<<

288 A. SMITH WOODWARD ON THE

THE HISTORY OF FOSSIL CROCODILEB.

By ARTHUR SMITH WOODWARD, F.G.S., Assistant in the Geological Department, British Museum (Natural History.)

1.-INTRODUCTION.

The rapid progress of every department of scientific research, and the unfortunate scattering of the various contributions to the same subject in the different journals and periodicals throughout the world. renders the task of the compiler and historian of no small importance, " Stock-taking" in science is as important in its own way as is the same procedure in commercial undertakings, and occasional reviews of the present state of any particular branch of inquiry are of the greatest value in indicating the most promising paths for further traverses, and the most profitable domains for future conquest. The abstruse nature of much of the literature, too, and the out-of­ the-way character of most of the older, and even some of the newer, works and memoirs, make it often advisable to possess a less de­ tailed and more accessible account of what they contain; and it is the remembrance of such considerations as these that has induced the writer to offer the present communication as a slight attempt towards providing, in the form of a connected narrative, an outline of what is already known concerning one small group of fossil ver­ tebrate -that of the Crocodilia. It is proposed to enumerate, in order, the more important extinct types of this ancient reptilian class whose remains are met with in the various successive formations, and, while referring more par­ ticularly to those discovered in British deposits, to make usc of all available materials to elucidate the developmental history of the race. And if, at times, more attention is given to fragmentary fossils than their small importance may be deemed to justify, we plead, as an excuse, that such relics frequently fall into the hands of collectors, and the desultory notices of them to be found in ordi­ nary text-books are usually much too brief and indefinite to satisfy the curiosity of those who pursue their collecting in a thoroughly philosophical manner. HISTORY OF FOSSIL CROCODILES. 289

n.-RECENT CROCODILES.

To render the fossils intelligible, however, and to bring out their features of greatest interest, it is necessary, in the first place, to take a general glance at the tribes of Crocodiles as they exist in the liv­ ing fauna of the globe;* or, in the words of modern biology, to consider the culmination of the Crocodilian" phylum" as it is to be observed at the present day. With regard to the number of species still surviving, it is almost impossible to make any definite statement, on account of the great differenees of opinion that prevail as to what really constitute specific characters; and the same remark applies to the generic divisions; t but some or other members of the class are still to be met with in nearly all tropical and subtropical regions of the earth's surface t-notwithstanding their rapidly approaching in many parts through the interference of man-and there are at least two, if not three, well-marked types or families still to be distin­ guished. There are the Gavials or Gharials specially adapted for an aquatic existence; and the Crocodiles and Alligators, more accus­ tomed to an amphibious mode of life. According to Dr. Gray, the true Gavials (Gharials) are at present exclusively confined to the rivers of the interior of India, and are there found only in the lower regions where the climate is hot. Their most prominent and striking superficial peculiarity con­ sists in the great elongation and slenderness of the snout (Fig. 1), which is armed with a numerous series of sharp, equal-sized teeth, admirably adapted for the capture of slippery prey; this peculiarity,

* For good illustrations of the osteology of living Crocodiles, the follow­ ing works may be consulted :-G. Ouvier, 'Ossemens Fossiles;' De Blain­ ville, 'Osteographie j' d'Alton and Burmeister, 'Del' Fossile Gavial von Boll,' Halle, 1854 j' and" Synopsis," by J. E. Gray, in 'Traus. Zool. Soc., Vol. vi, quoted below. t T. H. Huxley, "Notes on the Specific and Generic Characters of Recent Crocodilia."-' Proc. Linn. Soc.,' Vol. iv (Zoology), 1860, p. 1; J. E. Gray, "Synopsis of the Species of Recent Crocodilians or Emydosaurians."­ 'Trans. Zool. Soc.,' Vol. vi (1862), pp. 125.169, pls. xxxi- xxxiv ; J. E. Gray, " On the Change of Form of the Heads of Orocodiles."-' Brit. Assoc. Re­ ports,' 1862, p. 109; J. E. Gray, "Catalogue of Shield in the British Museum," Part 2, 1872. t For the geographical distribution of recent Crocodilia, see A. Strauch, "Synopsis del' gegenwart.ig lebenden Crocodiliden."-' Mem. Acad, Imp. Sci. St. Petersbourg ' [7J, Vol. x, No. 13 (1866). 290 A. SMITH WOODWARD ON THE in fact, being one distinguishing the Gavial family from those of the Crocodiles and Alligators, and the other most essential distinctions being also found in tbe same parts. In the lower jaw of the Gavial (Fig. 1), for example, the two halves (or "rami") are united together by a very long suture, and scarcely any of the teeth arc fixed to the jaw behind this" symphysial area;" again, the two bones that contribute largely towards roofing in the front part of the nasal cavity in most vertebrates, and hence called" nasals," are fixed quite at the commencement of the snout, and do not reach the external nostril at all; and there are also some other minor features that may be equally looked upon as distinctive.

2.

FIG. 1. Mandible of Gavialis gangeticus. FIG. 2. Mandible of Crocodilus palustris.

Turning next to the true Crocodiles, we find their geographical range much more extensive; they occur both in India, Africa, and the tropical parts of America, and at least two species, also, have their haunts in the rivers and lakes of Australia. They are easily distinguished from the Gavials by the form and character of the skull (Figs. 3, 9) ; the jaws are relatively shorter, and gently rounded in front, while the edges, instead of being straight, have a " fes- HISTORY OF FOSSIL CROCODILES. 291 tooned" arrangement, and the powerful teeth are much more unequal in size. The nasal bones, too, reach-or closely approach -the hinder border of the external nostril, and the two rami of the lower jaw (Fig. 2) are only united in front for a very short dis­ tance. The form and proportions of the head, in fact, are more adapted for the seizure of great mammalian quadrupeds than for the slippery fishes upon which the Gavial feeds. The Alligators-with the exception of one lately discovered in. China-are exclusively confined to America, and the single species inhabiting the Mississippi is particularly noteworthy on account of its being occasionally met with in the oft-frozen regions of the North. Two other genera, occurring chiefly in the tropics of the Southern Continent, are also of great interest to the palreontologist, because, of still living forms, they alone are completely encased all round in the bony armour that formed such a conspicuous feature in the majority of extinct Crocodilian genera. On a superficial glance, it is by no means easy to distinguish the members of this family from the true Crocodiles, and, as has already been remarked, some zoologists are inclined to unite them together much more closely than has long' been customary. The Alligators differ, how­ ever, in the circumstance that the fourth tooth of the lower jaw on each side-often called the" canine," from its great relative size­ is received in a socket in the upper jaw, while the corresponding tooth of the Crocodile simply fits in a groove and is visible when the mouth is closed. The dentition of the Alligators is further characteristic from the peculiarity of the lower teeth passing on the inner side of the upper teeth in biting, instead of fitting into the interspaces of the upper row. Summarising the more important of these differential char­ acters," it may be convenient to have them arranged in tabular form, and the following brief comparative statement will, perhaps, facilitate future reference. (See next page.)

* T. H. Huxley, "Notes on the Specific and Generic Characters of Recent Crocodilia."-' Proo, Linn. Soc.,' Vol. iv (Zoology), 1860, p.T, 292 A. SMITH WOODWARD ON THE

CROCODI L ID.E. GAVIAL I D.E . ALLIGATORI D.E . C n o c o Dl LI D.-E PROP ER.

Skull short and broad. Skull more or less Snout very long and short and broad. slender.

Nas al bones reach ing Nasal bones r eaching Nas al bones short, and external nostr il, external nostril, or not r eaching the pre­ sometimes di viding (at least) the pre­ maxillre (except in it. maxillas. Tomist oma),

Teeth stout, and vari­ Teeth stout, and more Teeth slen de r, shar p' able in size. or less var ia ble in edged, a nd subequal. size.

Anterior pair of m an­ Anter ior pa ir of man­ An terior pair of man­ di bular teeth, and dibular teeth re ­ di bul ar teeth, and fourth pai r, receivcd ceived in pits of four th pair, pass in. in pit s in the mar­ pramaxillas, but to gr ooves of pre' g ins of the premax­ fourth pair passing maxillm. Illeoand maxillee, only in to grooves a t junction of pre . maxillas wit h max. Illes,

Mandibular teeth be. Mandihular teeth be­ Mandibular teeth be­ hind the fourth pass hi nd th e fourth pass hind the fourth pass inside the maxillary between the maxil­ bet ween the maxil­ t eeth. lary t eeth. lary t eeth.

Mandibul ar symph y­ Mandibular sy mphy­ Mandibular symphysis sis shor t, and den ti­ sis short, and den­ very long, splenial tion exte nding con­ tition extendi ng bones ente ring into sidera bly behind the considerably be­ it, and dentition al, symphysial area. hind the symphy­ most confined to sial area. symphysial area.

Ventral scutes some­ Ven tral scutes absen t. Ventral scutes absent. times present.

It must be remembered, however, th at th ere still survive some curious intermediate forms, which-as always happens in natural history-interpose to create difficulties in our would-be simple classifications, and among fossil Crocodilia there are also many oth ers of a similarly trouble some nature. The living "false Gavial" of Born eo, for exampl e, seems quite on the boun­ daries of the Crocodiles and Gavials, and rejoices in no less than five different generic title s,* variously employed according to

... Tomistoma, Mecistops, RflY " CflOSllCflUS, Gavialis, and Croeodilus, HISTORY OF FOSSIL CROCODILES. 293 the special views of each individual author; and certain Eocene fossils from Hordwell (Hampshire) appear to show that the pecu­ liarities of the Alligator, likewise, are not so distinctive as some­ times imagined. But we have omitted, hitherto, to determine precisely what con­ stitutes a Crocodilian. Such a definition is of all the more im­ portance to us, as palreontologists, since it often happens that fossil animals, when of any considerable antiquity, are of a remark­ ably generalised type, and exhibit a most cnrious mixture of pecu­ liarities that are now only found in separate families or orders. It is necessary, therefore, to be able to decide when a certain set ot features so predominate as to justify the reference of these primi­ tive composite forms to any particnlar group, and we shall next proceed accordingly to take a brief glance at the more important points in the organization of living Crocodiles so far as concerns their skeletal structures. Commencing with the skull,· and taking the common Indian " Muggar" (Fig. St) as typical of the order, the mode of attach­ ment of the lower jaw is first worthy of notice. Like all other reptiles and birds, this creature ensures as wide a gape as possible by using as a prop for its mandible on each side a bone which we (and the other mammalia) employ in a different manner to form part of the mechanism of hearing;:j: but whereas this" quadrate,"

.. For a detailed description of the cranial osteology, see an admirable little work by Professor L. C. Miall, entitled, ' Studies in Comparative Anatomy.-No. I. The Skull of the Crocodile.' (Macmillan and Co., ]878.) We may also add that some slight differences between the nomenclature employed by Owen and that of Huxley and others, render the various accounts of fossil skulls at first rather confusing; it should be remembered that the term tympanic (Owen) is equivalent to quadrate (Huxley and others) mastoid " squamosal "" squamosal" "" quadrato-jugal " alisphenoid " prootic" orbitosphenoid " alisphenoid " t For the loan of this woodcut, and also Figs. 6, 7, 8, 9, 10, we arc indebted to the Council of the Geological Society and Sir Richard Owen. The others have been drawn by Mr. W. H. Hill, a Member of the Associa­ tion. t This is the commonly accepted interpretation; see, however, P. Albrecht, "Sur la valeur morphologique de I'artioulation mandibulaire, du cartilage de Meckel, et des Osselets de l'oute " (Brussels, 1883), and L. Dollo " On the Malleus of the Lacertilia, and the Malar and Quadrate Bones of Mammalia,"-' Quart. Journ. Micros. Sci.,' 1883, p. 579. 294 ASMITII WOODW ARD ON THE as it is term ed, is loosely atta ched to the skull in most other living members of the reptilian class, in the Crocodilia it is firmly united. All the cranial bones, in fact, have very firm sutural connections. The posterior aperture of the nostril s is also very peculiar; for instead of opening about the middle of the roof of the mouth, plates of bone prolong the canal backwards almost as far as the

FIG. 3. Skull of Crocodilus palustris, I, side view ; 2, upper view; 3, under view; 0, orbits; T, supra -temporal fossre; P, palato-pter ygoid vacuity ; N, posterior nares; E, aperture of eustachian canal. junction of the neck-vertebrrewith the skull ; but the fossil forms ex­ hibit such interesting deviations from this arrangement that further particulars may be reserved until those forms are considered. The teeth, again, are different from those of other living reptiles, since they are all implanted in distinct sockets, and not anchylosed with th e jaw; and as long as the lives there is a constant HISTORY OF FOSSIL OROCODILES. 295 dental succession, the new teeth rising beneath and pushing out the old ones.* Other peculiarities, common to the skulls of all living Croco­ diles, are the absence of any perforation in the "parietal" bone on the top, and the undivided character of the external nasal aperture, which is placed towards the end of the snout. With regard to th e backbone, all the existing members of the order agree in having most of the vertebras " proccelian "-that is, in having the component bones united by cup-and-ball joints, the hollows being anterior. The only exceptions to this arrangement occur in the "atlas" and" axis," which are specially adapted for their connection with the skull j in the two "sacral" vertebrre which require a particularly firm union, from their support of the hind limbs j and in the first tail-vertebra, which has a ball at each end to connect the posterior hollow of the second sacral with the normally placed anterior hollow of the following tail-vertebra. But while some other living reptiles have the majority of the vertebree procealian, and are also characterised by a sacrum con­ sisting onlyof two vertebrre,the manner in which theanteriorribs are fastened to the backbone in the Crocodiles is quite different from anything to be observed among other existing orders, and can only be paralleled by referring to extinct types of reptilian life. Each rib bifurcates more or less at its upper end, and has thus a double articulation, or, to use technical terms, has a distinct head and tubercle. Turning next to the appendages, we find only two bones on each side at the origin of the fore limbs, there being no trace of the collar bones (or " clavicles "), The upper of the two bones is the shoulder blade (" scapula "), and the lower the " coracoid "-the latter apparently having no vernacular name, perhaps because it occurs only as a process of the shoulder-blade in the more familiar higher animals, the Mammalia (except in the Monotremes), where it would escape notice except by a comparative anatomist. Into the socket formed by the meeting of these two bones there fits, as usual, the head of the upper arm-bone (" humerus "), which has a peculiar twisted form with very little expansion of the ends, and then follow the two components of the fore-arm (" radius " and " ulna ")

* See R. Owen, • Odontography,' Vol. i, pp. 293.295, Vol. ii, pl, 75 j also R. Owen, ' Paleeontology,' 2nd edit. (1861), pp. 301.303. 296 A. SMITH WOODWARD ON THE the latter without any prominent elbow process; and to these are added the bones of the wrist (" carpus") and the five digits of the foot.

