Water Reptiles of the Past and Present

CHELONIA

No order of reptiles of the past or present is more sharply and unequivocally distinguished from all others than the Chelonia or Testudinata. No order has had a more uniformly continuous and uneventful history. None now in existence has had a longer known history, and of none is the origin more obscure. The first known members of the order, in Triassic times, were turtles in all respects, as well or nearly as well adapted for their peculiar mode of life as are those now living, and were they now living they would attract no especial attention from the ordinary observer and but little from the naturalist. From time to time some have gone after better things, and have come to grief, but the main line has remained with fewer improvements, fewer evolutional changes, than any other group of higher vertebrates. The turtles seem very early to have adapted themselves so well to their peculiar mode of life, to have intrenched themselves so thoroughly in their own province, that no other creatures have been able to overcome them, or to drive them from it. The remains of no other air-breathing vertebrates are so omnipresent in the rocks as those of the turtles; they may be expected wherever fossils of air-breathing animals are found, though unfortunately often only in scattered and broken fragments. The loose union of their skeletal bones and their general habits of life in shallow waters left their bodies as food for scavengers, or for dismemberment by the tides and currents.

Relationships with other reptiles they really have none. Some have thought that the plesiosaurs were their first cousins, others the Placodontia, an indeterminate group of extinct reptiles usually placed with the Anomodontia. But their relationship with neither of these is closer than with the crocodiles, dinosaurs, or pterodactyls. They are the only reptiles that we know, besides the cotylosaurs, which have no holes in the temporal roof of the skull, and as the cotylosaurs were the most primitive and the oldest of reptiles, this fact incontestably proves that the turtles had a very ancient origin, though we know them no farther back than the later Triassic. They are the only order of reptiles of which not a single member is known to have teeth, or even vestiges of them. Until recently only a single specimen has been known from the Trias, and of that only the casts of the shell; but the shell was as fully developed and as complete as that of a modern alligator snapper, which it resembled much in form and in size. And doubtless the habits of this ancient Proganochelys were similar to those of the alligator snapper. The early cotylosaurian reptiles were all littoral-or marsh-loving animals, and more or less aquatic, and doubtless the early turtles continued in the same environments and with the same habits after acquiring a shell for protection and losing their teeth, which for some inexplicable reason they seemed no longer to need. Until near the close of the Jurassic period probably all turtles were amphibious animals of the marshes, spending much, perhaps the larger part, of the time in the water, good swimmers, and yet good crawlers. With the beginning of the Cretaceous, however, some of them became ambitious for new and untried modes of life. Various ones went down into the sea and became marine animals, reaching the zenith of their prosperity and the maximum of size before the close of the period, but continuing in diminished size and numbers to the present time, if we may consider the leather-back turtle as really their descendant. Others in the Cretaceous took to the rivers and ponds, and became almost as thoroughly aquatic in their thin shape and soft covering; and their lineal descendants still continue in the rivers of the Northern Hemisphere. Still others, in the Age of Mammals, took to the upland, and competed with the mammals in the open places and prairies, reaching their maximum in Miocene-Pliocene times, when for some unknown reason the giants among them were driven from the mainlands to continue a precarious existence to the present time in some of the larger islands.

Were there no turtles living we should look upon the fossil forms as among the strangest of all vertebrate animals—animals which had developed the strange habit of concealing themselves inside of their ribs, for that is literally what the turtles do. The box or shell of an ordinary turtle is composed of the backbones and ribs, to which are soldered a shell of bony skin plates above, with the clavicles, interclavicle, and ventral ribs below. Except in the strange leather-back turtle described farther on, these plates form definite series. Ten of them cover the spines of the dorsal vertebrae, in the mid-line, one over each, of which the turtles have the smallest number of any known reptiles. There are eight on each side over the ribs, united by suture with each other and with the middle series; and, in addition, there are twenty-six bones surrounding them and attached to them. All these bones compose what is called the carapace, which forms a complete roof in the more terrestrial types, more or less imperfect, with vacuities between the bones in the marine forms. On the under side, in addition to the clavicles and the interclavicle, there are three pairs of enlarged ventral ribs that go to form the plastron, solid and complete in land turtles, with openings in the water forms. And in the land forms the plastron is more or less firmly united with the carapace.

