JTn7TIT|l,l,l,l.l|l|l|l|l|TIITrMil il I|l|l|l| I'l'lilUiiit^ I^n^K^VHY Storrs lEricnltiiral Collep. Vol .. . jj^. . ■- .•.. .< Class J^O. . SSJ: Cosi -^ 'Date ./!y'J, ?< n^i S i^r. • E-KESKNTBD BY TnmCTTItTL''' ' '■'■' I.lilili' I'l' I'gHi Please handle this volume with care. The University of Connecticut Libraries, Stons J Digitized by the Internet Archive in 2009 with funding from Boston Library Consortium IVIember Libraries http://www.archive.org/details/fossilfishesfossOOnewb [Monograi>h XIV.] The publicatiouS of the Uuited States Geological Sarvey are issued in accortlauce witli the statute approved March 3, 1879, which declares that — • "The publications of the Geological Survey shall coasist of the annual report of operations, geo- logical and economic maps illustrating the resources and classification of the lands, and reports upon general and economic geology and paleontology. The annual report of operations of the Geological Survey shall accompany the annual report of the Secretary of the Interior. All special memoirs and reports of said Survey shall be issued in uniform quarto series if deemed necessary by the Director, but otlierwise in ordinary octavos. Three thousand copies of each shall be published for scientific exchanges and for sale at the price of publication ; and all literary and cartograjihic materials received in exchange shall be the property of the United States and form a part of the library of the organizatiou : Aud the moaey resulting from the sale of such publicatioas shall be covered into the Treasury of the United States." "7 On July 7, 1882, the following joint resolution, referring to all Goverumeut publications, was passed by Congress : "That whenever auy document or report shall be ordered printed by Congress, there shall be printed, in addition to the number in each case stated, the ' usual number' (1,900) of copies for blading and distribution amoug those entitled to receive them." Except in those cases in which an extra number of any publication has been supplied to the Sur- vey by special resolution of Congress or has been ordered by the Secretary of the Interior, this office has no copies for gratuitous distribution. ANNUAL EEPOETS. I. First Annual Report of the United States Geological Survey, by Clarence King. 1880. 8°. 79 pp. 1 map. — A preliminary report describing plan of organization and publications. II. Second Annual lieport of the United States Geological Survey, 1880-81, by J. W. Powell. 1883. 8°. Iv, 588 pp. 61 pi. 1 map. III. Third Annual Report of the United States Geological Survey, 1831-82, by J. W. Powell. 1883. 8°. xviii, 564 pp. 67 pi. and maps. IV. Fourth Annual Report of the United States Geological Survey, 1882-'83, by J. \V. Powell. 1884. . 8°. xxxii, 473 pp. 85 pi. aud maps. V. Fifth Annual Report of the United States Geological Survey, 1833-54, by J. AV. Powell. 1885. 8°. xxxvi, 469 pp. 58 pi. and mans. VI. Sixth Annual Report of the United States Geological Survey, 1884-85, by J. W. Powell. 1 83. 8°. xxix, 570 pp. 65 pi. and maps. VII. Seventh Annual Report of the United States Geological Survey, 1885-86, by J. W. Powell. 1888. 8°. XX, 656 pp. 72 pi. and maps. The Eighth and Ninth Annual Reports are in press. MONOGRAPHS. Monograph I is not yet published. II. Tertiary History of the Grand Canon District, with atlas, by Clarence E. Button, Capt. , U. S. A. 1882. 4°. xiv, 264 pp. 42 pi. and atlas of 24 sheets folio. Price $10.12. III. Geology of the Comstock Lode and the Washoe District, with atlas, by George F. Becker. 1883. 4°. sv, 42'2pp. 7 pi. aud atlas of 21 sheets folio. Price $11. IV. Comstock Mining and Miners, by Eliot Lord. 1883. 4*^. xiv, 451 pp. 3 pi. Price |1.50. V. The Copper-Bearing Rocks of Lake Superior, by Roland Duer Irving. 1883. 4°. xvi,464pp. 15 1. 29 pi. and maps. Price |1.85. VI. Contributions to the Knowledge of the Older Mesozoic Flora of Virginia, by William Morris Fontaine. 1883. 4°. xi, 144 pp. .54 1. ,54 pi. Price $1.05. VII. Silver-Lead Deposits of Eureka, Nevada, by Joseph Story Curtis. 1884. 4°. xiii, 200 pp. 16 pi. Price $1.20. VIII. Paleontology of the Eureka District, by Charles Doolittle Walcott. 1831. 4°. xiii, 293 pp. 2U. 24 pi. Price fl. 10, I II ADVERTISEMENT. IX. BracbiopodaandLamellibranchiataof theEaritau Clavs and GreensandMarlsof New Jersey, by Robert P. Whitfield. 1885. 4°. ss, 338 pp. 35 pi. 1 map." Price $1.15. X. Dinocerata. A MoDograpli of an Extinct Order of Gigantic Mammals, by Othniel Charles Marsh. 1886. 4°. xviii, 243 pp. 56 1. 56 p!. Price 52.70. XI. Geological History of Lake Lahontan, a Qnaternary Lake of Northwestern Nevada, by Israel Cook Russell. Ib'io. 4°. xiv, 2f^8 pp. 46 pi. and maps. Price §1.75. XII. Geology and Mining Industry of Leadville, Colorado, with atlas, by Samuel Franklin Em- mons. 1886. 4°. xxis, 770 pp. 45 pi. and atlas of 35 sheets folio. Price $8.40. XIII. Geology of the Quicksilver Deposits of the Pacific Slope, with atlas, by George F. Becker. 1888. 4°. xix, 486' pp. 7 pi. and atLas of 14 sheets folio. Price .«;-i.00. XIV. Fossil Fishes and Fossil Plants of the Triassie Rocks of New Jersey and the Connecticut Valley, by John S. Newberry. 1888. 4°. xiv, 152 pp. 26 pi. Price §1.00. In preparation: XV. Younger Mesozoic Flora of Virginia, by William M. Fontaine. XVI. Paleozoic Fishes of North America, by J. S. Newberry. XVII. Description of New Fossil Plants from the Dakota Group, by Leo Lesquereux. — Gasteropoda of the New Jersey Cretaceons and Eocene Marls, by R. P. Whitfield. — Geology of the Eureka Mining District, Nevada, with atlas, by Arnold Hague. — Lake Bonneville, by G. K. Gilbert. — Sauropoda, by O. C. Marsh. — Stegosauria, by O. C. Marsh. — BroutotheridEB, by O. C. Marsh. — ThePenokee-GogebicIron-BeariugSeriesof North Wisconsin and Michigan, by Roland D. Irving. — Report on the Denver Coal Basin, by S. F. Emmons. — Report on Silver Clilf and Ten-Mile Mining District, Colorado, by S. F. Emmons. — Flora of the Dakota Group, by J. S. Newberry. — The Glacial Lake Agassiz, by Warren Upham. — Geology of the Potomac Formation in Virginia, by W. M. Fontaine. BULLETINS. Each of the Bulletins contains but one paper and is complete in itself. They are, however, num- bered in a continuous series, and may be bouud in volumes of convenient size. To facilitate this, each Bulletin has two paginations, one proper to itself and another which belongs to it as part of the volume. 1. On Hypersthene-Audcsite and on Triclinic Pyroxene in Augitic Rocks, by Whitman Cross, with a Geological Sketch of Buffalo Peaks, Colorado, by S. F. Emmons. 1883. 8°. 42 pp. 2 pi. Price 10 cents. 2. Gold and Silver Conversion Tables, giviug the coining values of troy ounces of fine metal, etc., coioputed by Albert Williams, jr. 1883. 8°. 8 pp. Price 5 cents. 3. On the Fossil Faunas of the Upper Devonian, along the meridian of 76° 30', from Tompkins County, N. Y., to Bradford County, Pa., by Henry S. Williams. 1884. 8°. 36 pp. Price 5 cents. 4. On Mesozoic Fossils, by Charles A. White. 1884. 8°. 36 pp. 9 pi. Price 5 cents. 5. A Dictionary of Altitudes in the United States, compiled by Henry Gannett. 1884. 8°. 325 pp. Price 20 cents. 6. Elevations in the Dominion of Canada, by J. W. Spencer. 1884. 8°. 43 pp. Price 5 cents. 7. MapotecaGeologica Americana. ACatalogueof Geological Maps of America (North and South), 1752-1881, in geographic and chronologic order, by Jules Marcou and John Belknap Marcou. 1884. 8°. 184 pp. Price 10 cents. 8. On Secondary Enlargements of Mineral Fragments in Certain Eocks, by R. D. Irving and C. R. Van Hise. 1884. 8'\ 56 pp. 6 pi. Price 10 cents. y. A Report of work done in the Washington Laboratory during the fiscal year 1883-84. F\ W. Clarke, chief chemist; T. M. Chatard, assistant chemist. 1884. 8°. 40 pp. Price 5 cents. 10. On the Cambrian Faunas of North America. Preliminary studies, by Charles Doolittle Wal- cott. 1884. 8°. 74 pp. 10 pi. Price 5 cents. 11. On the Quaternary and Recent Mollusca of the Great Basin ; with Descriptions of New Forms, l)y R. Ellsworth Call. Introduced bv a sketch of the Quaternary Lakes of the Great Basin, bv G. K. Grlbert. 1884. 8°. 66 pp. 6 pi. Price 5 cents. 12. A Crystallographic Study of the Thinolite of Lake Lahontan, by Edward S. Dana. 1884. 8°. 34 pp. 3 pi. Price 5 cents. 13. Boundaries of the United States and of the several States and Territories, with a Historical Sketch of the Territorial Changes, by Henry Gannett. 1885. 8^. 135 i)p. Price 10 cents. 14. The Electrical and Magnetic Properties of the Iion-Carburets, by Carl Barus and Viueent Strouhal. 1885. 8°. 238 pp. Price 15 cents. 15. On the Mesozoic and Cenozoic Paleontology of California, by Charles A. White. 1885. 8°. 33 pp. Price 5 cents. 16. On the Higher Devonian Faunas of Ontario County, New York, by John M. Clarke. 1835. 8°. 86 pp. 3 pi. Price 5 cents. 17. Ou the Development of Crystallization in the Igneous Rocks of Wa.shoe, Nevada, with Notes on the Geology nf the District, by Arnold Hague and Joseph P. Iddings. 1885. H°. 44 pp. Price 5 cents. ADVEETISEMENT. Ill 18. On Marine Eocene, Fresh-water Miocene, and other Fossil Molhiaca of Western North America, by Charles A. White. 1885. 8°. 26 pp. 3 pi. Price 5 cents. 19. Notes on the Stratigraphy of California, by George F. Becker. 1885. 8°. 28pp. Price5cents. 'M. Contributions to tbo Mineralogy of the Rocky Mountains, by Whitmau Cross and W. F. Hille- brand. lf<85. 8°. 114 pp. 1 pi. Price 10 cents. 21. The Lignites of the Great Sioux Reservation. A Report on the' Region between the Grand and Moreau Rivers, Dakota, by Bailey Willis. 1885. 8°. 16 pp. 5 i^l. Price 5 cents. 22. On New Cretaceous Fossils from California, by Charles A. White. 1885. 8°. 25 pp. 5 pi. Price 5 cents. 23. Observations on the Junction between the Eastern Sandstone and the Keweenaw Series on Keweenaw Point, Lake Superior, by R. D. Irving and T. C. Chamborlin. 18!i5. 8°. 124 pp. 17 pi. Price 15 cents. 24. List of Marine Mollusca, comprising the Quaternary fossils and recent forms from American Localities between Cape Hatteras and Cape Roque, including the Bermudas, hy William Healey Dall. 1885. 8<^. 336 pp. Price 25 cents. 25. The Present Technical Condition of the Steel Industry of the United States, by Phineas Barnes. 1885. 8°. 85 pp. Price 10 cents. 26. Copper Smelting, by Henry M. Howe. 1885. 8°. 107 pp. Price 10 cents. 27. Report of work done in the Division of Chemistry and Physics, mainly during the fiscal year 1884-'85. 1886. 8°. 80 pp. Price 10 cents. 28. The Gabbros and Associated Hornblende Rocks occurring in the Neighborhood of Baltimore, Md., by George Huntington Williams. 1886. 8°. 78 pp. 4 pi. Price 10 cents. 29. On the Fresh- water Invertebrates of the North American Jurassic, by Charles A. White. 1883. 8°. 41pp. 4 pi. Price 5 cents. 30. Second Contribution to the Studies on the Cambrian Faunas of North America, by Charles Doolittle Walcott. 1886. 8°. 369 pp. .33 pi. Price 25 cents. 31. Systematic Review of our Present Knowledge of Fossil Insects, including Myriapods and Arachnids, by Samuel Hubbard Scudder. 1886. 8°. 128 pp. Price 15 cents. 32. Lists and Analyses of the Mineral Springs of the United States; a Preliminary Study, by Albert C. Peale. 1886. 8°. 235 pp. Price 20 cents. 33. Noteson theGeology of Northern California, by J. S. DiUer. 1886. 8°. 23 pp. PriceSconts. 34. On the relation of the Laramie Molluscan Fauna to that of the sucoeediug Fresh- water Eocene and other groups, by Charles A. White. 1886. 8°. 54 pp. 5 pi. Price 10 cents. 35. Phy.sical Properties of the Iron-Carburets, by Carl Barus and Vincent Strouhal. 1886. 8°. 62 pp. Price 10 cents. 36. SubsidenceofFineSolidParticlesiu Liquids, byCarlBarus. 1883. 8°. 58pp. Price 10 cents. 37. Types of the Laramie Flora, by Lester F. Ward. 1887. 8°. 354 pp. 57 pi. Price 25 cents. 38. P'eridotite of Elliott County, Kentucky, by J S.Diller. 1837. 8^. 31pp. 1 pi. Price 5 cents. 39. The Upper Beaches and Deltas of the Glacial Lake Agassiz, by Warren Upbara. 1887. 8-'. 84 pp. 1 pi. Price 10 cents. 40. Changes in River Courses in Washington Territory due to Glaciation, by Bailey Willis. 1887, 8°. 10 pp. 4 pi. Price 5 cents. 41. On the Fossil Faunas of the Upper Devonian — the Genesee Section, New York, by Henry S. Williams. 1887. 8°. 121 pp. 4 pi. Price 15 cents. 42. Report of work done in the Division of Chemistry and Physics, mainly during the fiscal year 1885-'86. F. W. Clarke, chief chemist. 1887. 8°. 1.52 pp. 1 pi. Price 15 cents. 43. Tertiary and Cretaceous Strata of the Tuscaloosa, Tombigbee, and Alabama Rivers, by Eugene A. Smith and Lawrence C. .Johnson. 1887. 8". 189 pp. 21 pi. Price 15 cents. 44. Bibliography of North American Geology for 1886, by Nelson H. Darton. 1887. 8°. 35 pp. Price 5 cents. 45. The Present Condition of Knowledge of the Geology of Texas, by Robert T. Hill. 1887. 8". 94 pp. Price 10 cents. 46. Nature and Origin of Deposits of Phosphate of Lime, by R. A. F. Penrose, jr., with an Intro- duction by N. S. Shaler. 1888. 8°. 143 pp. Price 15 cents. 47. Analyses of Waters of the Yellowstone National Park, with an Account of the Methods of Analysis employed, by Frank Austin Gooch and James Edward Whitfield. 1888. 8°. 84 jip. Price 10 cents. 48. On the Form and Position of the Sea Level, by Robert Simpson Woodward. 1888. 8°. 88 pp. Price 10 cents. Numbers 1 to 6 of the Bulletius form Volume I; Numbers 7 to 14, Volume II; Numbers 15 to 23, Volume HI; Numbers 24 to 30, Volume IV; Numbers 31 to 36, Volume V; Numbers 37 to 41, Volume VI; Numbers 42 to 46, Volume VII. Volume VIII is not yet complete. In press: 49. On the Latitudes and Longitudes of Certain Points in Missouri, Kansas, and New Mexico, by R. S. Woodward. 50. Formulas and Tables to facilitate the construction and use of Maps, by E. S. Woodward. 51. Invertebrate Fossils from California, Oregon, Washington Territory, and Alaska, by C. A. White. 52. On the Subafirial Decay of Rocks and the Origin of the Red Color of Certain Formations, by Israel C. Russell. 53. Geology of the Island of Nantucket, by N. S. Shaler, IV ADVERTISEMENT. lu preparation: — Notes on the Geology of Soutliwestern Kansas, by Robert Hay. — On the Glacial Boundary, by G. F. Wriglit. — The Gabbros and Associated Rocks in Delaware, by F. D. Chester. — Fossil Woods and Lignites of the Potomac Formation, by F. H. Knowlton. — Mineralogy of the Pacific Coast, by W. H. Melville and Waldemar Lindgren. — Report of work done in the Division of Chemistry and Physics, mainly during the fiscal year 18S6-'87. — A Report on the Thermo-Electvical Measurement and High Temperatures, by Carl Barus. — The Greenstone Schist Areas of the Menominee and Marquette R-,gious of Michigan, by George H. Williams; with an introduction by R. D. Irving. — Bibliography of the P.aleozoiC Crustacea, by A. W. Vogdes. — The Viscosity of Solids, by Carl Barus. — Author-Catalogue of Contributions to North American Geology, by N. H. Darton. — On a Group of Volcanic Rocks from the Tewan Mountains, New Mexico, and on the occurrence of Primary Quartz.in certain Basalts, by J. P. Iddiugs. — On the relations of the Traps of the Jura-Trias of New Jersey, by N. H. Darton. — Altitudes between Lake Superior and the Rocky Mouutains, by Warren Upham. — Mesozoic Fossils in the Permian of Texas, by C. A. White. STATISTICAL PAPERS. _ Mineral Resources of the United States [183.2], by Albert Williams, jr. 1833. 8=>. svii, 813 pp. Price .'■)0 cents. Mineral Resources of the Uuited States, 1333 aud 1831, by Albert Williams, jr. 1835. 8*^. xiv, 1016 pp. Price tiO cents. Mineral Resources of the Uuited States, 183.'>. Division of Mining Statistics and Technology. 1886. 8°. vii, 576 pp. Price 40 cents. Mineral Resources. of the Uuiced States, 1333, by David T. Day. 1837. S°. viii, 81.! pp. Price 50 cents. Miner.al Resources of the Uuited States, 1837, by David T. Day. 1833. S°. vii, 832 pp. Price 50 cents. In preparation : Mineral Resources of tlie United States, 1888, by David T. Day. The money received from tlie sale of these publications is deposited! in the Treasury, and the Secretary of that Department declines to receive bank checks, drafts, or postage stamps; all remit- tances, therefore, must be by POSTAL NOTE or MO^JEY order, made paj'able to the Librarian of the U. S. Geological Survey, or in currency for the exact amount. Correspondence relating to the pub- lications of the Survey should be addressed To THE Director of the United States Geological Survey, Washington, D. C. Washington, D. C, March 1, 1889. ADVEKTISEMENT. LIBRARY CATALOGUE SLIPS. United States. Deparlmeni of llie interior. {U. S. geological survey). Department of the interior | — | Monographs | of the | United States geological survey | Volume XIV | [Seal of the depart- ment] I ; Washington [ government printing olBce | 1888 I Second title: United States geological survey | J. W. Powell, i director | — | Fossil fishes and fossil plants | of the | triassic ' rocks I of I New Jersey and the Connecticut valley | by | John S. Newberry | [Vignette] | Washington | government printing office | 18-8 4°. xiv, 152 pp. 26 pi. NeT?rberry (John Strong). United States geological survey | J. W. Powell, director | — | ; Fossil fishes and fossil plants | of the | triassic rocks | of | New I Jersey and the Coniiectient valley | by | John S. Newberry | ; [Vignette] | ; Washington ] government printing office | 1838 I 4°. xiv, ]52pp. 20 pi. [United States. Department of the interior. {U. S. geological survey). Monograph XTY]. United States geological survey | J. W. Powell, director | — | Fossil fishes and fossil plants | of the | triassic rocks | of | New Jersey and the Connecticut valley | by | John S. Newberry | [Vignette] | Washington | government i:)rinting office | 1888 4°. :siT, 152 pp. 26 pi. [United States. Department of the interior. (U. S. geological survey). Monograph XIV]. DEPARTMENT OF THE INTERIOR MONOGRAPHS OF THE United States Geological Survey VOLUME XIY WASHINGTON CrOVERNMENT PRINTING OFFICE 18 88 UNITED STATES GEOLOGICAL SURVEY J. W. POWELL, DIRECTOR FOSSIL FISHES AND FOSSIL PLANTS TRIASSIC ROCKS NEW JERSEY AND THE CONNECTICUT VALLEY JOHN S. ISTE^V^^BERRY ? '" f WASHINGTON GOVEKNMENT PKINTING OFFICE 1S8S CONTENTS. Page. Letter of traxs.mittal ix Preface xi PAKT 1. Geological relations of the Triassic rocks op New Jersey and the Connecticut Valley. 1 Geological sketcli 3 Geological equivalents of our Triassic rocks 8 TART ir. Fossil fishes of the Triassic rocks of New Jeusky and the Connecticut A^alley ... 17 Fossil fishes IB List of Triassic fishes 23 Descriptions of genera and species 24 Genus Ischypterus Egerton 24 Iscbyptenis ovatns W. C. R 27 Marshii W. C. U 26 Agassizii W. C. K 30 micropterus, n. sp 31 tenuiccps, Ag. sp 32 fultus, Ag. sp 34 robustiis, n. sp 36 elegaus, n. sp 37 alatus, u. sp 37 modestus, n. sp 38 leuticularis, n. sp 39 lineatus, u. sp 40 niacropterus W. C. R 41 Braunii, u. sp 43 parvus W. C. R. (JIS.) 45 latus J. H. R 46 minutus, n, sp 48 gigas, n. sp 49 Genus Catopterus J. H. E '.. 50 Catopterus Redfieldl Egerton 53 gracilis J. H. R 55 minor, n. sp o" ornatus, n. sp 58 anguilliformis W. C. R 59 parvulus W, C. R 6u V VI CONTENTS. Page. Fossil fisbes — Continued. Geuus Dictyopyge Egcrton Gl DLctyopyge macriira Egertou 04 Genus Ptycholepis, Ag 65 Ptycbolepis Maisliii Newb ,. 66 Genus Acentropborus Traquair 67 Aceutropborus Cblcopensis, n. sp 69 Genus Diplurus Newb.;.. 70 Diplurus loiigioaudatas Newb 74 PARI III. Fossil plants op tuk Triassic rocks op New Jersey and the Connecticut Valley 77 Sketch of Triassic Flora 79 Descriptious of genera and species 82 Dendropbycus triassicus, n. sp 62 Baiera Miiusteriaua tlug 84 Equisetum Rogers! Schimper 65 Equisetuui Meriani (?) Brong 8G Scbizoneura planicostata Rogers sp 87 Pacbypbyllum simile, u sp 88 Pacbypbyllum brevifolium, n. sp... 89 Cbeirolepis Munsteri Scbiruper 90 Otozamites latior Saporta 90 Otozamites brevifolius F. Br 91 Cycadinocarpus Cbapini Newb., n. sp 92 Dioiiuites longifolius Emmons sp 92 Loperia simplex, n.sp 93 Clatbropteris platypbylla Brong 94 Palissya (?) sp 94 ILLUSTRATIONS, Page. Plate I. Ischypterus ovatus VV. C. R 100 II. Ischypterus Marshii W. C. R 102 III. Ischypterus Agassizii W. C. R 104 IV. Ischypterus micropterus New b 106 Fig. 1. Fish of average size 106 Fig. 2. Unusually largo individual 106 v. Ischypterus tenuiceps Ag 108 Figs. 1 and 2. Old individuals, showing extreme development of dorsal scales 108 Fig. 3. Young individual 108 VI. Fig. 1. Ischypterus robustus Newb HO Fig. 2. Ischypterus fultus Ag HO VII. Fig. 1. Ischypterus fultv-s Ag 112 Fig. 2. Ischypterus elegans Newb 112 Fig. 3. Ischypterus tenuiceps Ag 112 VIII. Fig. 1. Ischypterus alatus Newb 114 Fig. 2. Ischypterus alatiis Newb 114 IX. Ischypterus modestus Newb 116 Fig. 1. Young individual 116 Fig. 2. Mature iudividual 116 Fig. 3. Old individual " 116 X. Fig. 1. Ischypterus elegans Newb 118 Fig. 2. Ischypterus lenticularis Newb 118 Fig. 3. Ischypterus lenticularis Newh 118 XI. Ischypterus lineatus Newb 120 XII. Fig. 1. Ischypterus macrop!erus W. C. R 122 Fig. 2. Ischypterus micropterus Newb 122 Fig. 3. Ischypterus Brauuii Newb 122 XIII. Fig. 1. Ischypterus Braunii Newb 124 Fig. 2. Ischypterus Braunii Newb 124 Fig. 2a. Ischypterus Brauuii Newb 124 Fig. 3. Ischypterus latus J. II. R 124 Fig. 4. Ischypterus parvus W. C. R 124 Fig. 5. Ischypterus minutns Newb 124 Fig. 5a. Ischypterus minutns Newb 124 XIV. Fig. 1. Ischypterus elegans Newb Ii6 Fig. 2. Ischypterus elegans Newb 126 Fig. 3. Ischypterus gigas Newb 126 XV. Catopterus Redfieldi Egertou 128 Fig. 1. Broad form, old individual 128 VII VIII ILLUSTRATIONS. Page. Plate XV. Fig. 2. Normal form, youug individual 128 Fig. 3. Part of niatiivc lisli, s ho wing dot nils of iins IS.-^ XVI. Fig. 1. Catoptenis gracilis J. II. K 130 Figs. 3 aud 3. Catoptorns gincilis ,T. II. I?.; po.stcrior portions of bodies of two indi- viduals of sleudor form 130 Fig. 4. Catoptcinsparvuliis W. C.R 130 Fig. 5. Catoptenis iiarvuliis W. C. R 130 XVII. Catoptenis minor Kewb 133 XVIII. Fig. 1. Dictyopygo macrura Egertou 134 Fig. 2. Dictyopyge macrura Egertou 134 Fig. 3. Catoptenis ornatus Newb 134 Fig. 3a. Scales of dorsal line enlarged 134 Fig. 3b. Scales of side enlarged 134 Fig. 4. Catoptenis ? sp. young 134 Fig. 5. Catopterus anguilliformis W. C. R 134 XIX. Fig. 1. Ptycholepis Marsbii Xewb 13G Fig. 2. Ptycholepis Marsbii Xewb 130 Fig. 2a. Scales, enlarged 131') Fig. 3. Acentropliorus CUicopensis Ncwb ISO Fig. 4. Acentropliorus Cbicopensis Newb 13fi XX. Dijilurus longicaudatus Newb 138 Fig. 2. Supplemental caudal fin 138 Figs 3 and 4. Scales, outside 138 Fig. 5. Scales, inside - 138 XXI. Dendropbycus Triassicus Kewb 140 Fig. 1. Hasal portion of a frond 140 Fig. 2. Extremities of a group of braucbcs 140 XXII. Fig. 1. Baiera Miinsteriana '? Uug 142 Fig. 2. Pacbypbyllum simile Xewb -■ 142 Fig. 3. Pacbypbyllum brevi folium Ncwb 142 Fig. 3a. Twig with divergent leaves 142 Fig. 3b. Twig wilb apprcsscd leaves 142 Fig. 3c. Cone 142 Fig. 4. Clieirolepis Miiusferi, Scbeuk sp 142 Fig. 4a. Cone scale 142 Fig. 5. Pbragma of Equisetiim Rogersi Scbimp 142 Fig. 5a. Pbragma of Equisctum Rogersi Scbimp, under side 142 Fig. G. Clatbropteris platypliylla ? Brong 142 XX III. Baiera mnltifida Fontaine 144 XXIV. Fig. ]. Otozamites latior Saporta; base of frond, natural .size 140 Fig. 2. Otozamites latior Saporta ; summit of frond 146 Fig. 2a. Portion enlarged to sbow nervation 146 Fig. 3. Otozamites brevifolins Fr. Braun 146 Fig. 4. C'ycadinocarpus Chapini Newb 146 XXV. Fig. 1. Loperia simplex Ncwb.; main stem much flattened 148 Fig. 2. Loperia simples Newb. ; branch aud leaves 148 Fig. :i. Loperia simplex Ncwb. ; braucbcs springing from stem 148 Fig. 4. Dioouites longifolius, Emmons sp 148 XXVI. Trunk, of conifer (Palissya ?) - 150 LETTER OF TRANSMITTAL. Columbia Collegf, New York, August 1, 1887. Sir: I have the honor to transmit herewith a memoir prepared at your request on the Fossil Fishes and Fossil Plants of the Triassic Rocks of New Jersey and the Connecticut Valley. With great respect, your obedient servant, J. S. Newberry. Hon. J. W. Powell, Director U. S. Geological Survey. IS. PREFACE. It is hoped by the author that the following pages will do something to supply what has long been felt to be a want in American geology: a better knowledge of the fauna and flora of the Triassic rocks of eastern North America. These rocks probably fuj-nished the first fossils collected on this continent — fossil fishes from Durham and Sunderland, in the Connecticut Valley; fossil plants from the coal basin of Richmond, Va.; and, still more interesting, the wonderful series of so-called bird tracks fii'st noticed at Turner's Falls, Mass. A few of the fossil plants of Virginia Avere described by Prof. W. B. Rogers in the reports of the Association of American Geologists and Natu- ralists, 1843, and by Mi-. C. J. F. Bunbury in the Quarterly Journal of the Geological Society of London, volume 3, 1851, and some notices of the fossil fishes, with brief descriptions of certain species, were published by Agassiz, Sir Philip Egerton, and Messrs. W. C. and J. H. Redfield, at various times between 1838 and 18.56. Many figures and descriptions of the remains of both plants and animals were also published by Prof. Ebenezer Emmons in his Geological Report of the Midland Counties of North Carolina, in 1856, but, though deservedly eminent as a geologist, Professor Emmons had little acquaintance with paleontology, and this contribution rather increased than satisfied the desire for more thorough knowledge of the life of the At- lantic coast in Mesozoic times. No systematic collection nor thorough study of the fauna or flora of the formation as a whole was attempted until about 1880, when Prof. W. M. Fontaine, of the University of Virginia, began a careful review of the fossil plants of the Virginia and North Carolina XI I TRIASSIC FISHES AXD PLANTS. Mesozoic coal basins. His results were published in a memoir on The Older Mesozoic Flora of Virginia, which was issued in 1883, as volume 4 of the Monographs of the U. S. Geological Survey. This threw a flood of light upon the vegetation of the Atlantic coast in the Mesozoic ages and established beyond question the parallelism of our New Red Sandstone with the Keuper of Eui'ope; a matter which liad been much debated, with somewhat discordant conclusions, by Hitchcock, the brothers Rogers, Lyell, Marcou, and Emmons. Thus one of the wants which lias been referred to was satisfactorily supplied ; but the animal remains found in our Triassic rocks are still to be systematically reviewed. The immense series of tracks of terrestrial ani- mals found on the old beaches of the Triassic estuaries — the autographs, as I have elsewhere called tliem, of at least one hundred different kinds of bipeds and quadrupeds of diverse sizes and structures which inhabited the eastern coast of North America in the Triassic age, and left little other rec- ord of their existence — though beautifully illustrated by Hitchcock and Deane, are still as mysterious and tantalizing as ever. Comparatively few bones of the animals themselves have been met with up to the ^Ji'esent time ; but these confirm the conclusions, drawn from the remains of terrestrial Mes- ozoic animals found elsewhere, that the tracks referred to were not made by birds as first supposed, but by reptiles or amphibians. Doubtless in fu- ture years some Mesozoic cemetery will be discovered like those of Tilgate Forest in England and Bernissart in Belgium, where the abundance of ver- tebrate ramains and the perfection of their presentation will permit the re- habilitation of this interesting fauna. Tlie fossil fishes of our Triassic rocks have long needed a fuller ex- position than had thitlierto been given to tliem. The materials upon which the Messrs. Redfield based their important contributions to our knowl- edge of this group of fossils were incidentally collected from surface expos- ures and were necessaril}- limited in quantity, the fossils themselves were generally fragmentary and imperfect, and an interval of more than thirty years has elapsed since their last communication was made on this subject. Having long been interested in the Paleozoic fishes of Ohio, when I came to New York to reside and began to form a geological museum at Columbia PEE FACE. XIII College my attention was naturally directed to these, the most striking fos- sils to be found in the vicinity. After obtaining by purchase good speci- mens of all the species to be had in the valley of the Connecticut, men were employed to make excavations in the fish beds at Boonton, N. J., and from that locality many hundreds were obtained in a good state of preser- vation. These, with those procured elsewhere, gave me much more and better material for study than had been accessible to any one else who had been interested in the subject. The accumulation of this new material made it apparent that our Triassic rocks contained some genera not before found there and a larger number of species than had before been described. In order to identify these I examined the collections made by the Messrs. Red- field, all of which were courteously placed at my service by Prof 0. C. Marsh, as were those in the American Museum of Natural History by Prof. R. P. Whitfield and those in the cabinet at Amherst College by Prof B. K. Emerson. My first intention was simply to identify the species which had come into my possession, but I soon found that to do this satisfactoi'ily all the literature of the subject and all accessible material, both old and new, must be passed in review. AVhen this had been done it seemed to me that the facts I had gathered would be a valuable contribution to American geology if they could be put into shape and published, and the following memoir is the result of an effort in that direction. Many circumstances have rendered my task a difficult one. I have en- deavored with sincere loyalty to my old friends W. C. and J. H. Redfield, father and son, to secure to them as far as possible the fruit of their study of our Triassic fishes; but, from tJie limited amount and the imperfect pres- ervation of the material in their hands and the brevity of their descriptions • it has not always been possible to identify and accurately define their spe- cies. Besides this, many of the specimens which served as their types were burned with the other collections of the New York Lyceum of Natural History, and of the specimens remaining in the Redfield collection the greater part are without other labels than numbers to which no correspond- ing catalogue has been discovered. For these reasons I feel that in regard to