State of Connecticut State Geological and Natural History Survey Bulletin No. 18 TRIASSIC FISHES OF CONNECTICUT By CHARLES ROCHESTER EASTMAN, Ph.D. Professor of Paleontology at the University of Pittsburgh, and Curator in charge of Fossil Fishes at the Carnegie Museum \ S(c7 Fag 14273 BOOK 567.EA8 c. 1 EASTMAN f TRIASSIC FISHES OF CONNECTICUT 3 1153 001372fll M BULLETINS OF THE State Geological and Natural History Survey of Connecticut. i. First Biennial Report of the Commissioners of the State Geological and Natural History Survey, 1903- 1904. 2. A Preliminary Report on the Protozoa of the Fresh Waters of Connecticut: by Herbert William Conn. 3. A preliminary Report on the Hymeniales of Connecticut : by Edward Albert White. 4. The Clays and Clay Industries of .Connecticut : by Gerald Francis Loughlin. 5. The Ustilagineee, or Smuts, of Connecticut: by George Perkins Clinton. 6. Manual of the Geology of Connecticut :. by William North Rice and Herbert Ernest Gregory. 7. Preliminary Geological Map of Connecticut : by Herbert Ernest Gregory and Henry Hollister Robinson. 8. Bibliography of Connecticut Geology: by Herbert Ernest Gregory. 9. Second Biennial Report of the Commissioners of the State Geological and Natural History Survey, 1905-1906. 10. A preliminary Report on the Alga of the Fresh Waters of Connecticut : by Herbert William Conn and Lucia Washburn (Hazen) Webster. n. The Bryophytes of Connecticut: by Alexander William Evans and George Elwood Nichols. 12. Third Biennial Report of the Commissioners of the State Geological and Natural History Survey, 1907-1908. 13. The Lithology of Connecticut: by Joseph Barrell and Gerald Francis Loughlin. 14. Catalogue of the Flowering Plants and Ferns of Connecticut growing without cultivation : by a Committee of the Connecticut Botanical Society. [Out of print.] 15. Second Resort on the Hymeniales of Connecticut: bv Edward Albert White. i6. Guide to the Insects of Connecticut: prepared under the direction of Wilton Everett Britton. Part I. General Introduc- tion: by Wilton Everett Britton. Part II. The Euplexoptera and Orthoptera of Connecticut: by Benjamin Hovey Walden. 17. Fourth Biennial Report of the Commissioners of the State Geological and Natural History Survey, 1909-1910. 18. Triassic Fishes of Connecticut: by Charles Rochester Eastman. Bulletins I, 9, 12, and 17 are merely administrative reports, containing no scientific matter. The other bulletins may be classi- fied as follows : Geology : Bulletins 4, 6, 7, 8, 13, 18. Botany: Bulletins 3, 5, 10, 11, 14, 15. Zoology: Bulletins 2, 16. These bulletins are sold and otherwise distributed by the State Librarian. Postage, when bulletins are sent by mail, is as follows : No. I, $0.01 ; No. 2, .07 ; No. 3, .08 ; No. 4, .06 ; No. 5, .03 ; No. 6, .12 ; No. 7, .06; No. 8, .05 ; No. 9, .02 ; No. 10, .08 ; No. 11, 07; No. 12, ,C2; No. 13, .c8; No. 14! ; No. 15, .06; No. 16, .07; No. 17, .02; No. 18, .07. The prices when the bulletins are sold are as follows (including postage) : No. 1, $0.05 ; No. 2, .35 ; No. 3, .40 ; No. 4, .30; No. 5, .15; No. 6, .50: No. 7, .60*; No. 8, .20; No. 9, .05; No. 10, .35; No. 11, .30; No. 12, .05; No. 13, .40; No. 14!; No. 15, .35 ; No. 16, .35 ; No. 17, .05 ; No. 18, .25. Bulletins 1-5 are bound as Volume I. The price of this volume is $1.50. Bulletins 6-12 are bound as Volume II. The price of this volume is $2.45. Bulletins 13-15 are bound as Vol- ume III. The price of this volume is $2.50. Other volumes will follow. It is intended to follow a liberal policy in gratuitously dis- tributing these publications to public libraries, colleges, and scientific institutions, and to scientific men, teachers, and others who require particular bulletins for their work, especially to those who are citizens of Connecticut. Applications or inquiries should be addressed to George S. Godard. State Librarian. Hartford, Conn. * If map is mounted as a wall map, and sent by express, $1.60. t Bulletin No. 14 is out of print and can be furnished only in the bound volume. CATALOGUE SLIPS. Connecticut. State geological and natural history survey. Bulletin No. 18. Triassic fishes of Connecticut. By C. R. Eastman. Hartford, 1911. 77 pp., 11 pis., 8 figs. 23cm. Eastman, Charles Rochester. Triassic fishes of Connecticut. By Charles Roches- ter Eastman, Hartford, 1911. 77 pp., 11 pis., 8 figs. 23cm. (Bulletin no. 18, Connecticut geological and natural history survey.) CATALOGUE SLIPS. Geology, Eastman, C. R. Triassic fishes of Connecticut. Hartford, 1911. 77 pp., 11 pis., 8 figs. 23ca". (Bulletin no. 18, Connecticut geological and natural history survey.) Paleontology. Eastman, C. R. Triassic fishes of Connecticut. Hartford, 1911. 77 pp., 11 pis., 8 figs. 23cm. (Bulletin no. 18, Connecticut geological and natural history survey.) State of Connecticut PUBLIC DOCUMENT NO. 47 State Geological and Natural History Survey COMMISSIONERS Simeon Eben Baldwin, Governor of Connecticut (Chairman) Arthur Twining Hadley, President of Yale University William Arnold Shanklin, President of Wesleyan University Flavel Sweeten Luther, President of Trinity College (Secretary) Charles Lewis Beach, President of Connecticut Agricultural College SUPERINTENDENT William North Rice Bulletin No. 18 Hartford Printed for the State Geological and Natural History Survey IQII Publication Approved by The Board of Control. The Case, Lockwood & Brainard Co., Hartford, Conn. TRIASSIC FISHES OF CONNECTICUT By CHARLES ROCHESTER EASTMAN, Ph.lD. Professor of Paleontology at the University of Pittsburgh, and Curator in charge of Fossil Fishes at the Carnegie Museum 1 Die Weisheit ist nur in der Wahrheit. "— Goethe. HARTFORD Printed for the State Geological and Natural Histoby Survey 1911 h* CONTENTS. Page I. On the Study of Fossil Fishes in General ... 9 II. Geologic Correlation of the Connecticut Valley Fish- bearing Beds ...... 23 III. Geography of the Trias ..... 36 IV. Concerning Earlier Investigation of North American Tri- assic Fishes ...... 39 V. Systematic Descriptions of Upper Triassic Fishes . 4Z ILLUSTRATIONS. Plates. I. Semionotus agassizii (W. C. Redfield). Trias; Sunderland, Massachusetts. Type of Newberry's so-called Ischypter. us mars hi. Xtj II. Semionotus agassizii (W. C. Redfield). Trias; Sunderland, Massachusetts. Head and anterior portion of trunk- Xi III. Semionotus fultus (Agassiz). Trias; Boonton, New Jersey. Figured specimen. y.\ IV. Semionotus micropterus (Newberry). Trias; Durham, Con- necticut. X$ V. Semionotus elegans (Newberry). Trias; Boonton, New Jer- sey. Figured specimen. X^ VI. Semionotus nilssoni Agassiz. Rhaetic ; Hoegenaes, Sweden. Head portion of holotype, showing cranial plates and dentition. X{ VII. Ptycholepis mars hi Newberry. Trias ; Durham, Connecticut. VIII. Ptycholepis mars hi Newberry. Trias ; Durham, Connecticut. Showing dorsal aspect of vertically compressed cranium. IX. Catopterus gracilis J, H. Redfield. Trias ; near Middletown, Connecticut. Cotype. X| X. Catopterus gracilis J. H. Redfield. Trias ( Posterior shale ) ; Durham, Connecticut. Scales appearing whitish by reason of mineral replacement. Xf XI. Catopterus gracilis J. H. Redfield. Trias ; Durham, Connect- icut, xi Page. 8 connecticut geol. and nat. hist. survey. Figures in Text. i. Cephalaspis murchisoni Egerton. Lower Old Red Sand- stone ; Herefordshire. ( After A. S. Woodward) . 14 2. Outlines of Acanthodian Fishes. A, Climatius scutiger Egerton. B, Mesacanthus mitchelli (Egerton). C, Acanthodes sulcatus Agassiz. D, Acanthodes gracilis Roemer . . . ... . .15 3. Cladoselache fyleri Newberry. Cleveland shales (Upper Devonian) ; near Cleveland, Ohio. (After Bashford Dean) ....... 16 4. Dipterus valenciennesi Sedgwick and Murchison. Lower Old Red Sandstone ; Scotland. (After R. H. Traquair) 17 5. Catopterus redfieldi Egerton. Trias ; Durham, Connecticut. Reproduction of Newberry's original drawing of dorso- lateral aspect of the head ..... 54 6. Catopterus redfieldi Egerton. Trias ; Durham, Connecticut. Reproduction of Newberry's original drawing show- ing under side of the head . . . .54 7. Comparative diagrams showing types of cranial structure. 1, Rhabdokpis. 2, Nematopty chius. 3, Rhadinichthys. 4, Palazoniscus. 5, Semionotus . . . -59 8. Semionotus agassizii (W. C. Redfield). Trias ; Sunderland, Massachusetts. X£ . . • • .62 TRIASSIC FISHES OF CONNECTICUT. i. ON THE STUDY OF FOSSIL FISHES IN GENERAL. " There will we find laws which shall interpret, Through the simpler past, existing life." — Kingsley. I PALEONTOLOGY is the natural history of the Past. It is that branch of biological science which acquaints us with the endless succession of animate forms that has inhabited the earth since life first began. Primarily an extension of zoology and botany, as these subjects are commonly understood, it may be regarded also as an historical science, by virtue of the time element pervading it. Its aims and methods are akin to the historian's. The facts it deals with are vital facts, linked together by the principle of continuity and progressive development. The story it unfolds is one of world-wide changes, of silent, slow, and exceedingly gradual transformations wrought upon organic framework by an infinity of complex forces, strivings, tendencies, surroundings, all operating through immense cycles of time, and culminating finally in one supreme achievement — in the produc- tion of a race of beings possessed of self-conscious intelligence, and of a well-nigh unlimited development of that faculty. The historical aspect of paleontology is worth considering. What is the theme of human history, if it be not the development of mankind? Is it not a record of all the changes in the state of men which have occurred since the first evidences of " the sons of men " upon our globe ? Does it not, in a word, inform us of the progression of human events ? The theme of paleontology is similar, but broader. It is compassed by, and at the same time extends, the domain of universal history. It treats of the de- velopment of life in general, considers it in its grandes lignes, IO CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. in all its manifold manifestations. It is the record of all the changes and progressive modifications that have taken place among organic forms since their first appearance in remote geological antiquity. It seeks to ascertain the value of various adjustments to external conditions, of improvements in mechani- cal contrivances and other factors making for an advance; and, in tracing this line of progress, it aims to assign to different groups, or to different grades of the same group, their proper position and relative importance in the scheme of upward trans- formation. If the mystery of the beginning of all things must forever remain insoluble for us, as Darwin with his life-work behind him was obliged to admit,1 paleontology at least dispels for us some of the obscurity of former geological cycles, during which life existed on our planet and left memorials of its infinitely slow progress along the road to perfection. Perfection, that is, in the Darwinian sense: meaning the production of the higher animals, and their capacity for psychic advance. An impressive spectacle this ; no one can contemplate it seriously without feel- ing the sense of that infinity in contrast with which a man recog- nizes his own finitude. Then it is that one feels in accord with Keats' view: "Stop and consider! life is but a day; A fragile dew-drop on its perilous way From a tree's summit; a poor Indian's sleep While his boat hastens to the monstrous steep Of Montmorenci. ..." It is not only a just, but a truly ennobling conception to re- gard paleontology as an extension of human history. "La science des sciences, c'est done la science de fhomme/' Montaigne aptly remarks. Would we comprehend our own nature, and seek to know what man in his essence really is, what he has been, whence he came, whither bound, what destiny he may achieve, and, finally, what value attaches to his mortality — to acquire this self-knowledge we must study that larger nature of which man forms so insignificant a part. The soul grows in knowl- edge of itself as it realizes the contrast with the grandeur, the sheer massiveness of nature, and the eternity of the hidden forces iLife and Letters of Charles Darwin, edited by Francis Darwin. NO. l8.] TRIASSIC FISHES OF CONNECTICUT. II which are now and ever have been at work re-shaping the world since the very foundation of being. Knowledge of this sort vastly enlarges our consciousness, gives to our mortality a setting and perspective, dilates the mind and elevates the spirit by forc- ing them to range widely over the realm of universal history. It also enables us to form a relative estimate of ourselves and our career by applying a larger scale of life — the scale of in- finity. Clearly, therefore, the humanistic interest of paleontology is very great; and considerations of this nature help determine the value of any science quite apart from questions of practical utility. In every science there may be a twofold yield ; one that is expressed in material values, and another that is interpreted in terms of the spirit. Returns of both kinds are worth striving for. But, it may be objected by some, the facts of paleontology can at best only remotely affect our traditional outlook upon life. For those who are satisfied merely with the assembling of facts, and look no further than a connection between them, without being able to comprehend the life of thought in general, this objection may hold. But the thoroughgoing inquirer insists not merely upon an accumulation of dead knowledge; his mind aims at an interpretation of the results of investigation, and attaches to these things meanings and values. So far as relates to human or universal history, the supreme value lies in under- standing what has happened, in perceiving the meaning of events, in grasping the principles and laws that govern organic and social evolution. For this purpose the past must needs be re- constructed by means of the trained imagination out of all avail- able data. The more vigilant the imagination, the better his- torian, and the better scientific investigator, other things being equal ; for to the well-trained explorer in any science this faculty is never a hindrance, but a positive advantage. Obviously, if one lacks the power of transporting himself into the past, one can understand nothing of the past. But once that mental journey accomplished, and so soon as we acquire the habit of looking at experience objectively, without immediate relation to our own time and place, then, in the words of an English historian (Bury), "the modern age falls into line