In the skeleton contained within the box thus formed is the very peculiar pectoral girdle, composed of scapula and coracoid, the scapula so peculiar that the controversy as to its homologies is not yet quite settled. Most authors, until recently, have believed that its peculiar shape (Fig. 113) is due to the co-ossification of the procoracoid with the scapula instead of as usual its loss or union with the true coracoid, so called. We are now pretty sure that this is not true, since in reality there is no such bone as the procoracoid, the bone so called being the real or true coracoid; and because, in the second place, the long anterior projection called the procoracoid is really only an outgrowth of the scapula itself and not a fused, separate bone. Hence the bone is properly called the scapula-proscapula, and not the scapula procoracoid, as it usually has been. The coracoids are elongate and flattened and without the usual supracoracoid foramen, so generally present in reptiles. The only other reptiles having a similar structure of the scapula are the plesiosaurs, and it has been because of this apparent resemblance that some good paleontologists have thought the turtles and plesiosaurs were allied. The sacrum is composed of two vertebrae only, and the pelvis of the usual three bones, the ilium, the ischium, and the pubis, all covered over by the shell.

In every known turtle the neck is composed invariably of eight vertebrae, but they are peculiar in many respects. In the earliest known turtles the neck vertebrae were, as would be supposed, biconcave, but they soon became very variable in all; in each neck some are biconcave, some biconvex, some opisthocoelous, and some procoelous. And Dr. Hay tells us that the neck has increased in length in the later forms.

The skull also is very peculiar in that it has some very primitive characters and others very aberrant. The temporal roof, as has been said, has no holes through it, though it is often reduced by the emargination of the borders, whether from below or behind, until in some the whole temporal region is exposed, and not at all covered over. There is no parietal foramen, so constantly present in all the early reptiles and in the lizards and the tuatera of modern times. There are no teeth or vestiges of teeth, but the jaws have usually a horny cutting edge, which seems to be quite as serviceable; in the river turtle the lips are fleshy. There is no transverse or transpalatine bone. There is a single vomer only, not paired as in other reptiles, whence comes the doubtful theory that the vomers of other reptiles are not the real vomers originally so named in mammals, and hence often called prevomers. The vomer of the turtles under this theory is believed to be the real homologue of the mammalian bone. The palate is always slightly, sometimes nearly wholly, underfloored, as in mammals, carrying the internal nostrils far back in the mouth. In the occipital region of the skull there is a separate bone on each side called the paroccipital or opisthotic, which has been indistinguishably fused with the exoccipital in all other reptiles except the ichthyosaurs since Triassic times.

In the feet the numbers of phalanges—that is, the bones of the free digits—are like those of mammals, that is, two in the first and three in each of the other four digits. The land tortoises have lost some of these, while the river turtles have either gained one or two in the fourth finger and fourth toe, or else have enjoyed an uninterrupted descent from the primitive reptiles which normally possessed that number. All other reptiles, save those phylogenetically allied to the primitive mammals, that is, the Theriodontia and their allies, have normally the phalangeal formula 2, 3, 4, 5, 4. It was partly because of this similarity of the numbers of toe bones that the turtles have been classed in the great group of reptiles that includes the ancestors of the mammals; that is, under this theory, the turtles would enjoy a nearer relationship to the mammals and to man himself than any other living reptiles! But this classification has been shown to be quite artificial.

From what has been said, it will be surmised that the Chelonia represent in themselves one of the primary subdivisions of the class Reptilia, and that, unlike most others, the order has enjoyed a most remarkable longevity. And doubtless they are one of the primary branches of the reptilian stock, which has remained distinct since Permian times at least, if not since Carboniferous, isolated and remarkably homogeneous, giving off no branches which departed far from the main stock, and on the whole leading a singularly placid existence for ten or more million years.