specific distinctions my work is imperfect and is liable to modification with the gradual accumulation of more and better material. T have XIV TRTASSIC FISHES AND PLANTS. thought, however, that the fossil fishes from our Triassic rocks which have been lying in the cabinets of our colleges and collectors should if possible be made available for the use of teachers and students, and should contribute their part to illustrate what is perhaps the most interesting and j^et least known epoch in the geologic history of North America, that of the Trias. As a contribution to this history I have accumulated by far the largest col- lection of our Triassic fishes yet made, have studied them with some care, have labeled them plainly according to my lights, and have placed them in the Geological Museum of Columbia College, where they will be safe, since it is fire-proof, and where they will be accessible to those who shall carry to completeness the investigations of these fossils begun by the Messrs. Redfield, continued by me, and to be concluded by generations yet to come. In conclusion I desire to express my obligations for assistance and courtesies received from Professors Marsh, Whitfield, and Emerson ; to Mr. S. W. Loper, of Durham, Conn., Avhose enthusiasm in collecting and skill in developing such specimens as are found in his vicinity liave resulted in con- tributing to several museums the best examples of our Triassic fishes known to exist ; also to Mr. I. C. Russell, to whose intelligent supervision of the excavations made at Boonton, we owe the large amount and excellent con- dition of the material obtained there. I>^RT I. GEOLOGICAL RELATIONS OF THE TRIASSIC ROCKS OF NEW JERSEY AND THE CONNECTICUT VALLEY. MON XIV 1 GEOLOGICAL SKETCH. The rocks which inclose the fossils described on the following pages occupy a series of detached areas extending interruptedly from Nova Scotia to North Carolina. They are in the form of basins, with their longest diam- eters northeast and southwest, parallel with the bearing of the ridges of the Alleghany Mountain belt Indeed, they seem to have been deposited in troughs lying between the most easterly and lowest of these ridges ; troughs that were for ages occupied b}' fresh or brackish water lakes or estuaries, the surface drainage of the adjacent country. After the Carboniferous age the whole region between the Mississippi and the Atlantic was raised above the ocean level, where it has remained with little variation of altitude to the present time. Of this belt of elevated country, which reached from the Green Mountains and the Adirondacks to the Gulf, the most easterly portion was much tlie older. This was formed by the! Blue Ridge and the Hudson Highlands, with one or more parallel ridges on the east, which have since been depressed. The Alleghanies proper were added toward the close of the Carboniferous age. In Mesozoic times all this broad belt of highland was suffering erosion, and the material removed was carried away by the draining streams, both east and west, either in suspension or in solution. That which was dissolved flowed off into the somewhat distant oceanic basins, where it was deposited by organic agencies as limestone or flint, as it was lime or silica; while the suspended material was spread as sand, clay, and gravel ever what are now the plains of the Mississippi Valley and what were then water-filled basins along the Atlantic coast. As I shall show farther on, the first of the Mesozoic strata to which we have access — composed of materials removed fi'om the ancient 3 4 TEIASSIC FISHES AND PLANTS. land mentioned above and sjDread round its margins — were laid down during the last half of the Triassic age, and a subsequent subsidence caused these to be covei'ed by finer and more calcareous sediments during lhe last half of the Cretaceous age. With the Mesozoic deposits of the interior of the continent we have now no immediate concern, as the task before us is to trace the histoiy of those which accumulated on its eastern slope, and especially those formed during the Triassic age, viz : the shales, sandstones, and conglomerates, which were deposited in the lakes and bays already referred to. Of these Triassic areas the most northerly is that of Nova Scotia, about the Bay of Fundy, Prince Edward Island, etc. The second is that of the Connecticut Valley, which reaches from the north line of Massachusetts to Long Island Sound. The third, which may be called the Palisade area, extends from Rockland County, N. Y., to Orange County, Va., a distance of about three hundred and fifty miles. This area, which has the form of a long and narrow trough, is bounded on the west by the Blue Ridge, and on the east by the Archaean rocks of the Staten Island, Trenton, and Philadelphia axis. In Viro'inia and North Carolina are several distinct and smaller basins lying eastward of the Palisade area, some of which contain coal beds of economic value. The Triassic rocks which fill these basins are alike in this, that they consist chiefly of beds of conglomerate, sandstone (sometimes arkose), and shales, interstratified with heavy beds of diabase. The pre- vailing color of both sandstones and shales is red or reddish, but the Con- necticut area includes layers of nearly black shale charged with carbona- ceous matter, containing many remains of fishes and ijhuits, and even some thin films of coal. Also a small part of the series in New Jersey consists of dark or dove colored shales charged with organic matter, sometimes crowded with the remains of fishes, and exhaling a marked bituminous odor when struck with a hammer. It should also be said that a small detached basin of Triassic rocks in Southbury, Conn., includes a thin sheet of impure lime- stone. The sandstones of the series are frequently firm and massive, and are extensively used as building stone, important quarries having been for many years worked at Long Meadow, Mass., Portland, Conn., and Newark, N. J. GEOLOGICAL SKETCH, 5 These quarries have furnished much of the building material employed in our cities, and the fronts of fully one-half of the residences of New York are composed of " brownstone," as it is called, derived from them. The thickness of the Triassic series in the Connecticut and Palisade areas is 5,000 feet or more, and the arrangement of the strata among them- selves is peculiar and as yet not satisfactoril}^ explained. In the Connecticut area the rocks all dip toward the east, the outcropping edges of the trap sheets left in strong- relief by the erosion of the associated beds forming the bold escarpments of Exst Rick and West Rock, at New Haven, Conn., Mount Tom and Mount Holyoke, in Massachusetts. In New Jersey, on the contrary, the dip of the Triassic beds is generally towards the west at an angle of from three to fifteen deg'rees with the horizon, and the edges of the trap sheets form bold cliffs, which face the east and constitute the sum- mit of the ridges known as the Palisades, First and Second Newark Mount- ains, etc. In addition to these sheets of trap the strata are cut by many dikes of diabase, which cross them vertically or at a high angle. The origin of the singular structure I have-- described has been much discussed. Many of the beds show ripple-marks, sun cracks, and i-ain-drop impressions, which prove that they were once beaches or mud flats, some- times exposed to the air. The}' are also frequently impressed by the tracks of large and small animals, generally three-toed, but sometimes showing four or five digits. These Avere at first supposed to be for the most part the tracks of birds, but are now believed to have been made by reptiles and amphibians. We have here autographs of perhaps one hundred diff"erent kinds of animals of wliich scarce any other record has been discovered, a few scattered bones and one or two imperfect skeletons being all yet found of the creatures themselves. Everything indicates that these tracks were made by animals that frequented the shores of bays and estuaries where the retreating tide left broad surfaces which were their feeding grounds. Inasmuch as many successive beds show ripple-marks, sun cracks, and tracks, the conclusion seems inevitable that the areas where these strata were de- posited Avere slowly sinking and that the . land-wash spread by the tide constantly formed new sheets, iipon which fresh records were inscribed. The downward movement must have been very slow, for it apparently about 6 TEIASSIC FISHES AND PLANTS. kept pace witli the accumulation of material ; but it was not regular, as we find alternations of conglomerate, sandstone, and shale, which were deposited in water of different depth. The uniform dip of the Triassic strata in each basin and the opposite inclination of those in the Connecticut and the Pali- sade areas have been variously explained. One theory proposed by Prof H. D. Rogers^ is that the strata were laid down upon a slope. This theory might be true of the coarse material deposited along an inclined shore, but could hardly be applicable to the finely laminated shales and limestones which accumulated in deep water. Another view, advocated by Mr. I. C. RusselP is that the Triassic basins of the Connecticut Valley and New Jersey were once connected, and the strata were deposited continuously overall the area between them; that subsequently the central portion of this area was elevated and the Triassic rocks were eroded from- it, leaving the two sunken margins occupied by beds which rise respectively from the east and west towards the central elevated area. Mr. Russell has studied the structure of the Triassic area of New Jersey with much care, and in his paper he presents many facts and a strong array of arguments in favor of his theory. There are, however, two difficulties which suggest themselves and call for further investigation of the problem before we can accept this solution. These difficulties are — First. Not a trace of Triassic rocks is now anywhere visible in all the broad belt between the Hudson River and the Connecticut Valley, and it seems scarcely possible that if they once covei'ed this belt to the depth of many thousand feet they should have been so completely removed. Second. It has been found that on western Long Island, opposite the center of the New Jersey Triassic area, the crystalline rocks, which are a continuation southward of those lying between the Hudson and the Con- necticut, are covered unconformably by Cretaceous strata, with no Trias between them. This would indicate that the Trias of the New Jersey basin never reached over that portion of the divide. Still another theory has been proposed by Prof W. M. Davis to ac- count for the structure of the Triassic beds of the Connecticut Valley. This ' Final Report, Oi'ology of New Jersey, 1840, pp. 1(16-171. -Annals New York Academy of Sciences, vol. i, p. 220. GEOLOGICAL SKETCH. 7 was presented by liim to tlie geological section of the American Association at its meeting at Buffalo, in 1886, and was subsequently published in the American Journal of Science for November of that year. This theory sup- poses that the Triassic rocks were once deposited horizontally, or nearly so ; that the trap sheets were overflows and not intrusive, and that all were bi'oken and inclined by a series of faults, the result of lateral pressure ; this pressure affecting primarily the underlying crystalline rocks, which, stand- ing nearly vertical, slipped on each other, causing a series of fractures and uplifts along their lines of strike. Professor Davis has worked out this theory with much ingenuity and ability, and, if it shall be found upon further ex- amination that the series of faults which he depicts do really traverse the Triassic rocks, we shall be indebted to him for the solution of wliat has been one of the most difficult problems in American geology. But there are some facts which are apparently incompatible with its universal application. No such faults as Professor Davis supposes to exist are discoverable in the localities I have had an opportunity of exaniiniug since the promulgation of his views, viz : along the Palisades, and at East Rock and West Rock, New Haven; and some of the trap sheets are certainly intrusive, having baked the beds on both sides of them. The materials of which the Triassic beds are composed are all pre- sumably, and in part at least demonstrably, derived from the adjacent highlands. In New Jersey the conglomerates are made up of rolled frag- ments of the granitoid rocks of the neighboring hills, and the sandstone, arkose, and shale apparently represent the different stages of mechanical decomposition of the quartz and feldspar of the granite. The New Jersey highlands, as well as other portions of the Blue Ridge belt, are known to contain great quantities of iron ore, and the erosion of the gneiss which forms this belt must necessarily result in the distribution of a large amount of iron Hence it is not surprising that the shales and sandstones all con- tain enough of this element to give them a red or reddish color whenever it is in the form of the anhydrous peroxide. The fact that it is generally in this condition, and therefore that the rock is red, proves that it contained little or no organic matter when deposited ; for whenever decaying organic matter is present in any considerable quantity it reduces the peroxide of 8 TRIASSIC FISBES AND PLANTS. iron to protoxide, and makes the color, so far as influenced by the salts of iron, gray, green, or bUie. Where the organic matter is in very large quantity it imparts the characteristic color of carbon, and makes the shale or limestone which contains it black. The general absence of organic matter in the Triassic rocks is doubt- less due to the circumstances under which they were deposited ; that is, in brackish water, which is always unfriendly to life, and perhaps was sub- ject to high tides, which caused physical commotion, another unfavorable condition. We can imagine the circumstances attending the accumulation of the Triassic sediments to have been somewhat like those which now prevail in the Bay of Fundy, where the advance and retreat of a bore, or very high tide, keeps the water always in violent motion and turbid, and the alternating extremes of ebb and flow forbid the occupation of the littond zone by either animals or plants. The gray and blue shales of Boonton and Sunderland contain an abundance of organic matter, of which sufficient would be furnished by the fishes to partially deoxidize the iron deposited with them, while the black shales of Plainfield and Weehawken, N. J., and Durham, Conn., are colored simply by the abundance of carbonaceous matter. An illustration of the truth of the views here proposed is found in the dif- ference between the colors prevailing in the Palisade area and the Richmond basin. In the former the rocks are, as lias been stated, generally very barren of fossils and the color is mostly reddish, while in the latter the quantity of organic matter is large and the color of the rock is blue, gray, or black. GEOLOGICAL EQUIVALENTS OF OUR TRIASSIC ROCKS. The aofe of the series of rocks whicli have been called Triassic on the preceding pages has been mucli discussed. Maclure considered them the (Equivalent of the Old Red Sandstone of Scotland, being influenced by the similarity. of their lithological characters. Mr. Richard C. Taylor, for a time at least, entertained the opinion that the group of rocks we are now considering belonged to the Coal Measures, being led to this conclusion by the presence of coal beds in the Richmond basin and the general resemblance to the Coal Measures of Penns3'lvania exhibited by the associated rocks. A single one of the abundant fossil GEOLOGICAL SKETCH. 9 plants which occur in the Richmond coal basin would, however, have been sufficient to show the error of this opinion. Prof. Edward Hitchcock, who was one of the earliest to consider the subject, arrived at the conclusion that our Red Sandstone series was the equivalent of the New Red Sandstone of Europe. To this he was led mainly by the similai-ity of their lithological characters and their position relative to the Carboniferous rocks below and the Cretaceous above. He also mentions the discovery in these rocks of portions of a vertebrate skeleton which was not a fish, and he inferred from that fact that the series was Mesozoic, because at that time no animals of hio'her rank than fishes had been found in the Paleozoic rocks. We have since learned that am- phibians are common in the Coal Measures, and the remains of reptiles are not wanting. At the time Professor Hitchcock wrote, the Permian and the Trias were not separated, but both were included in the so-called New Red Sandstone. This term was used to designate the group which, containing much Red Sandstone, rests on the Carboniferous, and to distinguish it from the Old Red Sandstone below. So Professor Hitchcock supposed that we had in the rocks under con- sideration the equivalents of the Rotheliegende, as well as of the Bunter and the Keuper. The same view was taken later by Prof. Ebenezer Em- mons in his Geological Report of the Midland Counties of North Carolina, 1856, page 273, receding from an earlier opinion (1853) — when he called the whole series Triassic — for the reason that in the lower portion of the Dan River section. North Carolina, he found the remains of Thecodont saurians. On the other hand. Profs. W. B. and H. D. Rogers were led by the general resemblance of the ferns and cycads of the Richmond basin to those of tlie Lias of Whitby, England, to consider these rocks Liassic, that is. Lower Jurassic. This view was also shared by Sir Charles Lyell when, in 1845, he visited the Richmond coal basin and collected a series of fossil plants, which were examined by C. J. F. Bunbury, who had given much attention to fossil botany.-^ Prof. Jules Marcou, on his Geological Map of the United States, pub- lished in 1853, represents the New Red Sandstone of Virginia as Liassic, 'Quart. Jour. Geol. Snc. London, vol. 3, 1847, pp. 201-288. 10 TRIASSIC FISHES AND PLANTS. but in his Geology of North America, 1858, he claims that the Richmond coal basin is of the age of the Keuper or Upper Trias, although he accepts the view of Professor Emmons, that the coal series of Deep and Dan Rivers, North Carolina, is at base Permian. On page IG of the Geology of North America, Prof J. Marcou publishes a letter of Prof Oswald Heer, of Zurich, written July 25, 1857, in which he reviews the fossil flora of the Richmond and North Carolina coal basins, and regards it as contempo- raneous with that of the Keuper. In October, 1857, Sir Cliarles L5-ell, writing to Professor Marcou (loc. cit.) quotes a note from Mr. Bunbury upon this subject, in which, referring to his paper on the Riclmiond plants^ where he had expressed the opinion that the formation containing them might belong to the Jurassic or to tlie Tri- assic period, and that it might, with almost equal plausibility, be referred to either, he says: "At the time I wrote this the Basle and Baireuth beds were supposed to be Lias." Professor Marcou comments on this as follows: "As the Basle and Baireuth beds are now recognized by every geologist as belonging to the Keuper, it will appear that Bunbury never intended to put the Virginia coal field in the true Jurassic of England; so that we all ao-ree to regard the Red Sandstone of Virginia and North Carolina as Keuper." In 1883 Prof. W. M. Fontaine published the results of a careful study made by himself of the flora of the coal series of Virginia in a monograph issued by the U.S. Geological Survey, with the title "Contributions to the Knowledge of the Older Mesozoic Flora of Virginia." In this monogra])h (pp. 122, 123) he enumerates thirty-nine well-defined species of plants, of which 23 per cent, are peculiar to North Carolina, 41 per cent, are found in Virginia, 20 per cent, are allied to or identical with Jurassic forms while the number of species identical with or allied to Rhsetic plants amounts to 38 per cent.; or, as he sajs:^ Assuming with Feistmantel that the Eivjinahal group of India is of Liassic age, we have two species identical witli and six nearly allied to Jurassic plants, while seven species are identical with and eight closely allied to Rhajtic plants. This shows that the plant-beai'ing strata of Virginia and North Caro- lina maybe safely considered equivalent in age to the Rhfetic beds of Ger- ' Qnar. Jour. Geol. Soc. London, vol. 3, 1847, p. 288. =0p. cit., p. 123. GEOLOGICAL SKETCH. 11 many ; a result confirming and further illustrating the conclusions of Heer, Marcou, Bunbury, and others, who have regarded the RichmoFid coal series as Upper Trias. The Rhsetic, formerly included in the Keuper, is known to form beds of passage between the Trias and the Lias, though with a still prevailing Triassic facies. We are not yet in possession of the material necessary for making an exact comparison between the rocks of the southern Triassic areas and those of New Jersey and Connecticut. No considerable collection of the fossil fishes of the Richmond and North Carolina basins has been made, thousrh they are known to abound there ; but the few fish remains from North Carolina and Virginia which have come under my observation show that there are marked differences between the faunas of the northern and south- ern Triassic basins. On the other hand, the plants thus far collected in New Jersey and Connecticut are few — since they are not common in any locality yet known — while plants are by far the most striking and abun- dant fossils in the Virginia and North Carolina basins. They have been gathered by many collectors, and have been now studied by Professor Fon- taine and described in the monograph referred to above. More fishes from the southern areas and more plants from the northern must therefore be collected before a satisfactory comparison can be made. So far as they throw light upon this subject, the facts already gathered indicate a general parallelism between the northern and southern areas ; some differences, but many points of identity being discernible. For example, the Richmond coal basin has furnished to me one species of Catoptenis (C. gracilis J which is common in New Jersey, but by far the most abu ndant fish of the Rich- mond basin is Dictyoi)yc]e macrura Egt., which has not yet been found in tlie northern basins. Traces of tv.'o other genera and species unknown at the North have been obtained from Richmond. It is probable that the large fish of which a fragment is figured by Sir Philip Egerton,^ and called a Tetragonolepis, is an Ischypterus, identical with the large and broad spe- cies (/. ovatus) which occurs at Sunderland and Boonton. Whether the other species of fishes belonging to the genera Catopter vs, Isclyptenis, Bip- Inrus, etc., found in the northern basins, will be obtained at the South wlien ' Quart. Jour. Geol. Soc. London, vol. 3, 1S47, pi. 9. 12 TEIASSIC FISHES AND PLANTS. they shall be carefully sought for, remains to be seen. A conclusive com- parison cannot be made, however, till a thorough search for fishes is insti- tuted in the southern basins ; for, while we do not alwaj'S find the things we seek, we make full collections of those fossils only which are the objects of special search. A comparison of the fossil plants of the northei'n and southern basins is somewhat more satisfactory, and yet the limited number of species obtained at the North leaves the result of such comparison far from conclusive. Pro- fessor Fontaine, as has been mentioned, has enumerated about forty species of plants obtained from the Riclnnond and Nortli Carolina coal basins. Among these perliaps the most abundant is the large monophyllous fern Tceniopteris viagnifolia of Rogers, but this has not yet been found anywhere at the North, nor has any other similar fern been met with there. Another common plant at Richmond is Scliizoneiira lilanicostata {Calamites planicos- tatiis of Rogers), and this I have found at Milford, N. J.; Durham, Conn ; and Sunderland, Mass. An allied plant is JEqiiisetum Rocjersi, Schimper (Eqiiiscttim coluntnare of Brongniart and Rogers). Professor Funtaine says:^ "This plant is one of the most characteristic fossils of the Richmond coal field, and lias a wide vertical and horizontal range." He further says that it is almost everywhere found with Macroteniopteris magnifoUa, and that they form the only fossil plants of some localities. This plant is rare at the North, as I have obtained it from but one locality — Milford, N. J. Another common plant at Richmond is a fern, belonging to the genus Clathropteris, which Fontaine" identifies with G.plaUjphjUa var. expansa Saporta, and with ('. rcc- tiusculus of Edward Hitchcock, jr., described in the American Journal of Science for July, 1855. This plant occurs rather abundantly at Durham, Conn., and at East Hampton, Mass. We find also at the foi-mer place a delicate and slender Baicra, which may not be distinct from tha.t obtained by Ennnons in North Carolina and figured by him in his American Geology, Part 6, page 133. Another common plant at Durham has a simple flattened stem from half an inch to an inch in width and sometimes a foot or more in length. This is apparently identical with that figured by Emmons,^ and named by Fontaine Bamhiisiiim CaroUiicnsc. ' Mou. cited, p. 12. = Ibid , p 54. ^ Auieiicau Geology, ixut vi, i>. 132. GEOLOGICAL SKETCH. 13 Among the conifers apparently two or three are common to the northern and southern Triassic basins. Pallssya Braunil, Endl., occurs in North Car- oUna, and a fine specimen of it is figured by Fontaine on Plate L of his .monograph. I have a still finer specimen from the quarries at Newark, N. J., where it seems to be common; and, although the coarse sandstone has not often preserved the foliage, what I suppose to be portions of its trunks and branches are very numerous. The plant which Fontaine considers identical with Cheirolejns Mi'msteri, Schimper, is found at Durham and many other places in Connecticut, as well as in Massachusetts and New Jersey. As I have shown in my notes on the Triassic plants, this is probably not a Cheirolepis, but a Pachi/phi/Uum; but there is no question of its occurrence in all the northern and southern basins. On the other hand, among the small number of plants froni the Trias of New Jersey and Connecticut are two or three which have not yet been found at the South. Of these the most important is a species of Otozamites, which is rather common at Durham, but not yet found elsewhere. Its fronds- are one' to two feet in length by one to three inches wide. Wlien it was first found, many years ago, I was unable to distinguish it fr.im Otosamites brevi- Jolius Fr. Braun, one of the most characteristic plants of the Rhaetic beds of Bamberg, Baireuth, and other places. Recently Count Saporta has sep- arated the larger fronds with narrow pointed pinnules from the smaller with shorter rounded pinnules (all of which were formerly attributed to 0. brevifolms), and has made them the type of his species Olozamites latior} These correspond precisely in size, form, and nervation with our Durham plant, and we may therefore accept this as another species common to our Triassic beds and the Rhsetic of Germany and France, contributing an ad- ditional fact to the already sufficient proof of the parallelism that has been before reported. The relations of the Triassic beds of the Atlantic coast to those of the in- terior and the western margin of the continent can hardly be established without larger collections of fossils from western localities. The Triassic strata underlying the Indian Territory, northern Texas, New Mexico, etc., 1 Pal^ontologie fraiieaise, Veg6taux, vol. 2, p. 130, Pis. 97, 98.) 14 TEIASSIC FISHES AND PLANTS. are peculiarly barren of fossils. They are generall}^ reddish sandstones, con- glomerates, and shales below, with a series of highly colored indurated marls or fine-grained calcareous sandstones above, frequently charged with salt and sometimes including extensive sheets of gypsum. The sandstones are also conspicuously cross-bedded, and it is evident that the whole series was deposited in a shallow sea, swept by strong currents or high tides, and that bays, estuaries, or lagoons were formed at various times, in which the water was evaporated and its salt and gypsum were precipitated. These condi- tions were unfavorable to the presence of animal or vegetable life : as con- sequences, we rarely find any fossils in the beds, and the iron they contain is peroxidized, imparting to them their characteristic red color. This great sheet of Triassic rocks originally extended to the Wasatch Mountains, which formed the western shore of the sea in which they were deposited. Passing over hundreds of miles where these Triassic rocks were just beneath the surface and freely exposed in cliffs and stream beds, I have sought for months in vain to find in them any traces of life ; yet in two localities which I visited I was more successful, and from a third I have received a large collection of fossil plants. These localities are San Jos^, near Pecos, in New Mexico, the old copper mines above Abiquiu, and Los Bronces, on the Yaki River, in Sonora. At the first locality are found Walcliia and Ccdainites below, which mean Permian, and in softer beds of sandstone above — doubtless Triassic — impressions of fern fronds too indistinct for de- termination. In the roof shales of the old copper mines near Abiquiu plants are abundant, but the number of species is small. Of these the most common and conspicuous is an Otosamites with broad truncated pinnules, which I have called 0. Macomhii; another cycad less common is a Zamites (Z. occi- dentaUs Newb.), while twigs and cones of PachyphyUum are occasionally seen. At Los Bronces the number of species is much larger, and we find among them several which occur in North Carolina, and one of those ob- tained from Abiquiu {Otozamites Macomhii). The Carolina species are Pe- copteris buUatus Bunbury {3fertensides huUcdus Fontaine), Pecopteris falcatus Emmons (Laccopteris EmmonslFontame), and Tceniopteris magmfolia Rogers. These indicate a parallelism between the plant-bearing beds of the Atlantic GEOLOGICAL SKETCH. 