with its pre- decessors and loses its obtrusive prominence, and we come to 12 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. see our petty periods sub specie peremiitatis." World-facts con- templated in this way help us, so says yet another historical writer (Villari), to " gain a new consciousness of our own being, and to win deeper insight into recesses of our own nature." The essayist just quoted gives it as his opinion that the history of the whole universe is required to explain the individual man, " because," as he says, " more or less transformed, all history lives in us human beings. Therefore, as it lives in us all, why should we marvel at our power of transporting ourselves back into past times and living once more in them? In studying the history of Greece, we not only read the tale of a vanished past, but also that of a society and of a civilization that, although transformed, still endures within us as a constituent element of our mentality. Thus we are reading the history of a part of ourselves, and gain a clearer appreciation of that part on seeing it developed, magnified, and surrounded with its pristine glory, as it first flashed upon the world through the deeds of the Grecian people. . . . Thus in reading universal history we learn to recognize the process by which our own intelligence has been gradually built up. It has been justly remarked that, even as the geologist can trace the history of the transformations of the globe from any chance handful of earth, so too the philol- ogist, on analyzing some phrase you have uttered, will find in it the record of the transformations of tongues."1 It is an obvious truism that to every man the world as he sees it depends on his physical organization and upon the way he has been taught to look at it through education and years of experience. But the scientific conception of the world and of the value and meaning of life has become profoundly modified within comparatively recent times through the influence of re- organizing ideas. Men in all ages have shown the keenest interest in the problems of man's origin and past development. The first great step in advance was made by the shores of the TEgean more than two thousand years ago. But the positive results of ancient philosophy were inadequate and limited, as compared with modern, because of its more limited resources. After the time of the Greek poets and philosophers, more than two milleniums were to pass away before those new reorganizing 1 Villari, Pasquale, Studies Historical and Critical, 1907, p. 37 ff. No. 18.] TRIASSIC FISHES OF CONNECTICUT. 13 ideas — those which include the principle of continuity and the higher principle of development — became effectual in the world of thought, and enlightened mankind through the medium of Darwin's utterances. Both in history and in natural science the spirit of investigation was completely transformed by these ideas. So much by way of parenthesis. Enough has now been said to show that the natural history of the past possesses rich interest from a culturo-historical or humanistic standpoint. And from these matters we pass on to a consideration of our special theme, the study of fossil fishes in general. Before speaking, however, of the introduction and succession of the class of fishes in point of chronological sequence, it may be useful to insert at this point a table showing the principal divisions of the stratigraphic column, which will serve the double purpose of indicating the position of our Triassic rocks in the system, and marking the advent of successively higher classes of vertebrates. Forma- tional units, it is to be observed, are divided into systems, series, and groups; and the corresponding time-relations are expressed by the terms eras (or ages), periods, and epochs. Following is the commonly accepted arrangement: GEOLOGICAL TIME-SCALE. Eras Periods Life Cenozoic Mesozoic Paleozoic Archaean Quaternary Tertiary Cretaceous Jurassic Triassic Carboniferous Devonian Silurian Ordovician Cambrian Man Mammals the dominant class Reptiles dominant throughout era Birds appear Earliest mammals Amphibians the dominant class Fishes dominant Invertebrates still dominant Fishes appear Leading groups of invertebrates Scanty and indistinct organic re- mains 14 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. The Cambrian system, at the base of the Paleozoic, has thus far failed to yield any indication of the presence of backboned creatures. Neither Chordates nor " Protochordates," that is, primitive forerunners of the vertebrate phylum, make their ap- pearance in the geological time-scale 'until the Ordovician, after which they continue sparsely throughout the Silurian. From what lower group of organisms the primitive progenitors of the vertebrate stem were descended, and during what period the hiatus between diverse phyla was bridged over, we have no means of knowing. The absence of transitional forms, or indeed even of Protochordates, in strata anterior to the Ordovician is not a very significant fact, when it is considered that the primitive forerunners of chordate animals were probably soft- bodied, and therefore incapable of preservation in the rocks. Although numerous indications of fish-like vertebrates have been obtained at different localities both in this country and abroad, as, for instance, from the Ordovician of Colorado, Montana, and Scotland, it is not until the Silurian that their remains are found satisfactory enough for discussion. The dominant forms of fish life that we are acquainted with from rocks of this age belong to the lowly group of Ostracophores — creatures which differ from Fishes proper to such an extent that they are usually included in a separate class (Agnatha). They have incompletely formed jaws, are destitute of paired fins, and are without calcified en- doskeletal parts (Fig. i). On the other hand, as their name implies, they are protected by a shell-like external covering, whose elaboration can be traced through a number of successive stages. Toward the close of the Devonian they become entirely extinct, without leaving descendants. Fig. i. Cephalaspis murchisoni Egerton. Lower Old Red Sand- stone ; Herefordshire. Head-shield seen from above, tail twisted to show dorsal fin and heterocercal tail mainly in side view. X i^ (after Smith Woodward). No. 18.] TRIASSIC FISHES OF CONNECTICUT. 15 Although forming the dominant feature of Silurian verte- brate life, Ostracophores are nevertheless accompanied in the Z.SJl Fig 2. Outlines of Acanthodian Fishes, illustrating their gradual elongation in shape and loss of intermediate spines during successive periods. A, Climatius scutiger Egerton. Lower Old Red Sandstone; Scotland. B, Mesacanthus mitchelii (Egerton). Ibid. C, Acanthodes sulcatus Agassiz. Lower Carboniferous ; Edin- burgh. D, Acanthodes gracilis. Roemer. Lower Permian; Bohemia. a, anal fin; d, dorsal fin; i. sp., intermediate spines; p, pectoral fins. (From Smith Woodward, partly after Traquair and Fritsch.) i6 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. latter part of that system by creatures which surpass them in grade, and are perfectly recognizable as true fishes, possessing as they do ordinary jaws and two pairs of lateral fins. These oldest remains of typical fishes — they are called Acanthodians after the name of the first described genus — are probably to be regarded as Elasmobranchs, and evidently have not diverged very far from the primordial stock which gave rise not only to the line of sharks and rays, but also to different grades of higher fishes. Acanthodians (Fig. 2) are a long-lived race, continuing throughout the Paleozoic. An allied primitive tribe that was less successful, and by reason of its long-bodied form is regarded by some writers (Woodward) as senile, is that typified by Cladoselache (Fig. 3), which is known from the late Devonian. Fig. 3. Cladoselache fyleri Newberry. Cleveland shale (Upper Devonian) ; near Cleveland, Ohio. Right lateral aspect, about one-tenth natural size. A primitive shark, illustrating the simplest kind of paddle-fins, which are supported by nearly par- allel bars of cartilage (after Bashford Dean). During the Devonian appeared two large groups of fishes with paddle-shaped fins. These groups are commonly known as Lung-fishes and " fringe-finned " Ganoids — Dipnoi and Cross- opterygii they are technically called. Their geological history is peculiar. Both groups early acquired dominance, spread over all regions of the globe, and seem indeed to have culminated in the Devonian, being numerically and specifically more abundant during that period than at any subsequent epoch. Only two modern survivors of Crossopterygii are known from African rivers (Polypterus and Calamoichthys). Of the long and archaic line of Lung-fishes represented by Dipterus (Fig. 4) *and its associates in the Devonian, only the most generalized Ceratodont type, represented nowadays by but three fresh-water genera, has been able to persist until our own time. That the Ceratodont type has had a continuous existence since the early Paleozoic follows as a logical necessity from regarding the Dipterine group No. 18.] TRIASSIC FISHES OF CONNECTICUT. 17 l8 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. as a specialized derivative. Recently described remains of Pro- topterus from the African Tertiary present an astonishing like- ness to Sagenodus, as regards the dentition. Yet another group of fishes, representing a still higher grade than any of the foregoing, makes its first appearance during the Devonian, but does not begin to acquire dominance until the Carboniferous. This is the great group of Actinopterygians or "ray-finned" fishes, to which by far the greater number of modern forms belong. From the Devonian onward until the close of the Permian, this higher grade of fishes was represented by a single family of primitive Chondrostei, whose degenerate descendants finally passed over into modern Sturgeons. It is probable that the short-lived family of the Catopteridas, which gained a wide distribution in the Trias, is an offshoot of the tribe of primitive Sturgeons ; and it is noteworthy that the de- cline of the latter began simultaneously with the rise of the next higher suborder, or Protospondyli. No links are known connecting this suborder with the Chondrostei, hence in the present state of our knowledge, the Sturgeon tribe and the Pro- tospondyli are distinctly demarcated. During the Trias the Protospondyli are represented by the important and truly cos- mopolitan family of Semionotidse, which, with the previously mentioned Catopteridse, form the chief constituents of our local Triassic fish fauna. The only modern representatives of this suborder are the bow -fin and garpike {Amia and Lepidosteus), both confined to the fresh waters of North America, and ex- hibiting the long-bodied shape of senile or decrepit derelicts. Associated with members of the preceding suborder (Pro- tospondyli) in rocks of the Upper Trias are found a few fishes having a remarkably modern aspect, and characterized not only by a complete vertebral column, -but also by a simplified lower jaw, which consists of but two pieces on each side. The fore- runners of the Isospondyli, as this suborder is called in allusion to the circumstance that the vertebrae are simple, without being fused into a group behind the head, scarcely differ in grade from the modern herring tribe. Among typical representatives may be mentioned the genera Pholidophorus and Leptolepis, ranging throughout the Triassic and Jurassic. The group displays rather feeble vigor until the beginning of the Cretaceous, when it NO. l8.] TRIASSIC FISHES OF CONNECTICUT. IO, rapidly multiplied, became dominant, and replaced the Pro- tospondyli. Living members of the suborder belong to that division of bony fishes known as physostomous Teleostei. One of the late Professor Beecher's generalizations, which seems to hold true throughout the animal kingdom, is that spines characterize only the latest representatives of the class. Applying this to the class of Pisces, Dr. Smith Woodward re- marks: " The Acanthopterygii ("spine-finned") are thus the highest and latest fishes of all, though they sometimes eventually descend from their high estate by degeneration. They exhibit all the peculiar changes in the skull, upper jaw, and pelvic fins noticed as first appearing in a variable manner in the Cretaceous Isospondyli. The spiny-finned fishes began by Berycoids and possibly Scombroids in the Chalk, closely resembling, but not identical with genera living at the present day. By the Eocene period, however, nearly all the modern groups of Acanthopterygii had become completely separated and developed, and their sud- den appearance is as mysterious as that of early Eocene mammals." The same eminent authority also recapitulates the main out- lines of the evolutionary history of fishes in the following passage r1 " Fossils prove that the earliest fish-like organisms strength- ened their external armour so long as they remained compara- tively sedentary ; that next the most progressive members of the class began to acquire better powers of locomotion, and con- centrated all their growth-energy on the elaboration of fins; that, after the perfection of these organs, the internal bony skeleton was completed at the sacrifice of the outer plates, be- cause rapid movement necessitated a flexible body and rendered external armour less useful; that, finally, in the highest types the vertebrae and some of the fin-rays were reduced to a fixed and practically invariable number for each family and genus, while there was a remarkable development of spines. As survivors of most of these stages still exist, the changes in the soft parts which accompanied the successive advances in the skeleton can be inferred. Hence palaeontology furnishes a sure basis for a * Woodward, A. S., The .Relations of Paleontology to Biology. Ann. Mag. Nat. Hist., 1906, ser. 7, xviii, p. 314. 