary. RECORDS OF MEETINGS Birdy SECTION OF GEOLOGY AND MINERALOGY. + NovEeMBER, 1912. Section met al, 8:45 p. m., Vice-President Woodman presiding, 23 members and visitors being present. In the absence of Secretary Berkey, Dr. Hovey was elected Secretary pro tem. The minutes of the last meeting of the Section were read and approved. On motion duly made and seconded, Professor J. E. Woodman was nominated chairman of the Section and Vice-President of the Academy for the year 1913. Professor Berkey expressed the earnest wish that he be relieved of the secretaryship which he had held five years. The Section acquiesced in the request and passed a most cordial vote of thanks to him for his long, faithful and efficient services, which have contributed so much to the success of the Section. Professor C. T. Kirk of the Normal College of the City of New York, was then nominated Secretary of the Section for 1913 and unanimously elected. The chairman then appointed Professor J. F. Kemp, Dr. George F. Kunz and Dr. E. O. Hovey to serve with himself as a committee, in response to a request of the Council, to consider the condition of the Section and to make recommendations for its future work, this Com- mittee to report to the Council. The Secretary pro tem then read a letter from R. B. Earle of the Department of Geology, New York University, asking for a grant of $200 to assist him in carrying on research work on the origin and history of certain types of interbedded iron ores. On motion, this application was approved and referred to the Committee on Grants from Research Funds for consideration. The following programme was then offered: F. S. Hintze, THE Fosstts AND HorIzON OF THE DRIFT PEBBLES. : Marjorie O’Connell, PRESENT OPINIONS ON THE HABITS OF TEE EURYPTERIDS. . A. W. Grabau, WAS THERE A FoRMER GoAT ISLAND AT NIAGARA GLEN ? Mr. Hintze’s paper was discussed by Professor A. W. Grabau, Pro- fessor J. F. Kemp, Mr. B. E. Dodge and Professor D. 8. Martin. Miss O’Connell’s paper was discussed by Professor Grabau. 378 ANNALS NEW YORK ACADEMY OF SCIENCES SUMMARY OF PAPER. Professor Grabau said in abstract: Foster’s Flat below the whirlpool of Niagara and the topography and cross section of the gorge show — that the falls of Niagara were once localized there in the same style as is now the case at Goat Island. The same kind of development of river work is well illustrated also at the falls of the Genesee. _ The paper was discussed by Professor D. W. Johnson, who cited what is probably a similar case from near St. Anthony’s Falls, Minnesota, of the Mississippi River. The Section then adjourned. : EpmMuND Otis Hovey, Secretary pro tem. i SECTION OF BIOLOGY. 11 NovrempBer, 1912. On this occasion the Section of Biology co-operated with the Academy as a ‘whole and with the American Museum of Natural History in welcoming Dr. Alexis Carrel, recipient of the Nobel Prize in medicine, 1912, who gave a lecture in the large auditorium, entitled “Results of the Suture of Blood Vessels and the Transplantation of Organs,” about 800 persons being in attendance. After the lecture an informal re- ception, attended by the officers, members and friends of the Academy was held in honor of the lecturer. . Witi1am K. Grecory, Secretary. SECTION OF ASTRONOMY, PHYSICS AND CHEMISTRY. 18 NovEMBER, 1912. Section met at 8:15 Pp. M., Vice-President Poor presiding. The minutes of the last meeting of the Section were read and approved. The following programme was then offered: Albert B. Pacini, THe DistrisuTion or FERRIC CHLORIDE BETWEEN ETHER AND AQqurous Hyprocutoric AcID AT 25> ©. Charles Lane Poor, THE Cause or THE TIDES. RECORDS OF MEETINGS Bis) SUMMARY OF PAPERS. . Dr. Pacini gave a study of the distribution of ferric chloride under the conditions which obtain in the use of Rothe’s method for the de- termination of aluminum, nickel and other metals in steel, the mixed chlorides in hydrochloric acid solution being shaken out with ether which removes the greater portion of the ferric chloride. No decisive knowl- edge was gained regarding the state of molecular aggregation of the ferric chloride in the ether solution insomuch as data concerning the degree of disassociation of ferric chloride in aqueous solution are not at present available. Graphical treatment of the constants obtained yields a curve of two distinct sections: in low concentrations up to about 0.38 mols per liter in the ether layer, a straight line; above this point a parabola satisfying the equation (C, — .38)?— KC,, where K == -+ 1.8. The application of the results to analytical separation lies in the fact that the percentage of iron extracted from a hydrochloric acid solution by shaking out with ether is greater relatively as the concen- tration is lower, that is to say, the more dilute the original hydrochloric acid solution of iron, the nearer complete is the extraction of ferric chloride therefrom by ether. Professor Poor gave a brief outline of the theories of the tides as developed by La Place, Darwin and others, and contrasted these older theories with recent investigations and theories of Dr. Harris, of the United States Coast and Geodetic Survey. Until the researches of Dr. Harris appeared, the tides were considered as a world phenomenon, and primarily due to a large wave which originates in the Pacific Ocean and travels around the world at varying speeds, due to the depth of the oceans. This wave was supposed to take some fifty hours to travel from the Pacific around Cape Horn to the shores of New York. Dr. Harris considers the tides as purely local phenomena; the tide of each ocean basin is primarily due to a standing wave or oscillation originating in that basin and practically independent of the oscillations or tides in other basins. The tides at New York and the Atlantic Coast, under this theory, originate in the North Atlantic basin and are wholly independent of the tides in the Pacific and Indian Oceans. The Section then adjourned. : F. M. PEDERSEN, Secretary. 880 ANNALS NEW YORK ACADEMY OF SCIENCES SECTION OF ANTHROPOLOGY AND PSYCHOLOGY. 25 NOVEMBER, 1912. Section met in conjunction with the New York Branch of the Ameri- can Psychological Association, Dr. R. S. Woodworth acting as chairman. The afternoon session was held at the Psychological Laboratory, Co- lumbia University, at 4:10 Pp. m. and the evening session was held at the American Museum of Natural History at 8:15 P. M. The following programme was offered: F. Krueger, DIFFERENT-TONES AND CONSONANCE. Raymond Dodge, T'H= ATrempr TO MrasuRE MENTAL WoRK AS A PsycHo-DyNAMIC PROCESS. Robert M. Yerkes, THE PsycHoLocy oF THE EARTHWORM. The Section then adjourned. F. Lyman WELLS, Secretary. BUSINESS MEETING. 2 DECEMBER, 1912. The Academy met at 8:15 Pp. M. at the American Museum of Natural History, President Emerson McMillin presiding. The following candidates for membership in the Academy, recom- mended by Council, were duly elected: ACTIVE MEMBERSHIP. D. W. Johnson, Columbia University. AssocIaATe MEMBERSHIP. Marjorie O’Connell, 616 West 182nd Street. The Academy then adjourned. EpmuNnpD Otis Hovey, Recording Secretary. SECTION OF GEOLOGY AND MINERALOGY. 2 DECEMBER, 1912. ~ Section met at 8:20 p. M., President McMillin presiding. The meeting was devoted to the following public lecture: RECORDS OF MEETINGS 381 Harry Fielding Reid, THe SrISMOGRAPH AND Wat rt THAcHEs. Professor Reid described the characteristics of the seismograph, which has now been sufficiently perfected to record strong earthquakes occur- ring at the antipodes. The revelations of the instrument and problems awaiting solution were discussed and what happens at the time of an earthquake was explained. The lecture was illustrated with lantern slides. The Section then adjourned. CHarLes P. BEerKEY, Secretary. LECTURE. 6 DECEMBER, 1912. (In co-operation with the American Museum of Natural History.) Hugo de Vries, EXPERIMENTAL EVOLUTION. SECTION OF BIOLOGY. 9 DECEMBER, 1912. Section met at 8:15 p. M., Professor Henry Fairfield Osborn presiding. The minutes of the last meeting of the Section were read and approved. The following programme was then offered: C.-E. A. Winslow, A Musrum or Livine Bacteria. A. J. Goldfarb, Tue INFLUENCE oF THE NERVOUS SYSTEM UPON GRowTH. A. J. Goldfarb, A New Mernop or Fusinc Eces oF THE SAME SPECIES. SUMMARY OF PAPERS. Dr. Winslow said in abstract: The American Museum of Natural History is the first museum of its kind to recognize that the relation between man and his microbic foes is fundamentally a problem in natural history and a problem of such interest and importance as to warrant the creation of a special Museum Department of Health. The prime function of this department is of course to present in the form of effective exhibits the main facts about the parasites which cause 382 ANNALS NEW YORK ACADEMY OF SCIENCES disease, their life history, the conditions which favor their spread to man, their relations to intermediate insect hosts and the means by which mankind may be protected from their attacks. In connection with this work of public exhibition there seemed to be a unique oppprtunity for maintaining, as a sort of study collection, a museum of living bacteria for the benefit of .working laboratories all over the country. American bacteriologists have heretofore been compelled to send. to Vienna for authentic stock cultures, and many important type strains have been lost because the laboratories in which they were isolated had no facilities for keeping them permanently under cultivation. The authorities of the Museum were quick to appreciate the importance of the public service that could thus be rendered to those engaged in bacteriological teaching and research and early in 1911 endorsed the establishment of such a collection and bureau for the distribution of bacterial culture. A circular was sent out, to which the various labora- tories quickly responded by sending in the cultures in their possession. On December 1, 1912, the collection included 578 strains representing 374 different named types and including most of the important patho- genic and non-pathogenic forms which have been definitely described. During the period of somewhat less than two years, from January 1, 1911, to December 1, 1912, the laboratory distributed to 122 different colleges and research laboratories of the United States and Canada 1700 different cultures, in every case without charge. It is the policy of the Department to send cultures free to all teaching laboratories of college and university grade and to all research laboratories, whether cultures are sent to it in return or not. Many cultures have been called for by teaching laboratories for use in their class work. The most im- portant service the laboratory has been able to render, however, has been in furnishing authentic cultures to investigators who have been making a study of certain special groups and the published papers which have resulted, in which various detailed characters of the museum. types are described, of course greatly increase the value of the collection. The paper was discussed by Professor Osborn. Dr. Goldfarb described a series of experiments upon certain annelid worms (Amphinoma Lumbricus) which showed that the presence of the central nervous system was not essential for growth and regeneration. Dr. Goldfarb, in his second paper, described a method of fusing embryos and larvae of the sea urchin (Toxopneustes). After fertilizing the eggs they were placed in sea water to which enough 5 molecular Na Cl was added to make a solution of 35 to 75 per cent. The eggs: RECORDS OF MEETINGS 383 were left in this solution about six hours and were then transferred to normal sea water. The unfused larvae floated to the surface, the fused ones, which were obtained in great numbers, remained at or near the bottom. : The Section then adjourned. WiLuiam K. GREGORY, Secretary. ANNUAL MEETING. 