15 Trias and those of New Mexico and Sonora, and go far to prove tliat all our Triassic rocks which have yet 3nelded plants belong to the uppermost division of the system. In New Mexico there are at least two tliousand feet of sandstones and shales belonging to the Trias beneath the strsxta whicli contain the fossil plants at the copper mines. Immediately above the latter lie the sandstones of the Dakota group, the basal member of the Cretaceous system as repre- sented in that region ; so .that we have proof that these plant beds form the extreme upper part of the Trias. The lower beds of sandstone and the con- glomerate which forms the base of the series in New Mexico and Arizona may represent the lower portions of the Trias in the Old World, but unfor- tunately no fossils have yet been obtained from them. Many writers upon the Triassic beds of the West have called the whole formation Jnra-Trias, either under the impression that both systems were represented in the group, or as a matter of precaution in case this should be found to be true. There seems, however, to be no good reason for think- ing that the series of rocks which I have described represents the Jurassic of Europe. Another set of beds overlying the Triassic and underlying the Dakota sandstones occur in Utah, Colorado, and Wyoming, and are proved by their fossils to be Jurassic. But these beds wedge out toward the south, and I have been unable to find any traces 0/ them south of En- chanted Springs, near the lower line of Colorado. They consist of — (1) Gray earthy limestone witJi marine Jurassic moUusks, best shown in Wyoming and Utah. (2) Light sandy and gypsiferous strata which succeed the limestone to- ward the south, and (3) Alternations of reddish sandstones and shales— fresh-water bed.s, con- taining unios and saurian bones — the Atlantosaurus beds of Marsh. On this series rest the Dakota sandstones, and below it are the red sand- stones and conglomerates of the Trias. F^RT II. FOSSIL FISHES OF THE TRIASSIC ROCKS OF NEW JERSEY AND THE CONNECTICUT VALLEY. . MON XIV 2 17 FOSSIL FISHES, The fishes of our Triassic rocks, though so far as yet known repre- senting but six genera and about twenty-five species, are locally very numerous and are found in many localities. They were among the first fossils which attracted the attention of American geologists, and were re- ferred to by Mitchill, De Kay, and Hitchcock half a century ago. Some imperfect specimens, obtained in the Connecticut Valley, near Amherst, were sent by the latter to Professor Agassiz when he was publishing liis great work on fossil fishes. He figured and described two species, one of which he referred to the genus Palcconiscus and named P. fuUus; the other he considered a Eurynotus and called it E. tenidceps. About this time Mr. W. C. Redfield, of New York, began the study of these fishes. In con- nection with his son, Mr. J. H. Redfield, he published many notices of them during the succeeding twenty years. The first formal description of any of them was in a paper read by Mr. J. H. Redfield before the Lyceum of Natural History of New York, December 12, 1836, and subsequently pub- blished.^ It contains figures and descriptions of two species, Catopterus gracilis and Pahwniscus latus, the former being made the type of a new genus. In 1841 Mr. W. C. Redfield read before the Yale Natural History Society a paper entitled "Short Notices of American Fossil Fishes." This contains descriptions of five species of Paloeoniscus, three of which were then for the first time characterized, viz: P. fiiltvs Ag., P. latus J. H. R., P. macropterus W. C. R , P. Agassisii W. C. R, and P. ovatus W. C R. ; also, 'Auiials Lyceum Nat. Hist. N. Y., pt. 7, vol. 4, 1848, i^- 35. 19 20 TRIASSIC FISHES AND PLANTS. four species of Catopterus were described, viz : C. f/raciUs J. H. R., C, ma- cruriis W. C. R, G. cmgulUiformis W. C. R., and C. panntlm W. C. R.^ Ill 1854 Mr. J. H. Redfield read to the Association of American Geologists and Naturalists a paper entitled "A Catalogue of tlie Fossil Fishes of the United States as far as Known, witli Descriptions of Those Found in the New Red Sandstone." In this paper four species of Catopterus and nine species of Pakeoniscus were enumerated; of the latter four, viz, P. clupeiformis, P. rodratus, P. tenuis, and P. parvus, had not been before de- scribed. This report has never been published, because, as I was informed by Mr. W. C. Redfield, it was agreed between Professor Agassiz and himself that the whole subject should be reviewed in a joint monograph. Unfortu- nately this was not done, and the death of Mr. Redfield in 1857 prevented further publication of the large amount of valuable information which he had acquired on this subject. In 1847 Sir Charles Lyell published a paper" on the Coal Field of Eastern Virginia, in which he gave some notes on the fossil fishes he ob- tained there, with two beautiful plates, drawn by Joseph Dinkel. These notes also include the results of an examination of these fishes by Sir Philip Egerton, Avho reported (1) that the fish described by W. C. Redfield under the name of Catopterus macrurus should be considered the representative of a new genus "because it was homocercal," and he called it Dktijopyge ; (2) that the fishes from the American Trias referred by x\gassiz and Redfield to Palcconiscus were generically distinct, and he named the new genus whicli he created to receive them Ischjpterus, because of the great strength of tlie fin rays. Sir Philip Egerton also decided that some of the specimens of Catopterus brought back by Sir Charles Lyell were diff'erent from C. gracilis, and suffsested the name of C. Redfieldi for one of them. A portion of the middle of the body of a large fish with quadrangular scales which Sir Charles Lyell brought from Blackheath, Va., he referred to the genus Tctragonolepis, but on scarcely satisfactory grounds. Taking up the study of the fishes of tlie Trias soon after coming to New York in 186G, I had excavations made at Boonton, N. J., from which many hundreds of fishes were obtained, in circumstances which apparently throw 'Am. .Jonr. Sci., vol. 41, 1841, p. 24. •Qtr:irt. .Inr. Gi'ol. Sou. Loiiilou, vol, :!, 1847, p. 2iU. FOSSIL FISHES. 21 some light on the manner in which they were entombed. Tliis locahty is near the western margin of the Triassic area where strata of shaly sand- stone rest upon coarse conglomerate, showing the different conditions which prevailed at the same locality within a limited interval of time. Certain layers of the sliales are crowded with fishes, a slab a yard square carrying sometimes a half dozen or more. Some of these are dismembered, consist- ing of a shapeless aggregate of scales and bones, but most are nearly per- fect; and the number found at about the same level, with their perfection of preservation, seem to show that tlie generation inhabiting that portion of the Triassic basin at a certain time were somewhat suddenly killed and sunk to the bottom, where they were soon covered with the accunuilating sediment and were thus preserved. The layers of the shale which contain the largest number of fishes are impregnated with bituminous matter, burn- ino- for a time when thrown into the fire, and when struck with a hammer giving off a peculiar odor. Similar fish beds are known to exist at Pomp- ton, Plainfield, and beneath the trap of the Palisades above Hoboken, and it seems probable that the great mortality which strewed the bottom of 'the basin at times with dead fishes was the result of some phase of the volcanic action which poured out the trap masses of the Palisades and Newark Mountains. Fishes seem to be equally abundant in the Connecticut River basin. At Durham, Conn., and Turner's Falls, Mass., they are particularly numerous and well preserved, while they have also been obtained at Middletown, Sud- bury, Chicopee, Amherst, and Hadley's Falls. Collections made at all these localities have been studied by me, and among them I have identified with more or less confidence about twenty-five species. To the list of the species of Catopterus and Isclujpterus enumerated by the Messrs. Redfield perhaps as many more have been added, and two genera which they do not seem to have met with, viz, DipUirus Newb. and Pti/chokpis Ag. These will be found figured and described in another part of this menioir. In the revision of the group of fishes studied by the Messrs. Redfield access has been had to the specimens left by Mr. W. C. Redfield, most of which were do- nated to the Peabody Museum at Yale College. The collection contains many types of the species described by the Messrs. Redfield, but unfortu- 22 TRIASSK! FISHES AND PLANTS. nately not fully labeled. It is also to be regretted that some of their types and many specimens which they had studied and labeled perished in the destruction by fire of the geological collection belonging to the New York Lyceum of Natural History. I erive below a list of the fishes of the North American Trias as far as yet made out. It probably includes nearly all the species which lived in the water basins from which the Triassic strata were deposited in New Jersey and the Connecticut Valley, but in the southern extension of the Triassic belt some new things are sure yet to be found. No one has given special attention to the fishes of the Richmond coal basin or those of North Caro- lina, but the few specimens which have been incidentally collected indicate considerable differences between the fishfixuna of this region and that which I have studied farther north. By far the most common fish in the Rich- mond basin is Bidyopyge macrura Egt. {Cat02)terus macrurus W. C. R.), which I have not found elsewhere. With this are associated fragments of some other genera and species which have not yet been described. One of these is apparently a Bictyopjge considerably larger than D. macrura, and distinguished from it by having the flattened fin rays ornamented with raised lines. One new genus of which I have seen fragments is strongly marked by its relatively large, rounded, and ornamented opercula. In the Triassic strata of the Far West very few fish remains have been found. Mr. E. E. Howell obtained from the Trias in southeastern Utah some detached ganoid scales, and recently Mr. R. C. Hills found at San Miguel, in southwestern Colorado, near the middle of the Triassic series of that region, several specimens of a Catopterus hardly distinguishable from C. gracilis, but too imperfectly preserved for accurate determination. Prob- ably when the calcareous beds which represent the Trias in Idaho shall be more carefully examined they will be found to contain the remains of fishes which may be expected to resemble those of the Muschelkalk of Europe. FOSSIL FISHES. 23 LIST OF TRIASSIC FISHES. Diplurus longicaudaliis Newb. Ptycholepis Marshii Newb. Dictyopyge macrura Egt. Acentrophorus Chicopensis, n. sp. Catopterus Reljieldi Hgt. gracilis W. C. R. anguiUiformis W. C. E. parvulus W. C. R. minor, u. sp. ornatus, n. sp. Ischypterus fultns Ag. sp. macroptertisW. C R. ovatus W. 0. R. Aqassizii W. C. R, Ischypterus latus J. H. R. micropterus, n. sp. alatus, n. sp. modestus, n. sp. parvus W. C. R. Marsha W. C. R. minutus, n. sp. Braunii, n. sp. robustus, n. sp. tenuiceps Ag. elegans, n. sp. lenticularis, n. sp. lineatus, n. sp. (/igias, n. fsp. DESCRIPTIONS OF GENERA AND SPECIES. SUBCLASS GANOIDEI. ORDER LEPIDOSTEID^^. FAMILY LEPIDOTID^. Genus ISCHYPTERUS Egerton. Under the names of Pakeoniscus fultus and Eiirynofiis tenuiceps two species of this genus wei'e described by Agassiz.^ Previous to that time a specimen had been sent from Massachusetts to M. Alexandre Brongniart, at Paris; this was examined by De Blainville, and was referred to his genus Pakeofhrissum, for which Pakeoniscus was afterwards substituted. All these specimens lacked important parts, were distorted and imperfectly preserved; hence it is perhaps not surprising that their anatomical structure was misun- derstood, and they were included in genera to which they do not belong. Eurynotus is a palaeoniscoid genus, restricted to the Lower Carbonifei ous strata, having a different arrangement of the head plates, a high, broad dorsal, and a ver}- heterocercal tail; features in which it differs essentially from the fishes now under consideration. Palceoniscus also has been clearly shown by Dr. R. H. Traquair, in his admirable studies of the family,^ to be as distinctly separated from them. lu' 1841 Mr. W. C. Redfield pubhshed an article with the title " Short Notices of American Fossil Fishes."^ In this paper he describes five species of Pakeoniscus^ accepting the classification of Agassiz, who had referred them to this genus. These were P. fultus Ag., 'Poissons Fossiles (vol. %, p. 43, pi. 8, flgs. 4, 5; p. 159, pi. 14c, figs. 4, .''), 1833-1843.) = Quart. Jour. Geol. Soc, London, vol. 33, 1877, pp. 548-578, and Trans. Eoy. Soc, Edinburgh, vol. 29, pp. 343-391. Am. Jour. Sci., vol. 41, 1841, p. 24. •21 FOSSIL FISHES. 25 P. latus J. H. R , P. macropterus W. C. R , P. Agassizii W C. R., and P. ovatus \Y. C. R. Mr, Redfield in this article pointed out some of the peculiarities of this group of fishes, and sugg'ested that they should perhaps be separated fi'om Pcdceoniscus. This was subsequently done by Sir Pliilip Egerton/ who, on account of the great strength of the fin rays, named the genus which he created to receive them Ischyptenis. Unfortunately no detailed description of the anatomical chai'acters of the genus was given by Sir Philip Egerton, as he scarcely had sufficient material for the purpose. This is nuich to be regretted, as with his great knowledge, if he could have made a careful study of good specimens, he would have been led to discover and report the true relationship of the group. This is plainly with Lepidotus and its allies, and not with Pakeoniscus, as supposed by Agassiz ; an error into which he was led by the imperfect preservation of the fishes he examined, none of which showed any details of tlie all-important structure of the head. The head was small, and all the bones were delicate ; hence the almost universal de- ficiences in this part of their structure when fossilized. Among the many hundred specimens of Ischyptenis I have passed in review I have found a few in which nearly all the details of the bony structure were preserved, and I am able to describe this more fully than has before been possible, and to deduce from it with certainty the zoological relations of this group of fishes. Where distinctly visible the structure demonstrates an intimate relationship with Lepidotus, Dapedius, s,nd PJioUdop)horus, but most of all with Semionotus. Here the affinity is so close, that it is probable that both Agassiz and Sir Philip Egerton would have united Is'cliypterus with that genus if the material at their command had been more abundant and better. In an examination of nearly all the specimens of Iscliypterus contained in the museums of the United States and a considerable number of individuals of Semionotus I have been unable to detect any characters by which they can be distinguished. The outlines, size, and proportions of the body are essentially alike ; both are elliptical or ovoid, with a relatively small, pointed head, and weak, scarcely lobate tail. The positions, form, and structure of the fins are so nearly alike that the differences can hardly be regarded as of more than specific value. • Quart. Jour. Geo!. Soo. Loudon, vol. 6, 1850, p. 8. 26 TRIASSIC FISHES AND PLANTS. The fin rays are few, broad, and undivided below in both ; the anteriorrays are characteristically strong; the fulcra few, strongs acute, and closely appressed; in both a row of large, sometimes erect and pointed scales marks the line between the head and the dorsal fin. The obliquity of the posterior ex- tremity of the body is about the same ; the mouth is small, the mandibles and maxillaries are weak, the premaxillaries united in a uniform arch, set with an even row of small, abruptly pointed teeth, as are also the max- illaries and mandibles ; and the eye is placed above the posterior margin of the mouth. The divisions and the forms of the head plates are apparently the same, though I have not been able to verify by personal examination the descriptions of the head plates of Semionotus given by European authors. I can not, therefore, assert that Semionotus and Iscliy])terus should be united. I call attention, however, to the close, general, and special resemblances between them, and leave to those who may have better opportunities for studying the structure of Semionotus the decision of the question.^ The specific division of the great group of fishes representing the genus Ischjpterus obtained from the Upper Triassic rocks of New Jersey and the Connecticut Valley is a matter of no little difficulty. The descrip- tions given by Mr. AY. C Redfield in the article mentioned above are ex- ceedingly brief, and in the absence of the type specimens, which cannot now be certainly identified, it often becomes a matter of much doubt as to what were tlie fishes to which he applied these names. Only one has yet been figured, Ischjpterus latus J. H. R, and this lacks both head and tail. In examining the collection of Triassic fishes left by Mr. W. C Eedfield, now at Yale College, and to which I have had access through the courtesy of Prof 0. C. Marsh, I found many without labels, and those which were named were in some cases so imperfectly preserved, that it was not easy to use them as guides in classifying the much larger number of specimens contained in the cabinet of Columbia College. ' Since tbe above notes were written two interesting papers have been pnblisbecl ou the Triassic fishes of the Old World, viz, "On the Remains of Fishes from the Keuper of Warwick," by E. T. Newton, Rev. P. B. Brodic, and Edward Wilson, Quar. Jonr. Geol. Soc., August, 1887, and "On Two New Lepl- dotoid Ganoids from the Early Mesozoio Deposits of Orange Freo State, Sonth Africa," by A. Smith Woodward, Quart. Jonr. Geol. Soc. London, May, 1833. In both papers these fishe.s are described as species of SemioiiohM, which, if found iu our Triassic rocks would be unhesitatingly referred to Ischjtp- ieriis. FOSSIL FISHES. 27 On the following pages, so far as I have been able, I have enumerated and defined all the species of the genus which have come under my obser- vation. I deem it necessary to say, however, that future observations will probably diminish rather than increase the number of forms in which the differences should be given specific value. For example, I. alatus may prove to be only a variety of /. lineatiis and I. modestusa phase of /. elegans; but with marked differences and without connecting links, so far as yet ob- served, it has seemed to me hardly justifiable without further evidence of identity to unite them iinder a common name. ISCHYPTEEUS OVATUS W. C. R. Pl.'I, Fig. 1. Palceoniscus ovatits W. C. E., Am. Jour. Sci., vol. 41, 1841, p. 26. The only published description of this species is that cited above. It reads as follows : Palceoniscus ovatus W. C. II. — Wide or round-shaped Palaeoniscus. This spe- cies is shorter than P. AgassizH, and exceeds all the known American species in the comparative width or roundness of its form, and is also remarkable for the large size of its scales. It is of rare occurrence, and, owing probably to its great thickness, is seldom obtained in perfect form. This fossil also exhibits the spiue-like erections of the dorsal scales which have been noticed above. Found at Westfield and Middlefield, Conn.; Sunderland, Mass., and Boonton, N. J. In the manuscript report of Mr. J. H. Redfield the following notes on this species appear : Fish ovate ; head rather small and narrow ; body widening rapidly from the head to the dorsal and ventral fins, expanding as far as the ventral tins, from which point the form gradually narrows to the pedicel of the tail. Scales large, anterior ones con- centrically striate, those of the dorsal ridge pointed and elevated as in P. tenuiceps; pectoral fins small, comparatively slender; veutrals small; dorsal large, rays strong; anal not well observed. This is the broadest and most ovate species of Palceoniscus that is known, and perhaps ought to be referred to a separate genus. In the size of the scales it resembles P. AgassizH, but its form will readily distinguish it. In the collections made at Turner's Falls, Mass., and Boonton, N. J., I find a large species of IscJiypterus, which agrees very well with the descrip- tions given above. The fish reaches a length of from ten to twelve inches, is ovoid in form, with a breadth at the dorsal fin of from four to five inches ; 28 TRIASSIC FISHES AND PLANTS. all tlie fins are quite strong, and the scales are large, broad, and thick. The concentric lines upon their borders, mentioned by Mr. J. II. Redfield, are not peculiar to this species, but are more or less distinctly visible in all the members of the genus in certain states of preservation, It is most notice- able where the scales are partially decomposed and where tliey were of considerable thickness. In outline the fishes of this species resemble some individuals of Ischjpicrus tenuiceps, but they are considerably larger, the scales are also relatively larger and more quadrate ; the arch of the back is usually regular, and the outline is comparatively smooth, never showing the extreme development of the dorsal scales which is so conspicuous a feat- xire in the older individuals of I. tenuiceps. Judging from some of the specimens obtained from Turner's Falls, I am inclined to think that this is the fish which is figured by Sir Charles Lyell in his paper on the Virginian coal field and referred to by Sir Philip Egerton as a species of Tetragonolepis. In that specimen nothing is shown but a portion of the side near the head, without tail or fins. If the fins had been present they would probably have shown the great development of the fulcra, which is characteristic of Ischjpterm and wanting in Tetrago- iiolepis. This I infer from the facts that no other traces of the latter genus liave been found in the Triassic rocks of North America, and the scales on the sides of the large and broad species of IscJiyptcrus could hardly be dis- tinguished from those of a corresponding part of the body in Tetragonolepis. With precisely similar scales, however, we have in several instances the characteristic fins of Ischgpterus. No such specimens were contained in tlie collections made by Sir Charles Lyell in America, and the inference of Sir Philip Egerton was therefore a natural one, though probably erroneous. ISCHYPTERUS MaRSHII AV. C. R. PI. II, Fig. 1. Fishes twelve inches or more in length by three or four inches in breadth ; body fusiform in outline; head conical, obtuse, contained four and one-half times in the entire length; fins strong but relatively short; anterior base of dorsal midway between muzzle and tip of tail ; fulcra strong and short ; rays FOSSIL FISHES. 29 eight (?) ; caudal fin strongly forked, three inches wide in fish twelve inches long, uns^anmetrical, upper lobe longest ; scales universally large and thick; those of dorsal line less strongly spined than in other large species of the genus; boat-shaped scale covering anterior base of dorsal fin relatively small, rounded before, pointed behind, not notched ; rows of scales on sides more oblique than in other large species ; those on the middle and anterior portions of the body square or oblong, slightly higher than broad. This large and fine species was named by W. C Redfield, but was never described. It is referred to in his paper,^ and I find a specimen from Sun- derland, Mass., bearing this name in the Redfield collection at Yale College- In pursuance of my plan to secure to W. C. Redfield all the fruit of his labor in this field I have adopted it, and now supplement the name with a detailed description. So far as known this species only occurs at Sunderland, Mass., where a number of fine specimens have been procured, one of the best of which may be seen in the American Museum of Natural History, New York, to which institution, with many other fossil fishes, it was presented by the late Robert L. Stuart. This, like a number of others whicli I liave seen, has a length of about twelve inches, but the species probably attained somewhat greater dimensions. The body is broadly or more narrowly fusiform, the widest portion being midway between the dorsal fin and occiput. In general form it resembles Ischijpterus Agassizii, but attains greater dimensions and may be distinguished at a glance by the larger size of its scales and the more oblique position of the rows on the sides In general aspect this fish has much resemblance to some species of. Lepidotus, all of which are characterized by their relatively large and thick scales. The resemblance of Ischjpterus to Lepklotus has been referred to, and it is evident that they are closely allied, but as a whole the species of the former genus are smaller and are distinguished by the more salient row of sjDiny scales along the dorsal line, and by a greater prolongation of the upper lobe of the tail. 'Proc. Am. Assoc. Adv, Sci., AIb;iiiy mcotiug, 1856; jit. 2, j). 18S, 30 TRIASSIC FISHES AND TLANfS. IscnYPTEEUS Agassizii W. C. K PI. Ill, Fig. 1. Palmonisciis Agassizii W. 0. R., Am. Jour. Sci., vol. 41, 1341, p. 26. The first notice of this species was given by W. C. Eedfield in the article cited above. The description vv^hich he there published is as follows: Its length in the specimens hitherto obtained varies from seven and one-half to ten inches, and its width from three to four inches. The tins, with their armatures and insertions, are also of more remarkable thickness than in the species alreadj- noticed. The large scales or plates which belong to the anterior portion of the dorsal line are commonly found doubled together at their lateral edges by the incumbent pressure, which gives them the appearance of short spines or flattened rays ; and hence these are sometimes mistaken for an anterior comb-like dorsal. Mr. J. H. Redfield, in the manuscript report to which reference has been so frequently made, adds the following notes upon this species : Head narrow and pointed, scales large and smooth, sometimes with faint concen- tric striiTe; those of the anterior portion of the dorsal ridge very much elongated, strong and pointed, and apparently erectile; when in an erect position much resembling rays, and giving the appearance of a comb-like dorsal flu; back arched, but not so abruptly as in P. tenuiceps. The widest portion of the fish is found just anterior to the ventral fin; pectoral fin moderate; anterio.r raylets rather short; primary rays six or eight; ventral fins small, anterior raylets about ten ; primary rays about five or six; dorsal fins large, triangular, preceded by erect, pointed scales ; anterior raylets very long, twelve or more in number ; primary eight to ten ; anal fin large, but not so much elongated as in P. tenuiceps or P.fuUus; antei'ior raylets very strong, about twelve in number, primary rays six to eight ; tail forked, lobes acute, anterior raylets rather stout, rays of lower lobe much stouter than those of upper ; length, seven to ten inches ; breadth, three to three and one-half inches. Occurs at Sunderland, Mass , Westfleld and Middlefleld, Conn., Pompton and Boouton, K J. Among the fishes obtained at Boonton, N. J., are a dozen or more of unusually large size, and manifestly distinct from the many small fishes with which they are associated. These I have supposed to be the fishes to which W. C. Redfield gave the above name, and indeed there are no others found at that locality to which his description is at all applicable. These fishes are from ten to twelve inches in length and from three to three and a half inches wide. The head is conical and pointed, and in an individual twelve inches long it has a length of three inches ; the back is uni- formly and rather strongly arclied anterior to the dorsal fin ; the roAV of dorsal scales is .strong, though usually depressed, and when erected would present FOSSIL FISHES. 31 the appearance of a comb-like crest described by W. C Redfield. This row of scales is, however, less strongly developed than in Ischyptenis tenuiceps, and the arch of the back does not show the hump which is so characteristic of that species ; the fins are very strong ; the fulcra of the dorsal and anal fins unusually broad and long, forming arches nearly half an inch wide at base, curving gracefully backward to a point ; the anal fin when appressed reaches quite to the base of tlie caudal ; the tail when expanded is three inches wide at its extremity ; the scales of the sides are large and thick, those near the head square or oblong. I have seen no such fishes as these anywhere except at Boonton. At Durham we find a species of Ischyjitenis of about the same size, but con- siderably broader — the mature form of Ischypterus micropterus N. At Sun- derland occurs another species (Z Marsliii) which in form and general aspect resembles those under consideration, but it is narrower, with less strong dorsal and anal fins, with thicker and relatively broader scales, which form more oblique rows on the sides. For these reasons I have thought it wise to regard it as distinct. '^fe'- Ischypterus micropterus, n. sp. PI. IV, Figs. 1, 2; Fl. XII, Fig. 2. Fishes of medium or large size, five to ten inches long by one and a half to three and a half inches wide ; form conical, greatest breadth at pec- toral fins, thence tapering uniformly to tail; back and abdomen about equally arched ; head conical, acute, contained four and a half times in total length, nearly horizontal and straight below, rapidly sloped above ; muzzle prolonged, acute ; mouth very small ; maxillary and mandible slender, teeth small, conical, acute ; cranial plates granulated ; operculum narrow ; anterior margin vertical, posterior rounded, supraclavicles and clavicles slender; scales smooth, polished, oblong, twice as high as long on the sides near the head, rhomboidal on posterior portions of sides and tail ; scales of ante- rior dorsal line about fifteen in number, rounded and emarginate at base, abruptly narrowed to smooth acute spines above ; fins all relatively small and weak; anterior base of dorsal midway between tip of tail and extremity of muzzle; fulcra eight, relatively small ; fin rays eight, narrow, delicate; 32 TEIASSIC FISHES AND PLANTS. large scales of posterior dorsal line eight, elliptical before, elongate and spiny behind, running into fulcra of upper margin of dorsal, which are ten in number, eight beyond scaled extremity of body, all slender and rod- like ; caudal fin narrow and weak, obhque, upper lobe longest, rays fifteen, slender; fulcra of lower margin fifteen ; anal fin narrow, just reaching base of caudal, rays eight, fulcra ten. The most striking diagnostic characters of this species are its pointed rostrate, depressed muzzle; conical narrow head, horizontal below; the wedge- shaped outline of the body, which is widest near the head ; the small and delicate fins, and the narrow and oblique tail. The largest specimen which I have is ten and a half inches long by three and a half inches wide, the smallest five and a half by one and a half inches; but I have seen one speci- men which shows distinctly all the characters of the species, and yet is only about three and a half inches long. This is the most common species of Ischyjiterus at Durham, Conn., but I have not certainly identified it elsewhere. S. W. Loper has good speci- mens in his cabinet, and has supplied a fine series of diff"erent ages to the cabinets of Yale and Columbia. The figures given on PI. IV represent old and half-grown individ- uals ; that on PI. XII, Fig. 2, is still younger. ISCHYPTERUS TENUICEPS Ag., sp. ri. V, Figs. 1, 2, 3; Fl. VII, Fig. 3. Uurynotus teiiuiceps Ag., Poiss. Foss. vol. 3, p. 159, PJ. 14o, Figs. 4, 5; E. Hitcbcock, Geol. Mass., vol. 2, p. 459, PI. 29, Figs. 1,2. This species has been more fully illustrated tlian any other from the American Trias. Two figures of it are given by Agassiz in his Poissons Fossiles (loc. cit.) ; two are given by Professor Hitchcock in his quarto Re- port on the Geology of Massachusetts; one in Emmons's Geological Report of the Midland Counties of North Carolina, PI. IX (reproduced in his American Geology, pt. 6), a wood -cut probably of this species in Em- mons's Manual of Geologj^, page 188, and in American Geology, pt. 6, i^age 144; also, three figures of it are given on PI. IXa of the latter work. Of these last cited figures only one has the normal form of ihc species, the FOSSIL FISHES. 