20 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. natural classification in complete accord with the development of the group." Concerning the matter of classification, it need only be re- marked that Pisces proper are divided into four subclasses, all of which have enjoyed a continuous history from the early Devonian onward to the present day. These subclasses are known under the following designations : ( I ) The Elasmobranchii, including modern sharks and rays; (2) Holocephali, or Chimseroids; (3) Dipnoi (Dipneusti), or Lung-fishes; and (4) Teleostomi, in- cluding ganoids and modern bony fishes. Only the last-named of these grand divisions is represented in the Triassic rocks of the Atlantic border region, and of the two orders embraced by it, the first (Crossopterygii) is represented by a solitary family and genus, and the second (Actinopterygii) by five genera be- longing to three different families. The taxonomic relations of these families and genera are graphically illustrated in the fol- lowing scheme : — Subclass Orders Suborders Families and Genera Teleostomi Crossopterygii Actinopterygii Actinistia f Chondrostei i 1 Protospondyli I Coelacanthidae 1 . Diplurus ( Catopteridse \ 2. Catopterus { 3. Dictyopyge [ Semionotidse | 4. Acentrophorus \ 5. Semionotus 1 | Eugnathidae L 6. Piycholepis It will be seen from the foregoing table that the fauna with which we have to deal is relatively undiversified, and consists of surprisingly few elements as compared with contemporaneous fish faunas of other regions. The inference to be drawn is that these peculiarities are in all likelihood dependent upon the nature of the environment — that is to say, upon the absence of marine conditions over the area inhabited by this fauna. The evidence NO. l8.] TRIASSIC FISHES OF CONNECTICUT. 21 which compels us to regard the sediments in question as a non- marine facies of the Trias does not proceed primarily from the fossil fishes themselves. On the contrary, the North American species of Semionotus, Ptycholepis, and Diplurus are so closely affiliated with European " geminate types," to employ Jordan's term,1 which occur in the marine Trias, that it is impossible to suppose that there were any great physiological differences be- tween them. Hence there would be no reason in the absence of other evidence to believe that that they were adapted to a different habitat.3. While there is nothing in the character of the fossil fishes I which would prove conclusively whether the deposits were formed in salt or brackish or fresh water, the physical character of the deposits and the fossils other than fishes found in them make it substantially certain that the deposits are not marine.3 No corals, echinoderms, or brachiopods have been found in the Triassic in Connecticut or in any other of the Triassic basins of eastern North America. Mollusks are very few, and most of those found are undoubtedly fresh-water forms. A very few marine mollusks, it is claimed, have been found in the Triassic of Pennsylvania. A few Crustacea, probably fresh-water or brackish-water forms, have been found in some of the southern Triassic basins, though not in Connecticut. A few insect larvae have been found. For the rest the fossils of the formation con- sist of land plants and tracks of reptiles and amphibians, with a few skeletons of reptiles. Such an assemblage of fossils makes it clear that the formation is not marine, though the presence of a few marine shells (if those shells are rightly identified) indicate conditions in part estuarine. Until recently the opinion has been generally held that the deposits of the Triassic of eastern North America were formed in tidal estuaries whose waters for the most part were brackish or nearly fresh. It seems probable, however, that the deposits 1 Jordan, D. S., The Law of Geminate Species. Am. Nat., 1908, xlii, pp. 73-80. 2 De-Alessandri remarks as follows regarding the conditions under which the strata at Besano were deposited: " I caratteri litologici infatti dimostrano come i depositi costituenti la formazione raibliana di Besano debbono in parte ascriversi ad azione organica e che essi si costituivano poco lungi dalle coste. E l'esame della sua ittiofauna .... conferma appunto la natura costiera del giacimento." 3 Rice and Gregory, Manual of the Geology of Connecticut, pp. 166-179 (State Geol. and Nat. Hist. Surv., Bull. 6.) 22 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. were not formed in continuous water bodies coextensive with the areas occupied by the deposits; but that they include confluent alluvial fans formed by torrents descending from the mountain walls at the margin of the area, fluvial deposits formed by rivers migrating over the lowland, lacustrine deposits in places where the drainage was obstructed, with probably estuarine deposits in parts of the area that were covered by tidal waters, and very likely eolian deposits over parts that were dry land. Geol- ogists have recently come to appreciate the importance of such " continental " deposits.1 We have next to consider the question of the position occu- pied in the series of Triassic rocks by the fish-bearing beds in the local section, and also the ancillary query as to the con- temporaneity of deposits in the Connecticut Valley and the New York-Virginia basin. For a solution of these problems extended comparisons are necessary with related faunas of other regions, wheresoever they may be found; and, as this phase of the dis- cussion has received very little attention heretofore, we may be permitted to inquire into the matter somewhat fully. This will be the object of the following section. ij. V. Lewis, Origin and Relations of the Newark Rocks, pp. 102-108 (Geol. Surv. N. J., Ann. Rept. for 1906). W. M. Davis, The Triassic Formation of Connecti- cut, pp. 29-34 (18th Ann. Rept., U. S. Geol. Surv., Part ii). J. Barrell, Mud-cracks as a Criterion of Continental Sedimentation (Journ. of Geol., xiv. pp. 524-568). In con- nection with these writings one may consult several recent articles by J. Lomas who interprets the British Trias as " filled-in desert lakes " (Proc. Liverpool Geol. Soc, 1907, p. 183); also the Trias Reports of the British Assoc. Adv. Sci.; and Professor Bonney's paper On the Origin of the Trias {Proc. Yorkshire Geol. Soc, 1906, xvi, p. 1). NO. l8.] TRIASSIC FISHES OF CONNECTICUT. 23 II. GEOLOGIC CORRELATION OF THE CONNECTICUT VALLEY FISH-BEARING BEDS. The general consensus of opinion among geologists who have studied the Triassic rocks of eastern North America is that the sediments were deposited more or less contemporaneously in a number of isolated basins or troughs, these depressions occurring at intervals (rather than extending continuously) along the Atlantic border from Nova Scotia to South Carolina. As to the period of deposition represented by these sediments, in the opinion of the best authorities they are referred to the uppermost division of the Triassic system, that which in European geology is termed the Keuper. This opinion, be it observed, has rested hitherto al- most exclusively upon the evidence of paleobotany. And not un- naturally, owing to the prevailing dearth of marine invertebrates, which always afford the most reliable indication of the age of strata. In view of the almost total absence of the latter class of fossils, it is pointed out by Professors Rice and Gregory in their " Manual of the Geology of Connecticut " (p. 182), that the best paleontological evidence for purposes of correlation which is here available is that " afforded by comparison of the fossil plants, which occur abundantly in some areas of the formation, particularly in the Richmond area, with the fossil plants of some of the European strata." The results of such comparison show, as stated by these authors, " that the flora of these sandstones finds its nearest equivalent in that of the Keuper, the uppermost division of the European Trias. The indications afforded by the fishes and reptiles, though more scanty, are in harmony, so far as they go, with the evidence of the plants." In like manner Professor Lester F. Ward, writing in 1891, expressed the view that the flora of the New York- Virginia area fixes the horizon of the so-called " Newark formation " " with almost absolute 24 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. certainty at the summit of the Triassic system, and narrows the discussion down chiefly to the verbal question whether it shall be called Rhaetic or Keuper. ..... The beds that seem to be most nearly identical, so far as the plants are concerned, are those of Lunz, in Austria, and of Neue Welt, in Switzerland. These have been placed by the best European geologists in the Upper Keuper. Our American Trias can scarcely be lower than this, and it probably cannot be higher than the Rhaetic beds of Bavaria."1 Professor J. S. Newberry reached a similar conclusion in re- gard to the homotaxial relations of the " Newark " series of the Atlantic border region,2 but it is rather singular that his reference of these beds to the uppermost Trias was based almost entirely upon the evidence furnished by fossil plants, whereas that af- forded by the fishes was in part neglected, in part misinterpreted. This accomplished student of paleichthyology was clearly in error, as will presently appear, in supposing that the fishes of the Newark system are " not nearly related to those of any European formation," and he failed even to recognize the identity between the so-called " Ischypterus" of Egerton and Agassiz's genus Semionotus. As early as 1850 Professor Louis Agassiz declared that in his opinion the fossil fishes from the Virginia coal field, and " from the so-called New Red Sandstone [of New Jersey and New England], indicate an age intermediate between the European New Red and the Oolite." 3 Subsequently he modified this view so far as to state that the fossils in question are the equivalent neither of the Triassic fishes of southern Germany, nor of those from the English Lias, wherefore he referred the 1 Ward, L. F., The Plant-bearing Deposits of the American Trias. Bull. Geol. Soc. Am., 1891, iii, pp. 23-31. — Idem, Principles and Methods of Geologic Correlation by means of Fossil Plants. Am. Geol., 1891, ix, pp. 34-47- Writing three years earlier than Professor Ward, the Austrian geologist, D. Stur, concluded from the evidence of fossil plants that the Virginia coal-field area is the precise equivalent of the German Lettenkohle, which, according to some geologists, immediately underlies the Keuper, while by others it is regarded as the lowest division of the Keuper, just as the Rhaetic is sometimes incorporated with the Keuper as its uppermost member. The title of his paper is as follows: " Die Lunzer- (Letten- kohlen-) Flora in den ' older Mesozoic beds of the Coalfield of eastern Virginia.' " Verh. Geol. Reichsanst., 1888, no. 10, pp. 203-217. 2 Newberry, J. S., The Fauna and Flora of the Trias of New Jersey and the Con- necticut Valley. Trans. N. Y. Acad. Sci., 1887, vi, pp. 124-128. 8 Agassiz, L., Proc. Am. Assoc. Adv. Sci., 1850, iv, p. 276. NO. l8.] TRIASSIC FISHES OF CONNECTICUT. 25 Newark rocks to a time interval between the Trias and Lias, for which there is no corresponding European expression.1 Most recently of all it has been claimed by the present writer,2 in contrast to the views of Agassiz and Newberry, that the Newark fish fauna of the Atlantic border region does, in reality, manifest rather close relationship to those of certain European formations ; and in particular that a marked correspondence exists between it and the assemblage that is known from the top- most member of the Alpine Muschelkalk — that is to say, the summital portion of the Middle Trias as developed in the Mediterranean region. The precise horizon from which the foreign assemblage in question has been obtained is the so-called Perledo limestone, on the eastern verge of Lake Como, in Lombardy, and usually correlated with the base of the Buchen- stein beds.3 By some authors the latter are referred to the base of the Ladinian, by others — and these would seem to be in the minority — to the summit of the Virglorian (zone of Ceratites trinodosus). Under these names, Ladinian and Vir- glorian, are understood respectively the upper and lower mem- bers of the Alpine Muschelkalk, where, as is well known, a three- fold division such as characterizes the deposits of Germany north of the Alpine region is not observable.4 The position of the Buchenstein limestone, and also of the higher (Keuper) fish-bearing beds of Besano, Raibl, Seefeld, etc., as compared with the Anglo-German Trias, is shown in the subjoined table. This is constructed more especially after the writings of Baron F. von Huene on the British, and those of Professors Haug, Arthaber and others on the Alpine Trias.5 1 Agassiz, L., Proc. Am. Acad., 1852-57, iii, p. 69. 2 Geol. Surv. N. J., Ann. Rep. for 1904 (1905), pp. 70, 72. 3 A convenient geological guide for the Como section will be found in chapter 1 of Dr A. Tornquist's " Fiihrer durch das oberitalienische Seengebirge," forming volume 9 of the Sammhtng geologischer Fiihrer (Berlin, 1902). 4 " Le Trias moyen comprend, comme on sait, deux etages: le Virglorien Ren. (= Recoarien Bittn., Anisien Mojs., Waag., Dien.), et le Ladinien Bittn." — E. Haug, in Bull. Soc. Geol. France, 1906, ser. 4, vi, p. 368. 5 A generalized section of the Alpine Trias is given at page 254 of G. von Artha- ber's treatise, incorporated with the second part of Freeh's " Lethsa Geognostica " (Heft 2, 1905), and a more detailed section of the Como district at page 399 of the same work. Heft 1 of this volume contains an introduction, by the editor, to the study of the Mesozoic and Trias, and a description of the Continental Trias by E. Phillippi in collaboration with other authors (1903). See also the following by F. von Huene: " Eine Zusammenstellung uber die englische Trias und das Alter ihrer Fossilien. Centralbl. f. Mineral. Geol. Pal., 1908, pp. 9-17. A slightly different arrangement is indicated in the table given at p. 29 of G. De-Alessandri's memoir on the Triassic Fishes of Lombardy, 1910. 26 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. Correlation Scheme of Anglo-German and Alpine Trias. TRIAS GREAT BRITAIN GERMANY ALPINE REGION Rhaetic Rhaetic Rhaetic Upper Keuper marl (Upper Keuper sandstone) Upper \ Middle V Keuper Lower ) u f Hauptdolomit (Stern- ly mergelkeuper) ' p g -{ St. Cassian beds jjjlW 1 Raibl beds (Salz- no. 3- Idem, La Ittiofauna della Dolomia principale di Giffoni (prov. di Salerno). Palaeontogr. Italica, 1895, »> pp. 169-210.— Idem, Elenco dei Pesci fossili degli schisti bituminosi triasici di Giffoni, nel Salernitano. Rend, della R. Accad. delle Sci. di Napoli, 16 Dec. 1899. No. 1 8.] TRIASSIC FISHES OF CONNECTICUT. 