15 DEcEMBER, 1912. _ The Academy met in Annual -Meeting on Monday, 16 December, 1912, at the Hotel Endicott, at the close of the annual dinner, President McMillin presiding. The minutes of the last Annual Meeting, 18 December, tei were read and approved. Reports were presented by the Corresponding Secretary, the Record- ing Secretary, the Librarian and the Editor, all of which, on motion, were ordered received and placed on file. They are published herewith. The Treasurer’s report showed a net cash balance of $1,555.51 on hand at the close of business, 30 November, 1912. On motion, this report was received and referred to the Finance Committee for auditing. The following candidates for honorary membership and fellowship, recommended by Council, were duly elected: Honorary MEMBERS. Jliya Metchnikof, Biologist and Bacteriologist, Pasteur Institute, Paris, _ France, presented by Dr. C.-E. A. Winslow. Sir John Murray, Oceanographer, Edinburgh, Scotland, presented - by Dr. F. A. Lucas. Sho Watasé, Zodlogist, Imperial University of Tokio, Japan, pre- sented by Prof. Henry E. Crampton. Frank D. Adams, Geologist, McGill pe aeuNe Montreal, presented by Prof. James F. Kemp. George E. Hale, Astronomer, Solar Observatory, California, pre- sented by Dr. John Tatlock. | FELLOWS. ' Felix Arnold, M. D., 824 St. Nicholas Avenue, New York. Dr. R. B. Earle, New York University, New York. 384. ANNALS NEW YORK ACADEMY OF SCIENCES Prof. Jesse E. Hyde, School of Mining, Kingston, Ontario. Prof. D. W. Johnson, Columbia University, New York. Dr. A. B. Pacini, 275 West 140th Street, New York. Mr. V. Stefansson, American Museum of Natural History, New York, Dr. F. Lyman Wells, McLean Hospital, Waverley, Massachusetts. The Academy then proceeded to the election of officers for the year 1912. The ballots prepared by the Council in accordance with the By- laws were distributed. On motion it was unanimously voted that Dr. Stevenson cast one ballot for the entire list nominated by the Council. This was done and they were declared elected, more than the requisite number of members and fellows entitled to vote being present: President, EMERSON MoMILLIN. - Vice-Presidents, J. EpmMunD WoopMaANn (Section of Geology and Mineralogy), W. D. MarrHrw (Section of Biology), CHarLes LANE Poor (Section of Astronomy, Physics and Chemistry), W. P. MontTacuE (Section of Anthropology and Psychology). Corresponding Secretary, Henry E. Crampron. Recording Secretary, EpMunD Otis Hovey. Treasurer, Henry L. DoHERTY. Librarian, RatpH W. Tower. Editor, EpMuND OT1Is Hovey. ~ Councilors (to serve 3 years), FrepERIc A. Lucas, R. S. WoopwortH. Finance Committee, Herson McMiuin, Freprric 8S. Les, G. F. KUNZ. At the close of the elections, Mr. Emerson McMillin gave his address as retiring President, in which, after reviewing the present condition of the Academy as derived from conference with a large number of the men who have long been active in carrying on its various lines of work, he made several recommendations regarding plans which might be adopted for enlarging the usefulness and interest of the organization and its meetings. Mr. V. Stefansson then gave a most interesting summary account of the expedition which he and Dr. R. M. Anderson made along the Arctic coast of western North America, from Point Barrow to Corona- tion Gulf during the years 1908-1912 inclusiye. At the close of his lecture, Mr. Stefansson outlined the plans of the second expedition which he is now organizing for geographical and ethnological work on Victoria, Banks and Prince Patricks Islands in the years 1913-1916 inclusive, and indicated the manner in which his expedition and the Crocker Land Expedition will supplement each other’s work. The Academy then adjourned. Epmunp Otis Hovey, Recording Secretary. . RECORDS OF MEETINGS 385 REPORT OF THE CORRESPONDING SECRETARY. We have lost by death during the past year the following Honorary Members: | Sir George H. Darwin, elected 1899. Sir Joseph D. Hooker, elected 1907. Franz Leydig, elected 1900. M. Jules H. Poincaré, elected 1900. Eduard Strasburger, elected 1908. Prof. Ferdinand Zirkel, elected 1904. and the following Corresponding Members: Paul Schweitzer, elected 1867. George Jarvis Brush, elected 1876. There are at present upon our rolls 44 Honorary Members and 125 Corresponding Members. Respectfully submitted, Henry EH. CRAMPTON, Corresponding Secretary. REPORT OF THE RECORDING SECRETARY. During the year 1912, the Academy held 8 business meetings and 26 sectional meetings, at which 65 ‘stated papers were presented as follows: Section of Geology and Mineralogy, 24 papers; Section of Biology, 16 papers ; Section of Astronomy, Physics and Chemistry, 6 papers; Section of Anthropology and Psychology, 19 papers. _ Seven public lectures have been given at the American Museum to the members of the Academy and the Affiliated Societies and their friends, as follows: “The Planet Mars.” By Edward E. Barnard. ° “Astronomy, Education and Culture.” By F. S. Archenhold. “Geology and Mineral Resources of Alaska.” By Alfred H. Brooks. “Sanitation of the Panama Canal.” By James F. B. Bowles. “Results of the Suture of Blood Vessels and the Transplantation of Organs.” By Alexis Carrel. “The Seismograph and What it Teaches.” By Harry Fielding Reid. “Experimental Evolution.” By Hugo de Vries. At the present time, the membership of the Academy is 488, which includes 468 Active Members, 22 of whom are Associate Members, 86 386 ANNALS NEW YORK ACADEMY OF SCIENCES Fellows, 90 Life Members and 11 Patrons and 20 Non-resident Mem- bers. There have been 10 deaths during the year, 12 resignations have become effective, 7 names have been dropped from the roll for non- payment of dues, 2 names have been transferred to the list of Non- Resident Members. Twelve new members have been elected during the year. As the membership of the Academy a year ago was 502, there has been a net loss of 14 during the year 1912. Announcement is made with regret of the loss by death of the following members: Col. John Jacob Astor, Active Member for 17 years. George Borup, Active Member for 4 months. Charles F. Cox, Active Member for 36 years. Prof. Morris Loeb, Active Member for 20 years. William Pennington, Active Member for 6 years. Edward Russ, Active Member for 5 years. Prof. John B. Smith, Active Member for 5 years. Isidor Straus, Active Member for 7 years. James Terry, Life Member for 30 years. Col. John Weir, Life Member for 6 years. Respectfully submitted, — EpMuND Otis Hovey, Recording Secretary. REPORT OF THE LIBRARIAN. The accessions to the library of the New York Academy of Sciences during the current year have been, by exchange and donation, 313 volumes and 1,670 numbers. Successful efforts have again been made to complete imperfect files of publications and special acknowledg- ments from the Academy are herewith extended to the Institutul Me- teorologic al Romaniei for the presentation of nineteen volumes (1892— 1911) ; to the Verein fiir Erdkunde zu Leipzig for the presentation of five volumes (1865-1871); to the Belfast Natural History and Philo- sophical Society for the presentation of seven volumes (1872-1881) and to the Société Nationale des Sciences Naturelles et Mathématiques de Cherbourg for the presentation of six volumes (1858-1864). I am also pleased to report that the books of the Academy have been more ex- tensively used than during any of the preceding years. Respectfully submitted, RaLtpu W. TowEr, — Labrarian. RECORDS OF MEETINGS 397 REPORT OF THE EDITOR. The Editor reports that during the past fiscal year there were issued pages 177-263 completing Volume XXI and pages 1-337 of Volume XXII. Respectfully submitted, Epmunp Oris Hovey, Editor. REPORT OF THE TREASURER. RECHIPTS. DECEMBER 1, 1911—NovEMBER 30, 1912. Cashvon hand December 1 19Id a, Se ooh. ce Soe chew eeew cae $1,356.58 MEME MPeLShip: KCCSsc eres sis sre ciao rere er eee lade axkiqee ors we eels 200.00 Income from investments: Interest on mortgages on New York City real estate.. $860.00 Interest on railroad and other bonds................ 1,375 .00 2,235.00 MMKELES FeO mes aia BATA CAG lars snsvene aise el oa iets See dns ee: 4% (eile coile) io -elrews whew euaneuseaue 48.29 Active membership dues, 1909...............20.. cee eeee $10.00 ff on Ce yA ret ee nn Opie ncarec Re nOn Gig Ss CRO ORE Roe 50.00 t me HO SION Lees Gagan ck ecceeci orctoeieaa isc ccna 245 .00 :; seamen T VINA anes crar terra haya! Sia S ences we feve tat ois 3,185.00 3,490.00 Associate membership dues, 1911..................2005- $6.00 se s Sra aan Dons nrekiay a iusnaveye creche SeMa esc 45.00 51.00 SALE SMO ses CALTOMS rapid ten seedoyevel Sik cieisiccei visieiene aise aie-ateersloieecmens 89.30 SUUSCLIPLONS MOPAMMUAlGINMeM see cases oe ce cles cle meee 178.00 PIU ete peter ewag eect ieiencuater te rehcirolel cers iesei sis cir: eleeiene/siae sve siotaileualacevs)s $7,648.17 DISBURSEMENTS. DECEMBER 1, 1911—-NoVEMBER 30, 1912. Publications on account of Annals... .. 5... 06cc.. ss ccenes corre see $1,511.01 IP MONCTON OIE. JVM Mee BOG B60 Boe ee Cae ce eons 597.75 Recording Secretary’S CXPeEMSES........ 