The supporting bones of the hind limb (forming the I( pelvis") are, of course, three in number on each side-the I( ilium" at the top of the socket for the thigh-bone, the" pubis" in front, pointing downwards and forwards, and the "ischium" behind; and the general form of each of these elements is shown in Fig. 4. The

FIG. 4. Pelvis of Crocodile arrangement is particularly remarkable from the circumstance that the pubis does not form part of the socket (" acetabulum ") for the thigh-bone, but is only attached to a process of the ischium, and this is a special character of all the known Crocodiles, except the very earliest. The thigh-bone (" femur") is very much twisted, and there is no great expansion at either end; and another noticeable feature is the very ill-defined character of the little prominence (" third trochanter") on the inner side below the head of this bone. The hind foot has only four toes, with merely a rudiment of the fifth. There is still another feature of special interest to the palseon­ tologist in the hard parts of the Crocodilia, namely, the character of the dermal armour; for there seems to be no known exception to the rule that these animals are protected more or less with strong bony scutes in the skin, and with horny scales in the super­ ficial layer of the epidermis. It seems, further, that the scutes are always marked by the well-known pittings upon the outer surface, so that they are easily distinguishable in a fossil state. HISTORY OF FOSSIL CROCODILES. 297

Now, remembering the usually fragmentary nature of fossilized relics, and realising how exceptional must be the conditions under which the complete dermal covering will remain intact nntil buried in sediment, it becomes an interesting problem to determine how far the form of the various scutes will afford any clue to their original position, and how far a few of them detached will yield any particulars as to the extent of the armour to which they belonged. This question appears first to have presented itself to Prof. Huxley, in 1858, when he had before him a few disjecta membra of the old Crocodilian of Elgin-the Stagonolepis­ the majority of which were simply dermal soutes. He was able to find very little definite information on the subject in published memoirs, so that the only alternative was to undertake an elaborate study of all available materials bearing upon it j and the research resulted not only in providing data for the determination of the fossil fragments, but also in a valuable accession to our morpho­ logical knowledge of the group.* In fact, however much some of our zoological friends may ridicule the trifling nature of many of the characters upon which the palreontologist is content to expend all his energies in the hope of attaining a satisfactory result, it cannot be denied that their branch of science is largely indebted to the close observations of minutire made by those who have only dilapidated scraps of animals to work upon, and the case of the Crocodilian exoskeleton is one small example in point. All the known members of the order agree in having a dorsal covering of scutes, but only two genera appear to be protected with anything beyond. the horny epidermic scales upon the ventral surface. Ordinarily, among living forms the main upper shield consists of several rows of more or less quadrate scutes, not very firmly united together, and each characterised by a keel-like pro­ minence placed in a longitudinal direction. More interesting, how­ ever, and more important for our present purpose, are the two curious Alligators of South America-the Jacare and the Caiman-

* T. H. Hnxley, " On the Stagonolepis Robertsoni (Agassiz) of the Elgin Sandstones," 'Quart. Journ. Geol. Soc.,' Vol. xv (1859), a paper containiug an elaborate account of the dermal armour of Crocodilians; see also the already quoted memoir in 'Proc. Linn. Soc.,' 1860, and another "On the Dermal Armour of Crocodilus Hastingsim," by the same author, in ' Quart. Journ, Geol. Soc.,' Vol. xv, pp. 678-680, pI. xxv, 298 A. B!flTH WOODWARD ON THE whose dermal armour has been carefully described by Professor Huxley. These correspond with the majority of fossil forms in possessing scutes both on the upper and the lower aspects of the body. As in all the other living Crocodilia, those on the dorsal surface are more or less keeled j and it is a noticeable feature that all those on the ventral surface are destitute of any such prominence. The component scutes of each of these shields, too, are very firmly united together, unlike anything now known among the Crocodiles and Gavials, but (just as they are in most extinct forms), being joined to one another laterally by means of jagged sutures, and overlapping one another like the tiles of a house in the other direction. Besides the elements of the great shields, there are also scattered upon the legs and the sides of the body numerous little irregular scutes, with a somewhat conical rising in the middle, and consequently we may readily recognise three distinct kinds:­ (1) Dorsal keeled * scntes, (2) Ventral keel-less scutes, and (3) irregular scutes, scattered over various parts of the skin. All the most perfect remnants offossil Crocodilian armour tend to show that these three types of scutes occupied the same relative positions in the extinct forms, and we shall refermore in detail to the variations in composition of the shields as the different genera and families come under review in the sequel. Such are the more important characters of the living Crocodilia that specially concern the palroontologist, although several others -such as those connected with the circulatory organs, and the ear, for example-would be necessary to complete the enumeration from a zoologist's point of view.] It must be sufficient now, however, merely to add that many of the internal soft parts ex­ hibit a remarkably high degree of organization, and to conclude with the following brief tabulated statement of what has already been noted in regard to the skeletal structures. Those peculiari­ ties which are met with in all known fossil forms, besides the recent, are printed in italics j and to each character, also, is

• Often simply "angnlated "-not keeled-in extinct forms. t The characters of the ear might, indeed, claim a brief notice, did we intend to enter upon more technical points, for fossil forms tend to throw light on the manner in which it has been developed; but reference can readily be made to Prof. Miall's excellent description in the little manual above cited. See also R. Owen, " On the Communications between the 'l'ympanum and Palate in the Crocodilia," , Phil. 'I'rans.,' 1850. HiSTORY OF F09SII, CROCODILES. 299 affixed the name of the other reptilian orders in which it is repeated.

Repeated in

PRINCIPAL OSTEOLOGICAL CHARACTERS 0]1 LIVING CROCODI LIA .

I. External nostril terminal and single .. . X

2. No parietal foramen ... ." ...... X 3. Quadrate bone immovable ...... X P X XX 4. Transverse bone present ...... X X most XX X P 5. Pterygoid bones with anterior median processes reaching the vomers ...... X

6. Nasal passage prolonged backwards by outgrowths of palatines and pterygoids

7. T eeth confined to maxillre, prem=illre, and dentaries .. . •...... , ...... X X XX

8. Teeth in sockets, and not anchy losed to jatC X x x

9. Vertebree mostly proccelous most X x

10. Two sacral eertebros ...... X X X X 11. Cervical and anterior th oraoic ribs, with head and tubercle ...... , ...... X XX X 12. I nt erclavicle present , but no clavicles

13. Pubis excluded from acetabulum

14. F emur without prominent inne r troch anter

15. Dermal scuies present: two longitudinal dorsal rows, at f ewest ...... •..

nI.- P ALlE ONTOLOGICAL HISTORY OF THE C ROCODI LI A.

The earliest traces of the Crocodilian type of structure at present definitely known are met with in rocks of age. Specimens of true reptiles, either rhynchocephalian or dinosaurian, 300 A. SMITH WOODWARD ON THE have been discovered even so low as the ,* but nothing has hitherto been ascertained concerning the Palreozoic forerunners of the Crocodiles; and the order makes its first appearance in the Trias of Germany, in the probably equivalent rocks of the Scottish Highlands, in India, and in North America. In these early times, however, the type as represented to-day is only just beginning to be foreshadowed; and, like all their congeners, the first Mesozoic Crocodilia possessed a most curious mixture of reptilian characters, rarely-if ever-found associated at more recent periods. Combined with the peculiarities that are looked upon as the essential features of the order, there are numerous structures recalling the lizards and the dinosaurs, and added to the whole are some that are almost unique. Of the Belodon (= Phytosaul'us), from the Upper Keuper Sandstone of Stuttgart, more is known than of any other form; and at least three species have been determined and described by H. von Meyer.] In general outward shape, it was a creature very much resembling the living Gavial, having a long snout armed with almost equally-sized teeth, and the largest examples attaining a length of perhaps fourteen or fifteen feet. The state of preservation of many of the fossil bones-considering the nature of the coarse sandstone matrix in which they are embedded -is really quite astonishing; and some specimens are so perfect. that it has been possible to elucidate almost the complete osteology of the skull. Of this important part, three diagram­ matic sketches (after von Meyer) are given in Fig. 5; and it requires only a very hasty comparison with the corresponding figures of the skulls of more recent and living Crocodilia (such as Figs. 6-9) to recognise some most remarkable peculiarities. The long snout, for example, consists almost entirely of the enormously­ elongated premaxillary bones, as in the Ichthyosaurs and living Dolphins; and more than half of the tooth-bearing edge of the jaw on each side is supported by them, while in the later members of the order the maxillaries take the greater share. This snout has a characteristic downward bend at its extremity, and the

* Proterosaurus (H. von Meyer, "Fauna del' Vorwelt-e-Saurier aus dem Kupferschiefer del' Zechstein formation "), t H. von Meyer, "Reptilien aus dem Stubonsandstein des obern Kenpers," 'Palooontographica,' Vols. vii and xiv; also" Del' Schadel des Belodon aus dem Stubensandstein des Obern Keupers," ibid., Vol, x, pp. '227-246, pls, xxxviii-xlii. HISTORY OF FOSSI L CROCOD II,ES. 301

OQQO O ••O QOO OOOOO.OOOQOO 00°0

FIG. I) Skull of B elodon (after H. von Meyer). 1, side view; 2, upper view; 3, palatal view. In tho second figure, the most anteri or pair of open ings ar e the nostrils ; th e next pai r, th e ant.orb it al vacu ities ; next, the orbits ; hindermost, the temporal fosste. In the first figure th e three latter vacuities nlone are shown. nostrils do not open just above, but are placed far back near the orbits, where th ey form two distinct openings, and are surrounded completely by th e nasal bones-a feature unknown in any other of reptiles, recent or extinct. \Ve observe, again, that there are two gr eat lateral holes, one on each side of tbe nostril s and just in front of the orbits, probably corresponding with th e "ant-orbital vacuities" of birds and pterodactyles; and it is II very interesting fact tbat-although the fashion has long died out in the Crocodile community- t here were still some tr aces of th ese openings to be found in th e Steneosaurs and Teleosaurs of times, and an American genus (Thoracosaurus) retained them even so late as the Upper period. There are also other 302 A. SMITH WOODWARD ON THE characters in the skull of very considerable importance, both to the systematic palseontologist and to those unravelling the problems of evolutional history; and we ought likewise to dwell upon the bicon­ cave vertebrre with their processes, and the arrangement of the limbs and their connections. These peculiarities, however, will be referred to again later on, when alluding to the subject of classification. The only British fossils hitherto discovered that can be de­ finitely referred to a near relative of the Belodon are from the much-discussed yellow sandstones in the neighbourhood of Elgin and Lossiemouth, on the Moray Firth. The first specimen made known to science was a fragment of the scaly armour of the creature, described by Agassiz in 1844, and considered by him­ though he had only examined drawiugs-to belong to a fish. It was, accordingly, figured in his great work on the "Poissons Fossiles,"* under the name of Staqonolepis Robertsoni, and placed in the order of Ganoids, Some twelve years later, however, through the assiduous researches of the Rev. Dr. Gordon, of Birnie, more perfect remains of the same form were obtained, and there could be no longer any doubt as to their being really of a reptilian nature. In 1858, the conclusions of their discoverer and Sir Roderick Murchison were confirmed by Prof. (Sir Richard) Owen, Sir Philip Egerton, and Prof. Huxley, and the latter published his first paper on the subject in the following year. t Moreover, at least one other -Hyperodapedon G01'doni­ was made known at the same time, and then began the great geological controversy as to the age of these fossiliferous sand­ stones, which has continued ever since. Palseontological evidence is distinctly in favour of the Triassic age of the beds, for Hypero­ dapedon is known both in England and in India, and only occurs as a Triassic form, while it is quite certain that the Crocodilian Stagonolepis is a close ally of the Continental Belodon; and the recent discovery of a Dicynodont reptile :j:-a type especially * L. Agassiz, «Recherches sur les Poissons Fossiles du VieuI Gres Ronge,' p. 139, pl. xxxi, figs. 13, 14. t Paper already quoted, 'Quart. Journ. Geol. Soc.' Vol. xv (1859), pp, 440.460, pl. xiv, See also T. H. Huxley, on " Stagonolepis Robertsoni, and on the Evolution of the Crocodilia," lac. cit., Vol. xxxi (1875), pp. 423-438, pl. xix; and" The Crocodilian Remains found in the Elgin Sandstones, &e.," 'Mem. Geol. Survey,' Mon. iii. (1877). +R. H. Traquair, "A Preliminary Note on a New Fossil Reptile recently discovered at New Spynie, near Elgin," « British Association Reports,' 1885. See also J. W. Jndd, "The Presence of the Remains of Dicunodon. in the Triassic Sandstone of Elgin," 'Natnre,' vol. xxxii, p. 573 (Oct. 15th, 1885). HISTORY OF FOSSIL CROCODILES. 303 characteristic of the earliest Mesozoic rocks of Russia, South Africa, and India-appears to afford no further cause for hesitation in the matter. The skull of Stagonolepis is not completely known, but it seems to have been very similar in general characters to that of Belodon; the teeth, however, instead of being sharp and pointed, were blunt and swollen (Fig. 11), like the hind teeth in many recent Croco­ diles, and they are sometimes worn down by attrition. All the more important parts of the skeleton. have been found, and they agree with Belodon in exhibiting the strange mingling of reptilian characters, but the British genus was evidently better protected with dermal scutes. The tail was completely encased in a segmented sheath: the back was covered by a double row of keeled scutes: the sides and limbs (probably) were defended by scattered irregular ossifications; and a considerable armour of flat scutes­ mistaken by Agassiz for the scales of a fish-extended over the ven­ tral surface. All are ornamented with the characteristic pittings, and their resemblance to rain-prints suggested the generic name. As an interesting proof of the wide prevalence of this primitive type of Crocodilian life in the Triassic stage of the earth's history, it is worthy of note that remains indistinguishable from those of Belodon have been met with in the rocks both of North America'[ and India,t while a nearly-allied genus, named Pal'asuc!lus§ by Huxley, has left abundant indications of its presence at the same period in India. Jurassic Crocodilio, Passing on to the Jurassic representatives of the Crocodilia, meet first of all the introduction of the most important Mesozoic family II-the Teleosauridre-in the Upper Lias. Here many of • See figures and descriptions in Huxley's "Geol. Survey Memoir," already cited. t E. D. Cope" On the Reptilia of the Triassic Formations of the Atlantic Region of the United States," 'Proc. Amer, Phil, Soc.,' VoL xi (1871), pp. 444.4<16. (Reprinted in' Ann. and Mag. Nat. Histo' [4], Vol. vi, pp. 498-500). t R. Lydekker, "The Reptilia and Amphibia of the Maleri and Denwa Groups," 'Palmontologia Indica,' ser, iv, Vol i, pt. v (1885), p. 21. § R. Lydekker, loco oit., p.22. II The members of this family being so abundantly represented in the Jurassic rocks of the Old World, it is a remarkable fact that scarcely any have been recorded from equivalent strata in America. The writer has only sncceeded in meeting with one such instance: see H. Burmeister, " Die Versteinernngen von Juntas im 'I'hal des Rio de Copiapo," Abh, Natnrforsch. Gesell. Halle,' vol. vi, p. 122, pl. i, figs. 1-3. Anterior thoracic vertebrte lire here described under the name of neogceus. VOL. IX., No.5. 22 804 A. SMITH WOODWARD ON THE the most typical characteristics of the order have already been acquired, and, as will be shown in the sequel, the present group forms a defiuite connecting link between the Triassic" Parasnchia,' as they are termed, and the modern types (or "El1suchia ") of to-day. The Teleosaurs were all admirably adapted for an aquatic mode of existence, and reasoning from the frequent association of their fossil remains with a marine fauna, it seems almost certain that they were much more accustomed to a sea-going life than any that survive at present. The majority, in fact, seem to have lived habitually in salt water, and the appearances presented by certain of their petrified remains suggest that they fed ordinarily upon the hosts of cephalopoda that tenanted the Mesozoic seas.* In general outward form they differed but little from the modern Gavials, having the same long, slender snout, and a similar armature of numerous sharply-pointed teeth. There was a little variation, how­ ever, in the relative size of the limbs, the front pair being much smaller and the hinder much larger than those in their present-day allies.] and this arrangement may have had special reference to the rareness of the occasions on which they would be required for terrestrial progression.j Further, the biconcave character of the vertebrre, by allowing of flexibility, would likewise conduce towards freedom of movement through water, and the extensive develop­ ment of dermal armour, both on the upper and under sides of the body, would prove an effective protection in most cases against the huge ichthyosaurs and plesiosaurs that shared those primeevaldeeps. Among the less superficial characters of these ancient reptiles, there are many features of considerable importance and well worthy of careful attention. In the skull, for example, the" supra-temporal fossee " are especially noteworthy on account of their great relative size. The" antorbital vacuities," compared with those of the Triassic Orocodilia, are reduced to quite small perforations, and there is a little foramen below the occipital condyle that indicates an advance in the organ of hearing. The opening of the" posterior