33 Gorjanovic-Kramberger,1 De-Alessandri,2 and others. The nu- merals indicate the number of species represented at the respec- tive localities, which are as follows : Lumezzane, in Val Trompia, Lombardy; Seefeld, near Innsbruck, Tirol; Hallein, in Salz- burg, Austria ; and GifToni, in the Province of Salerno, southern Italy. Comparative Table of Hauptdolomit Fish Faunas. Genera Giffoni Hallein Seefeld Lumezzane Crossopterygii Fam. Ccelacanthid^e i. Undina 1 Actinopterygii Fam. BelonorhynchidjE 2. Belonorhynchus 1 I Fam. Semionotid,e 3. Semionotus 4. Colobodus 5. Heterolepidotus 6. Dapedius 7. Sftaniolepis 2 1 I 2 2 I I 2 I I 2 Fam. Macrosemiid^e 8. Ophiopsis % 9. Orthurus I I Fam. Pycnodontid;£ 10. Mesodon I Fam. EUGNATHID.E 11. Eugnathus 1 I Fam. Pholidophorid^e 12. Pholidophorus 13. Peltopfeurus 14. Thoracopterus 3 1 1 2 4 1 4 Total number of species 11 II 11 7 A final word may now be said concerning the relations be- tween the Atlantic border or "Newark" fish fauna and the meagre indications of Triassic fish life that have been obtained 1 Gorjanovif-Kramberger, K., Die obertriadische Fischfauna von Hallein in Salz- burg. Beitr. Pal'dont. Geol. 1905, xviii, pp. 193-224. 2 De-Alessandri, G., Studii sui Pesci Triasici della Lombardia. Mem. della Soc. ltd. di S dense Naturdi e Museo Civico di Storia Nat. di Milano, 1910, vii, pp. 1-148. 3 34 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. from western North America. Owing to the well known in- timate connection between the Upper Trias of California and that of southern Europe — the region which Neumayr years ago called the " central Mediterranean," and for which the name " Thetys " was proposed by Suess in 1894 — we might expect on purely a priori grounds to find a certain resemblance in the character of the vertebrate faunas of the two regions.1 The scant evidence thus far obtained, however, at least as regards the fishes, fails to justify any such expectation. From the Lower Trias of Idaho have been described a single detached dermal spine, apparently belonging to Aster acanthus? and a few fragmentary remains of Crossopterygii, which are possibly late survivals of Paleozoic families (Rhizodontidae and Osteol- epidse).3 Elasmobranchs and effete Crossopterygii persist even as late as the Upper Trias of Shasta county, California, where re- mains of Hybodus, Acrodus, Holoptychius and Xenestes have been brought to light, whose number, however, all told, makes an inconspicuous showing.4 Yet another sprinkling of ichthyic indications is known from the Red Beds of supposed Triassic age (Shinarump group) in southwestern Colorado and in the Kanab Canyon region of Utah and Arizona. Little has been published on the fossil vertebrate remains from this section,5 but, so far as the fishes are concerned, it is clear that they display no intimate relations with those of the Atlantic border Trias. On the contrary, the general aspect 1 In regard to the invertebrate faunas, Dr. James Perrin Smith has the following: " The most interesting fact brought out by a comparison of the Upper Trias of California with that of India and the Alpine Mediterranean region is its near rela- tionship with the latter, most genera and many species being common to the two regions. . . . This relationship of the Californian to the European faunas per- sists until after the middle of the Jurassic formation, when the Boreal fauna comes in." — Journ. Geol., 1898, vi, p. 786. 2 This is described under the name of Cosmacanthus by H. M. Evans, in Bull. Dept. Geol. Univ. of Calif., 1904, i", P- 397- 8 Goddard, M., Fish Remains from the Marine Lower Trias of Aspen Ridge, Idaho. Bull. Dept. Geol. Univ. of Calif., 1907, v, p. 145. * Wemple, E. M.. New Cestraciont teeth from the West American Triassic. Bull. Geol. Dept. Univ. of Calif., 1906, v, no. 4, p. 73. — Jordan, D. S., The Fossil Fishes of California. Ibid., 1907, v, no. 7, pp. 95-144. * The chief literary references are collected by Dr. Whitman Cross in his article on " The Triassic Portion of the Shinarump Group, Powell," to be found in the Journal of Geology, 1908, xvi, pp. 97-123. See also the joint paper by the same author and E. Howe, entitled " Red Beds of Southwestern Colorado and their Cor- relation." Bull. Geol. Soc. Am., 1905, xvi, pp. 447-486. No. l8.] TRIASSIC FISHES OF CONNECTICUT. 35 of the Shinarump material suggests the foreshadowing of Jurassic conditions, and on that account the fauna announces itself as proemial, to employ Dr. Garke's expressive term. The condition of the remains from the Kanab Canyon region is ex- tremely fragmentary, and among them only the genus Pholid- ophorus and certain Lepidotidae appear to be tolerably well indicated. 36 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. III. GEOGRAPHY OF THE TRIAS. " It is the soul that sees : the outward eyes Present the object, but the mind descries." — Longfellow. In this section it is not at all within our purpose to enter into an elaborate account of the physical conditions prevailing during the initial period of the Mesozoic, the theme being in itself an intricate one, and moreover, that phase of it which applies to our local section having been already sufficiently discussed. Such being the case, it will be sufficient merely to direct attention to a con- tribution pertinent to this topic, and one which contains perhaps the most comprehensive review that has been written on the faunal geography of the Upper Trias. We refer to the chapter on " The Seas of the Trias Era," by the late Professor E. Mojsisovics, Edler von Mojsvar, in his memoir on Triassic Cephalopods from the Himalayas,1 wherein is collected practically all that is known of the distribution and relations of the in- vertebrate faunas of this era. From this memoir, owing to its relevancy to the present dis- cussion, we have ventured to extract a paragraph or two in re- gard to the principal regions of the " Central Mediterranean Sea " of Neumayr, or " Thetys " of Suess, from which Triassic faunas are known. So different from one another in character are these faunas that they are manifestly to be regarded as representing ancient geographic regions. These provinces are, in the language of their chief exponent and interpreter, as follows : " 1. Die Mediterranprovinz, 2. die germanische Flachsee, und 3. die indische Provinz." Concerning the limits of these provinces the author remarks : " Die germanische Flachsee bildet eine Dependenz der Medi- terranprovinz, und kann als ein grosses Aestuarium aufgefasst werden, welches dem ausgedehnten, heute im atlantischen Oc£an 1 Denkschr. k. k. Akad. Wiss., 1896, lxiii, p. 687. No. l8.] TRIASSIC FISHES OF CONNECTICUT. 37 versunkenen Continente vorgelagert war. Diese triadische 'Atlantis ' existirte wahrscheinlich bereits am Schluss des palao- zoischen Zeitalters.1 Sie reichte im Westen vermuthlich bis zum heutigen Nordamerica,2 welches bekanntlich auf seiner Osthalfte ausgedehnte triadische Binnensedimente vom Charakter des germanischen Bttntsandsteines und Keupers besitzt, wahrend pelagische Sedimente der Trias bios auf dem pacifischen Abhange dieses Continentes anzutreffen sind." Expressed in English, the meaning of the above paragraph would be that the second of these geographical provinces forms a restricted area of the first, and may be regarded as having formerly been a large estuary of an extensive land area now submerged beneath the Atlantic Ocean. This hypothetical Tri- assic continent, the so-called "Atlantis," probably became elevated above sea level at least as early as the close of the Paleozoic. It presumably extended westward to the present continent of North America, for along the eastern border of the latter are found non-marine Triassic deposits corresponding to the central European Buntsandstein and Keuper, while marine Triassic rocks occur only along the Pacific slope. [It is proper to point out that the theory of a submerged " Atlantis " is by no means universally held among modern geologists, but on the contrary many of the foremost authorities are firm believers in the permanence of con- tinental land masses.] To pursue this topic of paleogeography a step further, it is of interest to recall that the eastern and western boundaries of the Triassic Thetys are thus delineated by Professor James P. Smith, in his article on the " Classification of Marine Trias " (Journ. of Geol., 1896, iv, p. 387) : " Along the western borders of Thetys were deposited the Triassic sediments of the Alps, Spain, southern Italy, the 1 Suess, Antlitz der Erde., ii, p. 317. 2 Einen sicheren Anhaltspunkt fur die Annahme eines solchen Continents bieten auch die Pflanzenreste dar, welche in den Kohlenfeldern des ostlichen Virginiens gefunden und von Stur mit den Pflanzen des Lunzer Sandsteins (julische Stufe) identificirt wurden. Vergl. Stur, "Die Lunzer- (Lettenkohlen-) Flora in den 'older Mesozoic beds of the Coalfields of eastern Virginia.' " Verh. Geol. Reichsanstalt, 1888, p. 203. Besides the paper of Stur's just referred to, one may consult the following by F. Zeller, which contains a comparison with the Alpine Trias: Beitrage zur Kenntniss der Lettenkohle und des Keupers in Schwaben. Neues Jahrb. f. Mineral, u. s. w., Beilage-Bd. xxv, 1908, pp. 1-134. His correlation of the fish-bearing beds of the Alpine Trias is essentially the same as that adopted in the present Report. 38 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. Balearic Islands, Sicily, Hungary, and the Balkan Peninsula. This region was named by Mojsisovics the Mediterranean Trias province. Most of the faunas of the Trias, from near the base to the top, are represented in this region. To the east the Thetys spread out to the waters of the Indian region, in which the sediments of the Himalayas and the Salt Range were accumulated. The Indian waters joined on the north, east and south with the great Arctic-Pacific Trias ocean, or Arctis of Mojsisovics, along the borders of which were de- posited the sediments of northern and eastern Siberia, Spits- bergen, Japan, Rotti, New Zealand, New Caledonia, Peru and western North America. But in this ocean region there were many provinces as yet unknown, or only vaguely defined." NO. l8.] TRIASSIC FISHES OF CONNECTICUT. 39 IV. CONCERNING EARLIER INVESTIGATION OF NORTH AMERICAN TRIASSIC FISHES. " In den Wissenschaf ten ist es hochst verdienstlich, das unzulangliche Wahre, was die Alten schon besessen, aufzusuchen und weiter zu fuhren." — Goethe. American vertebrate paleontology may be said to have begun with President Thomas Jefferson's description of fossil elephant remains from Virginia1 in 1787, and the bones of Megalonyx, afterwards named M. jeffersoni, a dozen years later.2 One has to turn back a little more than a century earlier, however, for the first published figure of an American fossil, this being Ecphora quadricostata from the Maryland Miocene.3 The earliest records of all relating to the discovery of fossil vertebrate remains in the western hemisphere date from the time of Hernandez, court physician to Philip II, and other Spanish explorers of the seven- teenth century.4 We cannot be sure when fossil fishes first began to attract attention in this country, but the earliest notices regarding them in any scientific publication fall within the second decade of the last century, and relate to remains discovered in the Connecticut Valley region. Several titles are comprised in these early notices, and among their authors occur such names as S. L. Mitchell, B. Silliman the elder, Edward Hitchcock, A. Brong- niart, W. W. Mather, James E. Dekay, and others. Per contra, 1 Notes on the State of Virginia. London, 1787. 2 A Memoir on the Discovery of certain bones of a Quadruped of the clawed kind in the western parts of Virginia. Trans. Anver. Phil. Soc, 1799, iv, pp. 246-260. Dr. O. P. Hay is authority for the statement that this work is said by C. G. Giebel to have been issued in 1797. 3 Lister, M., Historia sive Synopsis Methodical Conchyliorum. London, 1685. PI. 1059, fig. 2. 4 References to old Spanish works in which these remains are attributed to' a race of human giants are given in the second volumes respectively of Cuvier's " Ossemens Fossiles " and Humboldt's " Cosmos." The vulgar interpretation, which is ap- parently common to all primitive society, ancient and modern, finds an apt illustra- tion in the Gigantomachia of classical antiquity. Consult the suggestive article by Dr. Th. Skouphos, in Comptes rendus Cong. Inter. d'Arch., Athens, 1905, pp. 231-236. Also one by E. von Lasaulx on the Geology of the Greeks and the Romans, in Abhandl. bayer. Akad. Wissensch., 1852, vi, pp. 517-566. 40 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. the occurrence of similar remains in the more southerly region does not seem to have become generally known until toward the middle of the nineteenth century. Those from the Virginia Coal fields were studied successively by the Redfields, father and son, Sir Charles Lyell, Sir Philip Grey Egerton, and Louis Agassiz ; and a brief mention of fragmentary remains from North Carolina, accompanied by a few figures, was contributed by Ebenezer Emmons during the late fifties. It is, however, to William C. and John H. Redfield, who wrote between 1837 and 1857, that we are indebted for the first really satisfactory account of the Triassic fish fauna of this country, these two having described nearly all the important species. Their results are embodied in ten publications, eight by the elder, and two by the younger author. These same pioneers also brought together an important collection, of which a good part is still preserved in the Peabody Museum at Yale University, and the rest is unfortunately destroyed or dissipated. By far the most signal contribution to our knowledge of Amer- ican Triassic fishes is that contained in Professor J. S. Newberry's " Monograph on the Fossil Fishes and Fossil Plants of New Jersey and the Connecticut Valley."1 Several new species of Semionotus (described, however, under the title Ischypterus), Ptycholepis and Diplurus were established by him upon the evidence of tolerably satisfactory material, and one doubtful form was referred with some reservation to Acentrophorus, a genus that is otherwise limited, so far as known, to the Upper Permian. This elaborate work of Professor Newberry still re- mains our chief repository of information in regard to the par- ticular subject before our consideration. Since Newberry's time comparatively little has been added to our knowledge of the Newark fish fauna, except in the way of rectifying some minor details. An important memoir on the genus Semionotus, by Dr. E. Schellwien,2 appeared in 1901, in which a few new anatomical points, accompanied by illustra- tions, are worked out for two previously known American species. A number of additional structural characters were made known in 1903 by Dr. George F. Eaton, of Yale University, in the case 1 Monogr. U. S. Geol. Surv., xiv. Washington, 1888. 2 Schellwien, E., Ueber Semionotus Ag. Schriften der Phys.-Oekonom. Gesellsch. zu Kenigsberg i. Pr., 1901, pp. i-34t pi- i-iii. No. l8.] TRIASSIC FISHES OF CONNECTICUT. 