In most textbooks the order Chelonia is divided into three suborders, the Pleurodira, the Cryptodira, and the Trionychoidea. In recent years, however, the earlier members of the older group of Pleurodira have been separated into a fourth suborder, the Amphichelydia, a group characterized by some not very important differences in the plastron and skull, and including those forms in which the cervical vertebrae are amphicoelous. This group continued to Eocene times before it became extinct, and consisted of archaic forms which persisted after all the other suborders had come into existence. The Cryptodira, especially characterized by the manner in which they withdraw the head and neck within the shell by an S-like vertical flexure, are known from the Lower Jurassic and are still the dominant group of today, with more than one hundred and forty living species. The Pleurodira in the narrower sense are first known from their remains in the Upper Cretaceous of North America and are still represented by about forty species, living in the Southern Hemisphere. They are distinguished from the other groups by the manner in which they withdraw the neck and head into the shell, by a horizontal, sidewise flexure. The third suborder, the Trionychoidea, also began in Cretaceous times, so far as we know, and are represented by about seventy living species, chiefly in the Northern Hemisphere. They are especially characterized by the absence of bony marginal plates and the soft epidermis.

With the exception of the land tortoises, all turtles from the beginning of their career as an order to the present time have been more or less at home in the water. In some, like the marine forms, the adaptation to aquatic life has produced marked changes in structure: in the loss of the horny dermal shields and in the loss of bone tissue; in the flattening of the shell, and in the development of the front legs into swimming flippers, with a loss of the claws. In the absence of a guiding tail, which is always small in the marine turtles, propulsion must of course be wholly by the aid of the limbs. As oar propellers the marine turtles show some of the peculiar characters of the plesiosaurs. With a like short and broad body, a more or less elongated and flexible neck, there could be no sinuosity of the body in swimming. As an oar-like organ the humerus became flattened, and its muscular attachments, as in the plesiosaurs, descended far down the shaft, giving greater mechanical advantage. Unlike all other aquatic vertebrates, the turtles never developed real hyperphalangy. Only in the river turtles is there a possibility of an increase in the bones of the fourth digit.

To discuss in general the structure and habits of the living chelonians would extend this chapter to an undue length, and would add nothing to the many excellent works on natural history accessible to the student. We have therefore contented ourselves with a brief outline of the geological history of the order, with especial reference to their aquatic habits.

SIDE-NECKED TURTLES.
PLEURODIRA

The suborder of Chelonia, generally known as the snake-necked or side-necked turtles or tortoises, comprises about forty living species, confined to South America, Africa save the northernmost part, Madagascar, New Guinea, and Australia. In Australia they are the only members of the order known—another instance of the peculiar isolation of the fauna of that region. In the past they lived in North America during Upper Cretaceous times, the earliest known forms of the group in its restricted sense, of which seven species are described by Hay. In Eocene times they are also known from Europe and Asia, from both of which regions they have long since disappeared.

The Pleurodira, as the term indicates, are easily distinguished from all other turtles by the way in which they withdraw the head within the shell. Instead of withdrawing it by an S-shaped flexure of the neck between the shoulder-blades, as do other turtles, these bend the neck laterally in a horizontal plane, bringing the head within the margins of the shell in front of one or the other foreleg; and the margins of the shell are produced here in an eave-like fashion for the greater protection of the head. In the structure of the shell, which is always fully developed into a box, these turtles do not differ very much from the Cryptodira, though there may be some extra bones in the plastron, as also in the skull. The nasal bones are always, the lacrimals sometimes, well developed; the latter never, the former rarely, found in other groups. The lower jaws articulate a little differently, and the external ear is always fully surrounded by bone. Very characteristic is the bony union of the pelvis with the plastron below, which never occurs in other turtles, unless it be the Amphichelydia.

The side-necked turtles are all of fresh-water habit, similar to that of the fresh-water tortoises spoken of farther on. The neck is often very long and snake-like, which accounts for one of the names given to these turtles; because it is withdrawn into the shell sidewise, it has more distinctively ball-and-socket joints between the vertebrae, with distinct transverse processes for the attachment of the necessary side-moving muscles. The feet in all are more or less webbed and armed with strong claws.