II A. Quenstedt, "Handbuch del' Petrefaktenkunde' (1852), p, 100 (3rd edit., pt. i, 1882, p. 161). t This character is well shown in the restored skeleton figured by Owen in his" Mon. Foss. Reptiles London Clay," pt. Ii ('Mon. Pal. Soc.,' 1850), pl. xi, fig. ii, t It is interesting to note that the living Sea Lizard of the Galapagos Isles (Amblyrkynchus or Oreocephalae) exhibits similar proportions of the fore and hind limbs. Here, likewise, the relative diminution of the fore pair tends towards greater powers of swimming. HISTORY OF FOSSIL OROCODILES. 305 nares" upon the palate is also placed somewhat differently to that of the earlier forms, having moved slightly further back; and the external nostril-now single-is placed near the end of the snout, and surrounded entirely by the premaxillse, The vertebrre, as has just been stated, are still more or less biconcave (" amphiccelian " 01' "amphiplatyan "), but the transverse processes are more like those of our living Crocodiles, and the three bones supporting the hind limb are arranged exactly as they have been in this order ever since. The dermal shield on the back is composed of a double row of angulated scutes, united by sutures in the median line, and overlapping one another by their bevelled posterior edges j and the ventral cuirass consists of several rows of flat scutes all firmly jointed together by serrated sutures, and forming a well-nigh im­ pregnable armour.s With regard to the systematic classification of the Teleosaurs, very great confusion still prevails, for their fossil remains already discovered are so numerous, and the various descriptive accounts of them so scattered and disconnected, that it seems almost hopeless to establish any order out of the chaotic mass of materials at dis­ posal; and the difficulty is further enhanced by the fact that some of the original type specimens have been proved to be unfortunate restorations.] while one or two of the most important species were practically undefined by their authors. Moreover, it has been an almost universal custom to take advantage of each successive dis­ covery for the manufacture of a new generic or specific name, how­ ever imperfect the materials, and a most intricate and perplexing synonymy has thus arisen, that it would require long-continued research by the most profound specialist to unravel. Of all students of this group, the distinguished French paleeontologists, MM. Deslongehamps (ph'e et fils), have contributed most towards its systematic revision.j' and in endeavouring now to set forth a few of * Fun descriptions are given in HUXley's important memoir in 'Quart. Journ. Geol. Soc.,' 1859. t Vide E. E. Deslongchamps, ' Notes Paleontologiques,' pp. 108.111. t J. A. Eudes Deslongchamps, " Memoires surles 'l'eleosauriens de l'Epoque Jurasaique du Departement du Calvados, Mem. 1." 'Mem. Soc. Linn., Nor­ maudie,' Vol. xiii, Eugene Eudes Deslongchamps, • Notes Paleontologiqnes,' 1863.1869. See also' Bull. Soc. Geol. France' [2J, Vol. xxvii, 1870, pp. 299.351, with plates, and' Mem. Acad, Imp. ScL, &c., Oaen,' 1869, pp~31-80, with plates. Eugene Eudes Deslongchamps, "Le Jura Normand: Etudes Faleontologi, ques des divers Niveaux Jurassiques de la Normandie," Mon. iv (1877-8), pp.37-60, pis. i-v, vii, xv. (unfinished). 306 A. SMITH WOODWARD ON THE the leading points, we shall make use of their latest published con­ clusions as a basis, and take into account the more important additions to our knowledge contained in the various memoirs of other naturalists that have subsequently appeared. The nomenclature and arrangement of the Liassic Teleosaurs is perhaps in greatest need of revision. In Britain, indeed, there has been no elaborate attempt at their systematic study since 1841,'" and the unfortunate manner in which Buckland published Konig's MS. name of Teleosaurus Chapmani in 1836 has led to much of the confusion both here and on the Continent. Among the numerous remains from the Upper Lias of Whitby, at least three distinct forms appear to have been already described, although a critical comparison of all the materials available would probably result in the satisfactory determination of more. The earliest specimen made known to science was figured and described by Chapman and W ooller in the 'Philosophical Transactions' so long ago as 1758,t and the authors of these quaint communications perceived the general resemblance of their fossil in outward shape to that of the living Gavials and Alligators, thus suggesting an affinity; but none received a name until the publication of the Bridgewater Treatise, in which Buckland devoted a platej to the illustration of three typical examples, and referred them all to an undefined species, Teleosaurus Chapmani, Konig MS. Five years later, in his second Report on British Fossil Reptiles, read before the British Association,§ Prof. (Sir Richard) Owen gave a detailed account of the specimen represented in Fig. 1 of Buckland's plate just quoted, and subsequently attached the MS. name of T. brevior to a fine skull in the British Museum; in 1854, Mr. Charlesworth II brought forward evidence of possibly another form, terming it T. iechnodon, but publishing no definition; in 1861, Owen's' Palseon- .. R. Owen, "Report on British Fossil Reptiles," part ii,' Brit. Assoc. Reports,' 1841. t William Chapman, " An Account of the Fossile Bones of an Allegator, found on the Sea-shore near Whitby, in Yorkshire," 'Phil. Trans.' Vol. I, pp. 688-9, pl. xxxii (b), and -. Wooler, "A Description of the Fossil Skeleton of an Animal found in the Alum Rock near Whitby," ibid., pp. 786· 790, pI. xxx. This specimen was presented to the Royal Society, and is now in the British Museum. ::: Rev W. Buckland, 'Geolog-y and Mineralogy, etc.,' Vol. ii, p. 35, pt. 25. An early figure of a skull from Whitby is also given by E. Charles­ worth, in' Mag. Nat. Hist.,' n.s., Vol. i (1837), p. 532, fig. 65. § 'British Association Reports,' 1841, pp, 73-81. II E. Charlesworth, "British Association Reports," 1854, 'Trans. Sections,' p. 80. HIST ORY OF FO SSIL CROCODILES. 307 tology' appeared, with th e figure of a detached vertebra designated 1'. brevirostrisrv in 1876 Prof. J. F. Blaket described Owen's (MS .) T. breoior under the name of breuior ; in 1880, Prof. H. G. Secloy[ applied the provisional name of T. encephalus to a fragmentary skull-also from Whitby-in the Woodwardian Museum, Cambridge; and quite lately, in his

I History of British Fossil Reptiles,'§ Sir Richard Owen has re­ printed his previous (1841) description of T. Chapmani, and also added particular s of 1'. brevior, without attempting to incorporate th e results of recent Contin ental research. 11 According to M:acr. Deslongchamps, whose views on the subject appear to be now generally adopted, none of these forms are refer­ able to Teleosaurus proper, and ought ratber to be placed in the genera Mystl'iosaul'tts of Kaup and P elaqosaurus of Bronn-the broad-faced types, with much depressed cranium and upwardly directed orbits belonging to the former, and those with long slender snouts and more laterally placed orbits, widely separated, to the latter. The pr esent Professor of Oaen, M. Eugene Deslongchamps, regard s the Whitby Museum specimen described by Owen (figured by Young and Bird,*'" and also in Buckland's Fig. 1, pI. 25), as the type of fiIy strios aw'us Chapmani, Konig sp., and th e instruc­ tive original of plate 15 (Crocodilia) in Owen's ' British Fossil Reptiles,'tt as a typical example of Pelaqosaurus Bronqniarti, Kaup sp.tt With the latter, also, he would associate the specimens

* R. Owen, I Paleeontol ogy,' 2nd edit., p. 299, fig. 103 [IJ. t R. 'rate and J. F. Blake,' The Yorkshire Lias' (1876), p, 244, pl. i, figs. 1·3. t H. G. Seeley, " On the Cranial Characters of a large Teleosaur from th e Whi tby Lias," Qnart. Joum, Geol, Soc.,' Vol. xxxvi (1880), pp. 627­ 634, pl. xxiv, § Op. ci t . Vol. iv, pp. 130-139. II Unfortu nately in this extensive work th ere are no particulars as to the date at which each succ essive part appear ed ; but as references on previous pages (e.g., on p. 66) relato to papers published so recently as 1880, the following description of th e Teleosaurs cannot ha ve been printed before that or II later year. , K E. Deslongc hamp s, 'Notes Pal.,' and ' Le Jura Normand,' Mon. iv, ** Young and Bird, " Geological Snrvey of the York shire Coast," 2nd edit., 1828, pI. xvi, fig. 1. tt It may be noticed that this (British Museum) specimen is not described in Sir Rich ard Owen' s lett erpress, although th e plate is cit cd at the com­ mencement of the Secti on (Vol. iv, p.1 30). Measurements. however. are given by :M. Eugene Deslongchamps, I Le Jura Normand,' Mon. iv, p, 11. tt To this species, also, ?II. Deslongchamps assigns the restored figure given in Owen's ":Mon. Fo ss. Rept . London Clay," pt. ii (' Mon. Pal. Soc.', 1850), pl. xi, figs. 2, 2a. 308 A. SMITH WOODWARD ON THE shown in Figs. 2 and 3 of Buckland's plate 25 j* and, under the same re-arrangement, Teleosaurus breoior would be relegated to Mystl'iosaw'us. Both these genera (or sub-genera) are represented by numerous specimens in the Continental Museums, chiefly from the Upper Lias of Boll, in Wiirtemberg, and of Curcy, in Normandy, and the dis­ tinctive features of Pelaqoeaurus were very definitely elucidated by the elder Doslongehamps.j- It is more difficult, however, to compre­ hend theprecise particulars in which Mystl'iosau1'Usdeparts from the generic type of Steneosaw'us,~ and though Winkler not long ago§ published a most elaborate study of its specific variations, he appears to have left the wider question of its affinities altogether untouched. The latter palreontologist endeavours to show that all the specimens hitherto described really belong to a single species, divisible into two varieties, one English and one German; and he accordingly proposes (loc. cit., p. 83) to apply the name of M. Stukelyill to both, with the varietal adjunct of Chaprnani in the case of the first, and Bollensis in that of the second ; but the ordinary rules of priority in nomenclature render it doubtful whether such a solution of the difficulty will meet with general acceptance. In addition to Pelaqosaurus Bronqniarti, of which the fine Whitby specimen has already been mentioned, a smaller species­ P. typus-occurs in a wonderful state of preservation in the Upper Lias of Normandy, and this forms the subject of perhaps the most exhaustive memoirs on any reptilian skeleton from so ancient a deposit ever yet attempted. MM. Deslongchamps, indeed, have * Vide Deslongchamps, 'Le Jura Normand.' Mon. lv, pp, 8-13, passim. t J. A. Eudes Deslongchamps, " Memoires sur les Teleosauriens de l'Epoqne Jurassique du Departement du Calvados, Mem. i," 'Mem. Soc. Linn. Normnndie,' Vol. xiii, [Pelagosau,'us typus is here described as Teleosaurus temporalis, but the mistake was afterwards rectified.] Eugene Deslong­ champs in' Le Jura Normand,' Mon. iv, p, 13, et seq., adds further detailed accounts. See also Huxley's Memoir on Stagouolepis, already quoted, pls, xi, xii, with explanations. :t In 1877 Eugene Deslongchamps (' Le Jura Normand.' Mon. iv, p. 8, note) announced a forthcoming paper on this subject, to be published in the • Bull. Soc. ZooL France.' The writer, however, has not been able to meet with it either in that jonrnal or in any of the Caen publications. § T. C. Winkler. "Etude sur Ie genre Mystriosaurus," 'Archives dn Muses Teyler' Vol. iv, fasc. i. (1876). II On the assumption that the Whitby fossil described by Stukeley at the beginning of the last century [' Philosophical Transactions,' 1719, pp. 963· 968 (No. 360), pl. i.] belongs to this genns. The original specimen, how­ ever, is now in the British Museum, and is certainly a Plesiosaur. HISTORY OF FOSSIL OROCODILES. 309 succeeded in elucidating (loc. cit.) its osteology almost as com­ pletely as that of any living member of the order, and a re­ production in plaster of one of their most beautifully perfect examples has lately been added to the collection of the British Museum. Our National Collection is also enriched by several of the original specimens described in these classical works, and among such the cranial fossils are particularly instructive. The more important distinguishing features of the genus consist in its slender snout, with swollen fronto-nasal region; the circular, later­ ally-placed orbits, widely separated; its broad and strongly-pitted " temporal arcades," and the posterior nares of a greatly elongated shape. (See Fig. 6.)

FIG 6. Skull of typus (side, upper, and under views). 0, orbits; T, supra-temporal fossse; P, palate-pterygoid vacuity; N, pos­ terior nares; E, aperture of eustachian canal.

Teleosaurus itself, as now usually restricted, appears first in the Lower Oolites, and is not known to range beyond the Kimmeridge Clay. T. Cadomeneie, Geoffr., from the Fuller's earth of Caen, in Normandy, is regarded as the type-species, and of this form, also, :M M. Deslongchamps have been able to publish a remarkably sio A. SM IT H W OODWARD ON THE

complete restoration. " A striking feature of the genus is the very small size of the head compared with the proportions of the trunk, and in the skull the snout tap ers abruptly just in front of the orbits, assuming much th e same shape as that of the living Gavial. The orbits have a more upward aspect, and are placed closer t ogether, than those of P elaqosaurus ; and the posterior nares are broader th an they are long, while the palatine bones attain a rela­ ti vely small development. The teeth are also peculiar, being invariably numerous and slender, and nearly always projecting out­ wards instead of being fixed upright in the sockets. The genu s, as thu s understood, appears to be well represented in certain horizons of th e E nglish Oolites, but the fossils met with are rarely of more than a very fragmentary character. Sir Richard Owen t has identifi ed t eeth and vert ebrre of the typical species, T. Cadomensis, Geoffr., from th e Great Oolites of Enslow (near W ood­ stock) and Stonesfield, and also considers'[ a cervical vertebra from the Lower Oolites of Chipping Norton to indicate a very closely allied form. The same distinguished palreontologist, in hi s 'British Fo ssil Reptil es,' also founds a new species, T . lati­ fi'ons,§ upon the gre ater portion of a skull from strata of Great Oolit e age in Northamptonshire, but the teeth and some oth er important parts are wanting ; and P rof. Phillips, in his' Geology of Oxford,' adds two more from th e well-known Oolitic Flagstones of Stonesfield, which he designates T . brepidens and T. subulidens (Fig. 12) respectively, in allusion to the shape of the teeth . The former species II is founded upon a remarkabl y complete skull and mandible, and th e latter ~ upon a nearly perfect mand ible, with several oth er fragments; and associated with the remains of both are numerous vert ebrse and limb-bones, of which the P rofessor figures an instructive series. Th e genus has not hitherto been recorded from th e Oxfordian or Corallian deposits i but Sir Richard Owen" gives the name of T. asthenodeirus to a few detached vertebrse found in the Kimmeridge Clay of Shotover Hill, and Mr. J. W.