33 others being distorted and narrowed, but the originals were all from the same place, Turner's Falls, where this is the most abundant species, and they show the peculiar erect, thickened dorsal scales, which are not devel- oped to the same degree in any other. Unfortunately all the figures of I. teniiiceps yet published are taken from imperfect specimens. That on PI. IX of the American Geology, pt. 6, represents the posterior half of the body fairly well, but the head is a shapeless mass, and the arch of the back is only partially shown. As mentioned in the remarks on the genus Isclujptems, the preserva- tion of the head is so generally incomplete that we must conclude its bony structure, was delicate and largely reenforced by cartilage. Out of the large number of specimens which I have, however, a few give the outlines of the head and much of its structure with considerable accuracy. From these we learn that it was conical, rapidly sloping from the high nuchal arch, and from the smallness of the mouth, pointed at the muzzle. The general form was ovate, in that respect resembling Ischjptenis ovatus W. C. R., but the species may be distinguished generally at a glance by the high, thickened, and often obtuse scales which crown the humped back. The length of this species in mature individuals is eight inches and the breadth immediately behind the head is two and a half to three inches. The dorsal scales are often strangely thickened and distorted in the nuchal region, where they are sometimes more than half an inch long, clavate and blunt. This I was at first disposed to regard as the result of pyritous concretionary distortion, but I have seen it in so large a number that I am compelled to regard it as a specific character. In some cases the form of every scale of the row is observable ; it is seen that those immediately back of the head are much elongated, and the terminal spine is depressed backward, so that the scale is blunt and club-shaped. Possibly this is the result of disease, but if so, it attacked a majority of individuals. More likely it is a character developed by age and only fully shown by those that were quite old at the time of their entombment. It is possible also that it is a sexual character ; but, by whatever cause produced, it is a mark by which, when present, the species can be immediately recognized. MON XIV 3 34 TRIASSIC FISHES AND PLANTS. Tliis species is found much more abundant at Sunderland, Mass , than elsewhere. Probably more than half of the individuals which have been taken from the Triassic rocks there belong to it. A few individuals have been obtained from Durham and Boonton which were regarded by W. C. Redfield as specifically identical with these. Of this there may be some doubt, since nowhere else are fishes found which have the back so highly arched immediately behind the head, and set with the long, divergent, acute or clavate scales. The figures given on PI. V represent two old individuals and one very young one. Fig. 3 of PI. VII represents a mature but not old individual. ISCHYPTEEUS FULTUS Ag. sp. PI. VI, Fig. 2; PI. VII, Fig. 1. Palwoniscus fultus Ag., Poiss. Foss., vol. 2, p. 43, PI. VIII, Figs. 4, 5. Palceoniscus fultus W. C. R., Am. Jour. Sci., vol. 41, 1841, p. 25. Ischypterus fultus Egerton, Quart. Jour. Geol. Soc. Loudon, vol. 3, 1847, p. 277. Two very imperfect- fishes from Sunderland, Mass., both wanting the head and one the tail, served as a basis for Agassiz's description of this species. All that can be said about them is that thev represent one of the smaller and narrower species of the genus Ischijpterus, as defined by Sir Philip Egerton. But no one could positively assert, even with the speci- mens in hand, that they belonged to one or another of several species found in the Connecticut Valley and New Jersey. We are, however, better in- formed in regard to the fish accepted by the Messrs. Redfield as represent- ing the species I. fultus. Mr. W. C. Redfield, in the article so frequently referred to^^ makes the following remarks upon this species : Palwoniscus fultus Ag., tbe specimen figured by Professor Agassiz is destitute of the dorscal and head, as well as the upper portion of the body. The length was prob- ably four and a half inches; but this is ofteu exceeded in other specimens. The fins and their bony iusertions appear stouter than in P. latus, but less stout than iu some other species. Fouud at Westfield, Middlefleld, and Durham, Conn., aud Boonton, N. J. In the report read before the American Association of Geologists and Naturalists at New Haven, in 1845, by Mr. J. II. Redfield, I find the fol- lowing description of this species : Fish fusiform, head small, rather more than one-fifth of the whole length ; back nearly straight, but slightly arched; scales of medium size, ofteu \vith concentric striie, which 'Am. Jour. Sci., vol. 41, p. 25. FOSSIL FISHES. 35 are most apparent ou the posterior edge; scales of the dorsal ridge pointed aad erect- ile, but in a umch less degree than in Isohypterus tenuiceps; pectoral fins small, narrow, and pointed ; ventrals small, very narrow, and pointed ; dorsal and anal flus both very long, with the anterior raylets very strong and rather numerous; primary rays of anal about seven, slender; anterior raylets about twelve, anterior raylets of dorsal about fourteen ; tail forked, lobes more acute than in P. tenuiceps; accessary raylets long and numerous; length five to seven inches, breadth one and a half to two and a quarter inches. The specimens from which Agassiz constituted his Palwoniscus fultus were so imperfect, that it is difficult to decide with certainty which of our specimens should be referred to it. The character which he seized upon as its chief diagnostic, and on which he founded its specific name, fultus, viz, the extraordinary size of the anterior ■ raylets of the fins, exists in all the known American species of this genus. We are not at all sure that we have rightly referred P. macroptenis of W. C. E. to this species; for the specimens' figured by Agassiz are represented with dorsal and anal fins which are far from having the length of these fins lu P. macroptenis. Those specimens were evidently imperfect, and it is well known how easily the frail and carbonaceous rem- nants of rays are detached from these fossils, sometimes leaving hardly a trace behind, and it is very possible that these portions were broken in the specimens which were figured by Agassiz. His name of P. faltiis shouhl in justice to him be retained, and since the long-pointed fins of the fish we have described above, strengthened as they are by large anterior raylets, will render the term fultus quite applicable, we think it advisable to restrict Agassiz's name to this species, and suppress P. macropterus . This species is characterized by the length of the dorsal and anal flus, which are even longer than iir P. tenuiceps, from which species it is also readily distinguished by its form, the back not suddenly rising from the head as in that. Among the fishes left by W. C. Redfield I find many which are labeled Ischypterus fultus. Most of these are from Boonton, and it is repre- sented bj him as the most common species found there. The form is rather narrow, the length from six to eight inclies, the breadth never more than two inches at the widest part, which is halfway between the dorsal fin and the head ; the fins are relatively large ; tlie tail is scarcely forked, but rather scalloped, with a broad and shallow sinus; the head is depressed, longer than wide, and about one-sixth of the entire length. From Durham and Sunderland I have specimens which I suppose must represent the fish named P. fultus by Agassiz, for his specimens were derived from the latter place. They are smaller than those from New Jersey, not over six inches in length by one and a quarter inches in breadth. It is quite possible that they represent a diff'erent species from that so common at Boonton, but that can only be shown by more extensive comparisons than I have been able to make. 36 TEIASSIC FISHES AND PLANTS. The fish represented on PI. VI, Fig. 2, is perhaps a fair example of the species so common at Boonton, and which W. C Redfield first described as Palceoniscus macropferus. He afterward snppi-essed that name in deference to Agassiz's opinion that it was not diff'erent from those to wliich he had given the name P. fultus. PI. VII, Fig. 1, represents a smaller fish, of which I have a large number of specimens, but I have considered these the young of the larger form referred to above. ISCHYPTERUS ROBUSTUS, 11. Sp. PI. VI, Fig. 1. Fishes of medium or large size, eight inches or more in length by three in breadth anterior to the dorsal fin; outline ovoid; head large, narrowed, muzzle produced; dorsal fin very lai'ge, Its anterior mai-gin about the middle of the entire length and nearly twice as far from the posterior scaled ex- tremity of the body as from the head; fulcra very numerous, strong, curved ; rays eleven, very strong ; caudal fin of moderate size, upper lobe longest ; anal of moderate size ; ventrals inserted nearly opposite anterior margin of dorsal; pectoral fins relatively long and broad; scales of dorsal line long, forming a prominent crest; those of sides broad and thick. This Is a robust and coarsely organized fish, most nearly allied to Iscliijpterus ovatus of Redfield, but distinguished by the great height, breadth, and strength of the dorsal fin and Its anterior position. The pec- toral fins are a,lso longer and broader than in any other species that I have seen. The ventrals and anal are not well shown in the specimens before me, but are apparently delicate; the caudal is relatively narrow, the lower lobe nearly horizontal, the upper strongly elevated and produced. The great height and breadth of the dorsal fin of this species bring it closer to Semionotus than any other of its congeners, and there Is little doubt that If It had been found in the Mesozolc rocks of the Old World It would have been referred to that genus ; indeed, it is now difficult to say by what characters it could be distinguished generically from some of the described species oi Semionotus. The line of spine like dorsal scales is somewliat more conspicuous, but this is only a matter of detail, since something of tiie kind is seen in all the species of that genus with which I have compared it. FOSSIL FISHES. 37 The scales of this species are rehxtively large and strong, and it is evident that the fish was firmly and robustly organized; hence the name given it. Up to the present time I have seen but two or three specimens, and these are all from Boonton, N. J. The type is in the geological museum of Columbia College. ISCHYPTERUS ELEGANS, n. Sp. ' PI. VII, Fig. 2; PI. X, Fig. 1 ; PI. XIV, Figs. 1, 2. Fishes small, length four to six inches, greatest breadth two inches ; length of head one to one and a quarter inches, contained four and a half times in the entire length; body long-ovoid, elegantl}' arched; teeth rela- tively large, conical, acute ; scales smooth, about twenty in each vertical row in broadest part of body, and tliii-t}'-two in a longitudinal series along the median line to the base of tlie triangle which extends into the vipper lobe of the tail ; erect scales along dorsal line anterior to dorsal fin about twenty, relatively small, first four or five unarmed ; head small, pointed, depressed ; fins small, weak. This is the neatest species of the genus known to me ; the curves of the outline of the body are graceful, the scaling crowded but exact. In form it most resembles I. Uneatus, but is smaller and broader, the back is more distinctly and regularly arched, and the scales are more numerous. Another peculiar feature in the outline is the sudden contraction of the body behind the dorsal fin. The scales are brilliantly polished, and each one usually retains its position, so that tlie surface and outlines of the fish are well preserved. From this it may be inferred that the scales were thicker and more firmly united than in most species of the genus. Collected at Boonton, N. J.; type specimens in the geological museum of Columbia College. IsCHYPTEEtJS ALATUS, n. sp. PI. VIII, Figs. 1, 2. Fishes robust, eight inches in length by two and a half inches in great- est breadth ; head large, nearly one-third of entire length ; fins relatively 38 TIMASSIO PISHES AND PLANTS. large, rays and fulcra strong; dorsal fin set at the middle of the entu'e length and midway between the occiput and base of caudal. These fishes resemble most those I have called Ischyptcrus Uneatus, and they may prove to be only a well-marked variety of that species ; but in this group the body is somewhat narrower, the head is larger, the fins are stronger and more conspicuous, and the dorsal is more posterior in position. Up to the present time fishes having the characters given above have only been foimd at Boonton, N. J. Tlie types are in tlie geologieal museum of Columbia College. ISCHYPTEEUS MODESTUS, n. sp. ri.JX, Pigs. 1,3. Fishes four to six inches in length by one and a half to two inches in width ; outline of body long-ovoid, symmetrically arched above and below anterior to dorsal and anal fins, rapidly contracted behind to half the ante- rior breadth ; fins broad, strong, and rounded; dorsal fin exactly in middle of entire length, opposite ventrals, fulcra strong, twelve in number, rays eleven ; tail relatively broad, slightly eniarginate lobes nearl}- equal, rays fifteen ; anal rounded, not reaching base of caudal, fulcra ten (?), rays seven ; head relatively large, one-fourth the entire length, rounded, some- what obtuse ; scales of dorsal line eighteen, of medium size, the one imme- diately anterior to the dorsal fin shield-shaped, not emarginate behind; scales of sides relatively large and thick. The fishes which have been included in this species are small, and have the outlines of the body and fins rounded so as to give a smooth and gentle aspect ; the curves of the body are all graceful and flowing ; the back and abdomen are uniformly arched to the dorsal and anal fins ; behind these the outline contracts rapidly by concave curves until the width at the base of the tail is less than half that of the anterior portion of the body. The fishes most nearl)- allied to these are those which I have included under the name /. eler/ans, and it is perhaps not certain they sliould be re- garded as distinct. The head is, however, more obtuse and rounded, the back less highly arched, and the fins apparently broader than in that species. Also the scales are larjjer and thicker and those of the dorsal line strono;er. FOSSIL FISHES. 39 Collected at Boonton, N. J. Types in geological museum of Colum- bia College. ISCHYPTERUS LENTICULARIS, 11. Sp. PI. X, Figs. 2, 3. Fishes six to six and a half inches long by two to two and a half inches wide ; general outline lenticular ; body widest at the middle, sloping gently to tlie muzzle and tail ; head pointed or obtuse, relatively large, a little less than one-quarter of the entire length ; fins all small and delicate for the size of tlie fish ; scales apparentl}^ thin, those of the dorsal line relatively small. Among several hundred fishes obtained at Boonton, N. J., there are a number which correspond to the above description. They are relatively broad and have a nearly symmetrical lenticular outline, the tail being small and the body at its base only about one-third as wide as before the dorsal fin. The fins are all small and weak, the fulcra slender, nearly straight and closely appressed. The general form is similar to that of /. ovatus, but these fishes are not half the size of- those to which W. C. Redfield gave that name, and the whole structure is much more delicate. In Z ovatus the scales of the dorsal line and sides would seem to have been very thick and strong, the fins are large, the fulcra strongly arched. The rela- tion of these smaller ovoid fishes is rather with those to which I have given the name 7. elegans, and here the differences may be those of age or sex. The group designated by the latter name consists of fishes which are much smaller, often not much more than half the length arid breadth, the lower line of the body being nearly straight, the upper liighl}^ arched before tlie dorsal fin, concavely narrowed behind. Hence I have supposed that they constitute a distinct species. Up to the present time I have seen no such fishes as those under con- sideration at any other locality than at Boonton. There are none such in all the collections made at Durham or Sunderland. In tlie first of these localities I. microptenis apparently takes their place, but this, though like in the small size of the fins, is distinguished by its depressed, pointed muzzle and the cuneate outline of the body, which is widest immediately behind 40 TEIASSIC FISHES AND PLAlS^TS the head. At Sunderland, I. tenuiceps is the prevailing species, and, though often ovoid in outline^ may always be distinguislied by its humped back and huge dorsal scales. The fins are also larger than those of either of the above mentioned species. ISCHYPTERUS LINEATUS, n. sp. PI. XI, Figs. 1, 2. Fishes six to eight inches in length ; outline when perfectly preserved uniformly arched above and below ; head relatively large, contained about four times in entire length, broadly conical in outline ; fins all large ; fulcra arched ; scales of dorsal line spinous and strong, but less developed than in I. tenukeps; ribs and interspinous bones frequently preserved ; scales on sides thick and strong, arranged in continuous rows from the head backward, so as to give a lined appearance, which has suggested the spe- cific name. The fishes of this group are not easily separated from some of their associates; some individuals resembling those of/. Jenticularis; but in these latter the outline is more symmetrical, the fins smaller, the scales more deli- cate, particularly those of the dorsal line. On the other hand, tliey approach through the smaller individuals the group to which I have given the name of /. eJegans; but these latter are smaller, the arch of the back is higher, the head more depressed and acute, the fins and scales are more delicate. Still another variet}', including the narrower forms, comes nearer to I. fnltus. On the whole, however, this group of long ovoid fishes, from two to three inches wide, are distinguishable at a glance from those which have the nar- row lanceolate outlines of I. fuJtus, a fish which, though attaining the length of six to seven inches, never passes a width of an incli and a half. The fishes to which I have given the name of IscJii/ptrnts alatus, and have represented on PI. VIII, are perhaps most like those under consid- eration, and I hesitated long before separating them ; indeed, it is probable they will be found to run into each other, so that they must be regarded as varieties of one species. By comparing tlie figures now given, however, it will be seen that in the fishes I have called I. alatus the fins are stronger, and the dorsal is placed farther forward, its anterior margin being just mid- FOSSIL FISHES. 41 way between the occiput and the base of tlie caudal fin. In the fishes named I. lineatus the body is shorter and broader, the sides are more dis- tinctly lined, and the dorsal fin is set farther back. Found in considerable numbers at Boonton, N. J., but up to the present time not obtained from any otlier locality. Type specimens iu the geological museum of Columbia Collee-e. ISCHYP'TERUS MACEOPTEEUS W. C. R. PI. XII, Fig. 1. Fishes six to eight inches in length by one and a lialf to two and a half inches broad, long-ovoid or fusiform in outline, symmetrically arched above and below; head large, one quarter the entire length, conical in outline; fins relatively large and strong; dorsal opposite the interval between the anal and ventrals, point of insertion nearer to the extremity of the tail than to the muzzle, fulcra fifteen, rays eight ?; caudal broad, rays and fulcra strong; anal reaching to base of caudal, fulcra fifteen, rays ?; scales rela- tively thick ; ribs and spinous processes strong, and often distinctly showing in the fossil state. W. C. Redfield' describes vei-y briefly a species of Ischypterus, which he calls PalcBoniscus macropterus, in the following words: PakeonisGus macropterus W. C. R.— Long-fiuned Palseoniscus. This species is dis- tinguished by the longitudinal extension of the dor.sal and anal fins; which thus seem to present :i remote resemblance to the wings or foiiced tail of the coiiiinon swallow. Its length is comuioidy from live to seven inches, and its widtli from one and a half to two inches. Among the large number of fossil fishes which have been collected at Boonton, N. J., the most abundant are such as were regarded by the Messrs. Redfield as representing Agassiz's species Ischypterus fiiltus. They are gen- erally fusiform in outline, six to eiglit inches in length, and all have in marked degree the strong fin-fulcra characteristic of the genus. Tliere are, how- ever, two groups of these fishes having about the same average size, one more slender and coming nearer to those which, sent from the Connecticut Valley, were described by Agassiz with the name o( Palceoniscus fuUus; the ' Sliort Notices of Amcricau Fossil Fishes, Am. Jour. Sci., vol. 41, 1841, p. 25. 42 TRIASSIC FISHES AND PLANTS. otlier group is mucli broader, the body being sometimes two and a half inches liigh anterior to the dorsal fin. These were relatively flat fishes, while the others were cylindrical or fusiform. As we compare most of the mem- bers of the two groups they seem so unlike tliat no one would hesitate about considering them distinct species, but it is also true that there are interme- diate forms, which serve to connect these groups, and which are apparently as near to one as to the other. Hence it is not easy to define accurately either of tlie two species which W. C. Redfield has founded upon tlieni. In most cases, however, there need be no doubt, the fusiform and slender fish standing for I.fultm, the broader one fin- /. nmcropterus. In ni}- notes on Iscliypterus fultm I have furtlier discussed this question, "and have shown how difficult it is to identify the species, which have been described very briefly from imperfect material, and where the type specimens have been lost sight of. Another reason why we may suspect that the fishes combined by Agassiz and subsequently by Redfield under the name of /. faltns should be referred to two species, is found in their distribution. As remarked else- where, the individuals figured by Agassiz and taken as the types of his species fultiis, are so imperfect that they cannot certainly be identified with any of the Triassic fishes obtained from the Connecticut Valley or from New Jei'sey. I have even suspected that they were only mutilated speci- mens of the most common speices, I. tenuiceps, found at Sunderland, where Agassiz's fishes were obtained ; but occasionall}^ a narrow, fusiform, and smaller fish is met with at Sunderland and Turner's Falls, which may be the same with those figured by Agassiz. Whether this is identical with any of the fishes found in New Jersey is yet uncertain, because the material we now have for comparison is inadequate ; but if identical with either of the New Jersey forms it is with the narrower one, which Avas adopted by W. C. Redfield as the representative of the species I. fidtus. Up to the present time none of the broader fishes which I have taken as representing Redfield's species or variety, 7. macropterus, have been found the Connecticut Valley ; a fact which justifies the inference that these in closely allied forms are specifically distinct. FOSSIL FISHES. 43 ISCHYPTEEUS BrAUNII, n. Sp. PI. XII, Fig. 3; PL XIII, Figs. 1, 2, 2a. Fishes three to five inches in length by one to one and a half inches broad; outline long-elliptical ; body compressed; head relatively large, contained three to three and a half times in total lengtli ; teeth large, pointed, acute ; cranial bones granulated ; operculum semicircular, large ; preoperculum long-elliptical, having much the form of the operculum, but very much smaller ; fins small, with delicate fulcra and rays ; dorsal and anal placed far back, dorsal midway between occiput and extremity of tail, very long from front to rear, fulcra small, rays ten ; anal reaching back to or beyond base of caudal, fulcra eight (?), rays five ; jointed rays of caudal fin fifteen ; scales rhomboidal or square, more uniform in size than any other species known, number along lateral line tliirty-tlu'ee, in vertical rows sixteen; scales of dorsal line rounded before, pointed or short-spined behind. This species is of peculiar interest as coming from the base of the Triassic rocks of New Jersey, from a horizon probably several thousand feet lower than that of the Boonton specimens, which are from near the top of the series. It may be distinguished from all other known species by the uniformity in the size of the scales and by the posterior position of the dorsal fin. The armature of the dorsal line is also less strong and con- spicuous than in most of the species of the genus ; in this respect it is inter- mediate between the strongly spined species oi Isclujpterus, such as I.tenuiceps and those which have been grouped in the genus Acentropliorus by Dr. Traquair, of which we have an example in A. chicopensis, described in this memoir. In that fish all of the median scales of the dorsal line anterior to the dorsiai fin are unarmed. The only locality from which fishes of the present species have been obtained is Weehawken, N. J. Here, beneath the trap of the Palisades, is a stratum of highly metamoi'phosed slate which was once a bituminous shale, but which has been baked by the efi'usion of the great mass of molten matter above it; the fishes are found in this slate. In some layers it also 44 TRIASSIC FISHES AISD PLANTS. contains great numbers of bivalve crustaceans (Estheria), which would seem to indicate that it was deposited in brackish water. But little exca- vation has been made in this stratum, and it is probable that it will hereafter yield other things new to our Triassic feuna A description of this locality and of the fossils found there was pub- lished by L. P. Gratacap.^ A wood-cut figure of a large specimen of fish found there is given, and it is regarded as identical with Paloeoniscus httus of Eedfield. Through the courtesy of Mr. Gratacap I have examined the original of his illustration, and I have been permitted to make a draw- ing of it, which is now published (Fig. 2). I found it essentially like a large number of fishes from Weehawken which are in my hands, except that it is larger and broader than any other specimen I have seen. All the fishes from this locality have the dorsal and anal fins set for ba(;k, tlie anal reaching to or beyond the base of the caudal. This would serve to dis- tinguish them from /. hdiis, but they also differ from that species in. the gi-eater uniformity in the size of the scales In most species of Isclujpterus four rows of scales on either side of the line of dorsal spines are nearly square ; the next eight rows are higher than long; then follow seven rows of smaller scales to the median line of the abdomen. In these fishes, however, the scales on the side are not conspicuously larger than the otliers, and there is also less difference in their size, going from front to rear. Hence I must conclude tliat they belong to a distinct species from Iscliypieriis httus, which also occurs much higher in the Triassic series. In Mr. Gratacap's figure the number of scales in the vertical rows of the side is represented as twenty- three; a number which I have found equaled in only one species of the genus, I. ovntus, in which it is twenty-four. The specimen does not permit the scales in the widest part to be counted, but immediately anterior to the dorsal and anal fins the number is apparently sixteen ; this renders il prob- able that the number in the anterior rows may reach nineteen, a number which may be considered as normal for the genus. The posterior position of the dorsal fin, the uniformity in the size of the scales, and tlie unarmed or short-spined character of those of the dorsal line clearly mark this species as distinct from any other known. 'Am. Naturalist, vol. 20, 188fi, pp. 243-246. FOSSIL FISHES. 