0.00 c cee cece ee eee eee eeee 298 .55 Recording Secretary’s and Hditor’s allowance...........-.20.+-s. 1,200.00 Lecture Committee........ SAIN horse Nayar ateai'd ail, cig ates cleners Sue ateseens 200.00 GE CME AGSX CH SOSH eee a Hater reese neilalievorcine sais evenareitrsiw is nceseie sie veihs eselions 300.65 Esther Herrman Research Fund corte) AR ya aiage eteteleae Pasahs lata: sieneat ete c ete 800.00 AMT Al eIMee bine sad edinnM er sere eee ieis/ets ere cis © 41a o eusherereye.o) eer eisiel elec 184.70 AUCH AS CR Otem WOM Caer emae ye crseue ties Givelescieie Sc © s aidus le Ciayelieleieielewe (lol cusiolels 975.00 Interest charge on bond purchased.............--2.eeeeceecceeeee 5.00 Section of Geology and Mineralogy.............2eeeeeeeeeeceeeees 25.00 Cals Ro ripe lie leprae ar iey oe, LisaNova er sis Sidley etsie 6) Silsie oiaiotaris elev e alesis) saier evets 1,555.51 NOW oocedad cacobaaednoougeadoudboooDGUOCUdUaddUODODEU dE © $7,648 .17 888 ANNALS NEW YORK ACADEMY OF SCIENCES BALANCE SHEET, NOVEMBER 30, 1912. Investments (cost)........ $42,631.25 Permanent Fund.......... $22,812.57 Cash on hand............. 1,555.51 Publication Fund.......... 3,000.00 Audubon Fund............ 2,500.00 Hsther Herrman Research BUG ase hess hele ta ee 10,000.00 John Strong Newberry Ah) 010 eee IER IGS 6 0.6 1,000.00 Income Permanent Fund... 2,624.32 Income Audubon Fund..... 334.75 Income Hsther Herrman D000 Wat erae rere Hi als iS craic 1,465.96 Income Newberry Fund.... 449.16 $44,186.76 $44,186.76 Henry L. DOHERTY, Treasurer. 8 JANUARY, 1913. Examined and found to be correct, GEORGE F. KUNZ, FREDERIC S. LEE, Auditing Committee. THE ORGANIZATION OF THE NEW YORK ACADEMY OF SCIENCES THE ORIGINAL CHARTER AN ACT TO INCORPORATE THE LYCEUM OF NATURAL HISTORY IN THE CITY OF NEW YORK Passed A pril 20, 1818 WHerzEAS, The members of the Lyceum of Natural History have peti- tioned for an act of incorporation, and the Legislature, impressed with the importance of the study of Natural History, as connected with the wants, the comforts and the happiness of mankind, and conceiving it their duty to encourage all laudable attempts to promote the progress of science in this State—therefore, é: 1. Be tt enacted by the People of the State of New York represented in Senate and Assembly, That Samuel L. Mitchill, Casper W. Eddy, Fred- erick C. Schaeffer, Nathaniel Paulding, William Cooper, Benjamin P. Kissam, John Torrey, William Cumberland, D’Jurco V. Knevels, James Clements and James Pierce, and such other persons as now are, and may from time to time become members, shall be, and hereby are constituted a body corporate and politic, by the name of Lyceum oF Natura History IN THE City oF New York, and that by that name they shall have per- petual succession, and shall be persons capable of suing and being sued, pleaded and being impleaded, answering and being answered unto, de- fending and being defended, in all courts and places whatsoever ; and may have a common seal, with power to alter the same from time to time; and shall be capable of purchasing, taking, holding, and enjoying to them and their successors, any real estate in fee simple or otherwise, and any goods, chattels, and personal estate, and of selling, leasing, or otherwise dispos- ing of said real or personal estate, or any part thereof, at their will and pleasure: Provided always, that the clear annual value or income of such real or personal estate shall not exceed the sum of five thousand dollars: Provided, however, that the funds of the said Corporation shall be used and appropriated to the promotion of the objects stated in the preamble to this act, and those only. 2. And be wt further enacted, That the said Society shall from time to time, forever hereafter, have power to make, constitute, ordain, and estab- lish such by-laws and regulations as they shall judge proper, for the elec- (389) 390 ANNALS NEW YORK ACADEMY OF SCIENCES tion of their officers; for prescribing their respective functions, and the mode of discharging the same; for the admission of new members; for the — government of the officers and members thereof; for collecting annual contributions from the members towards the funds thereof; for regulat- ing the times and places of meeting of the said Society; for suspending or expelling such members as shall neglect or refuse to comply with the by-laws or regulations, and for the managing or directing the affairs and concerns of the said Society: Provided such by-laws and regulations be not repugnant to the Constitution and laws of this State or of the United States. 3. And be tt further enacted, That the officers of the said Society shall consist of a President and two Vice-Presidents, a Corresponding Secre- tary, a Recording Secretary, a Treasurer, and five Curators, and such other officers as the Society may judge necessary; who shall be annually chosen, and who shall continue in office for one year, or until others be elected in their stead; that if the annual election shall not be held at any of the days for that purpose appointed, it shall be lawful to make such election at any other day; and that five members of the said Society, assembling at the place and time designated for that purpose by any by- law or regulation of the Society, shall constitute a legal meeting thereof. 4. And be it further enacted, That Samuel L. Mitchill shall be the President; Casper W. Eddy the First Vice-President; Frederick C. Schaeffer the Second Vice-President; Nathaniel Paulding, Correspond- ing Secretary; William Cooper, Recording Secretary; Benjamin P. Kis- sam, Treasurer, and John Torrey, William Cumberland, D’Jurco V. Knevels, James Clements, and James Pierce, Curators; severally to be the first officers of the said Corporation, who shall hold their respective offices until the twenty-third day of February next, and until others shall be chosen in their places. 5. And be it further enacted, That the present Constitution of the said Association shall, after passing of this Act, continue to be the Constitu- tion thereof; and that no alteration shall be made therein, unless by a vote to that effect of three-fourths of the resident members, and upon the request in writing of one-third of such resident members, and submitted at least one month before any vote shall be taken thereupon. State of New York, Secretary's Office. I certiry the preceding to be a true copy of an original Act of the Legislature of this State, on file in this Office. ARCH’D CAMPBELL, ALBANy, April 29, 1818. Dep. Sec’y. ORGANIZATION | 39] ORDER OF COURT ORDER OF THE SUPREME COURT OF THE STATE OF NEW YORK TO CHANGE THE NAME OF THE LYCEUM OF NATURAL HISTORY IN THE CITY OF NEW YORK TO THE NEW YORK ACADEMY OF SCIENCES WHEREAS, in pursuance of the vote and proceedings of this Corpora- tion to change the corporate name thereof from “The Lyceum of Natural History in the City of New York” to “The New York Academy of Sci- ences,” which vote and proceedings appear to record, an application has been made in behalf of said Corporation to the Supreme Court of the State of New York to legalize and authorize such change, according to the statute in such case provided, by Chittenden & Hubbard, acting as . the attorneys of the Corporation, and the said Supreme Court, on the 5th day of January, 1876, made the following order upon such application in the premises, viz: At a special term of the Supreme Court of the State of New York, held at the Chambers thereof, in the County Court House, in the City of New York, the 5th day of January, 1876: Present—Hon. Gro. C. BARRETT, Justice. In the matter of the application of the Lyceum of Natural History in the City of New York to au- thorize it to assume the corporate name of the New York Academy of Sciences. On reading and filing the petition of the Lyceum of Natural History in the City of New York, duly verified by John S. Newberry, the Presi- dent and chief officer of said Corporation, to authorize it to assume the corporate name of the New York Academy of Sciences, duly setting forth 392 ANNALS NEW YORK ACADEMY OF SCIENCES the grounds of said application, and on reading and filing the affidavit of Geo. W. Quackenbush, showing that notice of such application had been duly published for six weeks in the State paper, to wit, The Albany Evening Journal, and the affidavit of David 8S. Owen, showing that notice of such application has also been duly published in the proper newspaper of the County of New York, in which county said Corporation had its business office, to wit, in The Daily Register, by which it appears to my satisfaction that such notice has been so published, and on reading and filing the affidavits of Robert H. Browne and J. 8. Newberry, thereunto annexed, by which it appears to my satisfaction that the application is made in pursuance of a resolution of the managers of said Corporation to that end named, and there appearing to me to be no reasonable objection to said Corporation so changing its name as prayed in said petition: Now on motion of Grosvenor 8. Hubbard, of Counsel for Petitioner, it is Ordered, That the Lyceum of Natural History in the City of New York be and is hereby authorized to assume the corporate name of The New York Academy of Sciences. : Indorsed: Filed January 5, 1876, : A copy. Wm. WatsH, Clerk. Resolution of THE ACADEMY, accepting the order of the Court, passed February 21, 1876 And whereas, The order hath been published as therein required, and all the proceedings necessary to carry out the same have been had, There- fore: Resolved, That the foregoing order be and the same is hereby accepted and adopted by this Corporation, and that in conformity therewith the corporate name thereof, from and after the adoption of the vote and reso- lution herein above referred to, be and the same is hereby declared to be THE NEW YORK ACADEMY OF SCIENCES. AMENDED CHARTER Marcy 19, 1902 CHAPTER 181 oF THE Laws oF 1902 Aw Act to amend chapter one hundred and ninety-seven of the laws of eighteen hundred and eighteen, entitled “An act to incorporate the Ly- ceum of Natural History in the City of New York,’ a Corporation now known as The New York Academy of Sciences and to extend the powers of said Corporation. . ORGANIZATION 393 (Became a law March 19, 1902, with the approval of the Governor. Passed, three-fifths being present.) The People of the State of New York, represented in Senate and As- sembly, do enact as follows: Section I. The Corporation incorporated by chapter one hundred and ninety-seven