The largest and perhaps the best known of all living side-necked turtles is the giant Amazon turtle of South America, which sometimes has a shell nearly three feet in length. Its feet are broadly webbed, and the shell is rather flat in the adult; it is an excellent swimmer in the waters of the Orinoco and Amazon. Six or seven species of the genus to which it belongs are known, all of them South American except one that lives in Madagascar and one fossil found in the Eocene of India. This remarkable distribution is but one more of the many instances known in zoölogy and paleontology that seem to prove an early land connection between India and South America. Had the migration between the two continents occurred by way of Asia and Bering straits, as did that of hosts of mammals, one would certainly expect to find some evidence of it in the North American Tertiary rocks, which, so far, is lacking.

CRYPTODIRA

The chief families of the Cryptodira turtles are the Chelydridae, or snappers; the Emydidae, or marsh tortoises; the Testudinidae, or land tortoises; the Chelonidae, or sea-turtles; the Protostegidae, or ancient sea-turtles; and the Dermochelydidae, or leather-backs. Other doubtful or smaller groups, both living and extinct, may be omitted, or incidentally mentioned.

SNAPPING TURTLES

The family of snapping turtles, the Chelydridae, are of interest because of their peculiar geographical distribution at the present time. Only four species are known, three of them from North America, the fourth from New Guinea. The family is one of the most primitive of living turtles, though no members of it are known with certainty from earlier rocks than the Oligocene. In all probability, also, they have retained, more than have any other group of turtles, unless it be some of the fresh-water tortoises, the primitive habits of the earlier or earliest turtles, though of course there have been modifications, both in structure and in habits. The three species of the United States include two of the snapping turtles proper and the alligator turtles of the southern states, which sometimes reach a length of three feet. All the species are largely aquatic in habit, powerful and active swimmers, with webbed feet and strong claws, and both on the land and in the water they are bold and fierce. They have a relatively large head and very strong jaws. Agassiz saw one bite off a piece of a plank an inch in thickness, and they can usually be raised from the ground by any object which they seize. The carapace and plastron are much reduced, and are rather loosely united. The shell is not large enough for the complete withdrawal of the head and legs within it, and the tail is unusually large and strong. The common snapping turtle, Chelydra serpentina, is found from Canada to Ecuador, and its remains have been found with those of the mammoth and mastodon in Pleistocene deposits; and related species of the same genus have been reported from the Miocene of England.

FRESH-WATER OR MARSH TORTOISES

The family of turtles or tortoises (Emydidae) represented at the present time by the common terrapin, painted tortoise, and box tortoise of the United States, and commonly called fresh-water turtles or tortoises, comprises the largest group of living chelonians—nearly a third of all existing members of the order. They are widely distributed over all parts of the earth except Australia, and are of very varied habits. Some are almost exclusively aquatic; others, like the painted tortoise, are partially so, while others, especially the common box tortoise, are almost as exclusively terrestrial as the true land tortoises, dying even, if forced to live long in water.

The shell in the more aquatic forms is depressed or flattened, but in the terrestrial kinds may be as highly arched as in the true land tortoises. The feet are adapted primarily for walking, but nearly always have the toes webbed, and the digits are longer than are those of the land tortoises. Only the two or three middle toes have claws. Some species have developed hinges in the plastron, whereby they may be completely closed up after the head and legs are withdrawn within the shell. Most of the species are carnivorous in habit, but a few, like the box tortoise, are strictly vegetarian.

Geologically the fresh-water tortoises have a not very ancient history, going back no farther than do the land tortoises, that is, to the beginning of the Cenozoic or Age of Mammals. Fully fifty species are known from the Tertiary rocks of North America, or more than three-fourths as many as are now living upon the earth.

The family at most can be said to be only amphibious in habit, and has no striking aquatic adaptations, since the shell is well developed and is covered with horny shields. The flattened shell of the more aquatic forms is characteristic, as is also the greater degree of webbing between the toes.

LAND TORTOISES

Perhaps the last of the more noteworthy specializations of the Chelonia, and indeed among the last of the more important specializations of the Reptilia, are the upland tortoises, of which the common “gopher” of the southern states is almost the only remnant in North America. They formed a part of the great hegira of forest and marsh animals to the open prairies, away from the lowlands and water which the turtles had inhabited almost exclusively for millions of years.