45 ISCHYPTEEUS PARVUS W. C. R. (MS). PI. XIII, Fig. i. ■ In the manuscript report of J. H. Redfield, now in my hands, I find a description of a small species of Ischi/pferus to which he gives the above name, crediting it to W. C Redfield. He also refers to the figures given by Prof. Edward Hitchcock^ as illustrating the species. His description is as follows : Fish small and fusiform; bead small — less than one-quarter length of tisb ; scales minute concentrically striate, pectorals rather small, rays delicate ; ventrals very small; dorsal small and triangular, with anterior raylets stout and few in number; anal very small; tail forked, lobes rather obtuse; length three inches, breadth three- quarters of an inch. Occurs at Sunderland, Mass., Boonton, N. J., and [teriiaps at Westfield, Conn. This species is rare. Very few perfect individuals have been found. Its small size and the delicate character of its scales and fins will at once distinguish it. The above description is so brief and general that in the absence of type specimens it is difficult, if not impossible, to identify the species. The figures in the Geology of Massachusetts, to which Mr. Redfield refers, are evidently drawn from very imperfectly preserved fishes, of which little more can be said than that they belong to the genus Ischypteriis. They are, how- ever, quite distinct from the little fishes found at Durham to which I have given the name /. minutus, being narrower and more fusiform and with much smaller dorsal fins. A little fish found at Sunderland, much more like those figured by Hitchcock, is represented on Plate XVHI, Fig. 4. It is fairly well preserved, and we can see by its fusiform body and small dorsal scales that it is not the young of I. tenuiceps. There can be little doubt, therefore, that it represents the species figured by Hitchcock and cited by Redfield as representing his I. parvus. The figure now given may therefore be taken as the first truthful illustration of that species. Whether it is dis- tinct from any other described remains to be shown by further investigation. At Durham and Sunderland fusiform fishes of the genus Iscliyptenis con- siderably larger than this or that figured by Hitchcock occur, though rarely, and not often in good preservation. These have the general 'Geol. Mass., r^narto ed., vol. 2, j)!. XXIX, fig. 3, aucl iii'atlas accoiliiiauyiug octavo ed., pi. XIV, fig. 44. 46 TRIASSIC FISHES AND PLANTS. aspects and proportions of the much smaller fish now figured, and it is possible they are only older individuals of the same species. Of this, how- ever, we have no positive proof. The larger fishes referred to were con- sidered by W. C. Redfield as belonging to the species /. fuUus, and that is possible ; but judging from the material I have seen I should say the fishes of the Connecticut Valley were more delicate in structure, with smaller and weaker fins, and that they will probably prove to be distinct. ISCHYPTEEUS LATUS 3. H. R. PI. XIII, Fig. 3. Palwoniscus latus J. H. R., Annals New York Lyceum Nat. Hist., vol. 4, PI. II, without description. Palmonisciis latus J. H. R., Am. Jour., Sci., vol. 41, 1841, p. 25. The figure given by Mr. J. H. Redfield in the Annals of the Lyceum lacks the head and does not fully show the tail nor the fins. No description accompanies the plate in the article referred to above; the only mention of the species in the Journal of Science is exceedingly brief, and reads as follows : PalcEoniscus latus J. H. Red field— Broad Palteouiscus. The common length of this species is from four to five inches, and its width is fiom one and a half to two and a quarter inches. It is figured in the Annals of the New York Lyceum of Natural His- tory, vol. 4. Found at Westfleld, Middleficld, and Durham, Conn., and Boonton, N. J. In the manuscript catalogue of the fossil fishes of the United States, read before the Association of American Geologists and Naturalists by J. H. Redfield, I find the following description of Palceoniscus latus.- Fish ovate, fusiform, head obtuse, rather large, between ouethh'd and one-quarter the whole length of the fish ; scales small, those of the anterior portion of the body much deeper than long, concentrically striate, especially on the posterior edge; pectoral fins small and delicate ; ventrals small ; dorsal rather large, with anterior raylets very long, stout, and numerous; anal moderate, anterior raylets strong, tail forked, lobes rather obtuse, anterior raylets small; length four to five inches, breadth one and a half to two inches. Occurs at Sunderland, Mass.; Middletown, Conn.; Pompton and Boonton, N. J. The dorsal and anal fins of P. latus are far less elongated than iu the other species, though they still preserve the strong armature peculiar to the American species of FOSSIL FISHES. 47 Palceouiscus. The comparative breatUU of this species with the sraallness of its scales will also readily distinguish it from its Americau cougeuers. The scales of the anterior portion of the body are deeper in proportion to their length than iu any other species unless in P. ovatvs. The figure given of this species by J. H. Redtield was taken from a specimen which has unfortunately been hjst. I have not been able to find in the Redfield collection the original of this figure; it probably belonged to the New York Lyceum of Natural History, and was burned with the rest of its collections. The lack of it has made tlie identification of the species difiicult. There are, however, no small, short, and broad fishes found at the localities enumerated by Mr. Redfield that agree at all well with his figure and description. At Sunderland, Mass., and Plainfield, N. J., we have obtained a few small ovoid fishes which correspond better than any others with the definition of /. latus. These fishes have about the dimensions assigned to this species by J. H. Redfield, viz, a length of four to five inches and a width of one and a half; the head is i-elatively small and pointed, the scales of the dorsal line are prominent, and the broad shield-shaped scale which covers the base of the dorsal fin is relatively very large— as large, indeed, as in any of the large species of the genus — and is notched behind where it touches the first of the fulcra. The fins are all small and weak, the body immediately anterior to the caudal fin is narrowed to about one-third of the breadth between the head and dorsal, the scales are relatively small and crowded, eighteen in a vertical row between the median lines of back and abdomen in the broadest part of the fish. Of these, the six lower are small, square, and of nearly uniform size; the seventh row is the beginning of a series consisting of eie-ht, which are higher tlian long, the middle ones near the head being just twice as high as broad; above these higher scales are four rows of smaller square ones, of which the uppermost is excavated to fit the rounded base of the great spined scale which stands at the head of the row. These little fishes I have supposed might represent Mr. Redfield's species, but I have found none at Boonton or Durham which I could associate with them. At Boonton a somewhat similar species (J. elegans) is not un- common, but that is larger, has smaller dorsal scales, and a more arched back. 48 TEIASSIO FISHES AND PLANTS. ISCHYPTEEUS MINUTUS, 11. Sp. PI. XIII, Figs. 5, 5a. Fisbes three inches In length b}^ one inch broad ; long-ovoid in outline ; body widest at base of dorsal fin ; head pointed, one-quarter the entire length ; dorsal fin located at about the center of the body, relatively large and broad ; anal fin just reaching to base of caudal ; tail narrow ; caudal fin, like all the others,, delicate in structure. The little fislies upon which the above description is based have been found only at Durham, Conn. They diff"er from the other small species of the genus found elsewhere by their broader, more ovoid outline^ the large size and breadth of the dorsal fin, and the general delicacy of structure. It is .quite possible that we have here the young of some species of Ischypterus of which the mature form has been described under another name, but there is no fish found in the locality where these occur with which the resemblance is so close as to indicate this, and no connecting links have been found between these little fishes and those of larger size. Their structure was evidently very delicate, and they are so imperfectly preserved that a full description and satisfactory comparisons can not be made from any specimens )-et obtained. Till more material further illumi- nating the subject shall be procured we may consider the individuals of this species as distinguished by their small size, ovoid form, delicate struct- ure, and especially by the relatively great size and breadth of the dorsal fin. The small size is in itself, perhaps, a sufficient reason for the delicate structure, which permitted the destruction of most parts; but it will be noticed that in botb the specimens now fissured the body is unusually wide opposite the dorsal fin, and this fin is relatively larger and broader than in any other known species of the genus. Possibly this is simply the i-esult of immaturity, as the fins are abnormally large in many young fishes. The great breadth of the dorsal fin may, however, prove a constant character, and thus serve as a means of distinguishing the species. Small fishes occur at Boonton and Sunderland, but they are usually so badly preserved, that little can be said of their specific relations. Oijly at Durham do we find FOSSIL FISHES. 49 the details of structure retaiued, and all the small fishes of the genera Catoptenis and Iscliyptenis which I have thouglit worthy to be figured have been obtained there. It is somewhat remarkable that among the thousands of presumably mature individuals of all the six genera yet found in our Triassic rocks the young are so generally absent. Some difference in size is perceptible among those wdiich we suppose to represent the twenty-eight known species, but if young and old were in the habit of associating together we ought to have graded sizes of many of the species. From the facts that we do not find them and that tlie variation in size among those which we are able to distinguish by certain definite characteristics is limited, we must infer that the young of all or most of these species associated together in different localities from those where the mature individuals are now found. Probably some such nurseries of tlie Triassic fishes will yet be discovered and will help the paleontologist of the future to discriminate between the species, but we must conclude that in the material now before us we have onl}' mature or submature fishes, and this gives a probability to the distinctions we now make. Where we find twenty, fifty, or one hundred fishes which present common characters in size, outline, strength, shape of fins, etc, we ma}' fairly conclude that these represent one species. Like all similar work, however, this must be considered as only provisional and liable to modification by the accumulation of more and better material. ISCHYPTERUS GIGAS, n. sp. PI. XIV, Fig. 3. Among the fish remains which were the fruit of many weeks of quarry- ing at Boonton a few fragments were obtained which belong to a species of Iscliypterus much larger than any hitherto described. Unfortunately, the importance of these specimens was not appreciated by the quarrymen, and tliey did not take pains to preserve all the material which they brought to light. The remains of two individuals were found, both unfortunately much macerated and dismembered, the tails and posterior portions of the body, the most resistant parts, alone being well preserved. The heads and MON XIV — -4 50 TEIASSIG FISHES AND PLANTS. middle portions of the body were a mass of scales and bones apparently representing the place and area of the abdomen, shoulders, and head, but two much confused to admit of accurate description or representation. The length of the fish must have been eighteen to twenty inches and the breadth of the body at the widest part at least six inches. At the narrowest point, immediately anterior to the base of the caudal fin it is quite two inches wide; the tail when fully expanded must have been five or six inches broad. It consisted, apparently, of fifteen closely jointed rays, some of which are one-quarter of an inch in width; the fulcra are numerous above, still more so below ; the anal fin was about three inches long, having at least sixteen fulcra and seven rays ; the scales half an inch or more in diameter, thick, enamel covered, and shining. The general aspect of the fish is that of Lepidofus, as it is much larger and coarser than most species of Iscliijpterus. The tail, however, is considerably more heterocercal than in any species of Leindotus, and in fact in structure is precisely like that of Ischjpteriis. The dorsal line is very imperfectly shown in my specimens, and it is impossible to determine from them whether a row of spine-like scales extended from the head to the dorsal fin. This would be conclusive as to the relationship of this fish to Ischypterus, and doubtless that evidence will be forthcoming. Genus CATOPTERUS J. H. R. Tile-scaled ganoids of medium size, body fusiform or long- ovoid in out- line; head relatively small, obtuse or acute, all head bones highly orna- mented ; cranium opercula, maxillaries, and mandibles covered with tuber- cles of enamel ; clavicles bearing parallel or interrupted raised lines ; teeth numerous, conical, acute on premaxillaries, maxillaries, and mandibles ; fins broadly or narrowly triangular, acute, all bearing numerous closely-set, rod-like fulcra along the anterior margins; rays many-jointed, enameled, and polished ; dorsal fin placed far back on the body, generally opposite the middle of the anal ; caudal fin deeply and gracefully forked ; extremity of body obliquely rounded and extended a sliort distance into the upper lobe of the caudal fin ; anal fin reaching nc^arly to base of caudal ; ventrals midway between anal and pectorals; scales rhomboidal on the sides, toward FOSSIL FISHES. 51 tlie head quadrate, often toothed, near the tail long lozenge-shaped, acute. Along tlie middle line of the back runs a row of somewhat larger ovoid, or polygonal scales of peculiar form. The surface of most of the scales is smooth and polished, but in some species those on the sides near the head are marked with oblique raised lines, and in one species the surface is occupied by lines parallel with the margin and converging to the posterior point. The most striking peculiarity of this genus is the posterior position of the dorsal fin, a cliaracter which suggested the name given it by J. H. Red- field. The species of Catoptems are among the most beautiful of fossil fishes ; the outline is graceful, the head bones are crowded with ornamentation, the scales highly polislied, often serrate or toothed on the posterior margin, and decorated with parallel or concentric raised Hnes. The fins are long, o-race- ful, and flowing ; the pectorals are falcate and acute, the first rays very strong, and thickly set with short fulcra, which give it a serrate appearance The margins of the other fins are decorated in the same way, so that tlie genus may be recognized by even a fragment of a fin. The fin rays are very numerous and frequently articulated, the joints flattened and highly polished, so that in the fossil state the form and structure are often beauti- fully preserved and never foil to excite admiration in the observer. No species of Catopierus has yet been found in the Mesozoic rocks of the Old World, or, at least, no fossil fish has yet been identified as such In eastern America, however, during the latter part of the Triassic age, two or three species were exceedingly numerous in the lakes and estuaries of the Atlantic coast. In New Jersey and the Connecticut Valley the species (^f Catopterus are fewer than of Ischjpteriis, and the number of individuals is on the whole less, but in some localities the two genera are about equally represented. They may be distinguislied at a glance, even when minor dif- ferences are not shown, by the position of the dorsal fin. In Ischyptems this is always anterior to the anal, while in Catopterus it is either opposite or posterior. Since the above description was written I have i-eceived from S. W. Loper, of Durham, Conn., some specimens, which enable me to add some- 52 TRIASSIC FISHES AND PLANTS. thing to the generic description, as they show better than any before known the under and upper sides of the head. From these it appears, first, that on tlie under side of the body the scales extend in a V-sliaped point consid- erably forward of the pectoral fins, the extreme angle being under the center of the head. Secondly, the apex of the arch formed by the mandibles is occupied by a median jugular plate similar to that of Amia; its surface is covered with coarse, rounded, or elongated tubercles, and its sides are notched to receive the conical extremities of the interclavicles (?) by which it is bor- dered. These are covered with polished raised lines with a radiated arrange- ment at the extremity ; they are often broken into tubercles of enamel. Thirdly, the mandibles are narrow and slender and, like the other bones of the head, coarsely granulated. Fourthly, the under side of the pectoral fins shows about ten rays which, simple at base, soon divide into polished rods articulated only toward their extremities ; in this respect showing a structure very diff'erent from that of the upper surface, in which the articu- lations are short and mimerous, apparentl}^ metamorphosed scales ; a char- acter exhibited tlu'oughout the unpaired dorsal, caudal, and anal fins. The bones of the sides and top of the head are not quite as distinctly shown, but the following points of structure can apparently be made out : The cranial bones are all rather coarsely tuberculated ; they consist of a pair of large polygonal frontals, which are notched on the anterior lower border for the eye orbit ; the ethmoid is pentagonal, wedge-shaped poste- riorly, the point interposed between the diverging lines of the frontals ; the sides are straight, slightly inclined toward each other forward, the anterior margin apparently joining the premaxillaries, which are united to form a transversely oval bone bristling with teeth — the extremity of the muzzle. The posterior angles of the frontals are cut to receive small oblong or ovoid parietals. The middle line of the head terminates behind by a triangular supraoccipital, of which the rounded base fits into a sinus in the frontals. On either side of the supraoccipital are small, polygonal post temporals, of which the posterior edge is joined by the scales of the back. The max- illaries are spatulate, broadly rounded or truncated behind and anteriorly fitted to the premaxillary. The orbit is formed by a bony ring, but the number of pieces composing it is not shown. rossiL FisoEs. 53 The operculum is semilunar, anterior margin slightly concave. It appar- ently consists of two parts, which may be operculum and interoperculum, but this is not plainly shown. Joining the mandible behind seems to be a small, oblong quadrate, but this is also too obscure to be insisted upon. All the specimens which show the structure of tlie head fairly well belong to Catopterus RedJielcK. In these tlie first rows of scales next the head and in the gular triangle are ornamented with tubercles or ridges, and their posterior margins are notclied or toothed. Like tlie joints of the fin rays tliese are brilliantly polished, and confirm what has been said in regard to the great beauty of tlie external decoration of this elegant fish. Catopterus Redfieldi Egerton. PI. XV, Figs. 1, 2, 3. Among the Triassic fishes taken to England by Sir Charles Lyell and examined by Sir Philip Egerton were (1) three species of IschyiHerus ; (2) representatives of Catopterus gracilis J. H. R ; (.3) Catopterus Bedfieldi, " a broader fish than the preceding, and with scales not so long in proportion to their depth." ^ This is all the description we have of this species ; but as there are found at Durham, Conn., many individuals of a large and broad species of Catopterus, and one to which the name ffracilis is certainly inap- plicable, I have thought it probable that this was the fish referred to by Sir Philip Egerton, and I take pleasure in accepting his name, and by figures and more complete descriptions securing to the founder of the genus the dedication of its finest species. This may be characterized as follows : Fish of large size, ten inches in lengtli by three in breadth ; long-ovoid in outline, broadest between ventral and pectoral fins ; head small, pointed, about one-sixth of the entire length, or one and one-half inches long and deep ; bones of the head all thickly set with enameled tubercles ; clavicles ornamented with raised lines and elongated tubercles of enamel; dorsal fin opposite middle of anal ; caudal fin forked, though less deeply than in some other species ; anal fin broad, not reaching the base of caudal ; ven- trals midway between anals and pectorals ; scales on sides near head oblong > Quart. Jour. Geo!. Soc. London, vol. 3, 1847, p. 278. 54 TEIASSIC PISHES AND PLANTS. or quadrate, sometimes twice asliigli as long, surface partially covered with I'aised lines which project to form teeth on the posterior margin. In tlie middle of the body the scales are longer than high, plain or faintly striated, and beai-ing one or more posterior teeth ; scales near tail rhomboidal, smooth ; scales of median line of back transversely oval or somewhat poh'g- onal, faintly striated ; teeth numerous on premaxillaries, maxillaries, and mandibles, from one- eighth to one-quarter of an inch long, conical, sub- acute. The average size of the fish of this species may be said to be nine inches in length by three in breadth. Tlie general form and proportions were similar to those of our shad and the outlines were equally elegant. As we always find the fishes of this species lying on the side, we may infer that they were laterally compressed, the vertical diameter being greater than the transverse. The specimens for wliieh Sir Philip Egerton suggested the name now given were from Dorliam, Conn., and this seems to be the special home of the species, though it has apparently been found at otiier localities in the Connecticut Valley and in New Jersey. Fully one-half of all the fishes obtained by Mr. Loper at Durham belong to this species, and he has fur- nished me with a lai'ge number of beautifully preserved specimens. As in all the species of the genus the head seems to have been largely cartilaginous, and as a consequence is often defective or distorted in the fossils. Occasionally, however, as in the specimen represented in Fig. 1, on Plate XV, the outline of the head is accurately shown as well as the posi- tion of the eye and the form of several of the head bones. But even here they are somewhat confused, and it is difficult to compare bone by bone the structure of the head with that of the palaioniscoid fishes of the Car- boniferous, Avith which tlie relationship has been supposed to be close. So far as we can judge from the specimens before us branchiostegals are want- ing, the operculum is nearly vertical, and the eye surrounded by a bony ring comjDOsed of two pieces. Unfortunately the head bones are not only generally displaced, but they are covered with a coating which obscures the sutures, the matrix clino'ing to the granulated surfaces of the head bones much more closely than to the polished scales. FOSSIL FISHES. 55 Catopterus gracilis J. H. R. P]. XVI, Figs. 1, 2, 3. Catopterus gracilis J. H. R. (Am. Jour. Sci., vol. 41, 1841, p. 27) Fish elon- g'lited, fusiform ; covered with rhomboidal scales of medium size. Head rather small, one-fifth of the whole length, aud in well preserved individuals presents a flnely granulated surface. Operculum lunate, arched ; teeth small, obtuse, in numerous rows; back nearly straight, slightly arched, lateral line nearly parallel with back. All the fins, including the caudal, have a series of very short aud close raylets begin- ning afc a point just anterior to the fin and extending from the first or anterior ray to its extremity, giving a serrated appearance to the anterior border of the fin. In the dorsal, anal, aud caudal fins these raylets are preceded by Lubricated, pointed scales, which seem gradually to pass into raylets; the pectoral fin is long and narrow, in- serted very near the operculum; the first, second, or third rays very strong and conspicuous, the remainder more slender; all the rays except perhaps the first are articulated or subdivided toward their extremities; number of primary rays ten to twelve, anterior raylets about twenty. Ventral fins small, inserted midway between the pectoral and anal, rather near the pectoral. The rays are all slender, about eight iu number, anal fin large, midway between ventral fin and tail, and occupies about one-fourth of the distance between them ; the rays are twenty-five to thirty in num- ber, very slender and filiform and much articulated ; dorsal fin small and triangular, situated opposite the posterior part of the anal ; rays ten to twelve, decreasing in size from the first; tail forked, slightly heterocercal ; the scales of the body extending to about one-third of the upper lobe ; lobes long and acute; caudal rays thirty to forty, finely articulated and subdivided. The scales of the anterior part of the body are much broader than those of the posterior, aud iu old individuals are undulate and subserrate on the posterior margins. The scales become more ftnd more rhombic and decrease in size as they approach the tail ; the scales of the dorsal ridges are of an irregular polygonal shape, presenting a triangular form posteriorly, and are much more imbricated than those of the sides. One or two very large scales are found upon the ventral ridge posterior to the anal tin. There are usually fifty-two to fifty-five rows of scales iu length aud fifteen to twenty in breadth ; length of fish ten inches. Found at Middletown, Durham, and Southbury, Conu., and Boon ton, K J. The above is a description of Catopterus gracilis contained in the manu- script cop3^ of the Report on the Fossil Fishes of the United States, read to the Association of American Geologists and Naturalists at New Haven, Conn., ill 1845, by J. H. Redfield, and kindly communicated to me by liim. A briefer and earlier description of the genus and species, with a figure of C. gracilis, was published by J. H. Redfield in volume 4, page 37, of the Annals of the New York Lyceum of Natural History. The type specimen was then in possession of the Yale Natural History Society, and is now in 56 TRIASSIC FISHES AND PLANTS. the cabinet of Yale College. Unfortunately it is verticallv crushed and nar- rowed, and gives a very imperfect idea of the species to which it belongs. After examining the specimen I am convinced that it sliould be associated with the broad form, for which the name C. Itecljieldi was suggested by Sir Philip Egerton. That fish in its normal condition has nearly the outline of the shad, and with a length of ten inches has frequently a breadth of three ; the name Cato})tcrus cjracilis is therefore inappropriate, and conveys a false impre.ssion. If the specimen figured were accepted as the type it would be better to consider it a synonym of C. Redfieldi and abandon the name C. gracilis. But there is a species of Catopterus which 'is much more slender than C. Redfieldi, and of this numerous specimens Avere in the hands of the Messrs. Redfield, and doubtless influenced them in selecting the specific name gracilis. It is certain that specimens of this fish served as a basis, in part at least, for J. II. Redfield's description, and it is even doubtful whether any good specimen of C. Redfieldi was ever examined by either W. C. or J. H. Redfield. They have mostly been procured from Durham, Conn., by Mr. S. W. Loper, in the last ten years. From these facts it has seemed to me less liable to produce confusion and to do more complete justice to Messrs.. Redfield to retain the name gracilis for the more slender fish, to which the description of J. H. Redfield is not inappropriate, while it is not applicable to the broader form to which the specific name Redfieldi has been given. I will only add to the description of J. H. Redfield that Ccdopterus gracilis is always fusiform, often quite slender, the head never more than one-fifth of the entire length, the fins relatively long and narrow, the body widest at the ventrals, where it is sometimes, thougli rarely, an inch and a half in width, and behind the dorsal often not much more than half an inch wide ; the scales are quadrate near the head, oblong in the middle, and rhomboid at the posterior extremity of the body. They are sometimes finely serrate on the posterior margins, never deeply toothed as in C. Redfieldi, and the surface in all the specimens I have seen is essentially plain. On PI. XVI, Fig. 1, is represented an entire fish of this species, and one of the broader forms, while Fig. 3 shows the posterior half of the body of one of the more slender individuals. The difference of form between this and FOSSIL FISHES. 