of the laws of eighteen hundred and eighteen, entitled “An act to incorporate the Lyceum of Natural History in the City of New York,” and formerly known by that name, but now known as The New York Academy of Sciences through change of name pursuant to order made by the supreme court at the city and county of New York, on January fifth, eighteen hundred and seventy-six, is hereby authorized and _ empowered to raise money for, and to erect and maintain, a building in the city of New York for its use, and in which also at its option other scientific societies may be admitted and have their headquarters upon such terms as said Corporation may make with them, portions of which building may be also rented out by said Corporation for any lawful uses for the purposes of obtaining income for the maintenance of such build- ing and for the promotion of the objects of the Corporation ; to establish, own, equip, and administer a public library, and a museum having es- pecial reference to scientific subjects; to publish communications, trans- actions, scientific works, and periodicals; to give scientific instruction by lectures or otherwise; to encourage the advancement of scientific research and discovery, by gifts of money, prizes, or other assistance thereto. The building, or rooms, of said Corporation in the City of New York used exclusively for library or scientific purposes shall be subject to the pro- visions and be entitled to the benefits of subdivision seven of section four of chapter nine hundred and eight of the laws of eighteen hundred and ninety-six, as amended. Section II. The said Corporation shall from time to time forever hereafter have power to make, constitute, ordain, and establish such by- laws and regulations as it shall judge proper for the election of its officers ; for prescribing their respective functions, and the mode of discharging the same; for the admission of new members; for the government of offi- cers and members thereof; for collecting dues and contributions towards the funds thereof; for regulating the times and places of meeting of said Corporation ; for suspending or expelling such members as shall neglect or refuse to comply with the by-laws or regulations, and for managing or. directing the affairs or concerns of the said Corporation: and may from time to time alter or modify its constitution, by-laws, rules, and regula- tions. 394. ANNALS NEW YORK ACADEMY OF SCIENCES Section III. The officers of the said Corporation shall consist of a president and two or more vice-presidents, a corresponding secretary, a recording secretary, a treasurer, and such other officers as the Corporation may judge necessary; who shall be chosen in the manner and for the terms prescribed by the constitution of the said Corporation. Srction IV. The present constitution of the said Corporation shall, after the passage of this act, continue to be the constitution thereof until amended as herein provided. Such constitution as may be adopted by a vote of not less than three-quarters of such resident members and fellows of the said New York Academy of Sciences as shall be present at a meet- ing thereof, called by the Recording Secretary for that purpose, within forty days after the passage of this act, by written notice duly mailed, postage prepaid, and addressed to each fellow and resident member at least ten days before such meeting, at his last known place of residence, with street and number when known, which meeting shall be held within three months after the passage of this act, shall be thereafter the consti- tution of the said New York Academy of Sciences, subject to alteration or amendment in the manner provided by such constitution. Suction V. The said Corporation shall have power to consolidate, to unite, to co-operate, or to ally itself with any other society or association in the city of New York organized for the promotion of the knowledge or the study of any science, or of research therein, and for this purpose to receive, hold, and administer real and personal property for the uses of such consolidation, union, co-operation, or alliance subject to such terms and regulations as may be agreed upon with such associations or societies. Section VI. This act shall take effect immediately. STATE OF New YORK, OFFICE OF THE SECRETARY OF STATE. I have compared the preceding with the original law on file in this office, and do hereby certify that the same is a correct transcript there- from, and the whole of said original law. Given under my hand and the seal of office of the Secretary of State, at the city of Albany, this eighth day of April, in the year one thousand nine hundred and two. JoHN T. McDonoucH, | Secretary of State. ORGANIZATION 395 CONSTITUTION -ApopTEeD, APRIL 24, 1902, AND AMENDED AT SUBSEQUENT TIMES ArticLE I. The name of this Corporation shall be The New York Academy of Sciences. Its object shall be the advancement and diffusion of scientific knowledge, and the center of its activities shall be in the City of New York. ArticLE II. The Academy shall consist of five classes of members, namely: Active Members, Fellows, Associate Members, Corresponding Members and Honorary Members. Active Members shall be the members of the Corporation who live in or near the City of New York, or who, having removed to a distance, desire to retain their connection with the Academy. Fellows shall be chosen from the Active Members in virtue of their scientific attainments. Corresponding and Honorary Members shall be chosen from among persons who have attained distinction in some branch of science. The number of Corresponding Members shall not exceed two hundred, and the number of Honorary Members shall not exceed fifty. -ARTICLE III. None but Fellows and Active Members who have paid their dues up to and including the last fiscal year shall be entitled to vote or to hold office in the Academy. ' ArticLtE IV. The officers of the Academy shall be a President, as many Vice-Presidents as there are sections of the Academy, a Correspond- ing Secretary, a Recording Secretary, a Treasurer, a Librarian, an Editor, six elected Councilors and one additional Councilor from each allied society or association. ‘The annual election shall be held on the third Monday in December, the officers then chosen to take office at the first meeting in January following. There shall also be elected at the same time a Finance Committee of three. ARTICLE V. ‘The officers named in Article IV shall constitute a Coun- cil, which shall be the executive body of the Academy with general control over its affairs, including the power to fill ad interim any vacancies that may occur in the offices. Past Presidents of the Academy shall be ez- officio members of the Council. ARTICLE VI. Societies organized for the study of any branch of science may become allied with the New York Academy of Sciences by consent of the Council. Members of allied societies may become Active Members of the Academy by paying the Academy’s annual fee, but as 396 ANNALS NEW YORK ACADEMY OF SCIENCES members of an allied society they shall be Associate Members with the rights and privileges of other Associate Members, except the receipt of its publications. Each allied society shall have the right to delegate one of its members, who is also an Active Member of the Academy, to the Council of the Academy, and such delegate shall have all the rights and privileges of other Councilors. ARTICLE VII. The President and Vice-Presidents shall not be eligible to more than one re-election until three years after retiring from office; the Secretaries and Treasurer shall be eligible to re-election without limitation. The President, Vice-Presidents and Secretaries shall be Fel- lows. The terms of office of elected Councilors shall be three years, and these officers shall be so grouped that two, at least one of whom shall be a Fellow, shall be elected and two retired each year. Councilors shall not be eligible to re-election until after the expiration of one year. ArTICLE VIII. The election of officers shall be by ballot, and the can- didates having the greatest number of votes shall be declared duly elected. ARTICLE [X. Ten members, the majority of whom shall be Fellows, shall form a quorum at any meeting of the Academy at which business is transacted. ARTICLE X. The Academy shall establish by-laws, and may amend them from time to time as therein provided. ARTICLE XI. This Constitution may be amended by a vote of not less than three-fourths of the fellows and three-fourths of the active members present and voting at a regular business meeting of the Academy, pro- vided that such amendment shall be publicly submitted in writing at the preceding business meeting, and provided also that the Recording Secre-— tary shall send a notice of the proposed amendment at least ten days before the meeting, at which a vote shall be taken, to each Fellow and Active Member entitled to vote. BY-LAWS 4 As ADOPTED, OCTOBER 6, 1902, AND AMENDED AT SUBSEQUENT TIMES CHapTer I OFFICERS 1. President. It shall be the duty of the President to preside at the business and special meetings of the Academy; he shall exercise the cus- tomary duties of a presiding officer. 2. Vice-Presidents. In the absence of the President, the senior Vice- President, in order of Fellowship, shall act as the presiding officer. sae ORGANIZATION 397 3. Corresponding Secretary. The Corresponding Secretary shall keep a corrected list of the Honorary and Corresponding Members, their titles and addresses, and shall conduct all correspondence with them. He shall make a report at the Annual Meeting. 4. Recording Secretary. The Recording Secretary shall keep the minutes of the Academy proceedings; he shall have charge of all docu- ments belonging to the Academy, and of its corporate seal, which he shall affix and attest as directed by the Council; he shall keep a corrected list of the Active Members and Fellows, and shall send them announcements of the Meetings of the Academy ; he shall notify all Members and Fellows of their election, and committees of their appointment; he shall give notice to the Treasurer and to the Council of matters requiring their action, and shall bring before the Academy business presented by the Council. He shall make a report at the Annual Meeting. 