41 of four or five American species of Semionotus.1 The diagnosis of this genus was further emended by Professor Gorjanovio- Kramberger, still more recently, in the course of his description of the Upper Triassic fish fauna of Hallein, Salzburg.2 In this memoir the author set forth evidence intending to show that the family position of Heterolepidotus is with the Semionotidae rather than with the Eugnathidse, and that Allolepidotus of Deecke is identical with Semionotus proper. During the same year, 1905, some revised descriptions of the Triassic fishes of New Jersey were published by the present writer, with incidental mention of Connecticut Valley forms.3 Reference was made in this paper to the totally different char- acter of the Kanab Valley fish fauna (Triassic portion of the Shinarump group, Powell) as compared with that of the Atlantic border region, and it was pointed out that the former displayed a marked Liassic aspect. That the beds which carry this fauna are in reality anterior to the Lias, and probably belong to the late Trias, has been recently argued by Dr. Whitman Cross in the Journal of Geology for 1908. The few contributions that have appeared in regard to the Triassic fishes of the Cordilleran region have already been referred to in the preceding section. In regard to restorations of the leading genera Semionotus and Dictyopyge, figures of these were published as early as 1864 by J. Struver, which are fairly accurate in respect to form of body and fin-structures, but leave much to be desired in the representation of cranial and facial bones. These figures are reproduced by Freeh in his Introduction to the Mesozoic (Part II. of the "Lethaea Geognostica," Stuttgart, 1903), and two other illustrations of American Triassic fishes are copied in the same work from Newberry's Monograph (Texttafel vi, vii). No satisfactory restoration of Catopterus has yet appeared, but some figures of the head portion, prepared from original draw- ings by the late Professor Newberry, are now published for the first time in the present Report in the section devoted to that genus. (Figs. 5, 6, p. 54.) 1 Eaton, G. F., Notes on the Collection of Triassic Fishes at Yale. Amer. Journ. Sci., 1903, ser. 4, xv, pp. 259-268, pi. v, vi. 2 Gorjanovic-Kramberger, K., Die obertriadische Fischfauna von Hallein in Salz- burg. Beitr. sur Pal'dont. und Geol., 1905, xviii, pp. 193-224, pi. xvii-xxi. sGeol. Surv. N. J., Ann. Rept. for 1904 (1905), pp. 67-102. 42 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. V. SYSTEMATIC DESCRIPTIONS OF UPPER TRIASSIC FISHES. "Die Natur ist das einzige Buch, das auf alien Blattern gewissen Inhalt bietet."— Goethe. Order CROSSOPTERYGII. Family CCELACANTHID^E. " Body deeply and irregularly fusiform, with cycloidal, deeply overlapping scales, more or less ornamented with ganoine. Branchiostegal apparatus consisting of an operculum on each side and a single pair of large jugular plates. Paired fins ob- tusely lobate. Two dorsal fins and a single anal; the anterior dorsal without baseosts, the posterior dorsal and the anal with bas- eosts, obtusely lobate. Axial skeleton extending to the extremity of the caudal fin, usually projecting and terminated by a small supplementary caudal fin. Air-bladder ossified." As remarked by Smith Woodward, from whose Catalogue the foregoing definition has been taken, the members of this family have perhaps the most remarkable geological range of all known extinct fishes, persisting as they do practically unchanged from the Upper Devonian to the Upper Chalk. " The group is special- ized," says this author, " in the large symmetrical caudal fin, which exhibits a series of supports directly apposed to the neural and haemal arches, equalling in number both these and the over- lapping dermal rays. It is also specialized in (i.) the fusion of the bones of the pterygo-quadrate arcade, (ii.) the reduction of the infradentaries to one, (iii.) the reduction of the opercular apparatus to the operculum on each side and a pair of gular plates, (iv.) the loss of the baseosts in the anterior dorsal fin, and (v.) the ossification of the air-bladder." This family, first proposed by Louis Agassiz in the second volume of his " Poissons Fossiles " (1844, p. 168), and after- wards greatly restricted by Huxley in two important memoirs of the British Geological Survey (Decades X and XII, 1861 and 1866), is at present understood as comprising not more than six NO. l8.] TRIASSIC FISHES OF CONNECTICUT. 43 well recognized genera, among which the most satisfactorily known are Ccelacanthus proper, Macropoma and Undina. The typical genus enjoys the truly remarkable range from the Upper Devonian to the close of the Paleozoic, and, if the evidence of one or two doubtful forms be accepted, possibly even higher; the remaining genera extend throughout the Mesozoic, and ex- hibit such constancy of structural characters that the family has been frequently cited as one of the most distinct and well defined in the animal kingdom. Huxley, for instance, drew attention to its singular compactness and homogeneity in the following para- graph '} . " The Ccelacanthini, as thus understood, are no less distinctly separated from other fishes than they are closely united to one another. In the form and arrangement of their fins; the structure of the tail and that of the cranium ; the form and num- ber of the jugular plates; the dentition; the dorsal interspinous bones ; the pelvic bones ; the ossified air-bladder ; the Ccelacan- thini differ widely from either the Saurodipterini, the Glypto- dipterini, or the Ctenodipterini ; but, on the other hand, they agree with these families and differ from almost all other fishes, in the same respects as those in which the several families just mentioned have been shown to agree with one another, viz., the number of the dorsal fins, the location of the paired fins, the absence of branchiostegal rays and their replacement by jugular bones." Finally, concerning the extraordinary conservatism and per- sistence manifested by the group of Ccelacanth fishes ever since its introduction, the illustrious English biologist whom we have quoted expresses himself as follows:2 " Bearing in mind the range of the Ccelacanths from the Car- boniferous [since ascertained to extend from the Devonian] to the Chalk formation inclusive, the uniformity of organization of the group appears to be something wonderful. I have no evidence as to the structure of the base and side walls of the skull in Ccelacanthus, but the data collected in the present Decade 1 Huxley, T. H., Preliminary Essay upon the Systematic Arrangement of the Fishes of the Devonian Epoch, prefixed to the Tenth Decade of the Figures and Descriptions illustrating British Organic Remains (1861, p. 20). 2 Illustrations of the Structure of the Crossopterygian Ganoids. Memoirs of the Geological Survey of the United Kingdom, Decade xii, 1866. Reprinted in the supple- mentary volume of the Scientific Memoirs of Thomas Henry Huxley, 1903. P- 6S- 44 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. shows that, in every other particular save the ornamentation of the fin-rays and scales, the organization of the Coelacanths has remained stationary from their first recorded appearance to their exit. They are remarkable examples of what I have elsewhere termed " persistent types ;" and, like the Labyrinthodonts, assist in bridging over the gap between the Palaeozoic and the Mesozoic faunae." The earliest known American representative of this family is a typical Ccelacanth, described as Ccelacanthus welleri,1 from the Lower Kinderhook (base of the Mississippian series) of Iowa. Three other species are known from the Coal Measures of Ohio and Illinois, but none from later horizons until we meet with the very remarkable and in some respects degenerate (e. g., as regards loss of certain of its head bones and most of its tail) Diplnrus in the " Newark " rocks of the Atlantic border region. So far as known, this genus comprises but a single large species, D. longicaudatus, which is common to both the Connecticut Valley and New Jersey areas. A vicarious form, to use a Ger- man expression, or perhaps what President Jordan would call a " geminate species " or genus,2 is represented in the Perledo limestone of Lombardy by Heptanema paradoxum Bellotti. Genus Diplurus Newberry. Supplementary caudal fin prominent, with much elongated pedicle ; fin-rays robust, closely articulated in the distal half ; pre- axial rays of the first dorsal and caudal fins with spinous tubercles. Scales and head bones irregularly striated, and some of the latter finely granulated. Diplurus longicaudatus Newberry. 1878. Diplurus longicaudatus J. S. Newberry, Ann. N. Y. Acad. Sci., i, p. 127. 1888. Diplurus longicaudatus J. S. Newberry, Monogr. U. S. Geol. Surv., xiv, p. 74, pi. 20. 1 891. Diplurus longicaudatus A. S. Woodward, Cat. Foss. Fishes Brit. Mus., pt. 2, p. 409. 1905. Diplurus longicaudatus C. R. Eastman, N. J. Geol. Surv. Rept. for 1904, p. 101. 1 Journ. Geol., 1908, xvi, p. 357. 2 Jordan, D. S., The law of geminate species. Am. Nat., 1908, xlii, p. 73. No. l8.] TRIASSIC FISHES OF CONNECTICUT. 45 The type and only known species, attaining a total length of about 70 cm. to the tip of the supplementary caudal fin, and maximum depth of trunk about 20 cm. Anterior dorsal fin strong, supported by a single large laminar axonost; the lobate posterior dorsal nearly opposite the anal, and corresponding to it in form and size. Caudal fin much elongated, and separated from the supplementary caudal by a distinct interval. Paired fins obtusely lobate. Scales large, cycloidal, and deeply over- lapping ; the exposed portion marked with fine longitudinal rugse ; teeth unknown. This large Crossopterygian is of extremely rare occurrence, being known by a scant half-dozen individuals, of which two, in- cluding the type, were obtained from near Boonton, New Jersey, and the others, very imperfect, from Durham, Connecticut. Most of the remains are now preserved in the American Museum of Natural History in New York, but there is one distorted ex- ample of the lower jaw in the Museum of Wesleyan University at Middletown (Cat. No. 846), which was collected by Mr. S. W. Loper from the anterior shales. Unfortunately this speci- men shows no satisfactory indication of teeth, but appearances are at least suggestive that these were slender and conical. The external surface of the bone is finely granulated. Order ACTINOPTERYGII. Paired fins non-lobate, having an extremely abbreviated en- doskeletal portion, and the dermal rays prominent. Caudal fin abbreviate-diphycercal, heterocercal, or homocercal. A single paired series of transversely elongated rays, with or without an anterior azygous element, developed in the branchiostegal mem- brane between the mandibular rami. Suborder ChONDROSTEI. Sturgeons. In these fishes, the oldest and most primitive of the Actin- opterygii, the notochord is more or less persistent, the supports of the dorsal and anal fins are less numerous than the dermal rays apposed to them, the paired fins more abbreviate than in the Crossopterygian order, and the tail is completely heterocer- cal. Primitive sturgeons differ also from the fringe-finned 46 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. ganoids in the development of a paired series of transversely- elongated branchiostegal rays to replace the pair of jugular plates between the mandibular rami; infraclavicular plates, however, are retained in both groups. Nearly all the older forms have a well developed rhombic and ganoid squamation. So far as known, the chondrocranium is but little ossified, and the cranial bones are mainly dermal. The evolutionary history of the sturgeon tribe is thus sum- marized by Professor Bridge in the Cambridge Natural History volume on Fishes (1904, p. 485) : " The Chondrostei are first represented in the Lower Devonian by the solitary Palasoniscid genus Cheirolepis, a contemporary of the earliest Crossopterygii. They occur throughout the Mesozoic period, except in the Cretaceous, and also in the Eocene, and, while steadily diminishing in number and variety, they gradually approximate to their degenerate and in some respects highly specialized descendants, the sturgeons and paddlefishes of the existing fish fauna. Of the seven families included in the group, the Palaeoniscidse are the oldest and most generalized. The Platysomatidse are a specialized offshoot from the Palseoniscidae, and, if they are rightly to be considered as Chondrostei, perhaps the same may be said of the problematic Belonorhynchidse. On the other hand, there are certain features which indicate an ap- proach to Fishes of an altogether more modern type. Finally, the Chondrostei represent a stage in a career of degeneration, the climax of which is reached by the modern Polyodontidse and Acipenseridae." Family CATOPTERID^. Trunk elongate or elongate-fusiform; tail abbreviate-hetero- cercal. Head bones well developed, ganoid; no median series of cranial roof-bones; teeth slender, conical; eye far forward, and snout prominent; mandibular' suspensorium more or less obliquely directed backward and downward. A series of branchi- ostegal rays present. Dorsal fin single and not much extended. Scales rhombic, ganoid. This short-lived family, in which are comprised not more than three closely related genera {Catopterus, Perleidus, and Dictyopyge), appears in the early Mesozoic just as the large and successful group of Palseoniscid fishes are entering upon their No. l8.] TRIASSIC FISHES OF CONNECTICUT. 