They began their career, Dr. Hay thinks, at about the beginning of the Cenozoic, that is, with the great development of the mammals, and reached the maximum of their development in the Miocene; and they have been on the decline ever since. In the Northern Hemisphere, at least, the slowly cooling climate throughout the Eocene, and a decided decrease in moisture, brought about the prairies and prairie plants before its close. Just as the horses, rhinoceroses, camels, and other herbivorous mammals took to these open places for the better and more abundant food found therein, so also the lowland tortoises found better food and fewer enemies there, for they are all strictly herbivorous in habit. The mammals became more conspicuous to their enemies when they went into the open, and it was only by the development of speed, more sober coloration, and perhaps greater cunning that they found safety from them. The tortoises were handicapped by low intelligence, and they could not develop speed, for they were not constructed to that end. But they did find protection in their bony shell, which became thicker, higher, and more convex, and with smaller openings. To quote Dr. Hay: “We may suppose that it would be much more difficult for a carnivorous animal to effect an entrance into such a shell than into one depressed, and whose borders may be spanned by the jaws of their enemies.” Perhaps also the highly arched form of the shell gave greater capacity for the lungs, and the tortoises in general, it is said, do have better lung capacity than the more aquatic or lowland types of turtles. Possibly, also, the heavier shell lessened the evaporation of the body fluids, and made the tortoises less dependent upon the vicinity of water. Certain it is that the common box tortoise, of like form and habits, occurs not rarely on the arid plains, far from water.

The neck and legs became fully retractile within the shell; the digits were shortened up, without a vestige of webbing membrane between them; the phalanges were reduced in number to two in each toe, and nearly all the toes have well formed claws. The feet are placed squarely upon the ground, and the body is elevated in walking. They can swim, when by accident they are thrown into the water, only as any terrestrial mammal can.

About forty species of land tortoises are known throughout the world at the present time, though North America, the probable original home of the tribe, has but three, all small. The larger species are all now denizens of islands, especially the Galapagos Islands, where the giant tortoises have long been famous. And many of our living forms have changed but little since Eocene times. In the Oligocene and Miocene they inhabited western North America in enormous numbers. In the Bad Lands of South Dakota one can often see the remains of a dozen or more of these giant tortoises at one time, specimens varying from one to three feet in length of shell. In river deposits, those of the late Miocene or early Pliocene, the writer has seen areas of an acre or more literally strewn with their remains, as though droves of them had been overwhelmed and perished together. About fifty species of these land tortoises are known from the American Tertiary, thirty-two of them belonging to the modern genus Testudo, which comprises the giant tortoises of the Galapagos. The largest known species of the group is one of Testudo from the Pliocene of India, which had a shell six feet in length. Why the larger species became extinct in Pliocene times on the mainland to survive only in the islands is not known; possibly their carnivorous enemies became too cunning and too numerous.

SEA-TURTLES.
CHELONIDAE

The sea-turtles, or Chelonidae comprise five or six living species, inhabitants for the most part of tropical and subtropical oceans, of which the green or edible turtle (Chelone), the hawksbill turtle (Caretta), and the loggerhead (Eretmochelys [Fig. 119]) are the best known. They are all thoroughly aquatic in habit, and of large size, from three to five feet in length. The carapace is heart-shaped, and reduced, that is, with large openings between the ribs; the plastron also is reduced and loosely united to the carapace. The neck is short and the head is not retractile within the shell. The temporal region of the skull is roofed over. The four legs form large and powerful flippers, and the hind legs are relatively small. The body is flattened and the tail is small. The aquatic characters of the limbs are seen especially in the broad and strong humerus, with the radial crest for the attachment of powerful muscles situated far down on the shaft; in the relative shortness of the radius and ulna, and the large size of the latter bone; in the flattened carpal bones; and in the great elongation of the digits and the absence of all but one or two of the claws. Unlike the leather-back turtle and the Cretaceous sea-turtles, the carapace and plastron are completely covered with horny shields, from which indeed the tortoise shell of commerce is derived. Except the green turtle, all members of the family are carnivorous.