5. Treasurer. The Treasurer shall have charge, under the direction of the Council, of all moneys belonging to the Academy, and of their investment. He shall receive all fees, dues and contributions to the Academy, and any income that may accrue from property or investment ; he shall report to the Council at its last meeting before the Annual Meet- ing the names of members in arrears; he shall keep the property of the Academy insured, and shall pay all debts against the Academy the dis- charge of which shall be ordered by the Council. He shall report to the Council from time to time the state of the finances, and at the Annual Meeting shall report to the Academy the receipts and expenditures for the entire year. 6. Librarian. The Librarian shall have charge of the library, under the general direction of the Library Committee of the Council, and shall conduct all correspondence respecting exchanges of the Academy. He ‘shall make a report on the condition of the library at the Annual Meeting. | %. Editor. The editor shall have charge of the publications of the Academy, under the general direction of the Publication Committee of the Council. He shall make a report on the condition of the publications at the Annual Meeting. CHAPTER II COUNCIL 1. Meetings. The Council shall meet once a month, or at the call of the President. It shall have general charge of the affairs of the Academy. 2. Quorum. Five members of the Council shall constitute a quorum. 3. Officers. The President, Vice-Presidents and Recording Secretary of the Academy shall hold the same offices in the Council. 398 ANNALS NEW YORK ACADEMY OF SCIENCES 4. Committees. The Standing Committees of the Council shall be: (1) an Executive Committee consisting of the President, Treasurer, and Recording Secretary; (2) a Committee on Publication; (3) a Committee on the Library, and such other committees as from time to time shall be authorized by the Council. The action of these committees shall be sub- ject to revision by the Council. Cuapter III FINANCE COMMITTEE The Finance Committee of the Academy shall audit the Annual Report of the Treasurer, and shall report on financial questions whenever called upon to do so by the Council. | CHaAptTerR IV ELECTIONS 1. Active Members. (a) Active Members shall be nominated in writ- ing to the Council by at least two Active Members or Fellows. If ap- proved by the Council, they may be elected at the succeeding business meeting. (6) Any Active Member who, eine removed to a distance ise the city of New York, shall nevertheless express a desire to retain his connec- tion with the Academy, may be placed by vote of the Council on a list of Non-Resident Members. Such members shall relinquish the full privi- leges and obligations of Active Members. (Vide Chapters V and X.) 2. Associate Members. Workers in science may be elected to Associate Membership for a period of two years in the manner prescribed for Active ~ Members. They shall not have the power to vote and shall not be eligible to election as Fellows, but may receive the publications, At any time sub- sequent to their election they may assume the full privileges of Active Members by paying the dues of such Members. 3. Fellows, Corresponding Members and Honorary Members. Nomi- nations for Fellows, Corresponding Members and Honorary Members may be made in writing either to the Recording Secretary or to the Council at its meeting prior to the Annual Meeting. If approved by the Council, the nominees shall then be balloted for at the Annual Meeting. 4, Officers. Nominations for Officers, with the exception of Vice- Presidents, may be sent in writing to the Recording Secretary, with the name of the proposer, at any time not less than thirty days before the Annual Meeting. Hach section of the Academy shall nominate a candi- ORGANIZATION 399 date for Vice-President, who, on election, shall be Chairman of the sec- tion ; the names of such nominees shall be sent to the Recording Secretary properly certified by the sectional secretaries, not less than thirty days before the Annual Meeting. The Council shall then prepare a list which shall be the regular ticket. This list shall be mailed to each Active Mem- ber and Fellow at least one week before the Annual Meeting. But any Active Member or Fellow entitled to vote shall be entitled to prepare and vote another ticket. CHAPTER V DUES 1. Dues. The annual dues of Active Members and Fellows shall be $10, payable in advance at the time of the Annual Meeting; but new members elected after May 1, shall pay $5 for the remainder of the fiscal year. . The annual dues of elected Associate Members shall be $3, payable in advance at the time of the Annual Meeting. Non-Resident Members shall be exempt from dues, so long as they shall relinquish the privileges of Active Membership. (Vide Chapter X.) 2. Members in Arrears. If any Active Member or Fellow whose dues remain unpaid for more than one year, shall neglect or refuse to pay the same within three months after notification by the Treasurer, his name may be erased from the rolls by vote of the Council. Upon payment of his arrears, however, such person may be restored to Active Membership or Fellowship by vote of the Council. 3. Renewal of Membership. Any Active Member or Fellow who shall resign because of removal to a distance from the city of New York, or any Non-Resident Member, may be restored by vote of the Council to Active Membership or Fellowship at any time upon application. CHAPTER VI PATRONS, DONORS AND LIFE MEMBERS 1. Patrons. Any person contributing at one time $1,000 to the general funds of the Academy shall be a Patron and, on election by the Council, shall enjoy all the privileges of an Active Member. 2. Donors. Any person contributing $50 or more annually to the general funds of the Academy shall be termed a Donor and, on election by the Council, shall enjoy all the privileges of an Active Memter. 3. Life Members. Any Active Member or Fellow contributing at one time $100 to the general funds of the Academy shall be a Life Member 400 ANNALS NEW YORK ACADEMY OF SCIENCES and shall thereafter be exempt from annual dues; and any Active Mem- ber or Fellow who has paid annual dues for twenty-five years or more may, upon his written request, be made a life member and be exempt from further payment of dues. CHaptTer VII SECTIONS 1. Sections. Sections devoted to special branches of Science may be established or discontinued by the Academy on the recommendation of the Council. The present sections of the Academy are the Section of Astronomy, Physics and Chemistry, the Section of Biology, the Section of Geology and Mineralogy and the Section of Anthropology and Psy- chology. 2. Organization. Tach section of the Academy shall have a Chairman and a Secretary, who shall have charge of the meetings of their Section. The regular election of these officers shall take place at the October or November meeting of the section, the officers then chosen to take office at the first meeting in January following. 3. Affiliation. Members of scientific societies affiliated with the Academy, and members of the Scientific Alliance, or men of science intro- duced by members of the Academy, may attend the meetings and present papers under the general regulations of the Academy. CHapter VIII MEETINGS 1. Business Meetings. Business meetings of the Academy shall be held on the first Monday of each month from October to May inclusive. 2. Sectional Meetings. Sectional meetings shall be held on Monday evenings from October to May inclusive, and at such other times as the Council may determine. The sectional meeting shall follow the business meeting when both occur on the same evening. 3. Annual Meeting. The Annual Meeting shall be held on the third Monday in December. 4. Special Meetings. A special meeting may be called by the Council, provided one week’s notice be sent to each Active Member and Fellow, stating the object of such meeting. ORGANIZATION 401 CHAPTER IX ORDER OF BUSINESS 1. Business Meetings. The following shall be the order of procedure at business meetings: . 1. Minutes of the previous business meeting. 2. Report of the Council. 3. Reports of Committees. 4, Elections. 5. Other business. 2. Sectional Meetings. The following shall be the order of procedure at sectional meetings: 1. Minutes of the preceding meeting of the section. 2. Presentation and discussion of papers. 3. Other scientific business. 3. Annual Meetings. The following shall be the order of procedure at Annual Meetings: | 1. Annual reports of the Corresponding Secretary, Recording Secretary, Treasurer, Librarian, and Editor. 2. Hlection of Honorary Members, Corresponding Members, and Fellows. Election of officers for the ensuing year. 4. Annual address of the retiring President. So CHAPTER X PUBLICATIONS 1. Publications. The established publications of the Academy shall be the Annals and the Memoirs. They shall be issued by the Editor under the supervision of the Committee on Publications. 2. Distribution. One copy of all publications shall be sent to each Patron, Life Member, Active Member and Fellow; provided, that upon inquiry by the Editor such Members or Fellows shall signify their desire to receive them. ; 3. Publication Fund. Contributions may be received for the publica- tion fund, and the income thereof shall be applied toward defraying the expenses of the scientific publications of the Academy. 402 ANNALS NEW YORK ACADEMY OF SCIENCES CHAPTER XT GENERAL PROVISIONS 1. Debts. No debts shall be incurred on behalf of the Academy, unless authorized by the Council. 2. Bills. All bills submitted to the Council must be certified as to correctness by the officers incurring them. 3. Investments. All the permanent funds of the ‘Weadete shall be invested in United States or in New York State securities or in first mortgages on real estate, provided they shall not exceed sixty-five per cent. of the value of the property, or in first-mortgage bonds of corpora- tions which have paid dividends continuously on their common stock for a period of not less than five years, All income from patron’s fees, life- membership fees and donor’s fees shall be added to the permanent fund. 4. Expulsion, etc. Any Member or Fellow may be censured, sus- pended or expelled, for violation of the Constitution or By-Laws, or for any offence deemed sufficient, by a vote of three-fourths of the Members and three-fourths of the Fellows present at any business meeting, provided such action shall have been recommended by the Council at a previous business meeting, and also, that one month’s notice of such recommenda- tion and of the offence charged shall have been given the Member accused. 