*' E. E. Deslongchamp s, 'Notes Pal.,' pp.163.201, pI. xiii. t R. Owen, ' British Association Reports,' 1841, p. 81. ::: ius; p. 81. § Sir R. Owen, op. cit., Vol. iii, p. 141, pI. xvii, (Crocodili a). ~ Prof. John Phillips, op. cit., pp . 186·189. , Ib id, pp. 194, 195. ** R. Owen, ' Bri tish Association Reports,' 1841, p. 81. HISTORY OF FOSSIL OROCODII,E S. 311

Hulke" has made known the discovery of an undoubted Teleosaurus snout in the equivalent clays of Dorsetshire. This Kimmeridgian specimen is especially remarkable for the expansion of its terminal extremity, and accordingly received the name of T. meqarhinus, Bulke. The remaining 'I'eleosaur s of th e Jurassic period seem to belong to about six or seven generic types ; th e earliest are referable to th e Steneosaurus of Geoffroy St.- Hilaire, and Teleidosaurus of Deslongchamps ; and those of the Middle and Upper Oolites chiefly belong to ./Eolodon (H. von Meyer), M etl'i01'hynchus (H. von Meyer), (Q uenstcdt), and Mac himosaurus (H. von Meyer) . Mr. E. T. Newtont has also described a frag menta ry mandible, probably Crocodilian, from the Coral Rag of W eymouth , and this, if rightly dete rmined, may possibly indicate another genus. No very complete skeletons of Steneoeaurushave hitherto been met with , but numerous well-preserved crania (Fig. 7) are known, both in this country and on the Continent. According to MM. Deslong­ champs' diagnosis, t the genus is readily distinguished from T eleo­ Sa Ul'U8 by the gently tape ring character of its snout, the large de­ velopment of th e palatin e bones, and the shape of the posterior narinl opening , which is as broad as it is long. The teeth are usually fixed upright in th e sockets, and possess a more or less dis­ tin ct longitudinal keel on each side. In France about twenty Steneosaurs appear to be known, and among British fossils nine oth ers have already been named and described j at least t hr ee of th e latter, however, were originally placed with this genus on very questionable grounds, and there is abundant reason for suspecti ng that they ought rather to be referred to Metriorhqnchus, It is also unfortun ate that the three cranial modifications made known by Sir Richard Owen arc not compared or cont rasted with those previously described by Deslongcbamps;

* J. W. Hulks, "Note on a Fragment of a TeJeosaurian Snout from Kimmeridge Bay, Dors et," 'Quart. Journ. Geol. Soc.' Vol. xxvii (1871), p. 442 , pl. xviii. This spec ime n was discovered by Mr. Mansel-Pleydell, and presen ted by him to t he Briti sh Museum. t B. T. Ne wton, "Notes on a Crocodil ian J aw from th e Corallian Rocks of Wey mouth," I Quar t. J ouru, Geol. Soc.,' Vol. xxx iv (1878), pp. 398.400, pi xvi.

::: E . E. Deslongcham ps, I Notes Pal. ,' p. 126. 312 A. SMITH WOODWARD ON THE and, with the exception of Steneosourue Boutilieri, Deslong.,· Mr. Hulke's S. Stephani,t from the Cornbrash of Closworth, Dorset, is thus the only English species yet satisfactorily defined. Those described by Owent are S. Geojfroyi, from the Great Oolite near Oxford; S. laticeps, also from this formation and locality; and S. temporalis, from the Oolite Freestone of Bath; all are founded upon more or less complete skulls, but there are no particulars as to the collection in which the type-specimen is preserved in either case. Phillips§ records a fragment of jaw, from the Kimmeridgian of Shotover Hill, under the rather indefinite II name of " S. lonqirostris, Cuv.;" and a cranial fossil of Oxfordian age, in the Woodwardian Museum, Cambridge, is regarded by Seeley' as the type of a new species, S. dasycephalus. Teleidoeaurus does not appear to have been recorded as yet from British strata, and only two species are described by Deslong­ champs"'· in Normandy; it occurs chiefly in the zone of the Fuller's Earth, and constitutes a link between Teleosaurus and another genus, Metl'iorhynchu8, now to be considered. The lattertt is typically a member of the Upper Oolitic Fauna, and ordinarily occurs in the Kimmeridge Clay; but as to its repre­ sentatives in this country, there is at present very little definite information. Deslongchamps, indeed, seems to be the only palseon­ tologist who has hitherto attempted to determine any English frag­ ments of it, and he incidentally refers to the occurrence of M. super-

• Under this name M. Deslongchamps (' Notes Paleont.,' P- 230, pl, xvi, fig. 2) mentions a plaster cast of a skull aud mandible from the Cornbraah near Oxford, received from the Bristol Institution, and labelled" Grocodilus Oxoniensis, Conybeare;" and in describing his outline figures of the speci­ men, be further refers to it as "Steneosaurus Oxoniensis, De la Beche." But Mr. Edward Wilson, to whose kindness the writer is indebted for particulars of the Crocodilian fossils now in the Bristol Mnseum, is unable to discover any such label in the collection; and the present whereabouts of the original specimen seem to be unknown. t J. C. Mansel-Pleydell, "Note on a Gavial Skull from the Cornbrash of Closworth," 'Proc. Dorset Nat. Hist. and Antiqnarian Field Club,' Vol. i (1877), pp, 28-32, pI. i, A detailed descriptive paragraph is supplied by J. W. Hnlke. t Sir Richard Owen, 'History of British Fossil Reptiles,' Vol. iii pp. 144,145, pI. 18, 19 (Crocodilia). § J. Phillips, • Geology of Oxford,' p. 388 (with woodcut of tooth). II See Deslongchamps, • Notes PaL,' pp. 110, 111. ~ H. G. Seeley, 'Index to Reptilia, etc., Woodwardian Museum,' 1869, p.140. ** E. E. Deslongohamps, 'Notes PaL,' pp. 271·291, pI. xviii, figs. 4-6, pI. xix, figs. 1-5. tt See detailed descriptions, E. Deslongchamps, op eit., p. 271, et seq. HISTORY OF FOSSIL OROCODILES. 313 ciliosus, de Blainv. sp., in the Oxford Clay near Oxford.s and of M. hastifer, Desl., in the Kimmeridgiarl.j He further adds an inter­ esting statementj regarding the large mandible from Kimmeridge Bay, described by Sir Richard Owen under the name of Pliosaurus trochanterius, and expresses his opinion that it certainly belongs to the present Crocodilian form. To Met"i01'hynchus, also, we may safely refer several fossils from the neighbourhood of Oxford, described by Professor Phillips as Steneosaurian. Among its many important characters,§ this genus is especially remarkable for the great development of the pre-frontal bones which overhang its laterally-placed orbits; the elongated, oval form of the upwardly directed external nostril; the large relative size of the three or four anterior teeth in each jaw; and the general contour of the hinder cranial region. Fragments of the snont may also be easily distinguished by the palatal surface exhibiting two parallel longi­ tudinal grooves. Moreover, it seems even possible to recognise de­ tached vertebral by the peculiarities of the neural arch, for the pedicle on each side is slightly prolonged downwards, thus present­ ing the appearance of clinging to the centrum.] Now, nearly all these characters are well shown in Prof. Phillips figures of "Steneosaurus palpebrosus" and U Steneoeaurus qra-

* E. E. Deslongohamps, 'Notes Pal.,' p. 319. top. cit., p. 353. t While describing the mandible of Metriorh!JuchulI Mereli, M. Deslong­ champs writes as follows (op. cit., p. 329) :-" Cette forme de mschoire infe, rieure, tonte differente de celle des autres animaux de la memo famille, rap­ pelle assez la merne piece appartenant a un animal d'nne tonte autre famille, c'est-a-dire des Sauropterygiens ; je venx dire celle qui est deorite par S. Rich. Owen comme etant celie d'un Pliosaurus qui, d'ailleurs, s'ecart.e nota­ blement de la forme habituelle du Pliosaurus graudill, et qu'il nomme Plio­ lIaurus trochaute,·ius.* II est hors de doute que la machoirs decrite ioi ap­ partient au genre Motriorbynche; il serait en effet trop etrange que cette forme fftt venue precisement se rencontrer dans tontes les assisea ou se trouvent des Miitriorbynches, et que je puisse rapporter des pieces en tout semblables par leurs caracteres particnlisrs de taille, de force et de briilvetii ou d'allongement du museau, a ohacune de mes especes de Metriorhynches, aussi bien dans les couches calloviennos qu'oxfordiennes et kimmeridgiennes j et, en effet, je oonnais des maintenant des machoires mferieures se rapper­ tant parfaitement aux M. superoiliosus, Moreli, et haetiter;' * R. Owen, 'Monograph of the Fossil Reptilia of the Kimmeridge Olay, (' Mon. Pal. Soc.,' 1868), p. 7, pI. iii., figs. 3-5). § E. E. Deslongcbamps, op. cit., p. 132. II Character mentioned by H. E. Sauvage in 'Bull. Soc. Geol. France' [3J, vii (1879), p. 695. See also figures and descriptions of vertebral by E. E. Doslongchamps, in Lennier's 'Etudes geologiques et paleontologiques SU1' l'Embouchure de la Seine, etc.,' 1870, pp. 50-52,pl. x, 314 A. SltfI TH WOODWARD ON THE cililJ,"· and these species are also confined to the Up per Oolites, like the majority of the Metriorhynchs of the Oontinent. Of the first named form, a large series of remains, from th e Kimm eridge Olay at Shotover, are preserved in th e Museum at Oxford, and the Professor illustrates both skull and mandible, vert ebrre, shoulder girdle, and ischium, besides several limb bones more doubtfully referred to the same animal. M. gracilis was founded upon less perfect materials: the ty pe-specimen is a fragmentary skull, dis­ covered by Mr. W ood Mason, and especialIy interestin g on account of its being of P ortlandian age. In the same work (p. 319), Prof. Phillips further descri bes a numb er of vertebrre from the Oxford Olay of Long Marston, said to agree very closely with th ose of the Shotover species; and it may be added that an anterior portion of a mandible obtained by Mr. Cunnington from the corresponding beds of Ohippenham, and now in th e British Museum (No. 46,323), undoubtedly belongs to the same generic type. AiJolodon (R. von Meyer) is also an Upper Oolitic 'I'eleosauriar» being characteristic of the Lithogra phic Stone of Bavaria, and (probably) of Oirin in France, but there is no record of its occur­ rence in Britain, and it does not appear to possess any special features of importance ; the type specimen,'] however, discovered at the beginning of the present century, forms a conspicuous object in the National Collection, and is thu s worthy of mention here. An oth er generic type of Kimmeridgian age is the Dakoeaurus of Quenstedt . (Figs. 8, 10.) It was founded, in 1tl58,:j: upon a portion of mandible, conta ining six teeth implant ed in distinct sockets - each of t hese teeth (Fig. 10) being somewhat

FIG. 10. Tooth of Dakoeaurue. oval in section, and characterised by a smooth surface, with two slightly serrated edges or "carinre." Detached fossils * J. Phillips, Geology of Oxford,' pp. 380-388, with woodcu ts , The Kimmeridge specimens wer e pre vious ly no ticed by Owen (' Brit. Assoc. Repor ts,'1841, p, 82) und er the n a m~ of Steneosaurus ros tro-minor, Geoffr. t Crocodi lus priscus, Sommermg, ' Den ks chr . Akad, Wiss. Mu nchen,' Vol. v (1814), p. 9, figs. 1, 2, 3 ; .lEolodon p ris cus H. von Mey er, "Rep t. L ith. Schief. " (' Fauna der Vorw elt,' 1860), p. 91. ~ A. Quens ted t, 'Der Jura,' 1858, p. 785, pI. 97, figs. 8-11. HISTORY OF FOSSIL CROCODILES. 315 closely resembling the Continental specimens figured by Quen­ stedt are not unfreqnently met with in several English Kim­ meridge Clay localities, but (with the exception of certain derived teeth in the Potton deposits) none appear to have been definitely identified until 1869, when Mr. Wood Mason· presented a note on the subject to the Geological Society, and Prof. H. G. Seeleyt recorded other specimens from the well-known pits near Ely. Since then, however, all the more important additions that have been made to our knowledge of the form, have been rendered pos-

FIG. 7. Skull of Steneosaurue,

FIG. 8. Skull of Dakosauru« Manselii.

FIG. 9. Skull of Crocodilus, * J. Wood Mason, "On Dokosaur-u.• from the Kimmeridge Clay of Shot­ over Hill," , Quart. Jonrn, Geol. Soc.,' Vol. xxv (lS6!l), pp. 218-2:W. t H. G. Seeley, 'Index to Fossil Reptilia, etc., Woodwardian Museum,' p. 109. 316 A. SMITH WOODWARD ON THE sible by discoveries in British strata, and the portions already de­ scribed consist of the skull, mandible, teeth, vertebral, and frag­ ments of ribs and limb-bones. The whole of the latter-an asso­ ciated series disinterred from the cliffs of Kimmerid ge Bay, by Mr. Mansel-Pleydell-were described by Mr. Hulke." in 1869-70, under th e name of Steneosaurus 1J1 anselii, Hulke , and had been presented by their discoverer to the British Museum, where they havc since occupied a prominent place in the Crocodilian collection. The skull (Fig. 8) differs from that of any typical species of Steneosaurus in the short and stout configuration of the jaws, th e great relative size and consequent fewness of the tceth, and in the fact that the ante. rior pointed ends of the nasal bones enter the hinder border of the external nostril. All these characters tend to place the species much nearer the broad-faced Crocodilian types oflater dat e, and, on a reconsideration of the original specimens, Sir Richard Owen has ac­ cordingly proposed to place them in a new genus, Plesiosuchust-s-e» named in allusion to this circumstanc e (see Figs. 7-9). It is remark­ able, however, that in the paper in question there is no reference to the identity of the teeth with those previously termed Dakosaurus by Quenstedt, and-if Mr. Hulke's determination be correct :j:-the earlier generic name must undoubtedly be retained. The interesting remains of Dakosaurus in the W oodwardian Museum, Cambridge, recorded and briefly noticed by Prof. Seeley (loc. cit.), have not hitherto been fnlly described. They include a tooth from the Coral Rag of North Grimston, which may perhaps belong to a new form,§ and sati sfactory evidence of another species

•J. W. Hnlke, "Notes on some F ossil Remains of a Gavial-like Saurian from Kimmeridge Bay, collected by J. C. Mansel, Esq., establishing its identity with Cnvier's D euxieme Gavial d' Honfleur, Tete d M lIsealt p lus court [Steneosau rus rostra-m inor of Geoffroy St.-Hilaire) and wi th Quenstedt's D okos aurus-" 'Quart. Journ. Geol, Soc.,' Vol. xxv (1869), pp. 390.400, pls, xvii, xviii, and " Note on ll- Crocodilian Skull from Kimmeridge Bay, Dor­ set," op. eit., Vol. xxvi (1870), pp. 167-172, pl , ix. t Sir Richard Owen, "On the Cranial and Vertebral Characters of tbe Crocodilian Genus , Owen," 'Quart. Journ, Geol, Soc.,' Vol. xl (1884), pp. 153-159. See also Owen's' Brit. Foss. Reptiles,' Vol. iii, pp. 146-151 (Orocodilia), pI. xx, figs. 1-4. t M. H. E. Sauvage (" Sur Ie genre Dacosaurus, Quenstedt," 'Bull. Soc. GeoJ. France' [3J, Vol. i, 1873, pp. 380-385) has questioned the correctness of tbis identification. He end eavoured to prove the Mosasanrian affinit ies of the detached teetb met with in Continental deposits, but subsequent re search bas failed to support th e idea. § It would be interesting to compare th is tooth with certain Coralli an specimens in the British Museum, bearing the MS. name of Piiosaurus tere­ tidells, Owen. HISTORY OF FOSSIL CROOODILES. 317

-D. liseocephalus, Seeley- from th e Kimmeridge Clay of Ely. D. lissocephalue is founded upon a remark ably complete skull, but its distinctness from D. lrlanselii ha s yet to be demonstrated; and whenever a detailed study of the associated fossils is undertaken, we may look forward to interesting particulars regarding the rest of the skeletal str uctures, for Prof. Seeley's 'Index,' just quot ed, contains a list of numerous other remains from th e same prolific locality. A Teleosaurian genus, having a skull of somewhat similar pro­ portions to that just considered, is the M achimosaurus of H. von Meyer.· This also occurs chiefly in th e Kimmeridge Clay­ th ough possibly ranging from the Fuller's E arth and through intervening formations t-and until six years ago was only known from detached teeth and other unsatisfactory fragments. At present, the sole recorded instanc e:j: of its discovery in British deposits is a doubtful tooth from the Kimmeridgian of Hardwick, near Oxford, and it is almost impossible to distinguish such isolated fossils (Fig. 13) from the teeth of a W ealden Oroco-

u

J:; .

a .