Extinct members of the family are known from scanty remains in Cenozoic and late Cretaceous rocks. From the earlier Cretaceous deposits of the plains more primitive allied forms occur, often classed in distinct families of which Toxochelys (Fig. 120) and Desmatochelys are the more noteworthy. The latter genus, especially, might well have been an ancestor of all the modern forms. About three feet in length, it had all the essential characteristics of the sea-turtles, in its thin form, roofed-over skull, reduced carapace, loose plastron, and flipper-like limbs. The single known specimen, preserved in the museum of the University of Kansas, came from the lower rocks of the Upper Cretaceous of Nebraska. Yet earlier, at the close of the Jurassic, there were shore turtles of considerable size that had begun to develop a fondness for the open seas; to acquire a depressed form and lightened shell, the limbs still retaining, however, more of the terrestrial or crawling form. They are grouped as a separate family, the Thalassemydae, and include the first of the Chelonia to depart from the marsh and fresh-water habits which for long ages, perhaps, had limited the activities and evolution of the turtles.

ANCIENT SEA-TURTLES.
PROTOSTEGIDAE

Forty-four years ago the late Professor E. D. Cope, one of the greatest naturalists America has ever produced, in almost the earliest exploration of the great Cretaceous fossil deposits of western Kansas, discovered and collected a remarkable specimen of one of the most extraordinary turtles that is known even yet. By an error somewhat natural for those times, when the theory of evolution was just beginning to attain acceptance by naturalists, he thought that the specimen, notwithstanding its monstrous size, represented a very primitive kind of turtle, and gave to it the name Protostega gigas, meaning gigantic first roof! The late Professor George Baur, to whom paleontology owes so much, showed that, far from being a primitive turtle, Protostega was really one of the most specialized types of the order. Professor Cope’s account of the discovery of the specimen is of so much interest that it may be quoted here:

“In the very young tortoise or turtle the ribs are separate, as in other animals. As they grow older they begin to expand at the upper side of the upper end, and with increased age the expansion extends throughout the length. The ribs first come in contact where the process commences, and in the land tortoise they are united at the end. In the sea-turtles the union ceases a little above the ends. The fragments of the Protostega were seen by one of the men projecting from a ledge of a low bluff. After several square feet of rock had been removed, we cleared up the floor and found ourselves well repaid. Many long, slender pieces of two inches in width lay upon the ledge. They were evidently ribs, with the usual heads, but behind each head was a plate-like the flattened bowl of a huge spoon, placed crosswise. Beneath these stretched two broad plates, two feet in width, and no thicker than binder’s board. The edges were fingered and the surface was hard and smooth. All this was quite new, among fully grown animals. Some bones of a large paddle were recognized, and a leg bone. The shoulder-blade of a huge tortoise came next, and further examination showed that we had stumbled on the burial place of the largest species of sea-turtle yet known. But the ribs were those of an ordinary turtle just hatched, and the great plates represented the bony deposit in the skin, which, commencing independently in modern turtles, unite with each other at an early day. But it was incredible that the largest of known turtles should be but just hatched, and for this and other reasons it has been concluded that this ‘ancient mariner’ is one of those forms, not uncommon in old days, whose incompleteness in some respects points to the truth of the belief that animals have assumed their modern perfection by a process of growth from more simple beginnings.”

Later studies by Doctors G. Baur, E. C. Case, O. P. Hay, and especially G. R. Wieland, of the abundant and excellent material, preserved in the museums of Yale and Kansas universities and the Carnegie Institution, and especially the discovery by Wieland in 1895 of an allied and yet larger form which he called Archelon, have determined practically every detail of the structure of this remarkable group of sea-turtles. A surprisingly complete specimen of Archelon is mounted in the museum of Yale University.

About a half-dozen species and two genera of the family have so far been described, all coming from the Upper Cretaceous deposit of Kansas and South Dakota, the genus Archelon from later rocks than those which have yielded Protostega.

The general form and structure of Archelon will best be understood from the accompanying figures after Wieland (Figs. 123, 124, 125) and the restoration of the living animal as interpreted by the writer (Fig. 126). If the leather-back turtle, described farther on, is really the descendant of these or allied turtles, as many authors believe, it of course represents the very highest aquatic specialization of all Chelonians. If, on the other hand, as some believe, the leather-back is the end of a long and independent line of descent, then Archelon represents the highest aquatic specialization of all other turtles.