47 decline. Derived in all probability from the ancient Palseoniscid stock, and still retaining certain of its characteristics, these genera represent an advance over primitive sturgeons in the di- rection of the next higher suborder (Protospondyli), yet without marking a definite transition to that group. The upper lobe of the tail has become shortened, although still heterocercal ; and in Dictyopyge at least the supports of the anal fin are recorded as fewer in number than the apposed dermal rays. The family is accompanied in the Trias by other Chondrosteans which be- came eel-shaped (Belonorhynchidae) and died out during that period. Still others, which gradually lost their scaly covering and head bones (Chondrosteus), continued to survive, and are represented by the sturgeons of the present day. The relations of this family are, therefore, with modern sturgeons rather than with the two surviving genera of Protospondyli, Amia and Lepidosteus. Genus Catopterus Redfield. (Syn. RedHeldius Hay.) Trunk elegantly fusiform, head relatively small, tail hemi- heterocercal. External bones more or less ornamented with ridges and tubercles of ganoine; no median series of cranial roof-bones. Fins of moderate size, consisting of robust rays, more or less enameled, and distally bifurcated; fulcra well de- veloped, short and closely set. Dorsal and anal fins triangular, the origin of the former behind that of the latter; caudal fin forked. Scales large or of moderate size, nearly or quite smooth, and serrated along their postero-inferior margin ; dorsal ridge-scales not much enlarged. Teeth numerous, small, acutely conical. This genus appears to be restricted to the Atlantic Border Trias of North America, although a supposed Catopterid genus, named Perleidus by De-Alessandri, occurs in the Ladinian lime- stone of Lombardy, and the still more closely related genus Dictyopyge is of world-wide distribution. It is to be noted that remains of Catopterus are on the whole less abundant than those of the accompanying genus Semionotus, both in the Connecticut Valley area and in New Jersey, and as a rule they are less well preserved. Nevertheless, the characters 48 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. presented by the former genus are so well marked and distinctive that there is seldom any difficulty in determining even the most fragmentary individuals. The most obvious peculiarity of the genus consists, as the name implies, in the remote position of the dorsal fin. In Semionotus the dorsal is always anterior to the anal fin, in Catopterus it is either opposite or posterior. The margins of all the fins are closely set with fine fulcra, in such wise that they present a delicately fringed appearance, and the fin-rays themselves are very numerous, finely articulated, and enameled. Other noticeable differences consist in the orna- mented condition of the cranial bones, and serration of the hinder margin of the scales. Whereas the genus Semionotus is represented in this country by half a dozen or more species, its associates Catopterus and Dictyopyge comprise a much smaller number, in fact not more than one or two each. After a critical study of differential characters we are forced to admit that only two species of Catop- terus are capable of being clearly distinguished. These are C. gracilis Redfield and C. redHeldi Egerton, both founded on large and nearly complete fishes which differ from one another chiefly as regards proportions of body and scale characters. The so-called C. parvulus Redfield is probably to be regarded as the young of C. gracilis, and the species named by Newberry C. minor and C. ornatus are supposed to stand in a similar relation to C. redHeldi. Catopterus gracilis J. H. Redfield. (Plates IX-XI.) 1837. Catopterus gracilis J. H. Redfield, Ann. Lyceum Nat. Hist. N. Y., iv, p. 37, pi. 1. 1841. Catopterus gracilis W. C. Redfield, Am. Journ. Sci., [1] xli, p. 27. 1841. Catopterus gracilis E. Hitchcock, Final Rept. Geol. Mass., ii, pp. 440, 460. 1888. Catopterus gracilis J. S. Newberry, Monogr. U. S. Geol. Surv., xiv, p. 55, pi. 16, figs. 1.-3. 1895. Catopterus gracilis A. S. Woodward, Cat. Foss. Fishes Brit. Mus., iii, p. 2. 1905. Catopterus gracilis C. R. Eastman, Ann. Rept. N. J. Geol. Surv. for 1904, p. 96. NO. l8.] TRIASSIC FISHES OF CONNECTICUT. 49 The type species, attaining a total length of about 25 cm. Length of head with opercular apparatus about equal to maximum depth of trunk, and contained five times in the total length of the fish ; depth of caudal pedicle somewhat less than one-half that of the abdominal region. Cranial bones finely granulated. Pelvic fins arising about midway between the pectorals and anal ; dorsal and anal fins subequal in size, and almost completely opposed. Scales smooth, none deeper than broad, those of the flank in the abdominal region very finely serrated. The fin-formula given for this species in the original descrip- tion by J. H. Redfield is as follows : D. 10-12; C. 30-40; A. 20-30; V. circa 8; P. 10-12. In the additional notes on this form drawn up by the elder Redfield, it is stated that " the pectoral fins are of an elongated form, and are strengthened on the anterior margin by one or two large and partly flattened rays, to the front of which the fringe of fine ray lets [fulcra] is attached. Owing to this peculiarity of structure, the smallest section of the pectoral fin will often serve to identify this species." Although the form of body in this species is usually more slender than in C. redfieldi, it sometimes happens that distorted specimens, in which the anterior part of the trunk has become " shortened up " by mechanical deformation, simulate the deeper- bodied species in outline and general proportions. Conversely, also, the greater depth of body in C. redfieldi as compared with the genotype is often obscured by the familiar hazard of vertical compression, a circumstance which has frequently led to a con- fusion of the two species. Indeed, this very circumstance happens to be illustrated in the case of one of the original cotypes upon which the species was established by the younger Redfield ; and so impressed was Newberry with the idea that the depth of body had become reduced by fortuitous agency that he actually proposed to cancel the specific name bestowed upon it by the original author, because, as he avers, for a fish which " in its normal condition has nearly the outline of the shad the name of Catopterus gracilis is inappropriate and conveys a false impression." 50 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. Arbitrary and captious as Newberry's procedure appears to the modern systematise such license was by no means uncom- mon, and, indeed, seems to have been not only tolerated but ap- proved by the best paleontological authorities of his day. Nowa- days, when nomenclatural codes are much more rigidly en- forced, it would be contrary to all rule to abandon a valid specific name because of either a real or imaginary incongruity of mean- ing, and in cases where the name originally bestowed upon a species has become displaced for no more cogent reason than this, the tendency is to reinstate it. In the present instance it cannot be said that Newberry's proposal has become generally adopted, nor has it the sanction of long established usage ; hence the only course open to us is to continue to recognize the original of Redfield's figure as one of the authentic cotypes of this species. We are indebted to the generosity of Professor Schuchert for the privilege of reproducing a photograph of this well pre- served exemplar (Plate IX), which is now the property of Yale • University Museum. In this will be noted inter alia the Pal- seoniscid-shaped head, forwardly placed orbit, and tolerably distinct outlines of facial and cranial plates. The mandible, un- fortunately, is missing, the striated opercular and tuberculated cheek plates are arranged after a different pattern from the corresponding parts in the Semionotidas, and there is no clear indication of either a circumorbital ring or of branchiostegal rays. Another nearly complete example of the same species is il- lustrated in Plate X. Like the first, it was obtained from near Durham, Connecticut, but from a somewhat higher level, the horizon being that known as the posterior shale.1 Mr. S. W. Loper, who collected it, remarks that this is the only good speci- men ever obtained from the beds in question, after many years of fruitless search. The specimen is remarkable for its well preserved squamation, and it also reveals the .outline of the head much more satisfactorily than the Redfield cotype. The mandible is 1 Most of the fossil fishes in the Connecticut valley have been found at two well- marked horizons. One stratum of black shale lies between the lower (anterior, Perci- val) and the thick middle or main lava sheet, another between the main and the upper (posterior, Percival) lava sheet. These two fossiliferous strata have been called accordingly the anterior and the posterior black shale. Davis and Loper, Two Belts of Fossiliferous Black Shale in the Triassic Formation of Connecticut. Bull. Geol. Soc. Am., ii, pp. 415-430. No. l8.] TRIASSIC FISHES OF CONNECTICUT. 5 1 Palseoniscid-like, and still carries a few minute teeth. Remains of the same species are common to both the New England and New Jersey areas. Catopterus redfieldi Egerton. (Figs. 5, 6.) 1847. Catopterus redfieldi Sir P. G. Egerton, Quart. Journ. Geol. Soc, iii, p. 278. 1888. Catopterus redfieldi J. S. Newberry, Monogr. U. S. Geol. Surv., xiv, p. 53, pi. 15, figs. 1-3. 1895. Catopterus redfieldi A. S. Woodward, Cat. Foss. Fishes Brit. Museum, pt. 3, p. 3. 1903. Catopterus redfieldi, F. Freeh, Lethaea geognostica, Part 2, Trias, p. 12, text-pl. 7, fig. 2. 1905. Catopterus redfieldi C. R. Eastman, Ann. Rept. Geol. Surv. N. J. for 1904, p. 98. This species is described by its founder as " broader than the preceding [C. gracilis], and with scales not so long in proportion to their depth." The original definition has been supplemented by a number of differential characters observed by Newberry, and the extended description given by the American author has been condensed by Smith Woodward into the following para- graph : "A comparatively robust species as large as the type. Length of head with opercular apparatus not more than two-thirds as great as the maximum depth of the trunk, and contained nearly six times in the total length of the fish ; depth of caudal pedicle equaling about one-third that of the abdominal region. Cranial bones finely granulated. Pelvic fins arising midway between the pectorals and anal; dorsal and anal fins nearly equal in size, and the former arising opposite to the middle of the latter. Scales mostly smooth, but sometimes in part longitudinally striated, the striae terminating in the coarse serrations of the posterior border which characterize the principal flank-scales; many of the flank-scales deeper than broad." Neither in this nor in any other species of Catopterus has the structure of the head and shoulder-girdle been satisfactorily worked out, these parts being as a rule too imperfectly preserved 52 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. for study. Nevertheless, an attempt was made in this direction by J. S. Newberry, and it is perhaps worthy of note that the material upon which his restorations were based, together with certain unpublished figures and manuscript notes, are now the property of the American Museum of Natural History. For the privilege of studying both the original material and the records of Newberry's interpretation of them, the writer is indebted to the courtesy of his friend Dr. Bashford Dean, Curator of fossil fishes in the American Museum of Natural History in New York. With his permission, two of Newberry's original drawings are reproduced for the first time in Figs. 5 and 6, one rep- resenting the head portion viewed from above and to one side, the other from below. With reference to the specimens serving as the basis of New- berry's restorations it may be remarked that the larger and more perfect (the one shown in Fig. 5, Cat. No. 2431) has the head portion preserved in the form of an impression, wherein certain sutural indications are plainly visible, others less clearly so. Three drawings of this specimen occur among Newberry's reli- quiae, all bearing explanatory legends in his handwriting. From a careful collation of these with the original it appears that our author was mistaken in his reading of several parts of the cranial osteology, more particularly as regards the cheek plates and opercular apparatus, and it is a question whether he has not sometimes mistaken grooves of the sensory canal system for suture lines. But at the same time it must be admitted that precisely in these particulars, owing to lack of definiteness in the impression, there is room for considerable latitude of interpretation, and that after all a final judgment cannot be based upon this single specimen standing alone, without the aid of well authenticated points of control derived from comparison of a large quantity of material. No single specimen has yet come to light which reveals the lateral aspect of th» head in thoroughly satisfactory manner, and any attempt to correct or improve upon Newberry's restoration must proceed from a mosaic built up of overlapping sections. Much effort has been expended by the present writer in this direction, and some progress has been made towards elaborating the complete cranial structure. Yet the work is NO. 1 8.] TRIASSIC FISHES OF CONNECTICUT. 53 still incomplete, and, owing to deficiency of reliable material, the time has not yet come when a thoroughly satisfactory and authentic restoration can be given of the head. For the present, however, we may content ourselves with calling attention to the general Palseoniscid-like arrangement of the cranial plates, as far as the details have been worked out; and, in default of a tentative figure showing these parts, we may refer the reader to the different types portrayed in Fig. 7 on page 59, with special emphasis upon the approach made by Catopterns to the early and more primitive models. A few words may be said in regard to the second of New- berry's drawings, which has this in common with the first, and indeed with all pioneer studies; that, however we may judge of its accuracy, it is at least an interesting historical document, and has a certain intrinsic value in so far as it acquaints us with a graphic presentation of the author's views at the stage he had then attained in his investigations. In Fig. 6 is represented Professor Newberry's idea of the structure of the under side of the head. A comparison of his drawing with the original speci- men (Cat. No. 635 G) shows that the head is much distorted, the clavicle and infraclavicle being displaced far forwards, and thereby producing a very deceptive appearance. It may be stated positively that no median jugular plate is present, nor is any trace to be seen of the branchiostegal apparatus. The space included within the angle formed by the mandibular rami ap- pears to have been covered in part by rhombic ganoid scales, in part by indurated skin ornamented with papillae, but not oc- cupied by distinct plates. Both of Newberry's originals were obtained from near Durham, Connecticut. By way of summarizing the few definitely known facts that have been gleaned from a comparison of very numerous cranial fragments belonging to this species, the following points may be noted: The head is in general Palaeoniscid-like. There is a pair of small parietals behind, in front of which are placed the narrow and elongate frontals, traversed longitudinally by sensory canals ; and these are succeeded in turn by a median ethmoid of the form shown in Newberry's drawing (Fig. 5, e). This median system of plates is bounded on either side by three pairs of lateral plates which may be designated as the squamosal, post- frontal, and prefrontal. There is no circumorbital ring, and the 54 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. suborbitals are apparently few in number. The preoperculum is inconspicuous, and the posteriorly enlarged maxillary re- sembles or at least suggests in form that of Paleozoic Chon- if £.