5. Changes in By-Laws. No alteration shall be made in these By- Laws unless it shall have been submitted publicly in writing at a business meeting, shall have been entered on the Minutes with the names of the Members or Fellows proposing it, and shall be adopted by two-thirds of the Members and Fellows present and voting at a subsequent business meeting. MEMBERSHIP OF THE NEW YORK ACADEMY OF SCIENCES HONORARY MEMBERS 31 DECEMBER, 1912. ELECTED. 1912. Frank D. Apams, Montreal, Canada. 1898. ArtTHuR Auwenrs, Berlin, Germany. 1889. CHARLES Barros, Lille, France. 1907. Wrtu1am Bateson, Cambridge, England. 1910. THropor Boveri, Wirzburg, Germany. 1901. CHaRLES VERNON Boys, London, England. 1904. W. C. Broécer, Christiana, Norway. 1911. Herrmann Crepner, Leipzig, Germany. 1876. W. Boyp Dawkins, Manchester, England. 1902. Sir JAmes Dewar, Cambridge, England. ° 1901: Emit Fiscuer, Berlin, Germany. 1876. Sir Arcuipatp Gurxre, Haslemere, Surrey, England. 1901. James Gerxiz, Edinburgh, Scotland. 1898. Sir Davip Grux, London, England. 1909. K. F. GéseLt, Munich, Germany. 1889. GrorcE LINCOLN GoopaLE, Cambridge, Mass. 1909. PauL von GrotH, Munich, Germany. ’ 1894. Ernst HAcKEL, Jena, Germany. 1912. Gzorce EH. Hate, Mt. Wilson, Calif. 1899. JuLius Hann, Vienna, Austria. 1898. GrorcE W. Hitt, West Nyack, N. Y. 1896. Amsrosius A. W. Husrecut, Utrecht, Netherlands. 1896. Ferix Kiser, Gottingen, Germany. 1909. Atrrep Lacrorx, Paris, France. 1876. ViKToR von Lane, Vienna, Austria. 1898. EH. Ray Lanxester, London, England. 1880. Sir Norman Lockyer, London, England. 1911. Ernst Macu, Vienna, Austria. 1912. Intya Metrounikor, Paris, France, 1912. Sir Jonw Murray, Edinburgh, Scotland. 1898. Friptyor Nansen, Christiana, Norway. 1908. WILHELM OstTWALD, Gross-Bothen, Germany. 1898. ALBRECHT PxENcK, Berlin, Germany. (403) 404 ANNALS NEW YORK ACADEMY OF SCIENCES wee ELECTED. 1898. WILHELM PFEFFER, Leipzig, Germany. 1900. EpwarpD CHARLES PICKERING, Cambridge, Mass. 1911. Epwarp BacnaLu Poutton, Oxford, England. 1901. Sir Witu1Am Ramsay, London, England. 1899. Lord RayLereH, Witham, Essex, England. 1898. Hans H. Reuscu, Christiania, Norway. 1887. Sir Henry Enrretp Roscoz, London, England. 1887. HzrtnricH Rosensuscu, Heidelberg, Germany. 1912. SHo Warass, Tokyo, Japan. 1904. KARL VON DEN STEINEN, Berlin, Germany. 1896. JosEPpH JoHN THomson, Cambridge, England. 1900. Epwarp Burnett Tytor, Oxford, England. 1904. Huco DE Vrizs, Amsterdam, Netherlands. 1907. James WarpD, Cambridge, England. 1909. AuGcusT WEISSMANN, Freiburg, Germany. 1904. WiLHELM WuwnpT, Leipzig, Germany. CORRESPONDING MEMBERS 31 DECEMBER, 1912. 1883. CHARLES ConraD ABport, Trenton, N. J. 1891. Jost G. AcuitERa, Mexico City, Mexico. 1890. Witi1am De Wirr ALEXANDER, Honolulu, Hawaii. 1899. CO. W. AnprEws, London, England. 1876. JoHN Howarp ApPLeton, Providence, R. I. 1899. J. G. Baker, Kew, England. 1898. Isaac BactEy Batrour, Edinburgh, Scotland. 1878. ALEXANDER GRAHAM BELL, Washington, D. C. 1867. Hpwarp L. BrertHoup, Golden, Colo. 1897. Hersert Bouton, Bristol, England. 1899. G. A. BouLencrr, London, England. 1874. T. 8. Branpscee, Berkeley, Calif. 1884. JoHN C. Branner, Stanford University, Calif. 1894. BonustAy Brauner, Prague, Bohemia. 1874. Wrii1am Brewster, Cambridge, Mass. 1898. T. C. CHAMBERLIN, Chicago, II]. 1876. FRANK WIGGLESWORTH CLARKE, Washington, D. C. 1891. L. Cuero, Ekaterinburg, Russia. 1877. THEoporEe B. Comstock, Los Angeles, Calif. ELECTED. 1868. 1876. 1880. LSI. 1895. SiC9. 1870. 1885. 1898. 1894. 1899. 1890. 1899. 1876. 1880. 1869. 1879. 1879. 1885. 1899. 1879. 1870. 1858. 1865. 1888. 1868. 1883. 1869. 1898. 1882. 1867. 1900. 1890. 1896. 1875. 1899. 1876. 1876. 1888. 1876. MEMBERSHIP M. C. Cooxs, London, England. H. B. CoRNWALL, Princeton, N. J. CHARLES B. Cory, Boston, Mass. JOSEPH CRAWForD, Philadelphia, Pa. Henry P. CusHine, Cleveland, O. T. Newtson Date, Pittsfield, Mass. Wiuii1am Hearty Datu, Washington, D. C. EpwWArD SALispurY Dana, New Haven, Conn. WitiiamM M. Davis, Cambridge, Mass. RuTHVEN DeEANzE, Chicago, Il. CHARLES D&PERET, Lyons, France. ‘ORVILLE A. DrrBy, Rio de Janeiro, Brazil. Louis Dotto, Brussels, Belgium. Henry W. Evuiorr, Lakewood, O. JOHN B. Exxiort, Tulane Univ., La. Francis EK. ENGELHARDT, Syracuse, N. Y. Herman Le Roy FarrcHitp, Rochester, N. Y. FRIEDRICH BERNHARD Firrica, Marburg, Germany. Lazarus FLetcHER, London, England. EBERHARD FRAAS, Stuttgart, Germany. REINHOLD FRITZGARTNER, Tegucigalpa, Honduras. Grove K. GILBERT, Washington, D. C. THEODORE NICHOLAS GILL, Washington, D. C. CuHarLes A. Gorssman, Amherst, Mass. Frank Austin Goocu, New Haven, Conn. C. R. GREENLEAF, San Francisco, Calif. Marquis ANTONIO DE GREGORIO, Palermo, Sicily. R. J. LecHMErE Guppy, Trinidad, British West Indies. GrorcE EH. Hatz, Mt. Wilson, Calif. Baron Ernst von Husse-Wartece, Lucerne, Switzerland. C. H. HitcxHcock, Honolulu, H. I. : WiLL1AmM Henry Houmeus, Washington, D. C. H. D. Hosxoup, Buenos Ayres, Argentine Republic. J. P. Ippines, Brinklow, Md. Matvern W. Itzs, Dubuque, Ia. Orto JAKEL, Greifswald, Germany. Davip STaRR JoRDAN, Stanford University, Calif. Grorce A. Kornic, Houghton, Mich. Baron R. Kuxt, Tokyo, Japan. JoHN W. LANGLEY, Cleveland, O. 405 406 ANNALS NEW YORK ACADEMY OF SCIENCES EXLECTED. 1876. S. A. Lartrmore, Rochester, N. Y. 1894. Wuti1am Lipsey, Princeton, N. J. 1899. ARCHIBALD LiversipGE, London, England. 1876. GrorGE Mactoskiz#, Princeton, N. J. 1876. JOHN WILLIAM MALLET, Charlottesville, Va. 1891. CHarLes Rrsore Mann, Chicago, III. 1867. Gzrorce F. MartrHew, St. John, N. B., Canada. 1874. CHARLES JOHNSON Maynarp, West Newton, Mass. 1874. THroporE LuquEER Meap, Oviedo, Fla. 1888. SrrH EH. Mesx, Chicago, Il. 1892. J. DE Menp1zABAaL-TAMBORREL, Mexico City, Mexico. 1874. Ciinton Hart Merriam, Washington, D. C. 1898. MANSFIELD MrrRIAM, South Bethlehem, Pa. 1878. CHARLES SEDGwiIcK Minot, Boston, Mass. 1876. WiLt1am GILBERT MixteR, New Haven, Conn. 1890. RicHarD MoLpENKE, Watchung, N. J. 1895. C. Luoyp Morea, Bristol, England. 1864. Epwarp S. Morss, Salem, Mass. 1898. GrorGE Murray, London, England. ——. Hucren NerTo, Giessen, Germany. 1866. ALFreD Newton, Cambridge, England. 1897. Francis C. NicHotas, New York, N. Y. 1882. Henry ALFRED ALForD NICHOLLS, Dominica, B. W. i, 1880. Epwarp J. Nouan, Philadelphia, Pa. 1879. FREDERICK A. OBER, Hackensack, N. J. 1876. Joun M. Orpway, New Silence, lian 1900. Grorce Howarp Parker, Cambridge, Mass. 1876. StepHEN F. PeckHAm, New York, N. Y. 1877. FREDERICK PRIME, Philadelphia, Pa. 1868. RAPHAEL PuMPELLY, Newport, R. I. 1876. B. ALEx. RANDALL, Philadelphia, Pa. 1876. Ira RemsSEN, Baltimore, Md. 1874. Ropert Ripeway, Washington, D. C. 1886. Witiiam L. Ross, Troy, N. Y. 1876. Samuet P. Sapruer, Philadelphia, Pa. 1899. D. Max Scutossrr, Munich, Germany. 1898. W.B. Scort, Princeton, N. J. 1894. W. T. Sepewicr, Boston, Mass. 1876. ANDREW SHERWOOD, Portland, Ore. 1883. J. Warp SmitH, Newark, N. J. ELECTED. 1895. 1890. 1896. 1890. 1876. 1885. 1893. 1899. 1877. 1876. issiral 1900. 1867. 1890. 1898. 1876. 1897. 1874. 1898. 1898. 1866. 1899. 1876. 1876. MEMBERSHIP 407 CHARLES H. SmytTu, Jr., Princeton, N. J. J. SELDEN SPENCER, Tarrytown, N. Y. Rospert STEARNS, Los Angeles, Calif. WaLTER LE CONTE STEVENS, Lexington, Va. Francis H. Storer, Boston, Mass. Rajah Sourrtnpro Mouun Tacore, Calcutta, India. J. P. THomson, Brisbane, Queensland, Australia. R. H. Traqguatr, Colinton, Scotland. JOHN TROWBRIDGE, Cambridge, Mass. D. K. Tutti, Philadelphia, Pa. Henri Van Heurcx, Antwerp, Belgium. CHARLES R. VAN Hiss, Madison, Wis. ADDISON EMERY VERRILL, New Haven, Conn. ANTHONY WAYNE VoebEs, San Diego, Calif. CHARLES DoOLITTLE WaALcott, Washington, D. C. Lronarp Watpo, New York, N. Y. Stuart WELLER, Chicago, II. I. C. Wuite, Morgantown, W. Va. Henry SHALER W1LLIAMS, Ithaca, N. Y. N. H. WInNcHELL, Minneapolis, Minn. Horatio C. Woop, Philadelphia, Pa. A. SmitH Woopwakrp, London, England. ARTHUR WILLIAMS WricHt, New Haven, Conn. Harry Cricy Yarrow, Washington, D. C. 408 ANNALS NEW YORK ACADEMY OF SCIENCES ACTIVE MEMBERS 1912 Fellowship is indicated by an asterisk (*) before the name; Life Mem- bership, by a dagger ({) ;.Patronship, by a section mark (§). *Abbe, Dr. Cleveland Abercrombie, David T. +Adams, Edward D. Agens, F. G., Sr. tAlexander, Chas. B. *Allen, J. A., Ph.D. Allen, James Lane *+ Allis, Edward Phelps, Jr., Ph.D. Ames, Oakes Anderson, A. A. Anderson, A. J.C. *+ Andrews, Roy C. + Anthony, R. A. Arctowski, Dr. Henryk Arend, Francis J. +Armstrong, 8. T., M.D. *Arnold, Felix, M.D. Ashby, George E. Astor, John Jacob? Avery, Samuel P. +Bailey, James M. + Barhydt, Mrs. P. H. *Barnhart, John Hendley Barron, George D. *Baskerville, Prof. Charles Baugh, Miss M. L. Beal, William R. *t Beck, Fanning C. T. *Beebe, C. William Beller, A. + Bergstresser, Charles M. *Berkey, Charles P., Ph.D. Betts, Samuel R. 1 Deceased. van Beuren, F. T. *Bickmore, Albert 8., Ph.D. *Bigelow, Prof. Maurice A., Ph.D. Bigelow, William 8. | Bijur, Moses {Billings, Miss Elizabeth Billings, Frederick Bishop, Heber R. Bishop, Miss Mary C. Bishop, Samuel H. *Blake, J. A., M.D. , *+Bliss, Prof. Charles B. +Blumenthal, George —*Boas, Prof. Franz Boettger, Henry W. Bohler, Richard F. Borup, George? +Bourn, W. B. Boyd, James Brinsmade, Charles Lyman *Bristol, Prof. Charles L. Bristol, Jno. I. D. *SBritton, Prot, Ne ii ene: *$Brown, Hon. Addison Brown, Edwin H. Browne, T. Quincy *Brownell, Silas B., LL.D. Bulkley, L. Duncan Burr, Winthrop *Bush, Wendell T. Byrne, Joseph, M.D. *Byrnes, Miss Esther F., Ph.D. Camp, Frederick A. MEMBERSHIP *Campbell, Prof. William, Ph.D. *Campbell, Prof. William M. Canfield, R. A. . Cannon, J. G. Carlebach, Walter Maxwell *SCasey, Col. T. L., U.S. A. Cassard, William J. Cassebeer, H. A., Jr. *+Cattell, Prof. J. McKeen, Ph.D. aeuandler Prot. ©. H PhD: §Chapin, Chester W. *Chapman, Frank M. +Chaves, José BE. *Cheesman, Timothy M., M.D. Childs, Wm., Jr. Chubb, Percy Clarkson, Banyer Cline, M. Hunt +Clyde, Wm. P. Cohn, Julius M. Collier, Robert J. +Collord, George W. Combe, Mrs. William - +Constant, S. Victor de Coppet, EH. J. Corning, Christopher, R. *Crampton, Prof. Henry E., Ph.D. +Crane, Zenas Crosby, Maunsell S. *Curtis, Carlton C. Curtis, G. Warrington *Dahlgren, B. E., D.M.D. Davies, J. Clarence Davis, Dr. Charles H. Davis, David T. ~ *+Davis, William T. *+Dean, Prof. Bashford, Ph.D. + Delafield, Maturin L., Jr. Delano, Warren, Jr. ! Demorest, William C. Devereux, W. B. 409 De Vinne, Theodore L. De Witt, William G. Dickerson, Edward N. Diefenthaler, C. E. Dimock, George E. Dodge, Rev. D. Stuart, D.D. + Dodge, Miss