In size, at least, Archelon attained the maximum of the order, reaching a length of more than twelve feet, and a weight of more than three tons. Except that the shell was not heart-shaped or elongated as in all modern sea-turtles, but nearly circular in outline, it had all the aquatic adaptations of the sea-turtle in a yet higher degree.

The shell was depressed; the dermal plates covering the ribs had almost entirely disappeared, remnants only of their upper ends remaining; the skull (Fig. 127) had the temporal region wholly roofed over; the neck was short and not retractile. The front legs were strong flippers, the humerus was long and stout, with the crest for the attachment of muscles far down on the shaft; the digits were greatly elongated and clawless, etc. The plastron only was less reduced than in the case of the modern sea-turtles. No traces of horny shields have been discovered. As to the nature of the covering and the general appearance of the turtle when alive, Dr. Wieland has kindly given the writer his views, as follows:

“After direct study or fairly close examination of all the fossil material of importance thus far collected representing the Protostegidae, it seems certain that in all the members of the group an external leathery layer was well developed. In no instance is there the slightest trace of horny shield sulci, or grooves; though it seems probable that there was some gradation toward a thin and perhaps even slightly horny hide. In Archelon ischyros the reduced condition of the carapace and the presence of the continuous row of large, median, supraneural elements render it quite certain that there was a development of leathery hide comparable to that of Dermochelys. The same may be said of Protostega gigas. But Archelon Marshii had a less reduced carapace, and the leathery skin was probably less well developed; and Protostega Copei, in which no trace of supraneurals remains, must have made some approach to the horn-shield condition. A more distinct suggestion of transition from the leathery to the horny shield covering may be seen in the very different contemporary Cretaceous form, Toxochelys Bauri, where ossified epi-or supraneurals occupy quite exactly the nodal relation of the five vertebral horn shields of later turtles, like Lytoloma, though there are not the slightest traces of sulci.

“From a purely anatomical standpoint I have suggested that Archelon had seven dorsal keels corresponding to those of Dermochelys. There is much excellent reason for regarding dermogene ossification as essentially double-layered throughout the Reptilia.

“In any restoration of Archelon ischyros only the mid-line should be accentuated as a series of rather sharp supraneural crests. These are shown to have been present by the characteristic groove-like median pits with radiating striae, which are such a prominent feature of epineurals. It is reasonable to believe that the pits mark the attachment of horny crests developed in the leathery hide. Such were doubtless projected, more or less keel-like, to a height of one or two inches, and thus gave to the mid-line of the carapace, when seen laterally, a distinctly sinuous outline not unlike that of Toxochelys.”

As regards the habits of these ancient sea-turtles, we may offer tolerably certain conjectures. In the opinion of the writer, the less reduced plastron indicates a bottom-feeding habit, a view that is strengthened by the more rounded form of the shell, like that of the river turtle. All in all it would seem that Protostega and Archelon lived habitually on the soft bottoms of the shallower seas, feeding upon the hordes of large shell-fish, for which their powerful parrot-like beak was admirably adapted. That the species of Protostega did not commonly frequent the deeper oceans is indicated by the general absence of their remains in the deeper water deposits. The writer, in a long collecting experience, always found their remains associated with those of the smaller Toxochelys, toothed birds, pterodactyls, and the smaller mosasaurs.

Perhaps no one can speak more authoritatively as to the habits of these gigantic sea-turtles of the Cretaceous than Dr. Wieland:

“With regard to the general habits and appearance of Archelon much might doubtless be said if the present-day sea-turtles were more familiar objects. Dr. Hay thought that Archelon ischyros was a clumsy or even a sluggish, mainly littoral animal, moving slowly about the bottom of quiet inlets in quest of shell-fish; I, on the contrary, much struck by the powerful flippers, and especially by the flattening of the humerus, with its low radial crest and obviously strong musculature, have held that unusual swimming power and adaptation to a strictly marine life were indicated. Perhaps, as usual where experts differ, it is probable that both views are in part correct, and that Archelon was only a moderately good swimmer. It may be noted that, notwithstanding the almost circular body, the femoral notch, that for the hind leg, lies far back, so that it is not necessary, on the score of bulk, to assume slowness of motion, or the inability to pursue a sea-going life. Furthermore, it is now known that the development of the digits fell little short of that seen in Colpochelys (Fig. 117) or Eretmochelys, truly marine turtles.