" ]5. ' -

T EE TH OF F OSSI L CnOCODI LIA. FIG. 11. Stago nolepis. F IG. 14. Goniopholis. FIG. 12. T eleosaurus, FIG. 15. Suchosaurus, FIG. 13. .

* Madrimosauras, H. v. Mey., I Nenes J ahrb.,' 1837, p, 560 i Machimo­ saurus, H. v, Mey., i bid., 1845, p. 310. t H. E. Sauvage , " Momoire sur les Dinosaur iens et lea Crocodiliens des Terra ins J urassiqu es de Bonlogne.s ur-Mer," • Mem. Soc. Gl!ol. France ' [2], Vol. x, Mem , ii (1874), p..49, ct. seq. t J. Phillips, • Geology of Oxford, ' p. 332. Recorded in list of fossils as Goniopholis sp., and suggested to bel ong to Machimosanrus by H. E. Sauvage, op. cit ., p. 50. 818 A. SmTH WOODWA RD ON THE dilian, Goniopholie (Fig. 14); * but of M. Mosa , MM. Sauvage and Lienard t have described the anterior portion of a skull, th e mandible, numerous vertebrre, ribs, dermal scutes, and fragm ents of limb-bones, etc., from the Kim­ meridgian of Daumont, near Issoncourt, and Selenka + also partially elucidates th e type-species, Jll. Hugii, by describing a few teeth and vertebras from corresponding beds in Hanover. Un­ fortunately, th e precise configuration of the nasal bones cannot be ascertained from the French cranial fossil, and other important parts are likewise wanting ; but in the fewness of its stout teeth Machimosaurus more approaches Dak oeaurus than any other de­ scribed Teleosaur, and the very short mandibular symphysis is especially noticeable. The characters of the skull and dentition, indeed, would almost suggest th e removal of these two genera to another Crocodilian family, but such a procedure is scarcely advis­ able upon present evidence.

P urbeck and W ealden Cracodilia . But the Teleosauridro are not confined exclusively to Jurassic strata. Although they attain th eir maximum development in that division of the Mesozoic Series, a few forms are still to be met with in the overlying W ealden Beds, and one or two scattered relics have also been discovered in deposits so late as the Cretaceous. Among British W ealden fossils, only one fragm entary skull has hitherto been described as possibly referable to a member of this aquatic family-namely, the Htjlaiochampsa vectiana of Owen,§ from th e southern cliffs of the Isle of Wight. The specimen, which was discovered by the late Rev. William Fox, M.A., and is now in the British Museum, consists solely of the hinder portion of the skull, and probably belongs to a young individual, so that

* Gon iQ]Jholisand Ma chimosaurlls were regarded as syuouym ous by H. E. Sa uvage, loco cit ., but the statement was withdrawn in 1879 (memoir quoted below), on the disco very of more comp lete re mains of th e last­ named genus. t H. E. Sauvage and F. Lienard, " Mcmoire sur le ge nre MacMmo­ 8aurus," 'Mem. Soc. Geol. France.' [ 3~, Vol. i, Mem, iv (1879). Abstract by H. E. Sauvage in ' Bull . Soc. Geol. FraDce' [3J, Vol. vii (1879), pp. 693-697. ::: Em il Selen ka, " Die f ossilen Kroko dilinen des Kimmeridge von Hannover," ' Pa lreontographic3,' Vol. xvi (1867), pp. 141-2, pI. xi, figs. 17­ 25. MM. Sa uvage and Lienard have evident ly overlooked this pa per. § R. Owen, " Monograph of th e Fossil Ileptilia of the Wealden and Purb eck For mations,' Supp l. vi (' Mon. Pal. Soo.,' 1873). HISTORY OF FOSSIL CROCODILES. 319 its precise affinities as yet can only be regarded as doubtful. But more perfect and satisfactory cranial evidence of a Wealden Teleosaur has been described by Dunker and Meyer * from beds of similar age in Germany, and is known by the name of Macrorhynchus MeYe1'i (Dunker). The vertebrre and scutes of the same form, probably, are also known, although these were originally described under tbe name of Pholidosaurus Schaumberqensi«.t Koken t adds another genus-Enaliosuchus-founded upon vertebrre, ribs, and fragments of limb-bones, from the Lower Chalk of North Germany; and the latest traces of a Teleosaurian hitherto recog­ nised are evidently the detached vertebrre from the Upper Green­ sand of New Jersey, made known by Sir Richard Owen in 1849,§ and generically designated Hyposau1'fJs. An interesting Crocodilian skull and mandible from the Middle Purbeck Beds of Swanage may be appropriately mentioned next, as the fossils already met with are scarcely complete enough to justify any definite conclusion as to their family relationships, and, so far as can be ascertained, indicate a form intermediate between the great group just considered and the Crocodiles of more recent times. This is the Petrosuchus levidens of Owcn 11­ a genus and species with the skull tapering almost as abruptly as that of a Gavial, or, at least, like the living Orocodilus cataphractus i with slender teeth, somewhat unequal in size, and arranged on a festooned alveolar border j and with the temporal fossse and posterior nares distinctly approaching those of the true Teleosaurs. The Wealden and Purbeck Beds, however, are more particularly characterised by Orocodilia of a type much nearer to the river and marsh-dwelling species of the present day, and at least two well-

It H. von Meyer, "Reptilien ans der Wealdenformation Norddeubsch­ lands" (in Dunker's 'Monogr. d. Nordd, Wealdeubildung,' 1846). t Ibid. p. 71, pls. xvii-xix, and 'Neues Jahrb.,' 1841, pp. 443-5. See also paper by E. Kokeu, quoted below, in which the generic identity is re­ garded as probable (p. 824, note). t Ernst Koken, "Die Reptilien der norddeutschen unteren Kreide," , Zeitschr. Deutsch. Gool, Gessll.,' Vol. xxxv (1883), pp. 792·8~4, pl. xxiv, fig. 5, pl. xxv, figs. 1-4. § R. Owen, "Notes on Remains of Fossil Reptiles discovered by Prof. Henry Rogers, of Pennsylvania, U.S., in Greensand Formations of New Jersey," 'Quart. Journ. Geol. Soc.,' Vol. v (1849), p. 383, pI. xi, figs. 7.10. See also J. Leidy, "Cretaceous Reptiles of the United States" (' Smith. sonian Contributions,' 1864), pp. 18.21, with plates. II R. Owen, "Mon. Foss. Rept. Weald. and Purb. Form.," Suppl. viii (' Mon. Pal. Soc.;' 1878), p. 10, pl. vi. VOL. IX., No.5. 23 320 A. SMITH WOODWARD ON THE marked families have already been established upon the abundant materials available. The first of these broad-faced forms that received a name was that constituting Owen's genus (Ioniopholis, 'fI and, though little was known of it at first, the original species has subsequently proved to have so many congeners, that it is now found convenient to group them all together in a family termed GONIOPHOLIDlE ; t Mr. Hulke, indeed, ventured, a few years ago,t to regard their peculiarities as justifying the erection of a new sub-order, META­ MESOSUCHIA, but the advisability of such a procedure has lately been called in question.§ Goniopholis-so named, in 1841,11 in allusion to the rectangular form of the (dorsal) scutes-was founded upon a number of detached teeth (Fig. 14) discovered by Dr. Mantell in the Wealden of Tilgate Forest, and upon a large portion of a skeleton from the Purbeck Beds of Swanage, which also exhibited one or two dental fragments of a similar type. Subsequent discoveries in the Wealden of Sussex, and the Purbecks of Dorset, have contributed further to elucidate the genus, adding two new species, and the recent fortunate acquisition of almost complete skeletons from the cele­ brated Wealden strata of Bernissart in Belgium, seems destined to snpply nearly all deficiencies in our knowledge of its osteology. The skull was not made known until 1878, when Hulke' (followed, in the case of one specimen, by Owen U) was able to describe all its more salient features; and it is only about two years since Dollo tt gave an outline of the general skeletal characters as revealed by the Belgian examples.

• R. Owen, "Report on British Fossil Reptiles," Part ii, 'Brit. Assoc. Rep.,' 1841, p. 69. t L. Dollo, " Premiere Note sur les Crocodiliens de Bernissart," 'Bull Mus. Roy. Nat. Bist. Belgique,' Vol. ii (1883), p. 334. :t J. W. Hulke, "Note on two Skulls from the Wealden and Purbeck Formations, indicating a new Subgroup of Crocodilia," 'Quart. Journ, Geol, Soc.,' Vol. xxxiv (1878), p. 381. § L. Dollo, op. eit., p. 329. II R. Owen, 'Brit. Assoc. Reports,' 1841, p. 69. , J. W. Hulke, paper already cited, 'Quart. Jonrn. Geol. Soc..' Vol. xxxiv, pp. 377'081, pI. xv. It should be added that Prof. Huxley previously described a portion of an indeterminable Crocodilian skull from the Wealden of Brook, Isle of Wight, in his paper of 1875, 'Quart. Jonrn, Geol. Soc.,' Vol. xxxi, p. 432, pl. xix, fig. 3; these later discoveries indicate its probable reference to Goniophoiis. ** R. Owen, "Mon. Foss. Rept. Weald. and Purb, Form.," Snppl, viii (' Mon. Pal. Soc.,' 1878), p. 7, pI. v. tt L. Dollo, note already quoted. HISTORY OF FOSSIL CROCODILE S. 321

Making use of these several sources of information, the leading peculiarities of Goniopholis may be readily summed up as follows. The skull, compared with those th at have hitherto occupied our attention, is short and stout, and in general form approximates to that of the living Crocodil es, which are adapted for th e capture of birds and terrestrial quadrupeds j while the nasal bones do not extend quite to the external nostril, and the situation of the posterior nares upon the palate is slightly further back than their position in the Teleosaurs, and yet more anterior than in the case of still existing Orocodilia, Th e supra-temporal fossre only slightly exceed the orbits in size, and the symphysis of the mandible is short. The body was protected, both dorsally and ventrally, by bony, pitted scutes, the upper shield consisting of only two longitudinal rows, and the lower of eight or ten. The former (Fig. 18) are rectangular, and in addition to repeating the characters of the corresponding Teleosaurian armour, are more firmly united hy a cnr10US "peg-and-socket" articulation, such as occurs among the scales of some ganoid fishes; and the ventral scutes (Fig. 19) arc polygonal, uniting exclusively by suture, without any imbrica­ tion. The vertebrre belong to the amphi ccelian or platyccelian type . It is also interesting to observe that the fore-limbs in the Belgian skeletons - are as long as (if not longer than) th e hinder pair­ -a fact suggestive of their powers of terrestrial progression rather than of swimming. The teeth of the " Swanage Crocodile," t and Mantell's speci­ mens from Tilgate Forest, being remarkable for their thick and stunted shape, the original type-species was named G. crassidens, A second form, described by Hulke and Owen in 1878,:1: is known as G. simus, Owen, and it is to thi s that the magnificent skeletons from Bernissart may be referred : the te eth are more slender than in the first-named species and th e head apparently less tapering. G. tenuidens is the name given by Owen,§ in 1879 , to a fragmen­ tary mandible from the Middle Purbecks of Durdlestone Bay, and indicates a small form, characterised by the slenderness of its

* Well shown in figures by L. Dollo, op. cit., pI. xii, figs. 2, 3. t G. A. Mantell, • Wouders of Geology,' 3rd edit. (1839), Vol. i, pp. 387. 389, pl. i, 1 Memoirs in • Quart. Journ, Geol. Soc.' and • Mon. Pal. Soo.,' already qu oted. § R. Owen, .. Mon. F oss. Rept. Weald. and Purb, Form.," Suppl. ix (' Mon. Pal. Soc.,' 1879), p. 2, pl. i, fig. L 322 A. SMITH WOODWARl> ON THE teeth; and four vertebne, from the Wealden of Cuckfield, are figured in Owen's 'British Fossil Reptiles '.-though not de­ scribed-under the name of G. carinatus. The Middle Purbeck Beds of Durdlestone Bay have also yielded an interesting series of remains of " Dwarf Crocodiles," some of which are undoubtedly referable to the same family. So remark­ able, indeed, are tbe resemblances of many of these diminutive fossils to the corresponding parts of Goniopholis, that an inex­ perienced anatomist might readily be led to look upon them as the young of the larger form; but Sir Richard Owen has conclusively shown that they must rather be regarded as adult individuals, and has accordingly designated them by new generic names.t Nanno­ suchus, (hneuiaeuchus (= Brachydectes, Owen), and Theriosuchus are the genera already determined; and among these there are apparently distinguishable about four species. No complete skeleton of Nanmosuchus is yetknown, but there are numerous fragments in the type-series in the British Museum, which includes, among others, a nearly perfect skull showing the upper aspect, and well-preserved portions of the mandible, with teeth. The cranium is about four-and-a-half inches in length, characteristically sculptured, and very similar in general shape to that of Goniopholis crassidens, The nasal bones do not extend so far forwards as the external nostril, and the supra­ temporal fossse appear to be quite equal to the orbits in size. A " supra-orbital" bone is present, and several teeth are also to be observed, some in their natural positions. The latter, both in the skull and mandible, are slender and sharply pointed, and usually marked with delicate longitudinal ridges; they are somewhat uniform in size-though scarcely as much so as in the Teleosaurs and Gavials-and are placed further apart than in these aquatic forms, thus making a nearer approach to the family now under con­ sideration. The vertebrre are biconcave (amphi ccelian) ; the dorsal scutes agree with those of Goniopholis in possessing a peg-and­ socket articulation; and the humerus betokens a fore-limb of rather slight build, like that of the living Gavialidse, N. g,'acilidens is the only species yet determined. Of Oioeniasuchus, nothing beyond the lower jaw has hitherto been described. It is a genus particularly remarkable from the '" Op, cii., pl. 14 (Crocodilia). t R. Owen, H Mon. Foss. Rept. Weald. and Pnrb. Form.," Snppl. ix (' Mon. Pal. Boc.,' 1879). See also' Quart. Joarn. Geol, Soc.,' Vol. xxxv, pp. 149.152, pl. ix. HISTORY OF FOSSIl, CROCODILES. S2S fact that only about one-third of the alveolar border is provided with teeth, the hinder portion being edentulous; and in allusion to this peculiarity, Sir Richard Owen originally proposed (loc. cit.) to give it the name of Brachqdectee-s ct "short biter." The latter term, however, had already been pre-occupied by Oope" for a Oarboniferous Labyrinthodont, and the present writer has recently t suggested to substitute for it the generic title under which it is here quoted. The portions of mandible of this form are also note-

,a.