Grace H. *Dodge, Prof. Richard E., A.M. Doherty, Henry L. Donald, James M. *Doremus, Prof. Charles A., Ph.D. *+ Douglas, James Douglass, Alfred Draper, Mrs. M. A. P. Drummond, Isaac W., M.D. *Dudley, P. H., Ph.D. *Dunham, Edward K., M.D. +Dunn, Gano +Dunscombe, George Elsworth *Dutcher, Wm. *Dwight, Jonathan, Jr., M.D. Dwight, Mrs. M. E. *Harle, R. B. *Hastman, Prof. Charles R. *+Hilliott, Prof. A. H., Ph.D. : Emmet, C. Temple Eno, William Phelps Estabrook, A. F. EKvarts, Allen W. *Hyerman, John Fairchild, Charles 8. Fargo, James C. Farmer, Alexander S. *Farrand, Prof. Livingston, M.D. Farrington, Wm. H. Fearing, D. B. Ferguson, Mrs. Juliana Armour § Field, C. de Peyster Field, William B. Osgood *Finley, Pres. John H. *Fishberg, Maurice, M.D. 410 ANNALS NEW YORK ACADEMY OF SCIENCES ‘Follett, Richard E. Foot, James D. + Ford, James B. Fordyce, John A. de Forest, Robert W. Friedrick, J. J. Frissell, A. 8. Fuller, Charles D. *Gager, C. Stuart, Ph.D. Gallatin, F. Gardner, Clarence Roe Gibson, R. W. *Gies, Prof. William J. *Girty, George H., Ph.D. Godkin, Lawrence Goodridge, Frederick G. Goodwin, Albert C.1 §Gould, Edwin §Gould, George J. *+Grabau, Prof. Amadeus W. *Gratacap, Louis P. Green, James W. Greenhut, Benedict J. *Gregory, W. K., Ph.D. +Grinnell, G. B. Griscom, C. A., Jr. Guernsey, H. W. Guggenheim, William Guinzburg, A. M. von Hagen, Hugo Haines, John P. Halls, William, Jr. Hammond, James B. Hardon, Mrs. H. W. +Harrah, Chas. J. + Harriman, Mrs. E. H. Hasslacher, Jacob Haupt, Louis, M.D. Havemeyer, J. C. Havemeyer, William F. 1 Deceased. Healy, J. R. *Hering, Prof. Daniel W. Hewlett, Walter J. *Hill, Robert T. Hirsch, Charles 8S. *Hitchcock, Miss F. R. M., Ph.D. Hochschild, Berthold Hollenback, Miss Amelia B. *Hollick, Arthur, Ph.D. + Holt, Henry + Hopkins, George B. *Hornaday, Wiliam T., Se.D. Hotchkiss, Henry D. *t Hovey, Edmund Otis, Ph.D. *Howe, Marshall A., Ph.D. + Hoyt, A. W. +Hoyt, Theodore R. +Hubbard, Thomas H. Hubbard, Walter C. Humphreys, Edwin W. _ Humphreys, Frederic H. + Huntington, Archer M. *Hussakof, Louis, Ph.D. Hustace, Francis + Hutter, Karl +Hyde, B. Talbot B. Hyde, E. Francis + Hyde, Frederic E., M.D. Hyde, Henry St. John *Hyde, Jesse E. tIles, George *Trving, Prof. John D. von Isakovics, Alois Iselin, Mrs. William E. +Jackson, V. H. *Jacobi, Abram, M.D. James, F. Wilton tJarvie, James N. Jennings, Robert H. *Johnson, Prof. D. W., Ph.D. MEMBERSHIP + Johnston, J. Herbert Jones, Dwight A. *S Julien, Alexis A., Ph.D. Kahn, Otto H. Kautz-Hulenburg, Miss P. R. *tKemp, Prof. James F., Sc.D. + Keppler, Rudolph + Kessler, George A. Kinney, Morris Kohlman, Charles *t Kunz, George F., M.A., Ph.D. +Lamb, Osborn R. Landon, Francis G. Lang, Herbert Langdon, Woodbury G. Langeloth, J. *Langmann, Gustav, M.D. Lawrence, Amos E. Lawrence, John B. +Lawton, James M. *Ledoux, Albert R., Ph.D. *Lee, Prof. Frederic 8., Ph.D. *$Levison, Wallace Goold Levy, Emanuel Lichtenstein, M. Lichtenstein, Paul lhieb, J. W., Jr. Lindbo, J. A. + Loeb, James *Loeb, Prof. Morris, Ph.D.? tLow, Hon. Seth, LL.D. *Lowie, Robert H., Ph.D. *Lucas, F. A., D. Se. *Lusk, Prof. Graham, M.D. Lydig, Philip M. Lyman, Frank _ Lyon, Ralph McCarthy, J. M. *tMcMillin, Emerson McNeil, Charles R. 1 Deceased. MacArthur, Arthur F. Macy, Miss Mary Sutton, M.D. tMacy, V. Everit Mager, F. Robert Mann, W. D. Mansfield, Prof. William Marble, Manton Marcou, John B.1 Marling, Alfred H. t Marshall, Louis Marston, H.S8. + Martin, Bradley *tMartin, Prof. Daniel 8S. *Martin, T. Commerford *+Matthew, W. D., Ph.D. Maxwell, Francis T. § Mead, Walter H. Mellen, C.S. *Meltzer, 8. J., M.D. Merrill, Frederick J. H., Ph.D. Metz, Herman A. ~ Milburn, J. G. . Miller, George N., M.D. *+ Miner, Roy Waldo Mitchell, Arthur M. Monae-Lesser, A., M.D. Morgan, J. Pierpont *Morgan, Prof. Thomas H. Morgan, William Fellowes Morris, Lewis R., M.D. Munn, John P. +Nash, Nathaniel C. + Nesbit, Abram G. Notman, George Oakes, Francis J. Ochs, Adolph S. Oettinger, P. J., M.D. *+ Ogilvie, Miss Ida H., Ph.D. tOlcott, E. E. Olmsted, Mrs. Charles T. * 412 ANNALS NEW YORK ACADEMY OF SCIENCES Oppenheimer, Henry S. *+ Osborn, Prof. H. F., Se.D., LL.D. Osborn, William C. +Osborn, Mrs. William C. *Osburn, Raymond C., Ph.D. +Owen, Miss Juliette A. = Pacmis A) beeehaD: Paddock, Eugene H.* + Parish, Henry Parsons, C. W. *Parsons, John E. +Patten, John Paul, John J. *Pedersen, Prof. F. M., Ph.D. *tPellew, Prof. C. E., Ph.D. Pennington, William * + Perkins, William H. Perry, Charles J. *Peterson, Frederick, M.D. Pettigrew, David L. Pfizer, Charles, Jr. Philipp, P. Bernard Phoenix, Lloyd Pierce, Henry Clay Plant, Albert Planten, John R.* Polk, Dr. W. M. *Pollard, Charles L., Ph.D. *Poor, Prof. Charles L. * Porter, Eugene H. Post, Abram S. *Post, C. A. *Post, George B. Preston, Veryl *Prince, Prof. John Dyneley +Pyne, M. Taylor *+Ricketts, Prof. P. de P., Ph.D. Riederer, Ludwig Robert, Samuel Roberts, C. H. 1 Deceased. + Roebling, John A. Rogers, E. L. Rosenbaum, Selig Rossbach, Jacob ~ tde Rubio, H. A. C. *tRusby, Prof. Henry H., M.D. Russ, Edward + Sachs, Paul J. Sage, Dean Sage, John H. +Schermerhorn, F. A. Schiff, Jacob H. Scholle, A. H. Schoney, Dr. L. +Schott, Charles M., Jr. Scott, George S.? Scoville, Robert Seaman, Dr. Louis L. Seitz, Carl E. Seligman, Jefferson _ Sexton, Laurence E. Shaw, Mrs. John C. Shepard, C. Sidney §Shepard, Mrs. Finley J. *Sherwood, George H. Shillaber, William Shultz, Charles 8. *Sickels, Ivin, M.D. Sleight, Chas. E. Sloan, Benson B. Smith, Adelbert J. *Smith, Ernest E., M.D., Ph.D. Smith, Frank Morse Snow, Elbridge G. *Southwick, Edmund B., Ph.D. Squibb, Edward H., M.D. Starr, Louis Morris *Starr, Prof. M. Allen *+Stefansson, V. Steinbrugge, Edward, Jr. MEMBERSHIP 413 +Stetson, F. L. *Stevens, George T., M.D. *+Stevenson, Prof. John J., LL.D. Stokes, James Stokes, J. G. Phelps +Stone, Miss Ellen J. Straus, Isidor + Strauss, Charles Strauss, Frederick +Streat, James Sturgis, Mrs. Elizabeth M. Taggart, Rush *t+Tatlock, John, Jr. Taylor, George Taylor, W. A. Taylor, William H. +Terry, James? Tesla, Nikola *Thatcher, Edward J., Jr. Thaw, A. Blair Thaw, Benjamin Thompson, Mrs. Frederick F. Thompson, Lewis S. +Thompson, Robert M. *Thompson, Prof. W. Gilman Thompson, Walter *Thorndike, Prof. Edward L. Thorne, Samuel *Tower, R. W., Ph.D. *Townsend, Charles H., Sc.D. Tows, C. D. *Trowbridge, Prof. C. C. Tuckerman, Alfred, Ph.D. Tuttle, Mrs. B. B. 1 Deceased. Ullmann, E. 8. + Vail, Theo. N. Vanderpoel, Mrs. J. A. + Van Slyck, George W. +Van Wyck, Robert A. Vreeland, Frederick K. Walker, William I. *tWaller, Prof. Elwyn, Ph.D. Warburg, F. N. Warburg, Paul M. Ward, Artemas + Ward, Charles Willis Ward, John Gilbert Waterbury, J. I. Watson, John J., Jr. + Weir, Col. John? *Wells, F. Lyman Williams, R. H. Wills, Charles T. *Wilson, Prof. E. B., Ph.D., LL.D. Wilson, J. H. Wilson, Miss M. B., M.D. *Winslow, Prof. Charles-H. A. *Wissler, Clark, Ph.D. Woerishoffer, Mrs. Anna Wood, Mrs. Cynthia A. Wood, William C. *Woodbridge, Prof. F. J. E. *Woodhull, Prof. John F., Ph.D. *Woodman, Prof. J. Edmund *Woodward, Prof. R.S. *Woodworth, Prof. R. 8S. Younglove, John, M.D. Zabriskie, George 414 . ANNALS NEW YORK ACADEMY OF SCIENCES ASSOCIATE MEMBERS Billingsley, Paul — Kellicott, W. E., Ph.D. Brown, Harold Chapman, Ph.D. Kirk, Charles T. Brown, T. C. McGregor, James Howard Byrne, Joseph P. Montague, W. P., Ph.D. Fenner, Clarence N., Ph.D. Mook, Charles Fettke, Chas. R. Moon, Miss Evangeline Gordon, Clarence E. Northup, Dwight Jalplena, Jt3 Ins, 1e)n.|D) O’Connell, Miss Marjorie Haseman, J. D. Rogers, G. Sherburne Hunter, George W. Stevenson, A. E: Johnson, Julius M. Wood, Miss Elvira NON-RESIDENT MEMBERS *Berry, Edward W. *Lloyd, Prof. F. EH. Buchner, Edward F. *Mayer, Dr. A. G. *Bumpus, H. C. Meyer, Adolph Burnett, Douglass Petrunkevitch, Alexander, Ph.D. *Davis, William H. 2 PratiseD ies) emele English, George L. *Ries, Prof. H. Finlay, Prof. G. I. Reuter, L. H. Frankland, Frederick W. *Sumner, Dr. F. B. Hoffman, 8. V. *van Ingen, Prof. G. Kendig, Amos B. “Wheeler, Wm. Morton GENERAL INDEX TO VOLUME XXII Names of Authors and other Persons in Heavy-face Type Titles of Papers in SMALL CAPS Active Members, Plection of, 339, 344, 355, 380 : Active Members, List of, 408 Adams, —, Reference to, 50 Adams, Frank D., Honorary Member, 383 AERIAL TRANSMISSION OF DISEASE, THE, C. V. Chapin [Abstract], 351, 352 Agassiz, —, Reference to, 86 AGE OF WALKING AND TALKING IN RELATION TO GENERAL PRACTICE, THE, C. D. Mead [Abstract], 359, 363 ° ALASKA, GEOLOGY AND MINERAL RESOURCES or, Alfred H. Brooks [Title], 371 Allen, —, Reference to, 50 Allorisma costatum, 4 gilberti, 5 (Pleurophorella ?) reflerum, 3 reflexum, 3 ALTERATIONS IN THE SNAKE RIVER BASALTS, Charles T. Kirk [Title], 356 Alula, 4 geinitzi, 3 gen. nov., 3 gilberti, 2, 3, 5 gilberti White?, 5 2 lanceolata, 3 squamulifera, 2, 3, 4, 5 AMALIA FARM METEORITE, THE, Edmund Otis Hovey [Abstract], 345 Ameginho, —, Reference to, 101 Andrews, Roy C., AN EXPLORATION OF NORTHEASTERN KoreEA [Title], 375 ANNUAL MEETING, MINUTES OF THE, Edmund Otis Hovey, 383 Ants of South America, 82 Aquatic migration in South America, 42 Arber, —, Reference to, 84 Archean rocks of South America, 17, 18 Archenhold, F. S., ASTRONOMY, EDUCATION AND CULTURE [Title], 353 Arctowski, Henry, Active Member, 339 Arges marmoratus from the Republic of ‘ Colombia, 327-333 Arnold, Felix, Fellow, 383 Associate Members, Blection of, 3389, 355, 380 Associate Members, List of, 414 Asst, Jappa, Reference to, 11 Astor, John Jacob, Death of, 370 ASTRONOMY, EDUCATION AND CULTURE, F. S. Archenhold [Title], 353 ATTEMPT TO MEASURE MENTAL WoRK AS A PsycHo-DyNAMIC PROCESS, THE, Ray- mond Dodge [Title], 380 AUDITORY AND VISUAL Memory, A. E. Chris- lip [Abstract], 346, 347 BACTERIA IN City Dust, THH NUMBER AND KINDS oF, C.