57 the preceding species will be seen by comparing the latter figure with that of a corresponding portion of tlie body of C. Bedfiddl given on PI. XV, Fig. 3. Catopterus minor, n. sp. n. XVII, Figs. 1-4. Fishes robust, fusiform, five to six inches in length by one and a half inches in diameter at widest part, which is immediately behind the pectoral fins ; head depressed, conical, pointed, all head bones covered with coarse granulations of enamel; clavicles marked with strong longitudinal plica- tions; fins triangular or falcate, sharp-pointed; caudal deeply forked, lobes gracefully arched, acute; dorsal and anal fins opposite; radial formula as follows: Pectorals— fulcra twenty-seven, rays six; ventrals— fulcra eight- een, rays five; anal — fulcra twenty-four, rays twenty long and three shorter on anterior margin; dorsal — fulcra three, rays fifteen; caudal — rays thirty long, three shorter above and below, thirty-six in all; lower lobe — fulcra thirty, with three large fulcral scales at base ; upper lobe — fulcra eighteen, with three fulcral scales, which are succeeded forward by four large peltate scales on dorsal line, reaching half way to base of dorsal fin; scales quadrangular, nearly uniform in size; on the lateral line forty- two, which are marked by mucous tubes; about twenty-four in the vertical rows on the side near the head; those of the median line ovoid or polygonal; surface of all the scales on the anterior portion of the body ornamented with raised lines; on the nape and abdomen part of the scales carry one or two raised lines parallel with the margin and. converghig to the pos- terior point; on the side near the head all the scales are obliquely trav- ersed by raised and often beaded lines, which terminate in acute denticu- lations of the posterior margin. From Durham, Conn., I have obtained, throuo-h Mr. Loper, quite a number of small specimens of Catopterus, which are of nearly uniform size —about five inches in length by one and a quarter in width — all lying partly upon the abdomen and showing the line of median scales upon the back. This proves that the body was round, or perhaps somewhat flattened vertically; the head is small, depressed, pointed; the scales of the posterior portion of the body highly polished; those of the anterior ornamented with raised lines and having the posterior 58 TEIASSIC FISHES AND PLANTS. margins coarsely denticulate ; the fins are of moderate size, ver j graceful in their outlines and beautifully constructed and preserved. On tlie whole, these are the handsomest fossil fishes of which I have any knowledge. I have been somewhat in doubt whether they may not be regarded as the young of C. BcdfiekU, with which they are associated and which they in some points resemble, but they present some distinct characters which the}- have in common among themselves, such as tlie pointed head, the round and vertically flattened body, the ornamented scales varying comparatively little in size, and the opposite position of tlie anal and dorsal fins. These char- acters have seemed to me snflficient to make this little group of fishes the representatives of a distinct species. The relationship of these fishes to that which I 'have called Catopterus ornatus is close; the size, form, position, and other features of the body are the same, the only difference being the pecul- iar ornamentation which covers most of the scales of C. ornatus, and is only faintly indicated in a very few scales of some individuals of C. minor. The radiating lines which mark the side scales in the present species are wanting or but faintly indicated in C ornatus, but we have some traces of them in the much larger fishes which I have supposed to represent C. Redjieldi. The theory that these fishes constitute a distinct species of Catopterus is confirmed by the fact that, so far as at present known, they are found at no other locality than Durham, Conn., although the larger species of the genus are abundant at Boonton, and are sometimes met with at Sunderland. Catopterus genatus, n. sp. PI. XVIII, Figs. 3, 3a, 3b. Fishes fusiform, fi^'e inches long by one and a quarter inches wide at the broadest part ; head bones unknown ; fins all delicate ; anal opposite dorsal ; scales rhomboidal or elliptical, of nearly uniform size, relatively large, external surface ornamented by raised lines parallel with the border and terminating in the posterior point or angle ; along the dorsal median line is a row of ovoid scales somewhat larger than the others, marked by the usual raised lines parallel with the margin, and in addition a single raised line, sometimes beaded, which passes from tlie center of the scale to the posterior point. On tlio sides near the head the scales, which all show FOSSIL FISHES. 59 more or less of the concentric lines, are also faintly marked with radiating, beaded lines terminating in sharp teeth at the posterior border. Only a single specimen of tliis little fish has yet been found. It has tlie form and size of Catopterus minor, but differs from that and all other species known, in the peculiar and pronounced ornamentation of the scales. Most of these are decorated with strong raised lines parallel with the mar- gins and running to the posterior point, which is often somewhat prolonged. On the sides near the head this ornamentation is joined to or superseded b}^ the radiating raised beaded lines terminating in teeth, often though not always seen in Catoptcrus Redfieldi and Catojiterus minor. The body must have been round or somewhat flattened verticallv, since it lies on the ab- domen with the middle line of the back uppermost, the position, generally assumed by the fishes which I have designated by the name of C. minor. The general aspect of these fishes is so similar, that I have been inclined to consider them as varieties of the same species, but the ornamentation of the scales in C. ornafus is so marked, that I do not feel authorized to unite them without better evidence than I now possess. The ornamentation described above is on the same plan with that of the scales of Ccelacanthus elegans from the Coal Measures, but the number of raised and converging lines is less in the Triassic fish. Figs. Sa and 3& represent the scales enlarged to show the ornamenta- tion, the former the ovate scales of the dorsal line, the latter the rhomboidal scales of the sides. Catopterus anguillifoemis W. C. R. PI. XVIII, Fig. 5. W. C. Redfield describes a species of Catopterus in the following words:' Catopterus ancjxdTtiformia W. C. R. (Eel sbapecl Catopterus).— This remarkable species, as liitlicrto found, is from seven to nearly ten inches iu length; width, half to three-fourths of an inch. It has a finely-forked and extended caudal fin of delicate structure; a well-extended dorsal; and all the fins are fringed with the fine raylets which pertain to this genus. The impressions of the fins are usually but faintly visi- ble, owing, probably, to their delicate structure. The scales are equally indistinct, and the impression of the head is seldom visible. Found at Westfield and Middletowu, Conn.; Boonton, N. J.; and, as I have been informed, at Sunderland, Mass. 'Amer. Jour. Sci., vol. 41, 1841, p. 27. 60 TEIASSIC FISHES AND PLANTS In llie collection of fossil fishes left by W. C. Redfield, and among- hundreds of specimens I have examined from Boonton, Durham, Smider- land, and elsewhere, I have seen but two which correspond with this de- scription. These are from Durham, and are contained in a shale that is quite metamorphosed, and in which the impressions of the fossils are indis- tinct. They are very defective in details of structure, but it is liardly pos- sible to avoid tlie conclusion that they represent a fish diff"erent from any other known. One of these is represented on PI. XVIII, Fig. 5 ; the other is larger, though scarcely wider, and the outline is less complete. It would be somewhat surprising if it should prove true that in the same locality lived two species of the same genus differing in form as much as these slender eel-like fishes differ from Catopterus Bedfieldi, which, when mature, was relatively as broad as a shad. It is therefore quite pos- sible that when better specimens of the slender fish shall be found they will present points of structure which will require reference to a new genus. So far as can be observed, however, they exhibit the characters of Catop- teriis ; the tail is deeply forked, and the caudal, like the pectorals, is bor- dered by the fine fulcra so characteristic of that genus; so that, till conflict- ing evidence is found, we must follow W. C. Redfield in the name he has given. Tlie absence of all details of structure in these fossil fishes is proof of great delicacy of organization, such as we find in the young of most fishes, but the great length of these specimens forbids tlie inference that they are young fishes, since no traces of larger individuals with anything like the same proportions have been discovered. Catopterus parvulcs W. C. R. PI. XVI, Figs. 4, 5. The description given of this species by Mr. Redfield will be found in the article' so frequently cited on the preceding pages. It reads as follows: Catopterus parvulus: W. 0, E. (Little Catopterus). — This small and delicate fossil is but obscurely developed in the few si)ecimens which have been obtained. The ex- tremely fine spread caudal and other fins, with their slender frontal raylets, serve to I Am. Jour. Sci., vol. 41, 1841. FOSSIL FISHES. 61 mark it as a member of the geuus, although these raylets are fewer iu uumber and of greater and more unequal length than in the other species. In the few specimens obtained the caudal extremity is commonly found iu a bent or half-twisted position. Found at Middlefield, Goon., Sunderland, Mass., and Boontou, N. J. Ill our excavations at Boonton, where we obtained several hundred fishes in better or worse condition, a few dehcate, imperfect, and usually distorted specimens were found whicli correspond fairly well with the above description, and yet it has seemed to me that they are probably the young of the larger species of Catopterus; if not, the species can only be satisfac- torily defined from material more perfect than any I have yet seen. These little fishes are generally from two and a half to four inches in length and very imperfectly preserved; that is, the scales are scarcely visi- ble and all details of head structure are wanting. The fins, especially the caudal, are sometimes fairly well shown, and consist of numerous extremely fine parallel rays bordered by fulcra of corresponding delicacy. These prove that they belong to the genus Catopteriis, but their minute size and their delicacy of structure are signs of immaturity, and it is therefore impos- sible to affirm that they constitute a distinct species. It may be said, how- ever, that with these little fishes somewhat larger ones are found which exhibit nearly equal delicacy of structure. They are from four to six inches in length, with a maximum width of perhaps an inch near the head. The fins are sometimes well shown, but the scales are almost invisible. Some- thing of their indistinctness may be due to decomposition or to imperfect fossilization, but the fin rays are much more slender than in the smaller individuals of Catopterus found at Durham, which I have designated by the name Catopterus minor. Hence I must conclude that they are not specifi- cally identical with these. For the present it may be perhaps as well to let these small, delicate, and imperfectl}^ preserved specimens of Catopterus stand for Redfield's species C. parvulus, but it is quite possible they will prove to be the young of C. gracilis. Genus DICTYOPYGE Egerton. Small heterocercal ganoids; body fusiform; head siiiall, conical, one- fifth the entire length; muzzle rounded, obtuse; opercula large, semicircu- 62 TRIASSIG FISHES AND PLANTS. lar; clavicles coarsely plaited; bones of the head all granulated; scales rliom- boidal, smooth, those of the lateral line strongly marked, oblong, some- what rounded above and below, emarginate behind, showing conspicuous mucous pores or tubes; scales of the under side of the body ver}^ numerous, narrow, elongated longitudinally; pectoral and ventral fins small; dorsal fin opposite to or a little in advance of the anal; caudal fin forked; anal fin broad, rounded, consisting of twenty-two long and two short rays, of which the central ones are broadest and are supported by strong interspinous bones; anterior rays of all the fins set with short, oblique, obtuse, polished fulcra. At the base of the caudal fin, above and below, these are succeeded by large, ovate, pointed, fiilcral scales, which reacli forward to the dorsal and anal fins. The type specimens of Dictijopyge were obtained by Sir Charles Lyell at Blackheath, Va., and were described by Sir Philip Egerton.^ This fish was previously described by W. C. Redfield under the uinme of Catoptems macrurus in the American Journal of Science, (vol. 41, 1841, p. 27), but Sir Philip Egerton, as cited by Lyell, considered it distinct from the genus Catop- teriis, because "the dorsal fin is more strictly opposite to the anal than in Catoxiterus Bedjieldi," and because, "having a homocercal tail, it can not be comprehended in it." Mr. Redfield did not accept the genus Dictyopyfje of Egerton, because, as he said, Catopterus macrurus was really no less hetero- cercal than the other species of the genus, and with the other common char- acters the slight difference in the position of the fins had in his judgment only a .specific value. There is something to be said on both sides of this ques- tion, and perhaps it cannot be settled until we have more material; but by a careful study of that now in hand I have been inclined to accept the genus Dktyopyge. In Catopterus macrurus of Redfield the operculaare larger, con- stituting one-half a circle, the scales of the under side of the body are much more numerous, the dorsal fin is more in advance, the anal fin broader, larger, and rounder, and the interha?.mal spines by which it is supported are much stronger, and finally the tail is less forked than in tlie other species of Catopterus. In my specimens, as well as in those figured by Sir Philip Egerton, the dorsal fin is decidely in advance of the anal, and both are so •Quart. Jour. Gcol. Soc. Loudou, vol. 3, 1850, p. 275. FOSSIL FISHES, 63 large and round as to give a pecidiar aspect, which will strike the most casual observer. On the otlier hand, tlie character of the posterior end of the body- is precisely the same as in Catopferus, and the anterior margin of each fin is set with the numerous divergent fulcra which are so characteristic of that genus ; but in the species under consideration they are more numerous, shorter, blunter, and more divergent. Hence we must conclude that if this fish represents a different genus it is still very closely allied to Cat02)terus. Sir Philip Egerton, as cited by LyelV alludes to fragments of another and larger species of Bidyopjge from Chesterfield County, Va. This I sup- pose to be the same fish as that represented by some fragments I have from that region. It was a much larger fish than D. macrura and the divisions of the fin rays were marked by several raised lines, constituting a peculiar style of ornamentation. Johannes Striiver in 1864 published^ a notice of the Fossil Fishes of the Keuper, of Coburg, Saxony, in wliich he describes and figures a spe- cies of Bictyopyge (D. socialis) and reviews the structure and relations of the genus. With this notice he also publishes a figure and (p. 305) a de- scription of another fish associated with the last, Semionotus Bergeri Ag., to which I have alluded elsewhere. These figures and descriptions are of special interest for comparison with the fishes of our American Trias; for it is probable that, if a few good specimens of Ischypterus and Catopterus had come into the hands of Agassiz, Berger, Egerton, or Striiver previous to the publications of Semionotus, Catopterus, and Bictyopyge, Iscliypterus would liave been united with Semionotus and Bictyopyge socialis have been included in Catopterus. Judging from Striiver's figures it is impossible to designate any important character by which these fishes could be genericall}^ dis- tinguished. Semionotus Bergeri lias a dorsal fin which is a little broader than that of any of our species of Iscliypterus, but in all other respects, even to the row of erect and pointed scales on the back, there is the greatest similarity between the two genera ; nor are there any differences to which we can give generic value between Bictyopyge socialis of the Coburg-Keuper Saudstein and Catopterus gracilis of Redfield. It is true that in the former ^ Quart. Jour. Geol. Soo. Loudon, vol. ;!, p. 277. 2 Zeitschrift Deutscb. geol. Gesellscliaft, Berlin, vol. IC, 18(34, p. 303-330. 64 TEIASSIC FISHES AISTD TLANTS. the dorsal fin is exactly opposite the anal, while in the latter the first rays of the dorsal are opposite tlie middle of the anal, but in another species of Catopterus {C. minor N.) the position of the fins is exactly that of D. socialis ; while in the Virginia species, which was taken by Egerton as the type of his genus Dictyopuffe, the dorsal fin is sensibly anterior to the anal ; so that this character can not be considered as diagnostic. Another distinction which Striiver makes between Catopterus macrurus (which he erroneously names macropterus throughout his article) and Didijopyge socialis, viz. "fulcra der Schwanzflossen ziemlich gross" and "fulcra sammtlicher Flossen fein,'' does not hold good, for the fulcra are quite as fine in the Virginia as in the Co- burg specimens. The only difi^erences which I can specify between our commonest species of Catopterus and Dictyopyge socialis are the broader operculum, the narrower scales of the belly, and the less deeply forked tail of the latter. In these characters Dictyopyge macrurus and D. socialis are dis- tinguished from all the species of Catopterus found in New Jersey or in Connecticut; and, as I have said elsewhere, these may perhaps afford a raison d'etre for Dlctyopyge. DiCTYOPYGE MACRURA Egerton. PI. XVIII, Figs. 1, 2. Catopterus macrurus W. C. R., Am. Jour. Sci., vol. 41, 1841, p. 27. Bictyopyge macrura Egertou, Quart. Jour. Geol. Soc, London, vol. 6, 1850, p. 4. Fishes small, fusiform; head rather small, surface finely granulated ; scales of medium size, those of the sides and back square or slightly rhomboidal, those on the under side of body very numerous and narrow; pectoral fins of medinm size, primary rays seven or eight, anterior raylets very tine, short, and close, over forty in number; ventrals small, triangular, and elongated, rays eight or nine, fulcra about thirty ; anal very large, quadrate, sometimes reaching as far as base of caudal, rays over thirty in luimber; tail forked, very long, acute, and spreading, lower lobe longest, rays thirty- five to forty, closely articulated and toward the extremity finely subdivided. Length five inches, breadth one and a quarter inches. The above description is abbreviated from that of J. H. Redfield. Fur- ther details will be found in the discussion of the generic relations. Up to the present tivne no specimens of this fish have been fomid else- where than in the Richmond coal basin. There it is locally very abun- FOSSIL FISHES. 65 dant ; one slab of shale formerly belonging to the Lyceum of Natural His- tory, though scarcely more than a foot square, carried impressions of over twenty individuals. Genus PTYCHOLEPIS Ag. Fusiform, tile-scaled ganoids of moderate size, from six to twelve inches in length ; head pointed ; fins all delicate and provided with minute fulcra, dorsal triangular in outline placed near the center of the back, pectoral fins pointed, anal fin nearer to the tail than to the ventrals, caudal but slightly heterocercal ; the posterior extremity of the body oblique, longer, and rounded on the upper side; scales quadrangular, generally much longer than high, and traversed by furrows which divide the surface into ridges or folds that suggested the name ; the posterior margin of the scales notched by the extremities of the furrows; head bones all highly ornamented Avith raised lines of enamel ; teeth small, conical, acute. Agassiz first described this genus (1843) from specimens found in the Lias at Boll, in Wiirtemberg. The type he called Ptycholepis Bollensis} This was a fish about a foot in length, which has been met with in England and at several places on the Continent of Europe. In 1852 Sir Philip Eger- ton described another and much smaller species, which he called P. minor, obtained from the Lias at Barrow-on-Soar.^ In 1853 he described and fig- ured still another species very much broader than the last, and called it P. curtus? The specimen upon which this description was founded was from the Lias near Lyme Regis. In 1878 S. W. Loper, of Durham, Conn, found in the Triassic beds, which have yielded so many fishes at that locality, several specimens of still another species of PtycJiolepis, which came into my possession and were de- scribed in the Annals of the New York Academy of Sciences, volume 1, p. 127. Since that time perhaps a dozen more or less complete individuals of this species have been obtained at Durham by Mr. Loper, all of which have passed under my observation. They vary considerably in size, the largest being eight inches long by two and a half broad; the smallest abofit 1 Poiss. Foss., vol. 3, p. 107, pi. LVIII bis. -Mem. Geol. Survey, United Kiugdom, British Organic Remains, Decade 6, 185'^, pi. VII. 'Ibid., Decade 8, 18.5.5, pi. VIH. MON SIV 5 68 TRIASSIC FISHES AND PLANTS. four inches long ; most of tliem being about six inches long by one and a quarter inches wide. Possibly these specimens represent more than one species, but the material yet found scarcely suffices for the accurate defini- tion of more than one, and it is quite possible that the differences they exhibit are only those of age. I give below the detailed description of this species taken with slight modification, from the paper referred to above. Ptycholepis Maeshii Newb. PI. XIX, Figs. 1, 2, 2a. Fish eight inches or more in length by two and a quarter in breadth, fusiform, robust ; head pointed, contained four and a half times in the en- tire length ; all the bones of the head marked with strong raised lines, those of the upper surface somewhat radiate; on the opercula, maxillaries, man- dibles, and gular plates more or less uudulately parallel and forked. The dorsal fin is of medium size and placed near the center of the back ; the anal is set far back, reaching nearly to the caudal; caudal small, forked, the scales and vertebral column reaching distinctly into the upper lobe. The scales on the anterior portion of the body are two or three times as long as high, and are marked with several longitudhial furrows and raised lines. In the middle and posterior portions they are five or six times as long as high, and are traversed by a superficial furrow, w.hich generally reaches from the anterior end half or two-thirds the length and is again resumed on the posterior margin; by this the extremities of the scales are forked. On the anterior portion of tlie abdominal surface the scales are exceedingly narrow, acute, and spine-like. Vertebral column partially ossified. On comparing our fish with the figure and description of P. Bolhnsis Ag. it will be seen that it differs from that species in the position of the dorsal fin (which is placed more anteriorly), in the details of the scales and head markings, and in the greater degree to which the tail is vertebrated and the spinal column ossified. In P. Bollensis the scales are covered with fine, simple, parallel ridges of enamel, but in P. Marslui the ridges are broadei', fewer, and are forked. From P. minor Egerton our species is easily distinguished by its greater size, narrower and notched scales, and FOSSIL FISHES. 67 more vertebrated tail. From P. curtus Egerton it differs in its more elon- gated form, in the plication of the scales, and the more heterocercal tail. The discovery in our Triassic rocks of a species of Ftycliolepis, a genus before found only in tlie Lias of Europe, might seem to open up again the long-debated question of the age of the New Red Sandstone of the Atlantic States, but in fact it does not seriously invalidate the conclusion, based on other evidence, that this series of strata is the equivalent of the Rhsetic beds of Europe. Tlie fish now described is a new species, and has the vertebral column prolonged to a greater distance into the upper lobe of the tail than its Liassic representatives. Without attaching too much importance to this character, we may fairly infer that it indicates a little earlier date. The two specimens now figured are perhaps about the average in size of all those yet found, but I have one which is eight inches long by two inches wide; another specimen is only four inclies long by five-eighths of an inch wide. As a whole our specimens are much smaller than the average of those of P. BoUensis. P. curtus, of Egerton, from the Lias of England, is no longer than our specimens, but it is much more robust. I have dedicated this species to Prof. 0. C. Marsh, of Yale College. All the specimens yet known have been obtained by S. W. Loper. at Dur- ham, Conn. Genus ACENTROPHORUS Traquair. Fi'om the Triassic rocks at Chicopee Falls, Mass., a considerable num- ber of little fishes have been obtained which are distinctly different from any others found in tliis country. Their affinities with Iscliypterus are so close, that I was for a long time disposed to consider tliem as belonging to a species of that genus. The structure is essentially the same throughout, with the exception that the crest of spinous scales which crowns the dorsal arch in Isclipyterus is here wanting and the median line is marked by a sei'ies of round or oval scales a little larger than the quadrangular ones which accompany them on either side. The body is fusiform or conical, widest near the head, tapering gradually with nearly straiglit lines above and below ; the fins are all weak, the dorsal placed far back, nearly as far, indeed, as in Gatopterus. The structure of the fin is like that of IscJiyp- 68 TEIASSIC FISHES AND PLANTS. ieriis, viz, relatively large spinous fulcra border the margins, the rays are few and widely separated, the caudal is narrower than in most species of Ischjptenis, and the inequality of the extremity of the body — i. e., the heter- ocercy — is about the same as in Ischypterus^ and considerably more marked than in Cutopterus. Searching for allies of these little fishes among the figures and descrip- tions which have been published I find in Agassiz's Pakeoniscus glapliyrus and in the three fishes described by J. W. Kirkby,' and called by him Palceonisciis alius, P. Abbsii, and P. varians, what perhaps may be members of the same genus. More material, and that in a better stata of preserva- tion will be required, however, before a satisfactory comparison can be made. It is perhaps not certain that the group referred to all belong to one genus. For example, Pakoniscus alias of Kirkby very closely resem- bles our small ovoid species of Lschpyterus (I. latus), diff'ering, so far as I can see, only in this, that none of the median row of dorsal scales in P. alius are spiny. Dr. R. H. Traquair — who has made a careful study of the fishes described by Mr. Kirkby — considers this a character of generic value, and it has led him to place all the group of Ischpj^terus-like fishes (Pakeoniscus glaphynis, P. alias, P. varians, and P. Abbsii) in a new genus, which, from the absence of spiny scales, he calls Acaiitrophor as? In the Chicopee fishes the structure, so far as can be made out, is altogether that of Ischjpterus, except that the median dorsal scales are all rounded or ovoid. Unfortunately the details of the head structure are ob- scured by the metamorphosis to which the inclosing' rocks have been sub- jected. The teeth are, however, distinctly shown, and they prove to- be. conical, pointed, and relatively strong. The form of the bod}' is more elongated than in most species of Ischjpkras; in that respect resembling Kirkby's Palceoniscus varians and P. Abbsii, but the dorsal fin is placed farther back than in those fishes or in any species of Ischypkrus known. It is, in fact, but little in advance of the anal. All the fins, including the caudal, have the structure of those of Ischypkrus, having few and many- jointed rays and long spiny fulcra, but all are relatively weak. 'Quart. Ji>ur. Geol. S'>c. Lonilon, vol. 20, 135t, p. 353. -Quart. Jour. Geol. Soc. London, vol. 33,1877, p. GGo. FOSSIL FISHES. 69 The size and number of the spiny scales of the dorsal line vary much in the diiferent species of Ischypterus, and it is not quite certain how far these scales can be accepted as a generic character. For example, in some specimens of I. tenuieeps, Ag. sp , the scales of the dorsal line immediately back of the head are enormously developed, being more than half an inch long, erect, radiate, and club-shaped, forming a salient crest, which gives a very striking aspect to the fish. On the other hand, in I. latus J. H. R. the anterior scales of the dorsal line are oval and not spined, wliile the scales back of these, though spined, are generally depressed and inconspicuous. In the figure of Palceoniscus alius published by Mr. Kirkby^ the form is almost exactly that of the specimens of Iscliypterus latus from Plainfield, N. J., and the size is but little less. The head bones, scales, and fins seem to be quite the same, only no spined scales appear in Mr. Kirkby's figures and descriptions. If this should be made a generic character, these two so- closely allied little fishes must be separated, but if the differences noticed above between P. tenuieeps and P. latus should be found to exist in equal or greater degree among other species, they should be given only specific value. The disparity in form between Palceoniscus alius Kirkby and P. varians Kirkby would seem to be associated with some other characters which sug- gest a generic distinctness; for example, the operculum is large and rounded in the latter, narrow and crescent- shaped in the former. That P. alius and P. glaphyrus are members of the same genus can hardly be doubted; but whether P. varians and P. Abbsii should be united with them may perhaps be questioned. The form of the body is certainly very different, and in P. alius the operculum is semilunar and the branchiostegals are scarcely visible (characters common to Iscliypterus), while in P. varians and P. Abhsii tlie opercula, including siibopercula, are nearly round and the branchiostegals arc very conspicuous. ACENTROPHOEUS CHICOPENSIS, n. sp. PI. XIX, Figs. 3, 4. Fishes six inches long by one and three-eighths inches wide, greatest breadth near head, from which point the body slopes equally above and iLoc. cit. 70 TRIASSIC FISHES AND PLANTS. below to the tail; head broad and obtuse, as wide as the body, about one- quarter the entire length; teeth conical, relatively large; dorsal line nearly straight, fins relatively small, dorsal and anal placed far back on body; dorsal midway between occiput and tip of tail and much behind middle of body; anal when depressed reaching nearly to base of caudal; ventrals nearer to anal than to pectorals; caudal narrow and weak; scales of medium size, apparently all smooth, those of the median doisal Hne round or oval without spines. A large number of fishes of this species are contained in the collection presented to Yale College by J. H. Redfield. They are from the same locality, Chicopee Falls, Mass., and are nearly of the same size. They are contained in a rather coarse sandy shale, which has been considerably met- amorphosed by the proximity of trap-rock. This has obscured some of tlie details of structure, such as the surface of the scales, the shape and mark- ings of the head bones, etc., but has left the outlines of the body and the position and form of the fins distinctly visible. The most striking charac- ters of these fishes are the narrow, wedge-shaped form of the body, the straightness of the dorsal and ventral lines, the smallness of the fins, the posterior position of the dorsal, and the rounded and unarmed margins of the median dorsal scales. As mentioned above, these seem sufficient to warrant our placing them in a distinct genus, and since they are in most respects very similar to the group of fishes upon which Dr. Traquair has founded his genus Acenfrojjhonis, it seems best to refer them at least pro- visionally to that. Order CROSSOPTERYGID.