&« S^* .3 <- 'C o g O g S\ tenuiceps or ,S\ fultus which accompany it, it is distinguished from the former of these by the following differential characters, as was first pointed out by Newberry: the dorsal ridge-scales, which are usually depressed, are less strongly developed than in 5. 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 the base, curving gracefully backward to a point." Remains of this species are common to both New Jersey and New England, the locality near Sunderland, Massachusetts, having furnished a number of excellently preserved specimens, including the type of the so-called Ischypterus marshi. A photo- graph of this particular individual is reproduced in Plate I of the present Report, and in Plate II is shown the head portion of the instructive example which served as the basis of Dr. Eaton's restoration, published in 1905. The originals of both plates are preserved in the Peabody Museum of Yale University, and equally perfect and important material is to be found in the American Museum of Natural History at New York. Other interesting specimens are the property of Amherst College and Wesleyan University, respectively. As long ago as 1845, the distribution of this species was stated by J. H. Redfield to be as follows : " Occurs at Sunderland, Mass. ; Westfield and Middle- field, Conn. ; Pompton and Boonton, N. J."1 1 Quoted by Newberry in his Monograph on Triassic Fishes, 1888, p. 30. 64 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. Semionotus fultus (Agassiz). (Plate III.) 1836. Palceoniscus fultus L. Agassiz, Poiss. Foss., ii, pt. I, PP- 4, 43, pl- 8, figs. 4, 5. 1841. PalcBoniscus fultus W. C. Redfield, Ami. Journ. Sci., [i]xli, p. 25. 1 841. PalcBoniscus macro pterus W. C. Redfield, ibid., p. 25. 1847. Ischypterus fultus Sir P. G. Egerton, Quart. Journ. Geol. Soc, iii, p. 277. 1850. Ischypterus fultus Sir P. G. Egerton, ibid., vi, pp. 8, 10. 1877. Ischypterus fultus R. H. Traquair, ibid., xxxiii, P- 559- 1888. Ischypterus fultus J. S. Newberry, Monogr. U. S. Geol. Surv., xiv, p. 34, pl. 6, fig. 2; pl. 7, fig. 1. 1895. Semionotus fultus A. S. Woodward, Cat. Foss. Fishes Brit. Mus., pt. 3, p. 58. 1 901. Semionotus fultus E. Schellwien, Phys.-okon. Ge- sellsch. Konigsberg, p. 29, pl. 3, figs. 4 (?), 5. 1903. Semionotus fultus G. F. Eaton, Am. Journ. Sci., [4] xv, p. 261, pl. 5, figs. i-4.# 1905. Semionotus fultus C. R. Eastman, Ann. Rept. N. J. Geol. Surv. for 1904, p. 83, pl. 2, figs. 1-4; pl. 9. The synonymy given above is that adopted by most recent writers. The two species, S. fultus and 6". macropterus, were first united by J. H. Redfield in his paper presented before the American Association of Geologists and Naturalists in 1845, but were afterwards held by Newberry to be distinct on account of slight, and, as a matter of fact, inconstant differences in their body proportions. It is now commonly recognized that minor dif- ferences of this nature are the result of accidental conditions of preservation. Following are the chief diagnostic features of this species : D. 10; C. 15; A. 10; P. 10. A gracefully fusiform species attaining a total length to the base of the caudal fin of about 15 cm., in which the length of No. l8.] TRIASSIC FISHES OF CONNECTICUT. 65 the head and opercular apparatus is contained three and one- half times. The maximum depth of trunk, which is equal to about one-fourth of the total length, occurs midway between the head and dorsal fin, where there are about twenty longitudinal rows of scales. Scales of lateral line about thirty-three. Dorsal fin arising at mid-length, pectorals nearer to the anal than to the pelvic fins, arising opposite a point directly in advance of the dorsal. Caudal not much forked. Anal with ten rays, partly opposed to hinder half of the dorsal, its origin being on the third oblique scale-row in advance of the dorsal fin. Dorsal fin-fulcra about twelve ; anal ten ; ventral and pectoral ten each. Apparently four dorsal fin-fulcra originate on the dorsal margin over the interneurals. The fifth dorsal fulcrum has its origin adjacent to that of the first ray, and is about equal in length to one-half the anterior margin of the fin. Scales smooth and not serrated posteriorly, the deepest ones occurring in the fourth row behind the clavicular arch ; these are about twice as deep as they are wide in their exposed portion. Dorsal ridge-scales acuminate. As has been stated, the sole criterion relied upon by Newberry for maintaining the so-called 5. macropterus as an independent species consisted in a supposed relatively greater depth of body, — " the fusiform and slender fish standing for /. fultus, and the broader one for I. macropterus." Curiously enough, it has been shown by Dr. Eaton, after a study of Newberry's originals in the American Museum of Natural History, that, whereas one of the specimens of S. macropterus in its compressed and flattened condition is deeper than a type of S. fultus, all the others are proportionally more slender.1 J. H. Redfield, after advocating the suppression of the trivial title macropterus, remarks that 5*. fultus is specially characterized by the length of the dorsal and anal fins, which are even longer than in S. tenuiceps? In the New Jersey area1, this species outnumbers all others in abundance, and in the Connecticut Valley Trias it is scarcely inferior in numerical importance to the ubiquitous 5. tenuiceps. The average length of body is stated by Newberry to be about six inches, the maximum rarely exceeding eight inches, including the 1 hoc. cit., 1903, p. 262. 2 Cited by Newberry, 1888, p. 35. 5 66 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. caudal fin. In Plate III is given a photographic reproduction of one of the original specimens serving for Newberry's description. Semionotus tenuiceps (Agassiz). 1836. Eurynotus tenuiceps L. Agassiz, Poiss. Foss., ii, pt. 1, pp. 159, 203, pi. 14c, figs. 3, 4, 5. 1837. Palceoniscus latus J. H. Redfield, Ann. Lyceum Nat. Hist. N. Y., iv, p. 38, pi. 2. 1837. Eurynotus tenuiceps J. H. Redfield, ibid., p. 39. 1841. Eurynotus tenuiceps E. Hitchcock, Geol. Mass., ii, p. 459, pi. 29, figs. 1, 2. 1841. Palceoniscus latus W. C. Redfield, Am. Journ. Sci., [1] xli, p. 25. 1850. Ischypterus latus Sir P. G. Egerton, Quart. Journ. Geol. Soc, vi, p. 10. 1857. Eurinotus ceratocephalus E. Emmons, Am. Geol., pt. 6, p. 144, pi. 9a. i860. Eurinotus ceratocephalus E. Emmons, Manual Geol., 2d ed., p. 188, fig. 164. 1877. Ischypterus latus R. H. Traquair, Quart. Journ. Geol. Soc, xxxiii, p. 559. 1888. Ischypterus tenuiceps J. S. Newberry, Monogr. U. S. Geol. Surv., xiv, p. 32, pi. 5, figs. 1-3, pi. 7, fig. 3. 1889. Allolepidotus americanus W. Deecke, Palaeontogr., xxxv, pp. 103, 114. 1895. Semionotus tenuiceps A. S. Woodward, Cat. Foss. Fishes Brit. Mus., pt. 3, p. 59. 1903. Semionotus tenuiceps G. F. Eaton, Am. Journ. Sci., [4] xv, p. 295. 1905. Semionotus tenuiceps C. R. Eastman, Ann. Rept. N. J. Geol. Surv. for 1904, p. 87. A species attaining a total length of about 20 cm., and readily distinguished from all others (except in young stages) by the excessive development of the dorsal ridge-scales ; these are very large and conspicuous, and, in mature individuals, comparatively obtuse. The anterior dorsal outline is considerably arched, usually forming a characteristic " hump " immediately behind the head. Length of head and opercular apparatus less than the No. l8.] TRIASSIC FISHES OF CONNECTICUT. 67 maximum depth of the trunk, and contained four times in the total length of the fish. Fins as in S. fultus. Scales smooth and serrated, those of the middle of the flank in part twice as deep as broad. The dorsal ridge-scale immediately in advance of the dorsal fin has its posterior border obtuse, and not pro- duced, and the corresponding ridge-scale in front of the anal fin is notched behind. Ribs more strongly developed than in any other species of the genus. This, the commonest form occurring within the Connecticut Valley area, is as a rule easily determinable, its most conspicuous features being the abrupt elevation of the dorsal outline im- mediately behind the head, and the spiny appearance of the back occasioned by its being set along the middle with long, thickened, and distally pointed or clavate ridge-scales. The ribs also are more strongly developed than in other species, their curved outlines being sometimes traceable even when covered with scales. Owing to the frequency with which this species has been illustrated, and the impossibility of mistaking it among collections of Triassic fishes, it has not been deemed essential to include a figure of it in the present Report. S. tenuiceps outnumbers all other species in the Connecticut Valley Trias, and is tolerably abundant also in New Jersey. At Turner's Falls and at Sunderland, Massachusetts, it is especially common, probably more than half of the individuals derived from the latter locality pertaining to this form. Semionotus micropterus (Newberry). (Plate IV.) 1888. Ischypterus micropterus J. S. Newberry, Trans. N. Y. Acad. Sci., vi, p. 127 (name only). 1888. Ischypterus micropterus J. S. Newberry, Monogr. U. S. Geol. Surv., xiv, p. 31, pi. 4, %s. 1, 2; pi. 12, fig. 2. 1893. Ischypterus newberryi S. W. Loper, Pop. Sci. News, March 18, and Pop. Science, May, 1899, p. 98. 1903. Semionotus micropterus G. F. Eaton, Amer. Journ. Sci., [4] xv, p. 263, pi. 5, figs. 6-8, 11, 13. 1905. Semionotus micropterus C. R. Eastman, Ann. Rept. N. J. Geol. Surv. for 1904, p. 87, pi. 2, figs. 6-8. 68 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. D. 8; C. 15; Ai 8. A regularly fusiform species attaining a total length to the base of the caudal fin of about 20 cm., the maximum depth oc- curring shortly behind the pectoral fins and amounting sometimes to nearly one-half the total length. The dorsal and ventral con- tours are more strongly arched than in S. fultus, but the relative position and size of the fins are about the same for both species. Dorsal, anal, and pectoral fin-fulcra relatively shorter than in S. fultus. Apparently three dorsal fin-fulcra originate on the dorsal line over the interneurals. The fifth dorsal fulcrum has its origin on the anterior margin of the anterior ray at a considerable distance from its base, and is about one-third as long as the anterior fin-margin. Pectorals with upwards of twenty fulcra. Ridge-scales moderate, spiniform, the one immediately in advance of the dorsal fin slightly produced into a point behind. Scales frequently serrated, those below the lateral line on the flanks tending to become bi- or tridentate on the postero-inferior angle. This is a deeper-bodied species than any thus far considered, its rather strongly convex outline marking a transition between the types presented by S. fultus, for instance, and 5. ovatus. Indeed, the approach to the last-named species in this respect is sometimes so close as to make a rigid distinction difficult without the aid of other characters. In the case of the specimen selected for illustration in Plate IV, Newberry himself appears to have been in doubt whether to refer it to 5\ micro pterus or 6\ ovatus, but finally decided in favor of the former, as shown by MS. records accompanying the original. The most reliable means for identifying the present species is furnished by fin characters, the details of which have been carefully worked out by Dr. Eaton in his paper of 1903, and are incorporated in the above definition. Remains of this species are fairly abundant in the Connecticut Valley Trias, and show considerable variation of size; that is to say, young individuals occur somewhat numerously, so that gradations may be traced up to the maximum recorded by New- berry. He states that the largest individuals known to him attain a length of ten and one-half inches, and the smallest are " only about three and one-half inches long." Corresponding with the last given dimension, and otherwise agreeing with the char- acters of this species, is the holotype of the late Mr. S. Ward No. l8.] TRIASSIC FISHES OF CONNECTICUT. 69 Loper's Ischypterus newberryi. The peculiarity to which Mr. Loper has called attention, namely, fine concentric scale-mark- ings, seems to have been occasioned by some form of chemical corrosion which has exposed the growth lines. A parallel in- stance has already been noticed in the case of Catopterus ornatus (supra, p. 55), and similar conditions are prevalent among fishes of the English Chalk. The original of Mr. Loper's description is of interest for showing a well preserved mandible beset with numerous slender teeth, and a very strongly developed support for the dorsal and anal fins. It is preserved in the Museum of Wesleyan University, and was obtained by Mr. Loper from the anterior shale near North Guilford, Connecticut. This species is not known to occur elsewhere than within the state of Connecticut, and is especially abundant in the vicinity of Durham. It is possible that the detached head figured by Dr. E. Schellwien in Plate 3, Fig. 4 of his memoir above cited belongs to 6". micropterus, since this is one of the few species in which the cheek plates are granulated. Semionotus ovatus (W. C. Redfield). 1842. Palaoniscus ovatus W. C. Redfield, Am. Journ. Sci., [1] «li, p. 26. 1847. ( ?) Tetragonolepis Sir P. G. Egerton, Quart, Journ. Geol. Soc, iii, p. 277. 1850. Ischypterus ovatus Sir P. G. Egerton, op. cit. vi, p. 10. 1888. Palceoniscus ovatus J. H. Redfield (quoted by New- berry), Monogr. U. S. Geol. Surv., xiv, p. 27. 1888. Ischypterus ovatus J. S. Newberry, loc. cit, p. 27, pi. i, fig. 1. 1903. Semionotus ovatus G. F. Eaton, Am. Journ. Sci., [4] xv, p. 266. 