-9 FIG. 16. Skull of Tkeriosuckus pusillus (after Owen). FIG. 17. Teeth of Tkeriosuckus pusillus, enlarged (atter Owen). FIG. 18. Dorsal Scute of Goniopholis. FIG. 19. Ventral Scute of Goniopholi«, worthy on account of the absence of any external vacuity towards the posterior end, such as is met with in nearly all other Orocodilia ; and there are probably two species-a smaller (0. minor) and a larger (0. major}, *E. D. Cope, "Synopsis of the Extinct Batrachia of North America," • Proc, Acad. Nat. Sci. Philadelphia,' 1868, p. 214. t A. Smith Woodward, "On the Literature and Nomenclature of British Fossil Crocodilis," I Geol. Mag.,' dec. iii, Vol. ii, 1885, p. 506. 324 A. SMITH WOODWARD ON THE

More perfect, however, are the remains of Theriosuchu«, Not only is the genus known from a well-preserved skull, and numerous detached scutes, vertebrse, and limb bones, but a nearly complete skeleton has also been discovered, and our knowledge of this dwarf form is thus comparatively extensive. Nevertheless, its precise affinities are even yet a matter of dispute, and its resemblances both to the Goniopholidee and the second family-Bernissartidoo-next to be considered, render it advisable to await the results of future research before forming a decided opinion. M. Louis Dollo, the founder of the latter family, suggesta " the possibility of its being correctly placed here; and, consequently, before entering into its structural peculiarities, a brief notice of the type-genus, Bernie sartia, may appropriately precede. This interesting Crocodilian t receives its name from the cele­ brated locality already alluded to above, whence have been obtained the remarkable series of Iguanodons and other vertebrate fossils now in the Brussels Museum. At least one complete skeleton has been met with, which shows that in certain respects Bemis­ sartia exhibits much closer affinities with the living Crocodiles than any other of similar age. The skull somewhat resembles that of Goniopholis in its broad and stout configuration, and also in the non-extension of the nasal bones so far forwards as the external nostril; but the supra-temporal fossse are distinctly smaller than the orbits, and there are also some peculiarities in the arrangement of the teeth. Again, instead of conforming to the Teleosaurian plan (repeated in Goniopholidoo), according to which the dorsal shield consists only of two longitudinal rows of scntes, this genus agrees with the Tertiary Crocodilia in possessing several of such series, and none are united by peg-and-socket joints, but all are imbricating, the hinder edge of one scute overlapping the front edge of that immediately posterior to it. The ventral soutes similarly overlap each other, instead of being exclusively united by suture, as in the Teleosaurs and Goniopholis, and this strong breast. plate-armour consists merely in a single continuous shield. Only one species, B. Fagesii, has yet been described, and in this the fore-limbs are very much shorter than the hinder pair, thus, per­ haps, indicating a chiefly aquatic mode of life. The skeleton already mounted 'at Brussels scarcely exceeds three feet in length. * L. Dollo, op. cit., p. 335. t L. Dollo, op. eii., p. 322, pl. xii, fig. 1, HISTORY OF FOSSIL OROCODILES. 325

Theriosuchus (Figs. 16, 17) seems to have rarely attained a length of more than eighteen inches, and is especially remark­ able from the characters of its skull and dentition. The facial portion of the skull, in advance of the orbits, consti­ tutes only about one-third of its entire length, thus retaining in the adult condition somewhat analogous proportions to those observed temporarily in the young of living Crocodilia; and the teeth (Fig. 17) are more varied in shape-more specialised-than in any other known member of the order. The nasal bones penetrate the aperture of the external nostril, almost dividing it into two, and the posterior nares on the palate are similarly placed to those of Goniopholis. The principal dorsal scutes also resemble those of the latter genus, having the same arrangement and mode of articulation; but the disposition of the remainder is not altogether obvious. The fore-limbs are very much smaller than the hinder pair, thus agreeing with Bernissartia rather than Goniopholis ; and' all the remains hitherto discovered may be referred to a single species, T. pusillus, If, therefore, the reference of Theriosuchus to the Bernissartidre eventually prove correct, the fact of its dorsal scutes being arranged on the Goniopholis plan, and not on the plan of Bernissartia, will become of especial interest; and the great relative size of the supra-temporal fossse will also be noteworthy. The present writer, indeed, is inclined to regard the genus as truly one of the Goniopholidse, SuchOSaUl'U8, Owen, is also a Wealden genus of Crocodiles, but at present it is too imperfectly understood to be relegated to any definite family position. It is founded 1Il upon detached teeth from the well-known strata of Tilgate Forest, characterised by their laterally compressed and slightly curved shape, and ornamented with prominent longitudinal ridges; they are especially remarkable from the fact that the compression is from side to side, and the opposite trenchant "carinre" hence anterior and posterior, while the corresponding edges in most other Crocodilian teeth that are known to possess them are distinctly lateral in position. (See Fig. 15.) Only one species, S. cultrideus, is as yet recognised, and

* R. Owen, "Report on British Fossil Reptiles, Part II," 'Brit. Assoc. Reports,' 1841, p. 67. See also Owen's" Mon. Foss. Rept. Weald. and Purb, Form." Buppl. viii, (' Mon. Pal. Boc.,' 1878), p. 12, pl. iv, figs. 5·8, and , British Foss. Reptiles,' Vol. i, p. 433, pI. 5 (Crocodilia), figs. 5-8. 326 A. RMITII WOODWARD OY THE with the teeth Sir Richard Owen has ventured to associate.a peculiar type of Wealden vertebra that can scarcely be referred to any other Saurian hitherto met with in those deposits. Only one other Crocodilian fossil of Wealden'" age remains to be noted-the so-called Crocodilus ? Saulii, from the neighbourhood of Hastings, Sussex. The evidence of the species consists in a small group of associated skeletal fragments, described and figured by Sir Richard Owen in 1851, t and doubtfully referred at the time to a "very young individual:' The conclusion as to the immaturity of the specimen may be correct, though the discovery of " Dwarf Crocodiles" renders it now uncertain, but subsequent research has failed to confirm the connection of this form with any living genus, and suggests its correct place as being in one or other of the two amphiceelian families just considered. The scutes are destitute of pegs and sockets, as in Bernieeartia. Cretaceous Crocodilia. As already remarked above (p. 319), the Teleosauridre finally dis­ appear in the Cretaceous Strata, and are only represented by a few very imperfect fragments. The vertebrre of Hyposauru8, in fact, are the latest evidence of an amphiccelian (or platy­ ccolian) Crocodilian hitherto discovered.j and in the upper beds of this great series we meet with the first traces of genera, having vertebrre of the same proccelian (concavo-convex) type as those of the present day.§ The Cambridge Greensand and the nearly equivalent Gosan beds of the Austrian Alps have yielded a few vertebrre, indistinguishable from those of the recent Crocodilus, which are accordingly referred to the still-existing form j and from the last mentioned locality some teeth are also recorded with more or less fragmentary limb-bones and a parietal bone, Prof. Seeley founds a new species, C.

II This fossil is said to be derived from the Greensand; Professor Rupert Jones, bowever, informs the writer that the horizon is truly Wealden, and the nature of the matrix strongly supports this conclusion. t R. Owen, "Mon. Foss. Rept. Cretaceous Formations" (' Mon. Pal. Soo., 1851), p. 45, pl. xv. Name given in Owen's' Brit. Foss. Rept.,' Index to Vol. ii, p. vi. t In the new edition of the 'Encyclopoodia Britannica,' art. Reptiles (vol. xx, p. 444), it is wrongly stated that all Amphiccelian Crocodilia are pre-Cretaceous. § The earliest known occurrence of this type of vertebra is in the Pterodactyles, which had already acquired it at the period of the deposition of the Lower Lias (e.g., Dimorphodon, from Lyme Regis) ; but among lizards, as well as crocodiles, the character docs not appear until Cretaceous times. HISTORY OF FOSSIL CROCODILES. 327 proavus,* upon the Gosau fossils, and considers that at least two others-named C. cantabriqiensi» t and C. icenicus:l: -are indicated by the vertebrse of the Cambridge Greensand. But more satisfactory remains of proccelian Crocodilia have been described from the French Cretaceous and from that of the United States. In each case remarkably perfect skulls ure known, besides numerous teeth, vertebrre, and other skeletal fragments j and Prof. Leidy regards some of the American specimens as indicating a new genus, most nearly allied to the living Gavial, but differing (1) in the less abruptly tapering form of the snout j (2) the absence of turned-up (" everted ") edges to the orbits, and (3) in certain details respecting the conformation of the upper cranial bones. Moreover, a curious "antorbital vacuity"-probably natural-is present on each side of the skull j and the distinguished American palreontologist just alluded to proposes § to name these fossils Thoracosaurus neocesa1'iensis, also suggesting that the Gavialis macrorhspichus of Mont Aime, France, probably belongs to the same extinct type. The latter was described by Gervais II in 1852, and the teeth and other relics appear to be comparatively abundant at the locality from whence they were first obtained. It must be added, however, that Prof. Cope' has made known other proccelian forms from the Greensand of New Jersey j and among these, perhaps, the most interesting are the remains belonging to his new genus, Holops. The sknll of this Crocodilian appears to differ but little from that of Thoracosaurus, except in the absence of "antorbital vacuities;" and it is quite possible

* H. G. Seeley, "The Reptile Fauna of the Goaan Formation, preserved in the Geological Mnseum of the University of Vienna," • Quart. Journ. Geol. Soc.,' Vol. xxxvii (l881), pp. 685-693, pl. xxviii, figs. 10,11 ; xxix, figs. 7-13,24,25; xxx, figs. 6.14. t H. G. Seeley, .. Index to Reptilia, etc., in Woodwardian Museum" (1869), p, xvi, and" On Cervical and Dorsal Vertebral of Orocodilas cantabrigiensis (Seeley) from the Cambridge Upper Greensand," 'Quart. Jonrn. Geol. Soc.,' Vol. xxx (1874), pp. 693·695. t H. G. Seeley, "On Orocodilus icenicus (Seeley), a second and larger species of Crocodile from the Cambridge Upper Greensand, contained in the Woodwardian Museum of the University of Cambridge," op. cit., Vol. xxxii (1876), pp. 437-439. § Joseph Leidy, "Cretaceous Reptiles of the United States" (' Smith. sonian Contributions to Knowledge,' 1864), pp, 5.12, with plates. II Paul Gervais, .. Zoologie et Paleontologie franeaises," 1st edit. (1852), p. 252, pl, lxx, figs. 14-24 (2nd edit., 1859, p. 447, pl, lix, figs. 14-24). , E. D. Cope, "Synopsis of the Extinct Babraohia, Reptilia, and Avos of North America," • Trans. Amer. Phil. Soe.,' Vol. xiv (1869), p. 62 (also • Proc, Acad. Nat. Sci. Philadelphia,' 1867, p, 143). 328 A. SMITH WOODWARD ON THE that Gavialis macrorhynchue may have more affinity with this type than with Leidy's genus already mentioned, though Gervais' figures are not sufficient to show the character of the bones immediately in front of the orbits. The Chalk of Britain has not hitherto yielded any undoubted traces of this reptilian order-for Polyptyclzodon is now known to be a Plesiosaur-but Gervais * mentions a Crocodilian vertebra from the higher beds of Maastricht, and Sauvage t has recorded a few indeterminable fragments from the Gault round the Paris basin. Eocene Crocodilia, Even the earliest of the Eocene strata in this country appear to be characterised by none but living genera. There are, indeed, some indications of these types being not quite so differentiated as happens to be the case at the present day, but altogether the general correspondence is remarkably close. Another feature is also especially noteworthy. Whereas, at the present time, there is no region of the globe in which all three families of Gavialidse, Crocodilidre, and Alligatoridre, are known to exist together, remains evidently referable to each group are found associated in the Lower Tertiaries, both of this country and of other districts on the Continent. In the Eocenes of the London and Hampshire Basins, for example, at least one Gavial, two Crocodiles, and an Alligator-like form have been met with; and Ludwig j also records abundant evidence of a Crocodile and an Alligator in the Oligocene beds of Mayence. The first notice of a Crocodile from the London Clay of Sheppey was by Baron Cuvier,§ who based his determination upon a detached cervical vertebra; an imperfect skull was subsequently discovered by Mr. Spencer, in 1831, and figured by Buckland II under the name of G. Spenceri >' and in 1850, the Palseontogra-

• P. Gervais, op. cit., 1st edit., p. 252; 2nd edit., p. 447. t H. E. Sauvage, "Recherches sur les Reptiles trouves dans Ie Gault de l'est du Bassin de Paris," 'Mem. Soc. Geol. France' [3J, Vol. ii (1882), Mem. iv, p. 20. A proecelian vertebra is figured in pI. iv, fig. 5. t R. Ludwig, " Fossils Crocodiliden aus del' 'I'ertiarformabion des Mainzer Beckens," 'Palreontographica,' Suppl. Vol. iii (1877), parts iv, v, § G. Cnvier, 'Ossemens Fossiles,' 2nd edit. (1824), Vol. v, pt. ii, p.165. II Rev. W. Buckland, 'Geol. and Min.' 2nd edit. (1837), Vol. i, p. 251 ; Vol. ii, p. 36, pl. 25', fig 1. HISTORY OF FO SSIL CROCODILES. 329 phical Society published an exhaustive monograph by Prof. (Sir Richard) Owen,· which embraced an account of all the fossil remains of this order, from British Eocene deposits, at that time available. With the exception of a brief, but valuable contribution to our knowledge of the Hordwell species, by Prof. Hnxley.] no further additions to the literature of th e subject seem to have been made since the appearance of this classical memoir j aud the subsequent discoveries of Crocodilian remains in the same beds have apparently afforded no evidence of any new specific form. Among the Sheppey fossils, two species are indicated not merely by the skulls, but also by numerous portions of th e vert ebral column j there is, however, no very definite information as yet con­ cerning the rest of th e skeleton, and isolated fragm ents of limb­ bones and dermal scute s are the only remains hitherto forthcoming for study . A very fine skull in the British Museum forms the type-specimen of Owen's C. ioliapicus, and is figured and elucidated in th e memoir already cited . Its general proporti ons agree most closely with those of th e skull in the living C. acutus, from the West Indies, and its alveolar borders are slightly festooned, though provided with teeth remark ably uniform in size. The symphysis of the mandible is also especially noticeable, being unusually long for a species of th e present genus. C. champsoides, Owen.j is characterised by a much more elongated skull-somewhat similarly shaped to that of the living T omistoma Sc hlegelii of Borneo-and resembling this latter species in the straighter contour of its alveolar border j th e teeth likewise approach those of T omistosui in slender­ ness and regularity, bnt the nasal bones differ in reaching forwards to th e external nostril, notwithstanding the length of the snout, instead of terminating some distance behind. Of the vertebral from Sheppey, Sir Richard Owen provisionally refers the stoutest to C. toliapicus, and the more slender to C. champsoides; and ex­ presses some doubt as to th e possibility of determini ng the pr ecise

* R. Owen, 'Monograph of tho Fossil Reptilia of the London Clay,' part ii (' Mon. Pal. Soc.,' 1850). t T. H. Huxley, "On tho Der mal Armour of Crocodilus Hastinqeiai;" 'Quart. Journ. Geol, Soc.,' Vol. xv (1859), pp. 678-680, pl. x xv, t A species closely allied t o this bas been described fr om tho Eocene of "Verona. by A. de Zigno, "Sopra nn oranio di Ooccodrillo," 'Atti R. Accad. Lincei,' (3) 1'01. 5, pp. 1-10, pIs. i, ii, 330 A. R?tIlTH WOODWARD ON THE position of Buckland's C. Spenceri, thus relegating this imperfectly characterised form to the rank of a synonym." A series of Crocodilian vertebroo, discovered by Mr. Frederic Dixon in the Lower Eocenes of Bognor, Sussex, probably belongs to C. toliapicus (Owen).t The most completely known of Tertiary Crocodiles hitherto met with in England is the C. Hastinqsice, Owen,t of which numerous remains are discovered in the celebrated Upper Eocene beds of Hordwell Cliff, Hampshire. Almost every portion of the skeleton, including remarkably fine skulls, have been found in this prolific locality. The species is especially noticeable on account of the manner in which it combines the peculiarities of both Crocodiles and Alligators; and while most of the skulls exhibit the Urocodilus­ modification in the so-called lower" canine" being exposed, one or two instances are known in which this tooth was received in a pit on each side of the upper jaw.§ and Searles Wood once described such a fossil under the new specific name of Alliqator hantoniensis.1I The general contour of the skull is very suggestive of that of Alli­ gator, and the position and aspect of the posterior nares is likewise somewhat similar j but most of its characters seem quite interme­ diate between those of the two genera; some, indeed, are quite peculiar to itself-as, for example the disposition of the short and broad nasal bones, which do not enter into the border of the exter-