^ Huxley. Family CCELACANTHINI Ag. Genus DIPLURUS Newb. Fishes of large size, fusiform in outline, having, in common with other members of the Coelacanth family, a depressed and pointed muzzle, some- what angular occiput, two dorsal fins supported on interspinous bones, a dyphycercal caudal fin traversed by the spinal column which bears at its extremity a small supplemental caudal ; the pectoral and ventral fins lobate, FOSSIL FISHES. 71 the anal nearly opposite the posterior dorsal. The scales are ovoid in outline, relatively thin, imbricated, with one-third to one-half the surface exposed, and this ornamented with raised enamel lines. The bones of the head and pectoral arch ai-e granulated, or ornamented with raised, tortuous, interlock- ing, and interrupted i-idges. The air bladder was ossified, the vertebral column cartilaginous, and having disappeared in fossilization, its place is rep- resented by a smooth band, which is continuous from the head to the ex- tremity of the tail. In the caudal and supplemental caudal fins the course of the spinal cord is marked by rows of scales of diminished size. Tlie neural and haemal spines were ossified, and are distinctly shown in the fossil state. The rays of the caudal fins were supported by interneural and interhsemal spines, to the extremities of which they are attached by sheathing splices. As in all the other members of the ftimily, the fin rays are hollow, and the sides are frequently crushed together in the fossil state, but in Diplurus the walls were strong and the rays generally retain their forms. As in Holopliagus and some of the living siluroids {Doras, Plecostomus, etc.), the fin rays are coated with short, closely-set, acute spines. The dentition is yet unknown, not being shown in any of the specimens found up to the present time. Whether the teeth were flat and obtuse, like those of Undina, or acute, as in Coslacanfhus, remains to be determined l)y further ob.serva- tion. Diplunis shows throughout its structure all the characteristic features of the remarkable family to which it belongs. Its resemblance to Conla- canthus, Hohpliagus, and Macropoma is so close, that if they all occurred in the same geologic formation Ave should hardly be justified, with our present knowledge, in regarding them as more widely separated than are diff"erent species of the same genus. This similarity among the members of the fixmily lias been noticed by Professor Huxley in his remarks on Ccelaccmthus, Holopliagus, Undina, etc.^ It is one of the most surprising and interesting facts in the history of fishes that this family should appear so suddenly, spread over the whole northern hemisphere, retain all the details of its highly specialized struct- ure through the Carboniferous and Mesozoic ages, and then disappear as ' Mem. Geol. Survey Uaifced Kiugaom, British Organic Remains, Decade 10, 18G1. 72 TRIASSIC FISHES A^T> PLANTS. suddenly as it came, leaving among tlie Tertiary and living fishes no de- scendants which can be affiliated witli it. In the desci'iption of Coilacanthus elegans^ I have referred to the close resemblance and possible identity of this little fresh-water fish with C. lepturus, which lived in the lagoons in the coal marshes of England. Not only are all details of internal structure the same, but the elaborate ornamentation of scales and head plates presents no tang'ible differences. From Tlolophagus (with which Diplurus seems to be most closely allied) its only obvious differences are the finer striation of the scales, the wider separation of the two caudal fins, and the fewer articulations of tlie fin rays. In Holopliagus and ^/wrZina and in the Jurassic species of Ccdacanthus described b}^ T. C. Winkler^ the supplemental caudal fin seems to spring directly from the extremity of the caudal. In Dipliirus there is a distinct interval between them; a character which suggested the name Dlpliims, or double-tail. Judging from the specimens of Holopliagus gulo which I re- cently had an opportunity of examining, both the original of Sir Philip Egerton's generic description and the more complete one figured by Pro- fessor Huxley,^ 1 should say that this was a shorter and broader fish than oiir Diplurus. Sir Philip Egerton's specimen is much smaller, but it wants the head, and can not be fully compared. In the body traces of another fish are visible, which had apparently been swallowed. This would show that Holopliagus was carnivorous. The scales of Holophagus are orna- mented with relatively few short, broad, divergent ridges of enamel, wliile our species of Diplurus has many fine parallel thread lines on the scales. In my description of the first specimens of Diplurus found I reported the fin rays to be smooth and the scales granulated, but that specimen was from Boonton, N. J., and was buried in a coarse, sandy shale, in which the minute spines of the fin rays were not discernible, and the thread lines of the scales were broken into granules by the grains of sand. Other and better specimens found later at Durham, Conn., show the characters now described. ■ Rept. Geol. Survey Ohio, vol. 1, pt. 2, Paleontology, 1873, p. 3159. '^Archives Miis^e Teyler, vol.3, pp. 101-116. 'Mem. Geol. Survey United Kingdom, British Organic Remains, Decade 13, 1673. FOSSIL FISHES. 73 T. C. Winkler has described two Coelacantli fislies which may be com- pared with Diplurm. Of these, the first is from tlie Solenhofen (Jurassic) hmestone, and he has called it Codamntlms Haarlemensis} This is in some respects imperfectly preserved, but is apparently distinct from any other described fish, although it deserves more careful comparison with Holo^h- agus gulo Egerton. The specimen upon which Mr. Winkler's description was based is a fish about one foot in length, of which the outlines are fringed and somewhat obscured by dendritic crystallizations of manganese common in the Solenhofen fossils. The scales have all disappeared, but some indications of their surface markings are visible at certain points. If correctly reported these consist of fine, parallel, nearly straight lines run- ning in an antero-posterior direction. This fish would also seem to be peculiar in the strength of the ventral fins, which are represented as fully equal in size to the dorsals. In this character it seems to be distinguislied from all other known members of the family. The surface markings of the scales are like those of our Dlplurus, but that fisli is much larger, the ventral fins are not so strong, and the supplemental caudal is more dis- tinctly separated. The second of the two species described by Mr. Winkler he called CoelacantJms giganteusr This was obtained from the Trias of Wiirzburg. Its derivation makes it of special interest for comparison with Biplums, be- cause no other Triassic Goelacanth is known. Unfortunately, however, the specimen described by Mr. Winkler is a mere fragment. Tliis indicates a fish of enormous size. The caudal fin rays, the only ones yet known, are six inches long and as large as goose quills ; the supplemental caudal is represented as small and as closely approximated to tlie anterior caudal. Whether these fishes described by Mr. Winkler should really be in- cluded in the same genus with the species oVCoelacantlms from the Coal Measures and Permian remains to be seen. The general structure of all the members of the family is so like, that much more material and that well preserved, will be necessary before exact comparison can be made. It is probable that the teeth will furnish tlie best diagnostic characters. The teeth of our Carboniferous CcelacantJms are certainly conical and ' Archives Musfe Teyler, vol. 3, p. 101. «Ibid., vol. o, p. 147. 74 TRIASSIC FISHES AND PLANTS. acute, as is shown in several specimens in my possession. Tfie teeth of Holopliarius can hardly be said to be known ; those of Undlna are stated b^' Count Miinster and by Professor Huxley' to be pavement-like and tuber- culated ; -finall}', those oi Macropoma are conical and acute. DiPLUEUS LONGICAUDATUS Newb. n. XX, Figs. 1-5. Biplurus longicaudatus Newb., Annals New York Acad. Sci., vol. 1, p. 127. Fish attainino- a leng'tli of three feet and a breadth of eiarlit or ten inches; body fusiform, symmetrical; head pointed, sloping rapidly down from the occiput; back gently arched, anterior dorsal fin strong, sup- ported by a semicircular bone ; posterior dorsal placed nearly opposite to the anal fin and midway between tlie anterior dorsal and the extremity of the body; caudal fin very long, supported by tliirty-two ? long and strong rays, which are spliced on to the interneural and interhaemal spines ; supplemental caudal separated from tlie caudal fin by a distinct interval ; in form it is an equilateral triangle about three inches on a side ; the web of this fin is supported by about nine simple fluted rays above and below, of which the bulbous bases were inserted into the cartilaginous extremity of the vertebral column, as posts are set in the ground ; paired fins strongly lobate ; anterior fin rays of these and the two dorsal fins roughened by many short, conical, acute spines ; teeth unknown ; scales ovoid, half an inch in diameter, the exposed poi'tion occupied b}^ fine parallel raised lines, running from front to rear. Only four specimens and the head of a fifth have yet been found of this the largest of our Triassic ganoids, and all these are now in the Geo- logical Museum of Columbia College. Two of tliese specimens were obtained in excavations made at Boonton, N. J.; the others were collected by S. W. Loper at Durham, Conn. The smaller of the Boonton specimens is figured on PI. XX of this memoir. This is interesting, as showing the general form of the fish, the position of the fins, etc., but the details of structure are not distinctly per- ceptible. Another and much larger specimen was found at Boonton, but it 'Mom. Geol. Survey, Uuited Kiugdom, BiitisU Organic Rouiiuns, Decade 10, 1801, p. 17. FOSSIL FISHES. 75 lay near the surface and was so much decomposed that it siipphed httle information, except as regards size; it was about three feet in length. The description of the genus and species published in the Annals of the New York Academy of Sciences was taken from the Boonton specimens, and, from their imperfect preservation, was in some respects erroneous. One of the specimens obtained at Durham, though wanting the muzzle and having the tail much decomposed, exhibits in other respects the beautiful preservation characteristic of the Durham fishes. This shows the oi'ua- raentation of the scales copied on PI. XX, Figs. 3-5, and spines on the fin raj's. Another quite imperfect specimen found at Durham was flattened vertically, and shows a broad rounded head like that of a salamander, but this outline is doubtless due to compression. Four out of the fi\'e specimens of Diplurus known were found lying on the side, from which we ma}^ infer that like most fishes it was broader vertically than transverseh", and that the rounded head of the specimen referred to above was the result of an unnatural position accidentally assumed. It is somewhat surprising- that no distinct teeth can be discerned in any of the heads of Diplurus yet found, thougli some impressions at the extremities of the mandibles of one of the Boonton specimens indicate but do not prove that the teeth were conical and acute. This is the character of the teeth of Coslacanfhus and Macropoma, and it is probable that Diplurus was the enemy and devourer of the many species of smaller ganoids with which it was associated. Numerous large coprolites are found in the same beds, and it would seem natural to refer them to Diplurus, but it is some- what remarkable that none of these coprolites have yet shown any traces of bones or scales such as we might expect to find in the excreta of fishes which lived on ganoids. In the absence of teeth we can not certainly determine whether Di/j^MrHS was carnivorous or herbivorous. The coprolites referred to afford good evidence that the Triassic estuaries contained in considerable numbers a large fish which did not feed on the various scaled ganoids that abounded in the same waters. On the other hand, it should be said tliat the head bones oi Diplurus, including cranium, operculn, maxillaries, and mandibles, were all well ossified, much more so indeed than those of Catopterus or 7G TRIASSIG FISHES AND PLANTS. Ischypterus ; from which we may infer that the dental apparatus was em- ployed in serious and severe work of some kind. The only vegetable re- mains found in the fish-beds of New Jersey and Connecticut are those of land plants — fronds of cycads and twigs of conifers — and it is hardly prob- able these could have formed the subsistence of Biplurus. Mollusks and Crustaceans are entirely absent; so, unless he devoured the scaled ganoids, of which the remains are so abundant, it is difficult to imagine of what his food could have consisted. His structure shows that he was a swift and powerful fish, and his congeners were carnivorous. We may expect there- fore that, when his dentition shall be discovered, that will solve the problem' by demonstrating his carnivorous habits. Pi^HT III. FOSSIL PLANTS OF THE TRIASSIG ROCKS OF NEW JERSEY AND THE CONNECTICUT VALLEY. 77 SKETCH OF TRIASSIC FLORA. The number of fossil plants obtained from the Triassic rocks of tlic valley of the Connecticut and from New Jersey is not large, and as a gen- eral rule they are not well preserved. A sufficient number of fairly good specimens have, however, been collected at Sunderland, Mass., Durham and Middletown, Conn., and at Newark and Milford, N. J., to demand a brief notice. These include the following species : SeMzoneura planicostata Hogers, si^. Otoza.Diiies hrevifolius Fr. Brauu. Equisetum Bogersi Schimper. Cyeadinocarpus Gliapini Newb. Eqiiisetum Meriani ? Brong. Pacliyphyllum simile Newb. ClatJiropteris platy])hylla Broug. Pacliyphyllum hrevifolium 'Sevih. Palissya Braunii Eudl. Cheirolepis Miinsteri Scliimper. Palissya diffusa'! Emmons, sp. Loperia simplex Newb. Baiera Milnsteriana Saporta. Bendrophycus Triassicus Newb. Baiera multijida Fontaine. Dioonitcs longifolius 'Emmons, sp. Otozamites latior Saporta. In addition to these are many ill-defined plant remains, some of which indicate genera and species new to science ; others are decorticated stems and branches, apparently of coniferous trees, probably Palissya. Some of these are quite plain and smooth, but others are marked with lozenge- shaped figures resembling a Lepidoclendron from which the bark was stripped, while the outlines of the rhomboidal leaf scars remained. These have been sometimes called Lcjndodendra, but without warrant, and we have no evi- dence that any species of Lepidodendron passed from the Paleozoic into the Mesozoic Age. 79 80 TllIASSIG FISHES AND PLANTS. No real plant-beds have been found in the northern Triassic areas, and all the remains of plants yet met with seem to be floated fragments, that here and there sank to the bottom of the basin and were buried in the accumu- lating sediments. In the sandstones, which were transported by rapid cur- rents or were formed by shore waves, delicate plants and the foliary ap- pendages of trees would naturally be triturated and destroyed, and in the quarries from which building stone is taken at Newark stems and branches, with an occasional cone, are all the plant remains that have survived the lough treatment which they have received. These are, however, so numer- ous in some of the layers, that they prove the former existence of land cov- ered with vegetation at no great distance At Milford, N. J., the plants are more numerous and somewhat better preserved. There we find the stems of Uqidsetum and Schimneura, with many twigs and some cones of conifers. The Equiseta not unfrequently show the diaphragms which partitioned off the stems at the joints, and, with other things, we sometimes meet with disks or flattened cones of which the surface is radiately striate and which have considerable resemblance to some of the woody fungi, PoJyporus, etc. These I have supposed may have been the diaphragms of Equisetum Bogersi, but they are not sufficiently well preserved to justify any positive assertion in regard to their botanical relations. At Durham, Conn., the fronds of cycads and ferns are not uncommon, and one specimen obtained by Mr. Loper shows a number of fronds of Otosamites radiating from what seems to have been the summit of a stem. The fern fronds, too, are grouped in such a way as to illustrate the radiate arrangement of the pinna; in Clathropteris. The quantity of carbonaceous matter in the shales here is large, and is so generally diffused that we must conclude it was largely derived from the decomposition of plant tissue. This indicates the proximity of a consider- able amount of growing vegetation at the time of the deposition of the shales, and it is possible tliat somewhere near this locality plant-beds will be found which will afford a better view of this vegetation. In the Portland quarries casts of the trunks and branches of trees are not unfrequently met with, but they are always imperfectly preserved, and SKETCH OF TRIASSIO FLORA. 81 we can only conjecture that tliey represent coniferous forests which grew on the highlands at no great distance. Here, too, the remains of what seem to be sea-weeds of a peculiar character are found in considerable abundance. These show a striking resemblance to. plants which have been obtained from the Umbral shales of Pennsylvania and which have been named Dendrophycus by Leo Lesquereux. The similarity is so great, that I have ventured to describe them as a species of this genus, and have called them Bendrophjcus Triasskus. They will be found figured and described more in detail in another part of this memoir. At Hadley's Falls, Sunderland, and, more rarely, at Boonton, the lay- ei's of shale are frequently covered with fragments of twigs of a conifer which has been sometimes referred to as a Voltzia, but, though the foliage is dimorphous, some of the twigs are clothed with closely appressed, scale- like leaves, while on others they are divergent, though always short and thick. One cone-bearing twig of this plant, found at Sunderland, shows distinctly that it is not a Voltzia, but is rather a Pachyphylhim. This plant is apparently the same with that which has been considered by Fontaine as identical with Cheirolepis Milnsteri of Schimper, but the cone referred to shows that it is not Cheirolepis, the scales being small and the exposed surfaces rhomboidal. At Milford, N. J., however, cones and detached scales are found which apparently do belong to Cheirolepis, and perhaps to C. Munsteri. Numerous leaf-bearing twigs associated with these cones show that the foliage was symmetrical and even elegant in character. The branches spread in the same plane and, terminating in twigs pinnately arranged, regularly diminishing in length, pi-esent somewhat the appearance of Thuja or Moriconia, but the form of the leaf is quite different, being short and triangular, similai to that of some species of Pachyphyllum. MON XIV 6 DESCRIPTIONS OF GENERA AND SPECIES. Denduophycus teiassicus, n. sp. PI. XXJ, Figs. 1, 2. In the Geology of Pennsylvania, by H. D. Rogers (vol. 2, PI. XXIII), is represented a fossil plant which is designated as an "algoid resembling a DesmarcstiaP This is fossilized in the red shale of the "Umbral," a part of the Lower Carboniferous formation in northern Pennsylvania. Tlie fossil consists of a number of branching stems from which are thrown off slender dichotomous branches from either side, and these branches, cylindrical like the stems, support numerous opposite or alternate simple branchlets. No reference to this plant is made in the text, but it is evident that it is what it is represented to be, a sea-weed, though its affinities with the living algse may be a matter of doubt. In the Report of tlie Second Geological Survey of Pennsylvania (vol. 3), Coal Flora, by Leo Lesquereux, a figure is given (PI. I) of what is appar- ently the same plant as that figured by Rogers. This is briefly described^ under the name Bendrophycus Desorii. It is said to occur in the red shale below Pottsville, and also abundantly in the bluffs of the Susquehanna above Pittston. Splendid specimens of this plant are reported by Mr. Lesquereux^ as occurring " near Davenport, Iowa, in a bed of clay and hardened sand- rock, traversing like a dike the Corniferous limestone overlaid by the Ham- ilton group." In his description of this plant Mr. Lesquereux says: The root.s or radicular appcndage.s, * » • are api)areiitly cjiindrical or tuba- louR, * * * often braiH'.liiiiy; at riylit angles, three or four mm. in diameter, of 1 Op. cit., pp. 699, 700. ^0]>. eit,, p. 701. R3 FOSSIL PLANTS. 83 coriaceous or horny texture shining on the surface. * * * The rhizoma is 1.20 to 1.50 m. long, perfectly cylindrical, ".i to 4 cm. in diameter, simple and regular iu its whole length, with a rough surface. * * * Tlie top of the rhizoma, abruptly en- larged into a globular shape resembling a cabbage head 17 cm. iu diameter, looked, when broken, like a convolute undeveloped frond, with branches densely rolled to- gether into a ball where the divisions or the relative disposition of the branches could not be distinctly observed. The fronds, very large, 1 to 1.25 m. long by 50 cm. broad, are composed of cylindrical divisions, the primary and secondary oues being thick, the larger 2 cm. in diameter, flattened on the surface, all gradually smaller from the base to their ends, closely distichous, dichotomous, flexuous, with oblique multiple subdivisions, the ultimate two-ranked being very closely pinnately distichous, cylin- drical, pointed, or gradually narrowing from the middle and effaced at the apex. I have copied this description nearly entire because it is ahnost liter- ally applicable to a plant represented on PI. XXI of this memoir and ob- tained from the sandstones of Portland, Conn. When we consider the vast interval of time between the deposition of the Umbral shale of Pennsyl- vania and that of the Rhaetic sandstone of Connecticut, one tlie base of the Carboniferous system and the other the summit of the Trias, it can not fail to be regarded as interesting and surprising that the resemblance should be so complete. But for the a priori improbability that a species of sea-weed should be so long-Hved I should hardly feel justified in giving even a new specific name to the Triassic specimens. Possibly a comparison of more material would show differences not now perceptible, but the pecu- liar mode of growth and the details of structure seem to be essentially the same. In the Portland sandstones, as in the Umbral shales, the fronds of Bendrophjcus are enrolled in masses that suggest cabbage-heads of large size and rather loose texture, while the mode of subdivision and the char- acter of the final ramifications of the frond are so like that, with the simi- larity of the inclosing rock, the specimens from the two localities and horizons are almost undistinguishable. Though a less conspicuous example of a " persistent type " than Stroplwmena rhomhoidalis or Atri/iJa reticularis the survival of a sea-weed of such strongly marked character through so great an interval is as unexpected as it is interesting. Spirophyton, which begins in the middle Devonian (Corniferous) and runs up into the Coal Measures, is another example of persistence in an allied group of algaj ; but that genus is represented at the different horizons by 84 TEIASSIO FISHES AXD PLANTS. quite different species. One found in the Waverly has the frond divided veiy much as in Dendrophjcus, while m other species in the same rock the fronds are simply plicate or swollen into buUse apparently by vesicles which served as floats. BaIERA MiJNSTEKIANA Ung. PI. XXII, Fig. 1. "We here give a representation of one of many specimens of a species of Baiera, f©und by Mr. S. W. Loper at Durham, Conn. Taken by itself, the larger of these specimens (Fig. 1 ) would seem to represent a species closely allied to B. Mmsteriana^ but somewhat taller and more slender than any described variety of that species. Other fragments, however, show that the fronds were sometimes much shorter and broader, and therefore more like the normal form, so that at present we have scarcely evidence that would justify us in separating them. On PI. XXIII is figured a fine specimen of a gigantic Baiera.^ described by Professor Fontaine in his monograph^ with the name B. multifida- This specimen is from Clover Hill, Va., and it is figured to show the close re- semblance between the Virginia and Connecticut plants 5 the former is much more robust, but the characters of the ultimate divisions are essentially the same, and the northern plant may only be a dwarf form of the soxithern. In my descriptions of the plants collected by Mr. A. Eemond, from the Triassic rocks of Los Bronces, Sonora, I liave noticed and figured another and quite different species of Baiera, to which I gave the name of JeanjMuUa radiata, the generic name being practically s}Mionymous with Baiera? Count Saporta has recently discussed at great length the probable rela- tions of the groups of plants which have been known by the names of Baiera, JeanjoauUa, etc' He regards them as belonging to a special line of gym- nosperms which have come down to us from Carboniferous times, and are now represented by the Ginglco (Salishuria). Schimper* also describes the relations of Baiera and Jeanpaidia to each • Mon. U. S. Geol. Survey, No. G, 1383, pp. 8", 88. '^ Report of the S.in Juan Expedition, p. 148; pi. VII, fi^j. (j. 'Pal^outologie Irangaise, V6g(?taiix, vol. :J, 183:!, p. '251. ^ Pal^oiitiologio vdg^tale, vol. 1, pp. 422, 68'3. FOSSIL PLANTS. 85 other and to other plants, and he takes Ihe apparently sensible view that the Wealden Cydojiteris digitata Brong. (Baiera digitata Schenk) and the Rhaetic Jeanpaidia Miinsteriana Ung., though perhaps members of the same botan- ical group, were generically distinct. Brongniart and Unger regarded Jeanpaidia as one of the Bhizocarpce, allied to MarsUia, but Schenk considers it a fern. He also supposes that it is allied to Hausmannia of the Wealden. Its relation with the later genus IS, however, very doubtful. From the Amboy clays of New Jersey, the basal portion of the Creta- ceous system, and resting immediately upon the Triassic beds, I have ob- tained many specimens of Hausmannia, and, though there is a remote re- semblance in the mode of the division of the frond, there is a radical differ- ence in the nervation, and they probably have notliing in common. Equisetum Rogersi Schlmper. PL XXII, Figs. 5, 5ft. As is mentioned in the preceding sketch of the plants of the northern Trias, Equisetum Rogersi occurs at Milford, N. J. It is quite abundant in the Richmond coal field and is mentioned by many writers who have refer- red to the plants of that region, Rogers, Brongniart, Bunbury, Schimper, and others. It is so fully described b}^ Professor Fontaine in his monograph (p. 10) that only a brief reference to it is needed here. Some specimens obtained at Milford, N. J., now in the collection of Lafayette College, which I have been permitted to examine through the courtesy of Prof. T. C. Portfer, exhibit fea tures which are worthy of remark. One of these, a compressed stem, 6°™' in diameter, has the joints below only 2'^"' apart, ar.d on these are set in spiral arrangement disks which mark the attachment of branches or roots such as we so frequently find in some spe- cies of Ccdamites from the Coal Measures. These disks are much distorted and obscured, but they would seem to have been elliptical in outline, 2™- long by 1 ^'^™' wide. In the same rocks and associated with stems of Equisetum are the dis- coid or low-conical, radiately striate bodies which I have already referred to 86 TIIIASSIO FISHES AND TLANTS. and have supposed ^\ere the diapliragms of Eqnisctnm. As these have not been before figured I give herewith representations of tlio flattened base of a small one and the conical, striated upper (?) surface of anotlier. Equisetum Meriani (?) Brong. Some years ago Mr. J. B. Woodworth, now of Boston, Mass., then liv- ing in Newark, N. J., sent to me, with a number of other fossil plants taken from the sandstone quarries near that city, several fragments of a large plant which gave no clew to its entire form, but of which the surface was differently marked from anything I had before obtained from our Trias. The fragments were flattened, only a few inches square, and the surface was deeply impressed by a series of parallel, angular furrows and ridges, three- eighths of an inch wide. The general aspect of the fossil was very like that of a fan-palm, such as we frequently find fossilized in the Tertiary rocks. It had also somewhat the aspect of a SigiUaria, but the sharply angular form of the folds and the absence of leaf scars forbade the supposition that we had here a relic of the great fluted trees of the Coal period. In reviewing the literature of the Triassic flora I have found' what are apparently representations of the same plant. The fossils figured by Sclienk are considered by him as portions of the stems of Equisdnm Meri- ani Brong. {Calamites Meriani Heer), a well-known plant of the Upper Trias of Europe, later placed by Schimper in his genus ScMzoneiira. Until the fructification of these Equisetoid plants of the Trias shall be found which will permit a better comparison with those of older and later formations, it is a useless expenditure of time to discuss the question whether they are species of Calamites which have survived from the Carboniferous age, are true Equiseta, or are species of an extinct genus of that family. It will be remembered that in the Permian rocks stems of Calamites have been found a foot or more in diameter (C. gigas, Brong.), on which the longitud- inal ridges are as broad as in the specimens before us, but in these the ribs are rounded and not angular, as are tliose of the specimens from Newark. 