1905. Semionotus ovatus C. R. Eastman, Ann. Rept. N. J. Geol. Surv. for 1904, p. 78, pi. 4-6. A large species attaining a total length of about 20 cm., with trunk very much deepened midway between the head and dorsal fin. Scales large and thick, becoming gradually deepened toward the middle of the flanks ; tail strong and considerably expanded. Number of dorsal and anal fin-fulcra greater than in any other JO CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. species, each fin having sometimes twenty or more. Length of the longest fulcrum of the dorsal fin nearly equalling one-half that of the anterior margin of the fin. In the original description of 5. ovatus by William C. Red- field, it is stated that " it 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 occur- rence, and, owing probably to its great thickness, is seldom ob- tained in a perfect form."' The younger Redfield, commenting on the same species in 1854, pronounced it " the broadest and most ovate species of Palaoniscus that is known," and added further, that " in size of the scales it resembles P. Agassizii, but its form will readily distinguish it." That is to say, the squama- tion is heavy, but the flank-scales are relatively deeper than in S. agassizii, and the form is also deeper-bodied, or more ovate. The Redfields, father and son, and also Newberry, agree in claiming for this species a distribution in both the Connecticut Valley and New Jersey Triassic basins. With this species New- berry also identifies a fragmentary individual from the Triassic Coal-field of Virginia, originally referred to Tetragonolepis by Sir Philip Grey Egerton. Noteworthy is the fact that all the more perfect examples have been obtained from a single locality near Boonton, New Jersey, and the recognition of this species in outlying areas depends upon the evidence of unsatisfactory material. The present writer has thus far failed to discover a single undoubted example of the species in question from the Connecticut Valley Trias, yet this is by no means equivalent to saying that its remains do not occur in this region. It may per- haps be worth mentioning that in the Museum of Wesleyan University is preserved the anterior half of a deep-bodied fish (Cat. No. 869) whose specific relations cannot be accurately determined. It is labeled as belonging to S. gigas, a " species " which can be maintained only in a provisional sense. The so- called Semionotus robustus of Newberry is but little better known, and is doubtfully distinct from 5. ovatus, which it ap- proximates in size. A certain resemblance between the published figure of S. robustus and the imperfect deep-bodied specimen at Wesleyan University just referred to cannot be denied. Further evidence, however, is necessary before we can positively affirm the presence of S. ovatus in the New England area. No. l8.] TRIASSIC FISHES OF CONNECTICUT. J\ Extra-limital Species of Semionotus. At least three other valid species of Semionotus, besides those already enumerated, have been described from the Trias of Eastern North America. These are, S. lineatus, elegans, and brauni of Newberry. They are all confined to the New Jersey area, so far as known, and the last-named is from the very base of the Trias in that state, being separated from the Boonton horizon by an interval of several thousand feet. The limits set to the present Report do not admit of elucidating the characters of these species, which can by no possibility be confused with the members of our local fauna. Nevertheless, it has been thought desirable to offer an illustration of the form which has been appropriately named .S\ elegans by Newberry (Plate V), and also to show the head-portion of the type specimen of 5". nilssoni (Plate VI), which enabled Agassiz to decipher the main elements of the cranial structure of this genus. To the list of imperfectly defined or doubtful species, the status of which is merely provisional, must be added the names of the so-called Ischypterus parvus, founded upon a figure pub- lished in Hitchcock's Geology of Massachusetts, in 1835 ; Ischypterus minutus Newberry, from Durham, Connecticut; and Ischypterus beardmorei Smith, from Boonton, New Jersey. Of uncertain position also are the obseure remains of a Semionotus- like form described by Newberry under the name of Acentro- phorus chicopensis, the few known examples of which have been obtained from metamorphosed sandy shales near Chicopee Falls, Massachusetts. It will be convenient to notice at this point the status of an imperfectly known European form, described in the first instance by Deecke as a species of Semionotus, and recently made the type of a distinct genus {Perleidus) by De-Alessandri, who places it in association with the Catopteridas. The type species, P. altolepis (Deecke), occurs in the Ladinian beds of Perledo, Lombardy, and the original specimen upon which it is founded is preserved in the Senckenbergian Museum at Frankfurt. Deecke, in describing the species, remarked that it appeared to him to denote a transitional stage between the genera Semionotus and Pholidophorus. Schellwien, who later examined the speci- men, doubted whether it could properly be included in the genus -7W-3 72 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. Semionotus, but did not attempt to fix its systematic position more precisely. The reasons which in Professor De-Alessandri's opinion justify a removal of this species, and with it the new genus Perleidus, to the group of Catopterids, are enumerated by this author as follows : " The arrangement of the cranial elements, the form of the maxilla, especially its expanded posterior portion, the absence of suborbital plates, and the presence of a large-sized postorbital, compel an assignment of this form to the family of Catopteridae, and make it necessary for us to regard it as the type of a new genus. Moreover, the position and form of the fins, the rather feeble fulcra, the configuration of the scales with their strongly denticulated posterior border, and the absence of a series of acuminate dorsal ridge-scales, are characters which warrant a separation from the genus Semionotus." x The new genus Perleidus is thus diagnosed by its founder: " Trunk elongate-fusiform, and head relatively small. Super- ficial ornament of the cranial plates consisting of rather fine tuberculations and rugae. Maxilla extended, and posteriorly en- larged. A series of circumorbitals present, but no suborbitals; one large postorbital plate present. Fins moderately developed, comprising robust articulated rays; fulcra small. Dorsal fin situated opposite the pelvic pair ; caudal slightly forked. Scales rhomboidal, deeper than long, smooth on their exposed portion, their posterior border denticulated."2 That the above-mentioned genus is well characterized there can be no doubt, and the reasons for excluding it from association with Semionotus are sufficiently valid. It must be admitted, however, that the form in question presents considerable resem- blance to Pholidophorus, and the position of the dorsal fin, which arises in advance of the anal, offers a marked contrast to the condition observed in the family Catopteridse, from which con- dition indeed is derived the name of the typical genus. We prefer to accept the Milanese author's determination of the family position of this genus in a provisional sense, rather than to assign it elsewhere without having had opportunity to study the actual specimens. 1 Studii sui Pesci Triasici della Lombardia. Mem. Soc. Ital. Sci. Nat., 1910, vii, p. Si- 2 hoc. cit., p. 49. NO. l8.] TRIASSIC FISHES OF CONNECTICUT. 73 Family EUGNATHID^E. Trunk fusiform or elongate, not much laterally compressed. Cranial and facial bones moderately robust, externally enameled, and opercular apparatus complete; gape of mouth wide, snout produced, marginal teeth conical and larger than the inner teeth. Fin-rays robust, articulated, and distally divided; fulcra con- spicuous. Dorsal fin short and acuminate. Scales rhombic, sometimes with rounded posterior angles. Genus Ptycholepis Agassiz. Trunk elegantly fusiform; snout acutely pointed and promi- nent; external bones highly ornamented with prominent waved ridges; marginal teeth very small and regular; dorsal fin in advance of anal, caudal fin forked ; scales all narrow and elongate, marked with deep longitudinal grooves. Fulcra biserial, con- spicuous on all the fins excepting the dorsal. Ptycholepis marshi Newberry. (Plates VII, VIII.) 1878. Ptycholepis marshi J. S. Newberry, Ann. N. Y. Acad. Sci., i, p. 127. 1888. Ptycholepis marshi J. S. Newberry, Monogr. U. S. Geol. Surv., xiv, p. 66, pi. 19, figs. 1, 2. 1895. Ptycholepis marshi A. S. Woodward, Cat. Foss. Fishes Brit. Mus., pt. 3, p. 324. 1905. Ptycholepis marshi C. R. Eastman, Ann. Rept. N. J. Geol. Surv. for 1904, p. 100. 1908. Ptycholepis marshi L. Hussakof, Bull. Am. Mus. Nat. Hist., xxv, p. 95. A species of slender proportions, attaining a length of about 20 cm. Head with opercular apparatus occupying somewhat more than one-fourth the total length of the fish. Ornamental rugse of cranial roof slightly radiating; those of the facial and opercular plates more or less parallel and forked. Dorsal fin far forwards, and pelvic fins arising opposite its hinder extremity. Scales exhibiting only longitudinal ridges and furrows, and the hinder border often deeply serrated. (Woodward.) 74 CONNECTICUT GEOL. AND NAT. HIST. SURVEY. [Bull. This gracefully formed and elaborately ornamented species is known by a dozen or more examples, all derived from a single locality near Durham, Connecticut. Among these are several excellently preserved individuals, including the type shown in Plate VII, material which might be expected under ordinary cir- cumstances to yield valuable enlightenment concerning cranial structure. Progress in this direction, however, is subject to the limitations imposed by the peculiar nature of the head bones themselves : that is to say, by the highly sculptured and heavily enameled outer surface which completely conceals suture lines. It is nevertheless permissible to draw certain inferences con- cerning the extent and arrangement of plates forming the cranial roof by noting the centers of radiation and territory traversed by the superficial radiating rugse; and the general pattern thus revealed has been found to agree with typical Eugnathidae. The dorsal aspect of the cranial roof, together with some of the facial bones and opercula, is favorably exposed for study in the specimen represented in Plate VIII, the original being preserved in the Yale Museum (Cat. No. 2608). The lateral aspect is even more favorably shown in the original of Plate VII, which is the property of Wesleyan University Museum (Cat. No. 907). This example, though of a young individual, is admirable for its presentation of fin and scale structure, and for showing the normal body contour. In connection with the distribution of this form, it should be recalled that its accompaniment by Semionotus, Catopterus and a Crossopterygian genus (Diplurus) is a fact of capital im- portance in assigning the fauna in question to a horizon equivalent to the Upper Muschelkalk and Lower Keuper of the European marine Trias. All the evidence derived from a study of the fossil fishes is in favor of establishing a correlation at a level embracing these two horizons, but probably not extending higher than the basal division of the Keuper in the Mediterranean region.1 For a recent review of the evidence for establishing an 1 That is, the Newark fauna cannot be regarded as younger than the faunas of Besano, Lombardy, and of Raibl, Carinthia (Lower Alpine Keuper), which mark the uppermost range of one of the intercommunal genera Ptycholepis. The Upper Muschel- kalk (Ladinian) terms of comparison are furnished by two Semionotid genera, one Captopterid, and one Crossopterygian, according to the revised determinations of Professor G. De-Alessandri (iqio). No. l8.] TRIASSIC FISHES OF CONNECTICUT. 75 inter-regional correlation of the Trias, based upon another class of remains than fossil fishes, we may be permitted to refer at this point to Dr. J. C. Merriam's elaborate memoir on "Triassic Ichthyosauria, with special reference to American forms " (es- pecially the chapter on Geologic and Geographic occurrence, pp. 12-20) .2 The evidence as to the age of the Triassic forma- tion of eastern North America which is furnished by reptilian remains (i. e., numerous footprints and a few skeletons of Dino- saurs) will be discussed in a forthcoming Bulletin of the Con- necticut State Survey by Professor R. S. Lull of Yale University. In conclusion, the writer of the present article desires to acknowledge his indebtedness and at the same time return hearty thanks to the following named friends and colleagues who have shown him many courtesies and placed numerous facilities at his disposal, thereby greatly aiding the preparation of this Report: Professor William North Rice and the late Mr. S. W. Loper of Wesleyan University; Professor Charles Schuchert and Dr. George F. Eaton of Yale; Professor B. K. Emerson and F. B. Loomis of Amherst; Professor Bashford Dean and Dr. E. O. Hovey of the American Museum of Natural History, New York ; and the authorities of the U. S. National Museum at Washington. 2 Memoirs of the Univ. of California, 1908, i, no. 1, pp. 1-196, pi. 1-18. Index of Genera and Species. Acanthodes sulcatus, 15. gracilis, 15. Acentrophorus chicopensis, 71. Acrodus, 34. Allolepidotus. americanus, 27, 66. Amia, 18, 47, 57. Asteracanthus, 34. Calamoichthys, 16. Catopterus, 32, 47, 74. gracilis, 48. macrurus, 55. minor, 48, 55. ornatus, 48, 55. parvulus, 48. redfieldi, 48, 51, 54. Cephalaspis murchisoni, 14. Ceratites trinodosus, 24. Cheirolepis, 46. Chondrosteus, 47. Cladoselache fyleri, 16. Climatius scutiger, 15. Ccelacanthus, 43. welleri, 44. Cosmacanthus, 34. Dapedius, 60. Dictyopyge, 55. macrura, 55. Diplurus, 31, 44, 74. longicaudatus, 44, 74. Dipterus valenciennesi, 17. Ecphora quadricostata, 39. Eurynotus ceratocephalus, 66. tenuiceps, 66. Heptanema paradoxum, 31, 44. Heterolepidotus, 27. Holoptychius, 34. Hybodus, 30, 34. Ischypterus, 24, 57, 61. agassizii, 61. beardmorei, 71. fultus, 64, 65. latus, 66 macropterus, 64. marshi, 61. micropterus, 67. minutus, 71. newberryi, 67, 69. parvus, 71. ovatus, 69. tenuiceps, 66. Lepidosteus, 18, 47, 57. Lepidotus, 60. Leptolepis, 18. Macropoma, 43. Megalonyx jeffersoni, 39. Mesacanthus mitchelli, 15. Nematoptychius, 59. Palseoniscus, 59. agassizii, 61. fultus, 64. latus, 66. macropterus, 64. ovatus, 69. Perleidus, 46, 47, 60, 72. altolepis, 27, 71. Pholidophorus, 18, 35, 71, 72. Polypterus, 16. Protopterus, 18. Ptycholepis, 31, 73, 74. marshi, 73. Redfieldius, 47. Rhabdolepis, 59. Rhadinichthys, 59. Sagenodus, 18. Semionotus, 24, 40, 57, 74. agassizii, 61, 62. bergeri, 58. brauni 71. capensis, 58. elegans, 71. fultus, 64. gigas, 70. lineatus, 71. marshi, 61. micropterus, 67. nilssoni, 58, 71. ovatus, 69. robustus, 70. tenuiceps, 66. Tetragonolepis, 69, 70. Undina, 43. Urolepis, 32. Xenestes, 34. Hi H < -J w < Oh XI O < w < -J w < Plate VI. Setm'onotus nilssoni Agassiz. Rhaetic; Hoegenaes, Sweden. Head portion of holotype, showing cra- nial plates and dentition. Original in Museum of Comparative Zoology, Cambridge, Massachusetts (Cat. No. 2685). X \. w H < -J to < X pa < ft* X w < a o p o +J