The rodents likewise were partially of immigrant and partially of native stock. To the former belonged the few mice and rats and a meadow-mouse (Microtus), a group not represented in present-day South America, and a rabbit. Very much more abundant and varied were the indigenous forms, all of which belonged to existing families and most of them to existing genera; the tree-porcupines, cavies, agoutis, spiny-rats, vizcachas, capybaras, coypus, etc., were abundantly represented, for the most part by extinct species.

The monkeys were of purely Neotropical type and several modern genera, such as Cebus and Callithrix, and one very large extinct genus, †Protopithecus, of the same family, have been found in the caverns of Brazil, but not in the pampas of Argentina, which would seem to have been a country of open plains.

In the South America of to-day one of the most striking and peculiar elements of the fauna is that formed by the Edentata, the sloths, anteaters and armadillos, and this was even more true of the same region in Pleistocene times. Anteaters and sloths are very scantily represented, but this is merely an accident of preservation; armadillos, on the other hand, were very numerous both in Brazil and in Argentina, and, in addition to many modern genera, there were several which are no longer in existence, such as †Chlamydotherium, which was a huge creature almost as large as a rhinoceros. Then there were the two extinct suborders of the †glyptodonts (†Glyptodontia) and the †ground-sloths (†Gravigrada) which were astonishingly abundant in Argentina and which, as was shown in a previous page (p. 205), were also well represented in North America.

Few more fantastic-looking mammals than the †glyptodonts have ever been found; the short, deep head, with its shield of thick, bony plates, the huge carapace made up of innumerable plates of bone firmly united at their edges and without the movable bands of the armadillo carapace, the enormous tail-sheath, the short legs and massive feet with broad hoofs, must have given these animals rather the appearance of gigantic tortoises than of mammals. The †glyptodonts were especially numerous and varied in the Argentine pampas, and a stately array of them is mounted in the museums of La Plata and Buenos Aires; in length, they ranged from six to twelve feet, including the tail. The skeleton and carapace did not differ very greatly in appearance among the various genera, but there were great differences in the form and size of the bony sheath enclosing the tail. In the genus †Glyptodon the sheath was composed throughout of movable overlapping rings, with prominent spines on them; in †Sclerocalyptus the hinder half of the sheath coalesced into a single piece, marked only by the elaborate ornamentation of the horny scales, while in †Dœdicurus the end had a tremendous, club-like expansion, which must have been set with great horn-like spines. The †glyptodonts were ponderous, slow-moving and inoffensive plant-feeders, almost invulnerable to attack, and probably used their massive tails, which could be freely swung from side to side, as redoubtable weapons of defence, much as the alligator uses his tail. In comparison with the bewildering variety in South America, the few that made their way into North America were quite insignificant.

Much the same statement applies to the †ground-sloths, and though these ranged far more widely through the northern continent than did the †glyptodonts, they were but few in comparison with the multitude which inhabited alike the forests of Brazil and the plains of the south. Two of the three genera of †ground-sloths which occur in the North American Pleistocene, †Megatherium and †Mylodon, are also found in South America; and though †Megalonyx has not yet been obtained there, the family of which it is a member was represented. In size, these creatures varied from a tapir to an elephant, though all were much shorter-legged than any elephant; the extremely massive tail, which the larger forms had, served to support the huge body, when erected to tear down the branches and leaves upon which these strange creatures fed.

Opossums were extremely numerous, especially in the Brazilian caves, where in half a cubic foot of earth 400 jaws were collected.

The Pleistocene mammalian fauna of South America was a mixture of modern forms with ancient, vanished types similar to that which we found in North America. The †ground-sloths and †glyptodonts, the †litopterns, †toxodonts and †typotheres, the antelopes, horses and †mastodons have all disappeared from the continent, or vanished altogether from the face of the earth.

II. Tertiary Faunas

1. Pliocene

North America.—No part of the Cenozoic history of North America is so imperfectly recorded and so unsatisfactorily known as that of the Pliocene, and the later portion of that epoch is especially obscure. If the Peace Creek formation of Florida is properly referred to the upper Pliocene, it would show that the mammals of that time were substantially the same as those of the early Pleistocene.

The only fauna, as yet discovered, which can be referred to the middle Pliocene, is that of the Blanco beds of northwestern Texas, which have yielded but a scanty list of mostly ill-preserved fossils. Obviously, these give us a very incomplete picture of the life of that time. The great †ground-sloths had already reached North America, and the genus †Megalonyx, so common in the forested areas of Pleistocene North America, was perhaps already in existence. The †glyptodonts were likewise represented by one genus (†Glyptotherium) which was distinguished by the simple rings of the tail-sheath. No rodents have yet been found and only a few of the Carnivora, though a large cat, a musteline and a large “†bear-dog” are known. There were no true elephants, but several species of †mastodons, all of which were different from those of the Pleistocene; and in some, grinding teeth, though still low-crowned, had become much larger and more complex, marking a stage of advance toward the elephantine dentition. Horses of primitive type, the feet having three functional toes instead of one, were relatively abundant. Very large llama-like animals were present, but nothing has been ascertained with regard to the deer and antelopes of the time, and the only other representative of the Artiodactyla yet recovered is a peccary, interesting as being a species of the genus (†Platygonus) which became so abundant and widespread in the Pleistocene. Scanty and incomplete as this fauna is, it suffices to show that the middle Pliocene mammals were much more primitive than those of the Pleistocene.