“Therefore, while there can be no doubt that Archelon was strictly carnivorous in habit, and well able to navigate the open seas, it is not likely that it fed on other than relatively slow-moving prey. Lydekker looked upon the broad mandibles and broad palate of Lytoloma as specializations for a mussel diet; and very similarly in Archelon, while the decurved beak would easily transform him into a most formidable enemy, the heavy premaxillaries and vomer, and the flat but deep lower jaw, suggest an adept crusher of crustaceans. The presence of vast quantities of Nautilus dekayi, which I found accompanying one of the specimens, was doubtless accidental, but it plainly suggests that this cephalopod was one of the teeming sources of food in the Archelon environment.

“The huge bulk of the mature Archelon might account for the shearing off and loss of the flippers of younger forms caught between the shells of the ‘elder boatmen of the Cretaceous seas,’ as Cope has called them, during any sudden rush while herding on the shores. But probably the young turtles did not much frequent the shores at either egg-laying or other times. Whence it is much more likely that it was a mosasaur or some of the gigantic fishes like Portheus which bit off the right hind flipper in the type-specimen of Archelon ischyros, well above the heel, as I have described it. That this happened rather early in life is shown by the arrested growth of the right femur and remaining portions of the tibia and fibula, which are all uniformly 10 per cent smaller than the corresponding bones of the left flipper.”

While there were many small fishes in the Niobrara seas which the Protostegas inhabited, the most striking thing in the fauna is the great abundance of molluscal shells, especially Ostrea congesta. And with them were great hordes of larger pelycypod mollusks, some of them of enormous size. Some of the largest reach a diameter of nearly four feet, with shells so thin that one can hardly understand how they could have supported such large, oyster-like creatures. One can imagine that such shell-fish might have afforded an almost inexhaustible source of food for the large turtles; and several times the writer has found remains of Protostega associated with such shells. From all of which evidence it seems very probable indeed that Dr. Wieland is right in imputing to these gigantic turtles a shell-feeding habit, a habit which required neither speed nor great prowess; and perhaps the formidable beak was used more in social quarrels than for food-getting. That these marine turtles departed from the usual reptilian habit of laying their eggs upon land is improbable. The tortoise shell turtles of the Bahamas lay three or four hundred eggs in a hollow scooped out in the sand and then leave the young to their own devices; certainly many a one is gobbled up by birds of prey or other enemies on their way to the water. Perhaps the young Archelon lost its hind leg in some such mishap.

LEATHER-BACK MARINE TURTLES

The most remarkable member of the Chelonia now living is Dermochelys coriacea (Fig. 128), the great leathery or leather-back turtle of the warmer parts of the Atlantic, Indian, and Pacific oceans, the sole member of the family Dermochelydidae. It is the largest of all living turtles and the most thoroughly aquatic of all, whether living or extinct. It sometimes reaches a length of six feet, or half that of the largest known extinct forms, and weighs a thousand or more pounds. Agassiz saw a specimen that he said weighed a ton. Unlike other turtles, it has a carapace quite peculiar to itself, composed of a layer of thin, irregularly polygonal bones forming a mosaic, completely hidden in the thick skin, and entirely free from the skeletal bones beneath them. The larger of these skin bones form seven rows above, which appear in the living animal as sharp keels running the whole length of the shell. On the under side there are five rows of smaller-sized bones, under which there are vestiges of bones representing the normal plastron of turtles. The limbs are powerful, flattened paddles, not unlike those of Eretmochelys, but wholly destitute of claws. The front paddles are much larger than the hind ones; the humerus is long and flattened, and the digits are elongated. The leather-back is a powerful and effective swimmer, going long distances. Its habits are not well known; its food is chiefly fish, crustaceans, and mollusks.