'Scbcnk's Beitriige zur Flora des Keupers uud dcr Rliatiscbeu Formation, Hamburg-, 18G4, pi. VIII, figs, lo, 16. FOSSJL PLANTS. 87 This latter character may not, however, be a constant feature, as Schenk gives a figure of one specimen' in which the ribs seem to be rounded. More material must be obtained before much can be said about the botanical character of these specimens, but they possess much geological interest from their evident similarity to those with which I have compared them, and their occurrence at the same horizon confirms the testimony of other fossils as to the Rhsetic age of our sandstones. SCHIZONEURA PLANICOSTATA RogCrS, Sp. Portions of the stems of this plant, so common in tlie Triassic rocks of Virginia, are occasionally met with in New Jersey, Connecticut, and Massachusetts. This was first described by Prof W. B. Rogers as a Calamites, of which it has very much the aspect, indeed, specimens of Calamites Cistii, Brong, from the Coal Measures could hardly be distinguished from the plant under consideration in some stages of preservation. Professor Fontaine, in the Monograph cited (p. 14), has described this plant from far better specimens than any which occur at the north, and has given reasons for uniting it with ScMzoneura of Scln'mper. Certainly the stems bear a close resemblance to those of the group of plants which have been described by Schimper, Schenk, Nathorst, and others, and found in the Trias and Lias of Europe. But we have never seen any foliary appendages which come very near to those of S. paradoxa figured by Schimper and Mougeot in their monograph on the fossil plants of the Grfes Bigarre'. It is probable that the leaves were narrowly linear, somewhat like those of ScMzoneura Virginiensis Fontaine, and were deciduous like the foliary ap- pendages of Calamites, which are so rarely found in connection with the stem. The only interest that attaches to the imperfect specimens yet found in the northern Triassic basins comes from the evidence they furnish of a close relationship between the deposits which contain them and those of Virginia. ' Op. cit., pi. VIII, fig. la. 88 TRIASSIC FISHES AND PLANTS. Pachyphyllum simile, n. sp. PI. XXir, Fig. 2. Foliaffe dimorplious, on the large branches appressed, sometimes scale-like, on the smaller twigs longer, crowded or open, leaves triangular or falciform, keeled, pointed. In the Triassic rocks of New Jersey, Connecticut, and Massachusetts slender detached twigs of coniferous trees are frequently met with, but they are usually fragmentary and not well preserved. They show two forms of foliage, the appressed and the divergent, and they vary much in their strength; some twigs being very slender, with comparatively remote leaves, while on others the leaves are longer and more crowded. These differ- ences are so marked that I have been led to think the specimens rep- resent two species. Both these forms are represented on PI. XXII, Fig. 2, the stronger and more leafy branches ; Figs. 3, 3a, 3b, the more slender twigs, with shorter leaves. To the first I have given the name Pachyphyllum simile from its resemblance to F. peregrinum of the Jurassic. The other I have called P. hrevifolium. As will be seen by comparing Fig. 2 with the representation of P. peregrinum given by Saporta,' there is a marked resemblance between them, but our plant never assumes the form shown in the figures of P. peregrinum given by Lindley and Hutton- or by Saporta on PI. XL VI of the volume just quoted. The two species are evidently allied but are quite distinct. This plant has been before found in America and has been figured and described, though from bad specimens and erroneously. Prof E. Emmons^ )-epresents a twig from Turner's Falls, Mass., the locality from which that now figui-ed was obtained. It is evidently the same thing, but is badl}^ figured and wrongly named Walchia. It does not belong to tliat genus and can be referred with confidence to Pachyphyllum. Professor Fontaine, in his Monograph, PI. XLVII, Figs. G, 7, represents twigs which are essen- tially identical with the form now figured from Turner's Falls. The larger of these two twigs is copied from Professor Rogers's papei', but no locality ' PaliSontologio franpaise, v^g^taux, vol. 3, 1883, jil. XCVII. 2 Fossil Flora, pi. LXXXVIII. "Aui. Geol., i)t. f), 1857, j). 108, lig. 76. FOSSIL PLANTS. 89 is given. By Fontaine it is referred to Gheirolepis Miinsteri, but I have elsewhere shown that it can not be a Cheirolepis ; we find C. Munsteri at Milford, N. J., but it is plainly distinguishable from this both by cones and by foliage. Pachyphyllum beevi folium, 1^ sp. n. XXir, Figs. 3-3c. Foliage dimorjjhous, on some branches closely appressed and scale-like, on terminal twigs divergent, though the leaves are always short and rela- tively broad. Cones ovoid, one inch in length ; scales rhomboidal, closely appressed. In many localities the Triassic rocks of New Jersey and the Connect- icut Valley, especially where they are fine, gray, or more rarely reddish shales, contain great numbers of slender coniferous twigs, generally short and much broken up. Of these two forms are now figured which may be recognized as typical. Figures 3 and 3ff the more leafy, and 36 the more scaly form. Sometimes we find twigs bearing leaves that are longer than those here shown ; leaves that are divergent, rather open, sometimes spat- ulate, and never really acute. These may be a phase of the foliage of the present species, but they more pi'obably belong to P. simile, showing a sim- ilar diversity of form to that seen in the figures of P. peregrinum given by Saporta.-^ The plant under consideration has been noticed by Emmons and Fon- taine ; by the former it was considered a Walchia, and given the two names W. hrevifolia and TK gracilis.^ Professor Fontaine^ refers it to Cheirolepis Munsteri Schimper ; but the discovery of ovoid cones having small rhom- boidal scales with twigs of this plant, and of digitate cone-scales with branches bearing more acute leaves, have shown its distinctness from the true CJieiroIepis Munsteri. The specimens represented in Figs. 3a and 3c are from Turner's Falls ; those in Figs. 3 and 3& from Durham, Conn. ' Pal^ontologie fratif aise, v^g^taux, vol, 3, pi. XLVIII, figs. 2, 3. 2 Am. Geol., pt. 6, 1857, pp. 107, 108, figs. 74, 75. = Mod. cited, pp. 88, 89. 90 TRIASSIC FISHES AND PLANTS. Cheirolepis Munsteei Schimper. PI. XXII, Figs. 4, 4rt. In my general sketch of the Triassic flora I iiave referred to this plant, and have said perhnps all that it is necessary to say in regard to it. I will only add that the foliage which I have described as expanded in the same plane, and having somewhat the general aspect of that of Thuja or Libocedrus, closely resembles that figured by Schenk in his Fossile Flora der Grenzschichten, and the cone scales are also nearly identical with those he represents ; so that there would seem to be no good reason why they should be regarded as distinct. Schenk calls tliis plant Brachyphyllum Miin- sferi, but Schimper has shown that the digitate cone scales separate this from all other species of that genus, and he makes it the type of his Cheiro- lepis. The pinnate arrangement of the branchlets of our plant, which must have given it the general aspect of Thuja, though the leaves are quite dif- ferent, is not shown in Schenk's figures, and it is therefore possible that this will constitute a specific distinction; but with so many other characters in common it is scarcely probable that this is not also shared by the Euro- pean and American plants. Otozamites latior Saporta. PI. XXIV, Figs. ], 2, 2a. On another page of this memoir I have referred to the geological im- portance of this plant, which is one of the several species common to our Upper Triassic rocks and the Rhfetic of Europe. Up to the present time we have found this only at Durham, Conn., but it is there quite abundant. The general character of the fronds is fairly represented in the figures now given. They are from one foot to perhaps two feet in length, and from two to three inches wide ; broadest in the middle, where the obliquely set pin- nules are two inches in length, narrow, linear, and pointed. Toward the sumnnt of tlie frond they are shorter and more crowded, while near the base they are still shorter and somewhat irregularly placed. On the upper side of the rachis the bases of the pinnules are elegantly adjusted to one another in alternate order, the line of contact between them being sinuous or FOSSIL PLANTS. 91 zigzag. The bases are auricled, the upper lobe is greatly developed, the attachment being at a single point, and from this the fine and parallel nerves radiate to all parts of the margin after the manner of the genus as shown in Fig. 2a. On a slab of slate now in the possession of Mr. S. W. Loper, at Dur- ham, several of the fronds are shown radiating from a central area, proba- bly the summit of the stem. Of the stem itself, which was doubtless some- what succulent, we have as yet found no traces. Localit}', black shales, associated with Ischypierus, Catoiiterus, etc., Durham, Conn. Otozamites beevifolius F. Br. PI. XXIV, Fig. 3. Among the many fi-onds of Otozamites obtained at Durham there are some of small size, set with short, crowded, rounded, or blunt-pointed pin- nules. These correspond perfectly to some of the specimens of Otozamites hrevifoUns figured by Saporta, Braun, Schimper, and Schenk. One of these is represented on PI. XXIV. We have not yet sufiicient material for comparing these fronds with those which I have referred to Otozamites latior, but it has seemed to me possible that these -two forms, which are here and in Europe so frequently intermingled, may be but varieties of the same species, the smaller fronds belonging to small plants or representing a spe- cial stage of growth. Whether this is true or not, it is a matter of much interest that we here find fronds which correspond so perfectly to those common in the Rluetic of Germany and France. In Schenk's admirable memoir, Die fossile Flora der Grenzschichten, Pis. XXXI, XXXIII, and XXXIV, a series of excellent figures are given whicli perfectly represent our Durham cycads, both the larger and the smaller forms (Otozamites latior and 0. hrevifolius). By Schenk they are considered to be ferns, as they were by Lindley, and he unites the two forms under the name of Otopteris Bucldandi. Prob- ably they should be united, but it is hardly possible tliat they are ferns. LocaHty, Triassic shales, Durham, Conn., collected by S. W. Loper. 92 TEIASSIC FISHES AND PLANTS. Cycadinocarpus Chapini Newb. PI. XXIV, Fig. i. Fruit broad-ovoid, nearly orbicular, 1.5 mm. wide by 18 mm. long, compressed, consisting of an ovoid nucleus bordered by a wing or margin which is emarginate or notched above, narrowed and becoming obsol'ete below; nucleus excavated in a broad sulcus, extending from the base to the center of the fruit, and traversed centrally by a depressed line. This interesting- specimen was obtained at Durham, Conn , by Rev. J. H. Cliapin, to whose courtesy I aa;i indebted for the opportunity of exam- ining it. It is plainly the fruit of a cycad, and perhaps of Otozamites latior, which is quite common at the locality where it was found. It probably consisted originally of a hard, ovoid, compressed nucleus, surrounded b)^ a sarcocarp covered with a leathery rind. When compressed in fossilization the nut is shown in relief, and the envelope forms the margin about it. The fruit of Cycas revohda Avould present much the same appearance if subjected to compression in clay. A large number of cycad fruits are known, but there is none described with Avhich I have been able to identify this. DiooNiTES LONGiFOLius Emmous, sp. PJ. XXV, Fig. 4. One specimen only of this plant has yet been found in the Trias of New Jersey, and that was taken from the quarries at Newark by Prof C. H. Hitchcock, It apparently represents the basal portion of a frond of large size, the rachis being ver}' strong. The pinnules of only one side are shown. These diverge at a large angle, are linear, attached by the entire base, and are decurrent. Their complete length is not shown, but they must have been at least two inches in length. They are separated by in- tervals of about twice their breadth. The nervation is obscure, but appar- ently fine and parallel. In many cycads the basal pinnules of the fronds are shorter, and in some more widely separated than those above. They also diverge from the rachis at a larger angle. It is almost certain that if we were to obtain the upper portion of the frond of which we here have the base we should find the pinnules more approximated and diverging at a more acute angle than is the case with those now before us. Hence we FOSSIL PLANTS. 93 should have specimens closely resembling those figured by Dr. Emmons. Of these, one (his Fig. 83) represents the middle part of the frond; the other (Fig. 82) is from a higher portion. This is inferable from the fact that in one the rachis is stronger and the pinnules are more separated and diverge at a larger angle than in the otlier. Dr. Emmons does not describe the nervation, but represents it as fine and parallel. This would exclude it from the genus Cycadites, in which the pinnules are traversed by a strong midrib. Both Dr. Emmons's species and that now figured belong clearly to Diooniks, as defined by Schimper.^ LOPEEIA SIMPLEX, n. Sp. PI. XXV, Figs. 1, 2, 3. One of the most common plants found in the Trias at Durham, as usually seen, is a straight, smooth, unjointed stem, once cylindrical, but now much compressed and replaced by jet. Of some of these stems por- tions have been obtained an inch or more in width and twelve or fifteen inches in length, but the plant was evidently a large one, and these are mere fragments of it. Recently Mr. S. W. Loper has found specimens which show more of this organism than was before known, and some of these are represented in our plate. Of these, that best preserved consists of a stem such as I have described, but which divides above into a number of branches, all springing from tlie same point. These branches are slender and flexuous, and bear what seem to be alternate, linear, acute, grass-like leaves, but in their state of preservation showing no nervation. This is ap- parently the same plant as that figured by Emmons ^ and copied by Fon- taine.* Professor Fontaine refers to these specimens on pages 119 and 120,* and "for convenience of reference" gives them the name of Bambusium Carolinense. I venture to substitute for that name the one now given, as it is quite certain that, the plants under consideration had no close botanical relationship with Bamhusa (the Bamboo), which is a grass, and, like all the Graminece, has jointed stems. Without more material it will be impossible to determine with any certainty the botanical relations of this plant, but it was most probably monocotyledonous, perhaps aquatic — a kind of gigantic iPal^ontologie v^g^tale, vol. 2, p. 147. s Monograph cited, pi. LII, figs. 1, 2, 2 Am. Geol., pt. 6, 1857, pp. 131, 132, figs. 90, 100. ■• Op. oit. 94 TRIASSIC FISHES AND PLANTS. SchoUera. The plant figured by Emmons^ is perhaps the summit of a stem which divides into five branches, and his figure 102^ represents a smaller specimen with four divisions. This he compares with Baiera, but I am led to doubt its connection with that genus, both from its manner of branching and from the fact that, associated with the larger stems described above, I have one even smaller than that represented by Emmons, in which the stem terminates above in five nearly equal branches Clathropteris platyphylla. Brong. PI. XXII, Fig. 6. At Sunderland, Westfield, and Durham, in the Connecticut Valley, fronds of Clntliropteris have been frequently met with. Much more rarely fragments of the same fern have been obtained from the coarser beds of Newark and Milford, N. J. In 1855 Edward Hitchcock, jr., described^ a portion of a frond of Cla- thropteris found near Easthampton, Mass., about the middle of the Triassic series. To this plant he gave the name of C. rectiiiscuhis; but it has the radiate arrangements of the lobes or pinna; which is characteristic of C. platiiphylla Brong., and its details furnish no characters, judging from his figure and description, by which it can be distinguished from that species. Clathropteris jjlati/phylla is a ver}' widely distributed fern in the Liassic and Rhsetic strata of the Old World, from England to India and China, and it has been collected by Professor Fontaine in the Virginia coal series. Fronds which I can not distingui.sh from those of this species also occur not unfre- quently at Durham, Conn. Tliese are always imperfect, but were evidently of large size and had a digitate or radiate arrangement of the pinnae. The fragment now figured is a portion of the upper part of a pinna, from the sandstone of Newark, N. J. Palissya? sp. PI. XXVI, Figs. 1, 2. I give herewith representations from photographs of two views of a coniferous trunk such as is frequently found in tlie sandstone quarries at > Am. Geol., p. 131, fig. 99. ^Op. cit., p. 133. =Am. .Joiir. Sci., 2(1 series, vol. 20, ia5r>, p. 22. FOSSIL PLANTS. 95 Newark, N.J. The decorticated surface of these stems is marked by rhom- boidal elevations, which somewhat resemble tlie markings on the trunks of Lepidodenclron when denuded of their coaly envelope. This resemblance has led to the announcement that Leindodendron had been found in our Triassic rocks, but this is a manifest error. Lepidodendron did not pass from the Carboniferous to the Mesozoic age, and these are plainly casts of the trunks of coniferous trees. Since this was written my attention has been drawn to a figure of the trunk of VoUmci Cobiirgensis Schaur (Palasontographica, vol. 11, p. 308, PI. XLVJ, Fig. 2). This is so much like the trunk now figured and the smaller ones not uncommonly met with in the quarries at Newark, that if we had anywhere found in the Trias of this region any certain traces of Voltzia I should have little hesitation in referring our specimens to that genus ; but no VoJtzia has yet been found in the New Jersey sandstones, while Palissya is rather common. Hence, I have been led to believe that these trunks and branches bearing lozenge-shaped markings belonged to the latter genus. As it was doubtless closely allied to VolUia it would not be at all surprising if we should find that the decorticated trunks of trees of the two genera were much alike. PLATES MON XIV 7 PLATE I PLATE I. Page. ISCHYPTEKUS OVATUS W. C. R., Booutou, N. J 27 100 o < > -I PLATE II. PLATE II. Page. IscHYPTERUS Mausiiii W. C. R., SuDcleiiaud, llass '^>i 102 U. B. GEOLOGICAL SURVEY MONOGRAPH XIV PLATE I) ISCHVPTERUS MARSHII. PLATE III PLATE III. Page. IscHYPTERUS Agassizii W. C. R., Boonton, N. J 3U 104 U. S, GEOLOGICAL SURVEY MONOGRAPH XIV PLATE III ISCHYPTERUS AGASSIZII. PLATE IV. PLATE IV. Page. IscHYPTERUS MICROPTERUS Newb., Durliam, Couu 31 Fig. 1 . Fish of average size 31 2. Unusually large individual - 31 106 PLATE Y. PLATE V. Page. ISCHYPTERUS TENUICKPS Ag., Suuderlaud, Mass 32 Figs. 1, 2. Old iudividiials, showing extreme development of dorsal scales 32 3. Young 32 108 U. S. GEOLOGICAL SURVEY MONOGRAPH XlV PLATE V ISCHYPTERUS TENUICEPS. PLATE VI. PLATE VI. Page. Fig. 1. IscHYPTERUS EOBUSTUS Newb., Boouton, N.J 36 2. IscHYPTEEUS FULTU.S Ag., Boontou, N. J 34 uo 1 - - V PLATE VII. PLATE VII. Page Fig. 1. IscilYPTERCS fultus Ag., Boonton, N. J. Small form 34 2. ISCHYPTERLS ELEGAXS Newb., BoontOD, N. J 37 3. IscHYPTERus TENUICEPS Ag., Sunderland, Mass. Mature but not old individual 32 112 U. S. GEOLOGICAL SURVEY MONOGRAPH XIV PLATE VII 1. ISCHYPTERUS FULTUS. 2. ISCHYPTERUS ELEGANS. 3. ISCHYPTERUS TENUICEPS. PLATE VIII MON xiy 8 ]>I.A'ri'^. VTTI. I'k;. I. iNCIlYPrKlllIM Al, ATI'S N(^\v)),, llDDiiInn, N..I . 'J. iMI'llVI'r'Um'M Al, ATI'S N - 1 ^^■^•^ ~^— r.sJSS-5*^<^^ ' \ 1. ISCHYPTERUS ELEGANS. 2.3. ISCHYPTERUS LENTICULARIS. PLATE XI PLATE XI. Page. Fig. 1. IscHYPTERUs LiSKATus Newb., Boouton, N. J 40 2. ISCHYPTERUS LINEATUS 40 120 PLATE XII PLATE XII. Page. Fig. 1. IscHYPTERUS MACROPTERDS W. C. R., Boontou, N. J 41 a. IscHYPTERUS MICROPTERUS Newb., Durham, Conn. Young individual 31 3. IscHYPTERUS Braunii Newb., Weehawken, N.J. Young 43 1^2 o I < PLATE XIII. PLATE XIII. Page. Figs. 1,2,2a. Ischypteeus Braunii Newb. , Weehawken, N. J 43 3. IscHYPTERUs LATUS J. H. E., Plainfield, N. J 46 4. IscHYPTERUS PARVUS W. C. E., Turner's Falls, Mass 45 5,5(1. IsCHYPTERUS MINUTUS Newb., Durham, Conn 48 124 U. S. GEOLOGICAL SURVEY MONOGRAPH XIV PLATE XIII 1. 2. ISCHYPTERUS BRAUNll. 3. ISCHYPTERUS LATUS. 4. ISCHYPTERUS PARVUS. 5. ISCHYPTERUS MINUTUS. PLATE XIV. PLATE XIV. Page. Fig. 1. ISCHYPTERUS ELEGANS Ncwb., Booutoii, N. J. iLnIividiial of iiieilium size, sliowiog broad dorsal fin 3" 2. IscHYPTERUS ELEGANS Newb., BoontoD, N. J. Largest indi vidual Seen 37 3. IscHY'PTERU.s GIGAS Ncwb., BoontoD, N. J. Tail of a fish nearly two feet iu length.. 49 126 U. S. GEOLOGICAL SURVEY MONOGRAPH XIV PLATE XIV 1, 2. ISCHYPTERUS ELEGANS. 3. ISCHYPTERUS GIGAS. PLATE XV. PLATE XV. Page. Catopterus Eedfieldi Egerton, Durham, Conu ■. 53 Fig. 1. Broad form. Old individual ". 53 2. Normal form of young individual - 53 3. Part of mature fish, showing details of fins 53 128 U. S. GEOLOGICAL SURVEY MONOGRAPH XIV PLATE XV CATOPTERUS REDFIELDI. PLATE XVI MON XIV 9 PLATE XVI. Page. Fig. 1. Catopterus gracilis J. H. R., Boonfoii, N. J. Mature individual, broad form 55 2, :3. Catopterus gracilis J. H. E., BooDtou, N. J. Posterior portions of bodies of two 55 indi vidals of slender form 55 4, 5. Catopterus parvulus W. C. R., Boouton, N. J 60 130 PLATE XYII. PLATE XVII. Figs. 1, 2, 3, 4, Catopterus minok Nowb., Durham, Coqu 57 132 U. S. GEOLOGICAL SURVEY MONOGRAPH XIV PLATE XVIl CATOPTERUS MINOR. PLATE XYIII. PLATE XVIII. Page. Figs. 1, 2. Dictyoptge macrura Egerton, Clover Hill, Va G4 3. Catopterus ornatus Newb., Durham, Conn 58 3a. Scales of dorsal line enlarged 58 36. Scales of side enlarged 58 4. Catopterus sp., young, Durham, Conn 50 5. Catopterus anguilliformis W. C. R., Durham, Conn 59 134 U. S. GEOLOGICAL SURVEY MONOGRAPH XIV PLATE XVIII ^S^%:^^-^-^^&J 3a 3b aSE - 1,2. DICTYOPVGE MACRURA. 3. CATOPTERUS ORNATUS. 4. CATOPTERUS. 5. CATOPTERUS ANGUILLIFORMIS. PLATE XIX. PLATE XIX. Page. Figs. 1, '2. Ptycholepis Marshii Newb., DuiUam, Conn 6(j 2a. Scales, enlarged C6 3, 4. AcENTROPHORUS CHicoPENSis Newb., Chicopee Falls, Mass 69 136 U. S. GEOLOGICAL SURVEY MONOGRAPH XIV PLATE XIX 1,2. PTYCHOLEPIS MARSHII. 3,4. ACENTROPHORUS CHICOPENSIS. PLATE XX. PLATE XX. Page. DiPLURUs LONGiCAUDATUs Newb., Boontou, N. J 74 Fig. 1. Fish, complete - 74 2. Second caudal fin, Boonton, N. J ' 74 3,4. Scales outer surface, showing ornamentation, natural size, Durham, Conn "4 5. A group of scales showing inner surface, Durham, Conu 74 138 , GEOLOGICAL SJRVEV DIPLURUS LONGICAUDATUS. MONOQRAPM «iv Plate x OIPLURUS LONGICAUDATUS. I I PLATE XXI. PLATE XXI. Page. DEXDROPnYCCS TRiAS.sicus Ncwb., PortlancT, Conn 62 Fjg. 1. Basal portion of a frond i^2 2. Extremities of a group of branches 82 140 U. S. GEOLOGICAL SURVEV MONOGRAPH XIV PLATE XXI DENDROPHYCUS TRIASSICUS- PLATE XXII PLATE XXII. Page. Fig. 1. Baiera MUNSTERlANA (?) Ung., Darham, Conn 84 2. Pachypiiyllum simile Newb., Sunderland, Mass 88 3. Pachyphyllum BREViFOLiUM Newb 89 3(7. Twig -n-itli divergent leaves 89 Sl>. Twig with appressed leaves 89 3e. Cone 89 4. C'HEir.OLEPis MuxSTERl Schimper, sp., Milford, N. J 90 4a. Cone scale , 90 5. Phragma of Eqoisetum Rogersi SoLimper, Milford.N.J 85 5a. Phragma of Equisetuji Eogersi, underside 85 6. Clathropteris platyphylla (?) Bi-ong., Milford, N.J .._ 94 142 U. S. GEOLOGICAL SURVEY MONOGRAPH XIV PLATE XXII 1. BAIERA MUNSTERIANA. 2. PACHYPHYLLUM SIMILE. 3. PACHYPHYLLUM BREVIFOLIUM. 4. OHEIROLEPIS MUNSTERI. 5. PHRAGMA OF EQUISETUM ROGERSI. 6. CLATHROPTERIS PLATYPHYLLA. PLATE XXIII. PLATE XXIII. Baiera multifida Fontaine, Clover Hill, Va 144 Page. 84 U. S. GEOLOGICAL SURVEY MONOGRAPH XIV PLATE XXlll BAIERA MULTIFIDA. PLATE XXIY. HON XIV 10 PIRATE XXIV, Page. Fig. 1. Otozajiites latior Saporta, Durbam, Conn. Base of frond, natural size 90 2. Otozamitiks LATioi! Saporta, Durham, Conn. Summit of frond DO 2a. Piirtiou enlarged to show nervation 90 3. Otozamites BREViFOLius Fr Br., Dnrliam, Cnnn 91 4. Cycadixocarpus Chapini Newb., Durham, Conn - 92 146 U. S. GEOLOGICAL SURVEY MONOGRAPH xrv PLATE XXIV 1, 2. OTOZAMITES LATIOR. 3. OTOZAMITES BREVIFOLIUS. 4. CYCADINOCARPUS CHAPINl. i PLATE XXY. PLATE XXV. Page. Fig. 1. LoPERii. simplex Newli., Durham, Coun. Main stem ranch flattened 9:^ 2. Branch anil leaves 93 3. Branches springing from stem, Dnrliam, Conn 93 4. DiooNiTES i,oNGii'Oi,ius Emnions, sp., Newark, N. J . 92 148 MONOGRAPH XIV PLATE XXV U. S. GEOLOGICAL SURVEY 1, 2, 3. LOPERIA SIMPLEX. 4. DIOONITES LONGiFOLIUS. PLATE XXVI. PLATE XXVI. Figs. 1,2. Trunk of couifer (Palissija'!) decorticated, showing rhoraboidal markings on sur- face, Newark, N. J 94 150 INDEX. Page. Acentroplioras, generic description of 67-G9 aflfinities of, with Iscbypterua C7 alUancea of, with Palosoniscus G8, 69 chlcopenisis, description of C9 Agassiz, L,, cited 20,41 Baiera 12 Miinsteiiana, description of 84 multifida, resembles Miinsteiiana 84 Eird or reptile tracks ■. 5 Bunbury, C.J. F., cited si, 9-11 Catopterus, generic description of 50, 51 most beautiful of fossil fishes 51 not found iu Mesozoic rocks of Old "World 51 species of i 19-21 gracilis, allusions to 1 1, 19, 22 description of 55-57 Redfieldi, structure of head of 53 description of 53 form of tlO naacrurus 20, 21 minor, description of 57, 58 akin to C. ornatus 58 ornatus, description of 58 of form and size of C. minor 59 anguilliforruis, description of 59 parvulus, description of 59 Cbapin, J. H., acknowledgments to 92 Cheirolepia Miinsteri, account of 90 Clathropteris platy phylla 12, 94 Coal measures of Pennsylvania 8 Colorado, fossils from 22 Connecticut, fossils from xi, xiii, xiv, 4^8, II, 12, 19-95 Cycadinocarpus Chapini 92 Davis, W. M., cited 6, 7 Dendropbycus triassicus, description of 83 iu Portland sandstones 83 Dictyopyge, generic description of 61-64 previously considered Catopterus macrurus 62 differences of the genera 63 comparison of fisbos of American Trias 63 macrura, description of 11, 64 Diplurus, generic description of 70 comparison of, with living siluroids 7J resemblance of, to Ccelacanthus,Holopliagus, and Macropoma 71, 72 lougicaudatus, description of 74r-7G Diuonit.es longifolius, description of 92,93 Egerton, P., cited xi, 11, 20, 25, 28, 34, 53, 56, 63, 65. 72 Kmerson, B. K., cited xiii, xiv Emuious, E.. cited xi, 9, 10, 12, 32, 88, b9 Equisetum Rogersi, description of 12,85 Meriani (?) 86 resemblance of, to Calamites gigas 86 Eurynotus tenuiceps 19,24 Page. Fontaine. W.M., cited xi, 10, 12, 13, 87-89, 94 Fossil fishes, Triassic, list of, from Trias of New Jer- sey and Connecticut 23 Fossil plants, lists of, fiom Trias of New Jersey and Connecticut. 79 Gratacap, L. P. cited 4i Geologic equivalents of Triassic rocks S Heer,0., cited 10 Hills, K.C., cited 22 Hitchcock, C. H,, acknowledgments to 92 Hitchcock, E., cited 9,12,32,45 Hitchcock, E., jr., cited 94 Howell, E. E., cited 'd2 Huxley, T. H., cited 71,74 Iscliypterus, generic description of 24 allied to Semionotus 25 allied to Lcpidotus 25 characteristics of genus 25 species of 20,21,21,25 Agassizii, description of 30 latus, description of ^6, 40 ovatus, descripiion of 11,27 Marshii, description of 28 resembles I. Ag 29, 31 micropterns, description of 31 tenuiceps, description of 32 resembles I. ovatus 33 fultus, description of 32 allied to P. macropterus 35 robustus, description of 36 allied to I. ovatus 36 allied to Semionotus ^ - 36 elegans, description of . 37 resemblance in form to I. lineatus - - 37 alatus, description of 37 modestus, description of 38 allied to I. elegaus 38 lenticiilaris, description of 39 similar in form to I. ovatus 39 diflferentiated from I. microp terns 39 lineatus, description of 40 differences from I. tenuiceps, I lenticularis, 1. elegans, and I. fultus 40 closely allied to I. alatus 40 macropterus, description of 41 resemblance to P. macropterus, P. fullus, and I. fultus 41,42 Braunii, description of 43 intermediate between T. tenuiceps and Acen- tropborus cbicopensis 44 reasons for classifying, as a species distinct from P. latus and I. latus 44 parvus, description of 45 reasons for classification 45 151 152 INDEX. Page. lechjpteras, minatus, deacriptioD of 48 gigas, description of 49 Kirkby, J.'W".,citea 68, G9 Lesquereux, Leo, cited 82 Loperia simplex 93 Loper, S. "W., acknowledgmenta to xiv, 51, 5i,56, 65, 67, 74, 80, 81, 91, 93 Lyell.C, cited 9,10,20,28,53 MacrotODiopteris magn ifolia 12 Marcou, J.,cited 9-11 Marsh, 0. C, cited xiii.xiv, 26 Marsh, 0. C.actnowledgments to 67 Massachusetts, bird tracks in xi, 12 fossils in -...- 21,27,29,30,34,45,47.67.70,79 New Jersey, fossils from xiii, sir, 5-8, 11, 12, 19-95 New Mexico, fossils from 14 New York, fossils from 4 North Carolina, fossils from xi, 4, 9-12, 22 Ohio, fossils from xii Otozamites latior 13 geological importance and general character of 90 bre vifolius 91 Macombii 14 Pachypbyllum simile, description of 88 resemblance of, to P. peregrinum 88 brevifolium, description of 89 Palseoniscus, species of 19, 20, 24 faltus 19,24 latns 19 ovatas, description of 27, 28 resemblance to Ischypterus tenuiccps 28 Palissya (?) 94,95 Braunii 13 Pecopteris bullatus 14 Page. Pecopteris falcatus 14 Pennsylvania, coal measures of 8 fossils from 82 Plants, fossil, list of, from Trias of New Jersey and Connecticut 79 Porter, T. C. , acknowledgments to 85 Ptycbolepis, generic description of 65 Marshii, description of 66, 67 Bollensis, how differentiated 66 curtus, characteristics of 66, 67 Redfield, J. H., cited si-siT, 19, 20, 27^ 29, 30, 34, 46. 47, 55, 5G, 64, 70 Redfield.^V. C, cited, .si, 19, 22, 26, 30, 34-36, 41, 42, 45. 59, 6U, 62 Eedfield, TV". C. , ackno wledgments to 60 Kemond, A., acknowledgments to 84 Kogers, H. D.. cited 6,9,82 Rogers, W.B., cited sl„9, 87,88 Russell, I.e., cited 6 Russell, I. C, acknowledgments to xiv Saporta, G., cited 12, 88 Schenk, A., cited 91 Schizoneura planicostata 12, 87 Semionotus, resemblance to Ischypterus 25,26 Stiiiver, J., cited 63 Stuart, R.L., cited 29 T0enioi)teri3 magnifolia, 12, 14 Taylor, R.C 8 Thecodont saurians - 9 Traquair, R. H., cited 24,68,70 Triassic fishes, list of 23 Virjiiuia, fossils from si, 4, 10-12, 20, 22, 62, 63, 84 Whit fit^ld. R. P. . cited siii, xiv Winkler, T.C., cited. 72,73 "Woodworth, J. B., acknowledgments to 86