-E. A. Winslow and I. S. Kligler [Abstract], 351, 352 Balta, José, Reference to, 226 : Barnard, Edward E., THrE PLANET Mars [Title], 339 Baskerville, Charles, TUNGSTEN [Abstract], 373 Bassler, —, Reference to, 160. Bateson, —, References to, 70, 114, 116, 119, 13311 BEDFORD SHALE OF OHIO, GEOLOGICAL AGE OF THE, George H. Girty, 295-319; _ [Title], 374 Berkey, Charles P., IS THrRE FAULT CoN- TROL OF THE HUDSON RIVER COURSE? [Abstract], 350, 351 SECTION OF GEOLOGY .AND MINERALOGY, 350, 355, 371, 374, 380 Boas, Dr. RADOSAVLJEVICH’S CRITIQUE OF PROFESSOR, Robert H. Lowie [Ab- stract], 354 Boas, Franz, A YEAR IN Mexico [Title], 376 Boettger, —, Reference to, 29 Booth, Garret and Blair, Reference to, 336 Borelli, —, cited, 268 Borup, George, Death of, 370 Bowles, James F. B., SANITATION OF THE PANAMA CANAL [Title], 371 Branner, J. C., cited, 30, 84, 91 References to, 10, 17, 31, 42, 86 Brazilian Coast, Topography of the, 25 Brogger, W. C., cited, 136, 137, 139, 142 References to, 137, 138, 139 Brooks, Alfred H., GmOLOGY AND MINERAL RESOURCES oF ALASKA [Title], 371 Broom, , References to, 96, 97, 99 Brown, —, Reference to, 115 Brown, Barrington, Reference to, 29 Burke and Pinckney, cited, 181 Burroughs, W. G., cited, 296 (415) 416 BUSINESS MEETINGS, MINUTES oF, Edmund Otis Hovey, 339, 344, 350, 355, 370, 373, 376, 380 Butters, Roy M., References to, 1, 2 Butts, —, cited, 298, 309 References to, 307, 308, 312 By-Laws of the New York Academy of Sci- ences, 396 Calandruccio, —, Reference to, 321 Callograptus grabaui sp. noy., 142 Campbell, William, Somn NoTES ON IRON AND STEEL [Title], 342 Carboniferous fossils of South America, 18 Carrel, Alexis, RESULTS OF THE SUTURE OF BLOOD VESSELS AND THE. TRANS- PLANTATION OF ORGANS [Title], 378 CAUSE OF THE TIDES, THR, Charles Lane Poor [Abstract], 378, 379 CEMENT, METAMORPHISM OF Albert B. Pacini, 340 CHANGES IN THE BEHAVIOR OF THH HEL DuR- ING TRANSFORMATION, Bashford Dean, 321-326 Chapin, C. V., THE ARIAL TRAN SAE SS TON or DisHAsE [Abstract], 351, 352 CHEMICAL ART, PRODUCTS oF, Louis H. Friedburg [Abstract], 353 Chemical composition of Portland cement, 164 PORTLAND, 161-224; [Title], Chrislip, A. E., AUDITORY AND VISUAL MEM- ory [Abstract], 346, 347 Christman, Erwin S., Reference to, 269 Clarke, —, cited, 298, 309 Clarke, F. W., cited, 217 Climbing catfish from the Republic of Co- lombia, 327-333 Coastal plains of Peru, 228 Collie, —, Reference to, 143 COLORADO, HASTERN, INVERTEBRATE FOSSILS FROM, George H. Girty, 1-8 Conrad, —, Reference to, 29 Constitution of the New York Academy of Sciences, 395 Cope, —, cited, 268, 269 References to, 54, 78 Corresponding Members, List of, 404 CORRESPONDING SECRETARY, REPORT OF THE, Henry E. Crampton, 385 Cox, Charles F., Death of, 344 Crampton, Henry E., REPORT OF THH CORRE- SPONDING SECRETARY, 385 Crandall, R., cited, 21 References to, 10, 17, 42 Crustacea of South America, 79 CuBAN MARINE FisuEs, NoTres on, John T. Nichols [Abstract], 358 Culler, A. J.. RELATION OF INTERFERENCE TO ADAPTABILITY [Abstract], 359, 363 Cumings, E. R., cited, 144 ANNALS NEW YORK ACADEMY OF SCIENCES Cunningham, —, Reference to, 321 Cushman, A. S., Reference to, 114 Dall, —, quoted, 30 Darwin, —, Reference to, 226 Davenport, —, cited, 122 References to, 114, 126, 127, 128 Davis, —, Reference to, 230 Davis, Frank, Reference to, 11 Day, Francis, cited, 321 Dean, Bashford, Do DEVELOPING EMBRYOS GivE REAL CLUES TO LINES OF DE- scent? [Abstract], 372 ON THE CHANGES IN THE BEHAVIOR OF THH HEL(Conger malabaricus) DURING Its TRANSFORMATION, 321-326; [Ab- stract], 372 References to, 11, 327 Deaths, 370, 374, 376 De Koninck, —, Reference to, 299 Derby, O. A., References to, 10, 17, 21, 30, 43 de Ribeiro, Alipo Miranda, Reference to, 38 Devonian fossils of South America, 18 de Vries, Hugo, EXPERIMENTAL HYOLUTION [Title], 381 References to, 54, 70 DICTYONEMA-FAUNA OF NAvy ISLAND, NEW BRUNSWICK, ON THE, F. F. Hahn, 135-160 Dictyonema flabelliforme Hichw. (sp.), 136 flabelliforme Hichw. var. acadica Mat- thew, 137 Dieckmann, G. P., cited, 209 Dieder, —, Reference to, 99 i] DIFFERENT-TONES AND CONSONANCH, F. Krueger [Title], 380 DISTRIBUTION OF FERRIC CHLORIDE BETWEEN ETHER AND AQUEOUS HYDROCHLORIC AcID AT 25° C., Albert B. Pacini [Abstract], 378, 379 DISTRIBUTION OF PETROLEUM DEPOSITS IN PERU, Vernon F. Marsters [Abstract], 350, 351 Do DEVELOPING EMBRYOS Give REAL CLUES AS TO LINES oF DuscuentT?, Bashford Dean [Abstract], 372 Dodge, Raymond, THE ATTEMPT TO MEASURE MENTAL WORK AS A PSYCHO-DYNAMIC Process [Title], 380 Doherty, Henry L., Treasurer, 350 REPORT OF THE TREASURER, 387 Doncaster, —, Reference to, 113 D’Orbigny, —, Reference to, 226 Drainage of Overlook Mountain, 264 Drirt PEBBLES, THE FOSSILS AND HORIZON oF THE, F. S. Hintze [Title], 377 Durham, Miss, Reference to, 113 Earle, R. B., Active Member, 344 Fellow, 383 Reference to, 161 Request for grant for research, 377 GENERAL INDEX TO VOLUME XXII Hast Andean. Sea, Topography of, 30 Echo Lake, 266 — EpiTor, RHPORT OF THE, Edmund Otis Hovey, 387 EDUCATION AND CULTURE, ASTRONOMY, F. S. Archenhold [Title], 353 EEL DURING TRANSFORMATION, CHANGES IN THE BEHAVIOR OF THE, Bashford Dean, 321-326; [Abstract], 372 Higenmann, —, cited, 36, 51 References to, 11, 16, 52, 58, 57, 59, 65, 67, 74, TT, 321 Emmons, —, Reference to, 140 Eschwege, —, Reference to, 17 Etheridge, —, References to, 29, 30 EURYPTERIDS, PRESENT OPINIONS ON THH HABITS OF THE, Marjorie O’Connell \[Title], 377 : Evans, —, References to, 18, 22 HXPERIMENT IN THE CATCHING OF PENNIES, E. S. Reynolds, J. T. Gyger and L. L. Winslow [Abstract], 359, 365 EXPERIMENTAL HVOLUTION, Hugo de Vries [Title], 381 EXPLORATION OF NORTHEASTERN Korea, AN, Roy C. Andrews [Title], 375 Ewald, —, Reference to, 281 Fearnside, W. G., cited, 136, 152, 157 Reference to, 158 Fellows, Election of, 383 Fettke, Charles R., Associate Member, 339 Fick, —, cited, 277 Reference to, 281 Fischer, O., Reference to, 268 Foerste, A. F., cited, 304, 318 Foote Mineral Company, Reference to, 336 Foote, Warren M., Reference to, 335 Forbes, —, References to, 101, 226 FORELANDS OF THE BRAS D’OR LAKES, CAPE BRETON ISLAND, Nova Scotia, J. E. Woodman [Abstract], 350, 351 FOSSILS AND HorRIZON OF THE DRIFT PEB- BLES, THE, F. S. Hintze [Title], 377. Frech, F., cited, 147 Friedburg, Louis H., PrRoDUCTS oF CHEMICAL Art [Abstract], 353 Gabb, —, Reference to, 29 Gaudry, —, cited, 269 Reference to, 100 GEM-BEARING PEGMATITES OF LOWER CALI- FORNIA, THE, George F. Kunz [Title], 371 ; GEOGRAPHICAL DISTRIBUTION IN SOUTH AMERICA, John D. Haseman, 9-112; [Title], 346 GEOLOGICAL AGE OF THH BEDFORD SHALES OF OHIO, George H. Girty, 295-319; [Title], 374 417 GEOLOGY AND MINERAL RESOURCES OF ALASKA, Alfred H. Brooks [Title], 371 Geology and topography of South America, 17-49 ; Gidley, J. W., cited, 286 Girty, George H., GEOLOGICAL AGE OF THE BEDFORD SHALH OF OHIO, 295-319; [Title], 374 ON SOME INVERTEBRATE FOSSILS FROM THN LYKINS FORMATION OF EASTERN CoLorRADO, 1-8; [Title], 345 Glaciation of Overlook Mountain, 262 Glenn, —, cited, 298, 309 GoAT ISLAND AT NIAGARA GLEN, WAS THERE A FormMmr, A. W. Grabau [Abstract], 377, 378 Goddard, Henry H., TH HEREDITY OF MEN- TAL TRAITS [Abstract], 346, 348 Goddard, Pliny E., NOTES ON THE JICARILLA APACHE [Abstract], 343 Goldbeck, A. T., cited, 208 Goldfarb, A. J.. THE INFLUENCH OF THE NERvVoUS SYSTEM UPON GROWTH [Title], 381 A New METHOD OF FUSING EGGS OF THH SAMBH SpEcIES [Abstract], 381, 382 Gondwana flora of South America, 83 Goodale, H. D., References to, 114, 116 Goodale, H. D., T. H. Morgan and, Sprx- LINKED INHERITANCE IN POULTRY, 113-133 Goodale, R. C., Reference to, 115 Goodsir, —, Reference to, 273 Goppert, —, cited, 137, 138 Reference to, 152 Gordon, John, Reference to, 11 Grabau, A. W., References to, 11, 136, 138, 148 Was THERE A FORMER GOAT ISLAND AT NraGarRA GLEN? [Abstract], 377, 378 Grants from research funds, 340, 377 Grassi, —, Reference to, 321 Gregory, —, cited, 55 References to, 11, 82, 99 Gregory, H. E., Reference to, 259 } Gregory, William K., NoTES ON CERTAIN PRINCIPLES OF QUADRUPEDAL LOCO- MOTION AND ON THE MECHANISM OF THD LIMBS OF HOOFED ANIMALS, 267- 298; Title], 372. SEcTion oF Brouoey, 340, 346, 351, 358, 871, 375,. 378, 381 Guyer, —, Reference to, 132 Gyger, J. T., E. S. Reynolds and L. L. Win- slow, EXPERIMENT IN THH CATCHING OF PENNIES [Abstract], 359, 365 Haacke, —, cited, 17 Reference, 101 Hadley, —, Reference to, 114 418 Hagedoorn, —, Reference to, 114 Hahn, F. F., Associate Member, 355 ON THD DICTYONEMA-FAUNA OF NAVY ISLAND, Nrw BRUNSWICK, 135-160; [Title], 372 Hale, George E., Honorary Member, 383 Hall, James, cited, 140 References to, 144, 159, 297, 304, 312 Hardening process of Portland Cement, 166 Hardon, Mrs. Henry W., Active Member, 355 Hartnagel, —, cited, 307 Hartt, —, References to, 17, 25 Haseman, John D., SoMmE FAcTORS OF GEO- GRAPHICAL DISTRIBUTION IN SoutTH AMERICA, 9-112; [Title], 346 Hatcher, —, References to, 17, 34 Hauthal, —, Reference to, 17 Haycraft, —, cited, 268, 276, 277, 280, 281 Hedley, —, Reference to, 101 Helmholtz, —, cited, 277 Henke, —, Reference to, 281 HEREDITY OF MENTAL TRAITS, THE, Henry H. Goddard [Abstract], 346, 348 Herrick, —, cited, 304, 305 Reference to, 306 Hickman, J. E., TH INFLUENCE OF NAR- COTICS ON PHYSICAL AND MBNTAL TRAITS OF OFFSPRING [Abstract], 346 Hintze, F. S., TH FOSSILS AND HORIZON OF THE DRIFT PEBBLES [Title], 377 Holland, W. J., Reference to, 10 Honorary Members, Election of, 383 List of, 403 Hopkinson, T., cited, 140 Horton, B. B., Reference to, 114 Hovey, Edmund Otis, THE AMALIA FARM MeErTEoRITH [Abstract], 345 THE KINGSTON, N. M., SIDERITE, 335- 337 MINUTES OF THE ANNUAL MEBTING, 383 MINUTES OF BUSINESS MEETINGS, 339, 343, 350, 355, 370, 373, 376, 380 RECORDS OF MEETINGS OF THE NEW YorRK ACADEMY OF SCIENCES, 339-414 REPORT OF THE EDITOR, 387 REPORT OF THE RECORDING SECRETARY, 385 ~ SECTION OF GEOLOGY AND MINERALOGY, 340, 344, 377 Hupson River Coursn, IS THERE FAULT CONTROL OF THE, Charles P. Berkey [Abstract], 350, 351 Hussakof, Louis, Reference to, 11 THE SPAWNING HABITS OF THE SEA LAMPREY, Petromyzon marinus [Ab- stract], 358 Huxley, —, Reference to, 101 Hyde, J. E., cited, 296 Fellow, 384 ANNALS NEW YORK ACADEMY OF SCIENCES ‘ ILLUSTRATIONS OF MINERAL ASSOCIATIONS BY MBANS OF COLOR PLATE AND OTHER PHOTOGRAPHS OF OPAQUE SPECIMENS, Wallace Goold Levison [Abstract], 356 INDIVIDUAL DIFFERENCES IN THE INTERESTS ' OF CHILDREN, Gertrude M. Kuper [Abstract], 359 : INFLUENCE OF NARCOTICS ON PHYSICAL AND MENTAL TRAITS OF OFFSPRING, TH, J. E. Hickman [Abstract], 346 INFLUENCE OF THE NERVOUS SYSTEM UPON GROWTH, THE, A. J. Goldfarb [Title], 381 INVERTEBRATE FOSSILS FROM HASTERN COLO- RADO, George H. Girty, 1-8 Invertebrates of South America, 79 Iron, SOME NOTES ON, William Campbell [Abstract], 343 Is THERD FAULT CONTROL OF THE HUDSON RIvEK CourSsn?, Charles P. Berkey [Abstract], 250, 351 Ives, P. P., Reference to, 115 Janda, F., cited, 183 Jay, James E., Reference to, 161 JICARILLA APACHE, NOTES ON THE, Pliny E. Goddard [Abstract], 343 Johannsen, —, Reference to, 54 Johnson, D..W., Active Member, 380 Fellow, 383 \ THE WESTWARD TRIP OF THE TRANS- CONTINENTAL WXCURSION OF THB AMERICAN GEOGRAPHICAL SOCIETY [Title], 374 Johnson, R. D. O., NoTES ON THH HABITS OF