* On this curious method of solving the difficulty, see the remarks of Dr. Leon Vaillant, "Etude zoologiqus sur les Crocodilfens fosniles tertiaires de St.·Gptand Ie Puy," 'Biblioth. I'Ecole Hantes Etudes,' Vol. vi, pp.lO, 11. It may be added that this memoir contains valuable information rospecting European Tertiary Orooodilia described before its date of publication (1872). t R. Owen, Mon. Foss. Rept, Lond. Clay,' pt. ii, p. 36. See also Dixon's , Geology of Sussex,' Ist edit., p. 207, pI. xv, figs I, 2 (2nd edit., p. 253, pl. xv, figs 1, 2), in which Sir Richard Owen names it " G. 8penceri, Buck. land." t R. Owen, ., On the Fossils obtained by the Marchlcness of Hastings from the Freshwater Eocene Beds of the Hordle Cliffs," •Brit. Assoc. Rep. 1847, Trans. Sections,' p. 65. See also, 'Mon. Foss. Rept. london Clay,' pt. ii, pp. 37~42 , pis. vi-Ix j xil, figs 2, 5. Some fragmentary fossils from the Swiss Eocenes are probably referable to this species. according to F. J. Pictet, 'Paleontologic Sulsse.c--Vertebres de Ill. Faune eocene,' p. 89, pl. vii. § It is interesting to notice that Prof. Cope has described a curious speci­ men of a South American Alligator, in which the lower" canine" on one side fits into a notch, while on the other the corresponding tooth is received in a pit (' Trans. Amer. Phil. Boc.,' Yol. xiv, 1869, p. 83). " Searles V. Wood, "On the discovery of an Alligator and of several new Mammalia in the Hordwell Cliff;' Charlesworth's' London Geological Jour. nal,' 1846, p. 6, pis. I, 6, 7. HiSTORY OF FOSSIL CROCODILES. 331 nnl nostril. Prof. Huxley (op. cit.) has further demonstrated that the dermal armour of this species was more complete than that in the living Crocodiles, and the Hordwell form resembled the South American Caimans and J acares in possessing a ventral as well as a dorsal shield. Reasoning from analogy, the ridged and keeled scutes form the dorsal armour, and the flat scutes-with lateral sutural edges, and an anterior smooth portion united by suture to the larger ornamented portion-constituted the covering of the ventral surface; and all the fossils hitherto discovered-whether teeth, limb-bones, vertebral, or scutes-appear to confirm the con­ clusion founded upon the skulls, that only a single specific type is represented. Indications of the Gavial in British Eocene formations appear to be very rare, but undoubted fragments of a mandible were dis­ covered many years ago by Mr. Dixon in the Bracklesham Beds of Sussex, and Sir Richard Owen has ventured to associate with these a gavial-like vertebra and femur from the same deposit and locality. The original specimens are now in the British Museum (Dixon Collection), and were described in 1850'" under the name of Gavialis Dixoni, Owen.

Miocene and Pliocene Crocodilia,

In Great Britain no undoubted remains of this reptilian order are known from deposits above the Hempstead Beds, t but on the Continent there is distinct evidence that a few species still survived even so late as the time when Pliocene strata were accumulating. At the latter period, probably, Crocodilians became extinct in this part of the globe. The Miocene Strata of France have yielded abundant relics of Crocodilue (=Diplocynodon) Ratelii,t etc., and evidence of addi­ tional species is met with in other European areas. Ludwig, as

• R. Owen, 'Mon. Foss. Rept. London Clay,' pt. ii. (Mon. Pal. Soc., 1850), p. 46, pI. x, Also F. Dixon, ' Geology of Sussex,' 1st edit., p. 208 (2nd edit., p. 253). Teeth possibly referable to this species have also been recorded from the Middle Eocenes of Belgium. Mr G. F. Harris kindly furnishes the writer with the following references:- Mourlon, 'Geologie de la Belgique,' Vol. ii (1881), p. 168; Rutot and Vincent, •Ann. Soc. GIlD!. Belg.,' Vol, vi (1879), p. 124 ; Vincent, 'Ann. Soc. Malacol. Belg.,' Vol. x (1875), p. 27. Winkler, however, questions the accuracy of these deter­ minations in 'Archives Musee Teyler,' Vol. iv, fasc. I, p. 29 (1874). t W. Boyd Dawkins, " Early Man in Britain," 1880, p. 42. :j: See paper by Leon Vaillant, already quoted. 332 A. SMITH WOODWARD ON THE already remarked (p. 328), has given an elaborate account of a Crocodile (C. Ebertsi) and an Alligator (A. Dm'wini) from the Lower Miocene (Oligocene) of Mayence, and the scattered papers containing accounts of minor remains are far too numerous to mention.v Interesting discoveries are also recorded from the well-known beds of the same age in Malta, and these probably belong to the gavial-like genus Tomistoma.i which is now confined to Borneo. One specimen, however, in the British Museum, has been quoted under the MS. name of Melitasaurus, Owen, and another, described by Mr. Hulke.j was originally assigned to a species of Orocodilus-C. gaudensis, Hulke, Undetermined Croco­ dilian remains from Malta are also recorded by Professor Leith Adams.§ In India, as might be expected, Crocodilians occur in most of the later Tertiary deposits-some forms almost identical with those now living there. From the Pliocene strata of the Siwalik Hills, Dr. Hugh F'alconcr] determined four species, and Mr. Lydekker has recently~ published a more elaborate study of the same specimens. It would appear from the later researches that these Pliocene fossils comprise two species of Crocodiles, and at least five Gavials, in addition to a remarkable gavial-like genns, termed Rham­ phoeuchue, One of the Crocodiles (C. si valeneie) is evidently the ancestral form of the existing C. palustrie, and a species of Gavial cannot be distinguished from the G. gangetieus of the modern

., See an excellent bibliograpby at the end of the paper by Toula and Kail, quoted below. Another recent contribution is by A. Hofmann, "Crocodiliden aus dem Miocaen del' Steiermark," in Mojsisovics and Nenmayr's' Beitr. Pal. Oest.-Ungarns,' Vol. v, pp. 26-35, pls, xi-xv, t R. Lydekker, "On the Occurrence of the Crocodilian Genus Tomisioma in the Miocene of the Maltese Islands," 'Quart. Journ, Geol. Soo.,' Vol. xlii, (1886), pp. 20-22, pI. ii, figs. 1, 2. A species of the same genus from the Miocene (P) of Eggenburg, in Austria, has recently been described under the name of Gaoialosuchus eggenburgensis (Toula and Kail, "Ueber einen Krokodil-Schadel aus den Tertiiirablagerungen von Eggenburg in Nieder­ osterreich," 'Dellkschr. k. Ak. Wiss. Wien,' Vol. 50, 1885, pp. 299·356 pls, i-iii), t J. W. Hulke, "Note on some Reptilian Fossils from Gozo," ibid., Vol. xxvii (1871), pp.30-32. § A. Leith Adams," On Remains of Mastodon and other Vertebrata of the Miocene Beds of the Maltese Islands," ibid., Vol. xxxv (1879), p. 527. II See Falconer's 'Palmontological Memoirs,' Vol. i, pp. 344-358, pls, 28,29. ,. R. Lydekker, " Siwalik Crocodilia," etc., "Paleeontologla Indica,' ser, x, Vol. iii, pp. 209-235, pls, xxviii-xxxiv (1886). HISTORY OF FO!lSIL CROCODILES. 333

rivers. Rhamphosuchus was a huge reptile, probably attaining a length of fifty or sixty feet, and is particularly inte resting from the fact th at although it possessed th e long snout of the Gavials, the teeth were arranged almost exactly as in th e living Alli gators -the first pair of lower teeth passing into notches in th e upper j aw, but the fourth pair received in pits, and the succeeding lower teeth biting on the inner side of the upper.

IV.-Ev OLUTION OF TH E CROCODILI.A.

Having thus briefly summarised the present state of our know­ ledge of the Crocodilian order, and pointed out the main sources of information, we are now prepared to proceed with a general sketch of the various Biological conclusions that appear to be justifiably deducible th erefrom. There are some groups of living beings whose origin dates from so remote a period that the very earliest of their fossil remains hitherto met with present all th e essential characters that have continued peculiar to the tribe ever since : the imperfection of th e geological record, in fact, or some other less evident circumstance, completely conceals t he primary stages of many groups from th e view of the scientific inquirer j and there are some whose begin­ ning and ending seem abruptly marked off and enshrouded in almost complete mystery. The Tur tle-and-Tortoise order, for example, was probably well-defined so long ago as the Triassic age j and, nevertheless-so far as known-is not represented by fossils of an earlier date. The Pterodactyles, too, are suddenly encount ered, with all th eir leading features, in the Lower Lias, and continue to occur, with remarkably few signs of structural modification, in all deposits until the Chalk, when they as mys­ teriously disappear. The Ichthyosaurs , as yet, are equally per­ plexing j and we might almost add the Plesiosaurs, except that in th is case light is ju st beginning to dawn.* But th e palreontological history of the Crocodiles is far more satisfactory, and affords much more evidence for determining the manner in which they have attained their present state of develop­ ment. On the whole, th ey may, perhaps, be looked upon as the

* H. G. Seeley, "On Ne usiicosaurus pusillus (Fraas), an Amphibious Reptil e having affinities with the Terrestrial Nothosau ria and with the Marine Pl esiosanria," • Quart. J ourn, Geol. Soc.,' Vol. xxxviii (1882), pp. 350-366, pl. xiii, 334 A. 8~IITH WOODWARD oN THE most « specialised" of existing reptiles-those in which there is the maximum departure from the primitive reptilian type-and the study of the successive steps by which this stage has been reached is thus surrounded with unusual interest. Whereas, among Lizards and Chelonians, it is only possible to glance at a comparatively un­ eventful history extending over long geological ages, we can glean from the known Crocodilia some most important facts bearing upon the question of the plasticity of organic types and the dascent of one form from another j and it is our purpose now to direct special attention to a few of their leading features, and finally con­ clude by showing how they provide a sure basis for the natural classification of the order. As already remarked in the detailed narrative (p. 300), the differentiation of this great reptilian tribe had advanced but slightly in the Triassic period. At that early age, in fact, Lizards, Dinosaurians, and Crocodiles were-to use a homely expression­ " very much mixed j" and it is almost impossible in some cases to refer the different genera to anyone of these orders as later de­ fined.- Our only resource is to make a general list of the char­ acters, and decide in favour of those that predominate j and, as regards Belodon, it may be noted that perhaps the most important Crocodilian resemblances are to be found in the situation of the posterior narial openings upon the palate. Although-as will be shown presently-a secondary posterior opening has been developed in later forms, the forward production of a process from the ptery­ goid bones to meet the vomers is a character not met with in any other known order, except the Rhynchocephalia,t and is especially well-marked in this very ancient genus. Belodon is further essentially Crocodilian in the "shoulder girdle," where an interclavicle is present, without any trace of clavicles j and the dermal armour is also most closely paralleled in the same great group. In having a fixed quadrate bone, however, it

* e.g., Aetosauru«, Fraas ("Aetosaurus f'erratus, Fraas, Die gepanzcrte Vogel.Echse ans dem Stubensand stein bei Stuttgart," Stuttgart, 1877). t The Rhynchocephalia (Gunther) constitute an order of especial in. terest to the palseontologist. The ir only living representative is the" Tnatera Lizard "-Spkenodon (= Hatteria) punctatum-of New Zealand, but in past times they were evidently an important tribe. The early Mesozoic reptiles, especially, whether referable to this order itself or not, display numerous signs of affinity with it. See A. Gunther, "Contribution to tho Anatomy of Hatteria (Rkynckocephatu8, Owen)," 'Phil. Trans.,' 1867, pp. 595·629, pls. xxvi-xxvii. HISTORY OF FOSSIl, CROCODILES. 335 agrees not only with Crocodiles, but also with Dinosaurs and Rhyn­ chocephalians. The ribs have a double articulation-as in Croco­ diles and Dinosaurs-and the position and upturned aspect of the transverse processes on the anterior thoracic vertebne are decidedly Dinosaurian characters. Finally, the shape and disposition of the pelvic bones are suggestive of similar arrangements among the Lizards. In tracing the successive modifications by which this primitive type has advanced towards the more specialised form of the existing fauna, it will be most convenient to consider a few of the leading characters separately, and collect the main points of our present information regarding each. Commencing with the skull, the early change in the situation of the external nostril is especially noteworthy j while in Belodon , from the uppermost Trias, the opening is double and placed just ill advance of the orbits, the Teleosaurs of the Upper Liassic Beds exhibit the ordinary undivided nostril, quite at the end of the snout. However, the alteration is not so extensive as might at first appear, for it consists in very little beyond the reduction of the immense premaxillre, and the corresponding increase of the maxillse ; and we may yet look forward to the discovery of inter­ mediate stages in the Rhsetics and Lower Lias, which have hitherto yielded no remains. The modification of the supra-temporal fossre has proceeded much more gradually, and-so far as is known-the modern type was only definitely approached towards the commencement of so late a period as the Wealden. All the Teleosauridse were char­ acterised by these fossse being of considerably greater extent than the orbits, and the Purbeckian family of Goniopholidse is the earliest yet discovered in which they exhibit any well-marked re­ duction in size. Here they are still little larger than the orbital vacuities themselves, but in a contemporaneous family-that of Bernissartidre-they become even further reduced, and are almost as small, compared with the orbits, as those of any living Orocodilus. At the present time they appear to be largest in the Gavials and smallest in the Alligators; and among some of the latter it is not unusual to find them completely covered with bone.

VOL. IX., No.5. 24 336 A. SMI TH WOODWARD ON THE

Moreov er, in the disposition of th e postfrontal bones,* th e Teleosauridre, and other Mesozoic forms, depart much less from the generalised reptilian type than do any Crocodilia of the pres ent day; and M. Doll o t considers that the peculiarities of their downwardly-directed processes indicate the absence of an external ear. The general contour of the hinder part of the skull was also less modified than happens to be the case now, and admirably suited, by it s perfect wedge-like form, for rap id prog ression through wate r. Ev en so aquatic a genus as the living Gavial ha s the sides of the cranium behind th e orbits approximately parallel -just as in all its contemporaneous allies-and the later Mesozoic members of tb e order appe ar to be somewhat intermediate between the two extremes in thi s interesting particular. A comparison of th e configuration of the brain in successive gene ra would also lead to instructive results; but only three exam ples appear to ha ve been hitherto described, and two of those are somewha t imperfect .j One made known by Prof. Seeley, § however, is enough to show th at the Teleosauri an cerebral char­ act ers differ considerably from those of existing Crocodiles, and bear an interestin g resemblance to th e modifications found in other early Mesozoic reptilian orders . But the most rema rkable character to be noted as we look back upon the historical records of the Crocodilian trib e is that relating to th e position of th e posterior nurial opening upon the palate. When briefly studying a typical recent skull (p. 294 ), we observed that th is aperture was placed quite at th e hinder end of the mouth -not in th e forward or median position ordinarily seen in the reptili an class; but a little closer exam ination will show that th is remote situation is only quite of secondary origin, th e true " pos-

* Pointed ant by Professor (Sir Richard) Owen in 'Ann . and }fag. Nat. Hist. ' [3J, Vol. i, 1858, p . 459. t L. Dall a, • Bull. Mus. R oy. d'Hist. Na t. Belgiqu e,' Vol. ii (1883), p. 317, note. :t: Morel de Glasvill e, .. Sur Ill. cavit6 cr anienn e et la position du trou optique d ans Ie Steneosaurus Heb erti;" r Bull. Soc. Geol. Franc e' [3], Vol. iv (1876), pp, 342.347, pls, viii, ix, H . G. Seeley, "On th e Cranial Char. aeter s of a large Teleosaur from th e Whitby Lias," •Quart. Journ. Geol. Soc. ,' Vol. xxxvi (1880), pp. 627-634, pI. xxiv. V. Lemoine... Note sur I'Encephale du Gav ial du :Mont Aime [ Gavialis macrorhynchus], etudie sur trois moulages na turels, " r Bull. Soc. Geol, France,' [3] Vol. xii (1884), pp. 158.162, pl. iv, § L oc. cit. HISTORY OF FOSSIL CROCODILES. 337 terior nares" being exactly in their usual place, and merely covered up by plates of bone that have grown out from the palatines and pterygoids, and thus prolonged the nasal passage." Whether this arrangement has special relation to the precautious against choking necessary to enable the Crocodile, with wide-open mouth, to drown its captured prey,t or whether there is some less evident cause, is a question for physiologists to decide; but the anatomical fact re­ mains the same, and it is exceedingly interesting to trace the successive stages by which so remarkable a specialisation bas taken place. Fortunately, our series of fossil remains is so extensive that we need very few more to demonstrate completely that there is an unbroken passage between the primitive and the modern type; and the following diagrammatic sketches from genera already known will indicate the more important intermediate conditions .

20

2 1.

* For an excellent detailed description of the arrangement of these bones in the Crocodilia, see Prof. Huxley's I Geol. .Survey Memoir on Stagonolepis,' or Prof. Miall's 'Skull of the Crocodile.' t Such an explanation will hardly suffice in the case of the Gavia.l, which feeds upon fish. 338 A. B~I1TH WOODWARD ON THE

.25.

FIGs. 20.25. Diagrams of Palates of Crocodilia, to show position of Posterior Nar es (P.N.). 20, B elodon , 21, P elaqosawr-us , 22, Steneosaurus ] 23, Theriosuchm«, 24, Young Orocodilus , 25, Adult Orocodi lus, [The maxillre are represented by cross-shading; palatines by oblique lines; pterygoids are dotted.] HISTORY OF FOSSIL CROCODILES. 339

In Belodon (Fig. 20)-from the Trias-the posterior nares are shown in their primitive situation, and there are no outgrowths either from the maxillaries, palatines, 01' pterygoids. In Pelago­ saurU8 (Fig. 21)-from the Lias-both maxillaries and palatines are produced into thin plates, and thus slightly prolong the nasal passage, causing the secondary "posterior nares" to appear between the hinder end of the palatine extensions. The palate of Steneosaurus (Fig. 22) and -of Oolitic age­ exhibits the secondary opening somewhat more remote in position. Theriosuchus (Fig. 23)-from the Wealden-makes a still nearer approach to living forms in this respect. And the Tertiary Croce­ dilia have the posterior nares almost, if not quite, as remotely placed as those of the existing members of the order (Figs. 24,25), in which the aperture is entirely surrounded by the pterygoids.*' It must be remarked, however, that the fact of the nasal passages in the old, Mesozoic forms being ossified for a less extent than in the later, Cainozoic members of the order, does not necessarily imply that the absolute posterior nares were placed much further forward than now. Upon this point, Prof. Eugene Deslongchamps has made some interesting observations,t and is inclined to believe that membranous extensions-in the Teleosaurs at any rate-were probably present, and not only served to prolong the passage, but formed small reservoirs for air, corresponding to the ossified " bullse" of existing Gavials. As might be expected, the various other parts of the skeleton confirm the conclusions suggested by the foregoing study of the modifications in cranial characters, and there are a few of these specially worthy of mention. In the Triassic genera Belodon and Stagonolepis, the vertebral all exhibit a well-marked amphicrelian (or biconcave) structure, and in many of the succeeding Teleosauridre the same feature is

* The facts of embryology usually agree in a remarkable manner with those of paleeontclogy-s-the history of the individnal repeating the history of the race. It is, therefore, interesting to note that the extensions of the palatine and pterygoid bones which prolong the nasal passage appear com. paratively late in the embryonic history of the Crocodile's skull, See observations of H. Rathke (' Untersuchungen iiber die Entwickelnng und den Korperban del' Krokodile,' edit. Wittich, 1866), translated by Prof. L. C. MiaH in his 'Skull of the Crocodile,' p. 49. Even in the hatched young of the living species the posterior nares are slightly fnrther forward (Fig. 24) than in the adult (Fig. 25). t E. E. Dealongohamps, 'La Jura Normand,' Mon. iv, p. 27. 340 A. SMITH WOODWARD ON THE very plainly evident; in later Mesozoic genera, however, the hollowing is often scarcely perceptible, and these centra, having thus almost flat ends, may be more appropriately termed" amphi­ platyan" (Owen). Hitherto, no forms intermediate between this and the well-defined procoelian modification seem to be forth­ coming, but we still know but little regarding the genera existing at the time when the lower Cretaceous beds were being deposited, and thus there is yet abundant hope. The pelvic bones suddenly attained their normal Crocodilian characters at a very early period-some time in the comparatively brief, but dark interval between the Upper Trias and the Upper Lias. Even the Pelagosaurs and Mystriosaurs, of Whitby and Wiirtemberg, show the peculiar exclusion of the pubis from the acetabulum (Fig. 4), and the corresponding process of the ischium i and the ilium has also changed from its more generalised shape, to become readily distinguishable as quite Crocodilian. Since the deposition of the Lower Oolites, indeed, there appears to have been no very marked alteration in the pelvis; and the limb bones have also remained substantially unaltered since this remote period, having considerably departed from the Belodon-type during the same mysterious interval. With regard to the dermal armour, it may be noted that" fashion" has changed but slowly. Belodon and Stagonolepis seem to have had a dorsal shield consisting of two longitudinal rows of broad carinated scutes, and the last named genus, at least, had also a ventral armour of flat scutes arranged in about eight longitudinal series. AU the Teleosaurs seem to have been characterised by a pro­ tective covering arranged upon the same plan, and only differing in the absence of keels on the dorsal scutes, which arc thus simply angulated. The Goniopholidre, of later times, still exhibited a very similar disposition of armour, and departed but little from the Teleoaaurian type, except in .having the component parts of the dorsal shield firmly united by peg-and-socket joints. (Fig. 18). The Bernissartidse, of Wealden age, made a much nearer approach to the" style" of the present day-just as they were more modern in every other character-and were armed on the back with several rows of imbricated scutes, and also protected somewhat similarly on the ventral aspect. The Upper Cretaceous Crocodilia probably resembled living forms even further still, and the earliest decided evidence of the same mode of disposition of the scutes as that HISTORY OF FOSSIL CROCODILES. 341 observed in the existing Caimans of South America, appears to be the case of Crocodilus HastingsiCB, from the Upper Eocenes in Hordwell Cliff, Hampshire. On the whole, therefore, there are abundant indications that the Crocodiles, Alligators, and Gavials of the existing fauna, are the terminal forms of a long chain of reptilian life that has been con­ tinually advancing from a generalised to a specialised condition as ages have rolled away; and the comparatively unbroken character of the series leads us to believe that the later type has always been developed from the older-that here, in fact, there is demon­ strative evidence of evolution. Beyond so general a statement, however, it is very difficult to proceed; and before it is possible to enter safely upon minor de­ tails, it will be necessary to await extensive accessions to our present incomplete knowledge. Among other perplexing factors that contribute to this difficulty of making out precise genealogies, there is the circumstance that for long periods-if not always­ there have coexisted Crocodilians adapted for a more terrestrial mode of living, and others fitted for a home almost exclusively in aquatic surroundings. These are the" longirostres " and the "breviros­ tres" of Dollo-the aquatic genera (like Teleosaurs and Gavials) having long snouts, and the more terrestrial genera (like Gonio­ pholids and Alligators) having a cranium short and strong. Such being the case, and a few intermediate forms, both past and present, being known, it becomes quite impossible as yet to determine mnch regarding the ancestral relationships of the various genera-if, indeed, there are ever materials for so doing to the satisfaction of all biologists concerned. As to the manner in which these changes have been brought about, and the nature of the immediate exciting cause, science at present can offer even less explanation. It has been enough, hitherto, to point out suggestive facts and hint at interesting coin­ cidences, and Sir Richard Owen's valuable remarks on the Purbeck CrocodiIia* (with the discussions that followed their reading) are especially worthy of study in this connection. This veteran

* R. Owen, " On the Influence of the Advent of a Higher Form of Life in Modifying the Structure of an Older and Lower Form," , Quart. Journ, Geol. Soc.' Vol. xxxiv (1878), pp. 421-430; also" On the Association of Dwarf Crocodiles (Nannosuchus and Theriosuchus pusillus, e.g.) with the Diminutive Mammals of the Purbeck Shales," ibid., Vol. xxxv (1879), pp. 148-155, pl. ix, 342 A. SMITH WOODWARD ON THE worker among fossil remains shows how the (already known) period of incoming of broad-faced genera agrees very closely with the time when mammalian life must have been becoming com­ paratively abundant; and not only that, but discusses the remark­ able association of the dwarf forms of this type with the remains of diminutive mammals in the well-known Middle Purbeck Beds of Swanage. A short and stout skull is particularly adapted for the capture of terrestrial prey and its firm prehension while drown­ ing, and the concomitant strengthening of the forelimbs also tended towards the greater facility of movement on land, requisite for successful raids upon the denizens of the river banks. The" everted "-or upturned-edges of the orbits, peculiar to later Tertiary Crocodiles, are also very interesting when con­ sidered in relation to enemies and prey This character appears to be unknown even so late as the Eocene period, and is probably connected with the circumstance of both foes and food-supplies being above them, instead of round about in the same watery element; for it is a character enabling them "to peer abroad and scan the banks of the stream with the least possible exposure of their head," 'it< and thus greatly to their advantage uuder these pnrticulur conditions. But, avoiding the realms of theory and speculation, it must be enough now merely to conclude our hasty sketch by referring to the ultimate effort of all biological research, namely, the natural arrangement of the group under consideration. The first attempt at a classification embracing both extinct and living Crocodilia, appears to have been made by Prof. (Sir Richard) Owen, in 1841.t Regarding the vertebral column as affording the most fundamental distinctive characters, he proposed to divide the order into three main sections (afterwards) known re­ spectivelyas AMPHIClELIA, PROClELIA, and OPISTHOCCELIA, in allu­ sion to the shape of the centrum. The genera with biconcave or flat-ended vertebrse belonged to the first sub-order; those of the ordinary concave-convex type, to the second; and those possessing

* See lecture by Prof. Sir Richard Owen, delivered on April 29th, 1858, and reported in • Ann, and Mag. Nat. Hist.' (3) Vol. i, pp. 41;6-463. This report contains much valuable aud suggestive information regarding the Teleosaurian skull and that of the later Crocodiles. t R. Owen, .. Report on British Fossil Reptiles," part ii, ' Brit. Assoc. Rep.,' 1841, p. 65, et seq. See also Owen's' Palesontology,' 2nd edit., 1861, pp. 298-306. HIST OR Y OF FOSSIL CROCODI LES . 343 cervical vert ebr as with th e "cup " posterior and the " ball " anterior, to the third. The two former sections still remain natur al, though th e pr ogress of research ha s rend ered t hem capable of considerable further subdivision; but th e Opisth occelia-repre­ sented by (letiosaurus and" Streptospondylus ""'- are n ow known to have been erroneously associated wit h the Crocodiles, and properly belong to the Din osauri an tri be.] Th e Oweni an system, however, was a decided step in the advancement of kn owledge , and did not.' g ive place to anything more elaborate unt.il 1875, when Prof. Huxley und ertook a revision of th e order and read his well-kn own memoir before th e Geological Society. :j: Provided with abunda nce of new materials, he was able to define three well-mark ed stages in the history of th e Crocodilian ph ylum-stages ind icat ed by a11 th e more important features in the skeleton of each successive type; and every subs equent dis­ covery mentioned in t he forego ing account either tend s to fill up a ga p, or affords additional evidence t hat we are not far now from an approx imate ly natural arrangement . 'I'iiis classificat ion (with a few additio ns) may be briefly stated as follows, and will sufficiently explain it self :-

ORDER,CROCODILIA.

SUB-ORDER I.-PARASUCHIA. P osterior nares opening toward s anterior end of mouth, the nasal pas sages not being prolonge d by osseous plates either from t he palatine or pte rygo id bones . Eus­ tac hian pass ages not enclosed by bone. Vcrtebrse amph iccolian, P ubis entering into eonstitntion of aceta bulum, and th e pelvis lizard-like. Two long itudinal seri es of articulated, carinated, dorsal scutes ; and in S taqonolepis (but appa rently not in B e/ndon) there is a ventral th oraco-abdominal shield, form ed of not mor e

* Th is is wrongly quo ted as an Amphicmlia n Crocodile in th e new edition of the ' En cyclopred ia Brita nn ica,' art. Reptiles (Vol. xx, p. 444). t A curious error occu rs in Claus' 'Text-book of Zoology,' recently t ra nslated into English (Sedgwic k and Heathcote, Vol. ii, p . 224), wh er e S teneosaurus is said to possess opisthoccelia n vertebrte. Thi s statoment is evide ntly copied from some old work on Paleeontology-i-e.jr., th at of Pict et (1853) or Giebel (1847}-for th e m istake had been rectified eve n before 1860, when Owen published his Pal zaontolog ieal Mannal; and Cetiosaura« was als o removed from th e Crocodilian order at least sixteen years ago. ~ T. H. Huxley, "On S taqono lepis Iiober tsoni, and on the Ev olution of t he Crocodilia," 'Quart. J onrn. Geol, s oo.,' Vol. xxxi, Pl'. 423.438. 344 A. SMITH WOODWARD ON THE HISTORY OF FOSSIL CROCODILES. than eight longitudinal series of articulated scutes, each of which consists of only one piece of bone. i. Belodontidre.-Belodon. ii. Stagonolepidre.-Stagonolepis. iii. Parasuehidre.-Parasuchus. SUB-ORDER II.-MEsoSUCHIA. Palatine bones produced into osseous plates, thus prolonging the nasal passages, and giving rise to secondary posterior nares, situated about the middle of the mouth: pterygoids excluded from the posterior nares. Middle Eustachian passage alone enclosed by bone-the two lateral canals only indicated by grooves. Vertebras amphicoelian or platyccelian, Pubis excluded from acetabulum: acetabular margin scarcely notched. Usually two rows of dorsal scutes, but sometimes more in later forms; ventral shield consisting of undivided flat scutes, united by sutures, and comprising not more than eight or ten longitudinal rows. i, Teleosauridre.- Teleosaurus, Pelaqosaurus, Teleidosaurus, Steneosaurus, Myst1'l'osaul'us, M'etriorhsmchus, LEolodon, Macl'ol'hynchus, Hqposaurue, Enalioeuchus, Dakosaurus, Machimosaurus, Hsjlceochampsa (7), Petrosuchus (7). ii. Goniopholidre.- Goniopholis, Nannoeuchue, Ouieniasuchue (7), Theriosuchus (1). iii. Bernissartidre.-Bernissa1,tia. SUB-ORDER III.-EusUCHIA. Both palatines and pterygoids produced into osseous plates, thus prolonging the nasal passages and giving rise to secondary posterior nares, situated beneath the hinder part of the skull. Vertebrse proccelian-except atlas and axis, two sacral (opposed faces flat), and first caudal (each face convex). Middle Eustachian passage enclosed by bone, and usually also the two lateral canals. Pubis excluded from aceta­ bulum : acetabular margin deeply notched. Always more than two longitudinal rows of carinated dorsal scutes; ventral armour, when present, comprising more than eight longitudinal rows of seutes, each of these scutes consisting of two pieces united suturally. i, Alligatoridre.-Alligat01·, Caiman, Jacare, ii. Crocodilidre.-Crocodilus. iii. Gavialidre.-ThoracosaU1·us, Holops, Gavialis, Tomistoma, Rhamphosuchue,