The fauna of the Snake Creek formation in western Nebraska and that of the presumably somewhat later beds of northwestern Nevada, which are referable to the lower Pliocene, may be considered together. The rodents, which are not very fully represented, were quite modern in character and belonged mostly to extinct species of modern genera, such as hares, pocket-gophers, beavers, forerunners of the †Giant Beaver, marmots, sewellels, etc. A remnant of a more ancient world, especially characteristic of the Miocene, is found in the remarkable burrowers, the horned †mylagaulids which have been extinct since the lower Pliocene. Carnivora were abundant, and members of all the families which inhabit North America to-day have been obtained; wolves, “†bear-dogs,” “†hyena-dogs” and forms like the Dhole of India were common. The terms “†bear-dogs” and “†hyena-dogs” are not to be understood as implying any relationships of these animals to bears or hyenas, but merely a certain superficial resemblance; these were very large members of the dog family (Canidæ), now extinct. Mustelines, large and small, are found, and possibly some bears had already made their way from the Old World, but this is still uncertain. †Sabre-tooth tigers and true cats, some as large as lions and one species fairly gigantic, were likewise characteristic of the time. There was a great wealth of horses, though the modern genus Equus was not among them; all the genera are now extinct and all were three-toed. Several distinct phyla were represented, some progressive and advancing toward the modern forms, others conservative and stationary. Browsing horses with low-crowned teeth, grazing horses with prismatic, cement-covered teeth, heavier and lighter, larger and smaller, must have covered the plains and thronged the woods. Ancestral tapirs were present, though far less common. A family which seems to be utterly exotic to North America, that of the rhinoceroses, was present, and of these there were three or four series, mostly without horns, or with a very small horn on the tip of the snout. The extremely aberrant perissodactyls (†Ancylopoda), in which the hoofs were converted into great claws, perhaps persisted, but the evidence is not conclusive.

The Artiodactyla were, for the most part, totally different from those of modern times, though several forms were ancestral to some now living. Peccaries more primitive than the living genus were the only representatives of the swine-like suborder; ancestral camels and llamas were among the commonest of the hoofed animals and an extinct phylum, that of the “†giraffe-camels” (†Alticamelus) continued over from the Miocene. The giraffe-camels are so called, not because of any actual relationships with the giraffes, but on account of certain likenesses in the proportions of the animals compared. †Alticamelus was a very large, camel-like creature, with remarkably elongate neck and limbs and comparatively small head, which no doubt resembled the giraffes in browsing upon trees which were above the reach of the ordinary camels and llamas of the time. It was the terminal member of a series, or phylum, which branched off from the main stock in Oligocene times and pursued a course of development which was independent of the principal series, but curiously parallel with it.

The deer of the lower Pliocene were little, graceful creatures (†Blastomeryx) which had no antlers, but the males were armed with sabre-like upper canine tusks, so that they must have resembled the Musk-Deer of Tibet, but were smaller and more slender. The remarkable group of “†deer-antelopes,” now extinct, was represented by †Merycodus, a dainty little creature, less than two feet high at the shoulder, which had the antlers and general appearance of a small deer, but the high-crowned grinding teeth which most antelopes have. True antelopes of two different lines were also present, though they are as yet known from little more than the bony horn-cores; of these, one is the flat-horned and the other the twisted-horned or strepsicerine type, such as is illustrated by the Eland and Kudu of modern Africa. The latter may, however, be related to the peculiarly North American Prong-Buck (Antilocapra) and not to the strepsicerine antelopes of the Old World. The last survivors of an exclusively North American family, the †oreodonts, which were wonderfully numerous and varied from the upper Eocene onward, are found here.

The †mastodons (†Gomphotherium) of this formation had well-developed tusks in the lower as well as in the upper jaw, and in one species the chin-region or symphysis of the lower jaw was greatly prolonged, an ancient feature.

That the South American edentates had already reached the northern continent is sufficiently proved by remains of †ground-sloths, which are, however, too incomplete to permit identification of the genus. †Glyptodonts have not yet been found, but this fact does not demonstrate that they had not accompanied the †ground-sloths in their migration, for at no time did they range so far north as Nebraska or northwestern Nevada, and the only mammal-bearing formation of lower Pliocene date known in the south, the Alachua Clay of Florida, has yielded too scanty a list of fossils to make its negative evidence at all conclusive on this point.

The mammals of the middle and especially of the lower Pliocene were much stranger and more primitive than might be inferred from the foregoing brief account. Except several of the Rodentia and perhaps one or two of the Carnivora, the genera are all extinct and such familiar terms as horses, rhinoceroses, camels, etc., can be employed only in a very comprehensive sense, as equivalent to families.

The Pliocene of South America is involved in some obscurity; not that there is any question as to the formations, or their order of succession, but there is much doubt as to the limits of the epoch both above and below. The latest Pliocene fauna, that of the Tarija Valley in Bolivia, was essentially the same as that of the Pleistocene and contained a similarly large proportion of migrant elements from the north, but it was evidently older and many of the species were different. The two divisions of the Araucanian fauna, contained in the beds of Catamarca and Monte Hermoso respectively, are very much alike and need not be given separate consideration. In one respect these presumably upper Pliocene faunas formed a very strong contrast to the mammalian assemblage of the Pleistocene, and that is in the quite insignificant part taken by the migrants from North America. Of the Carnivora there were but two representatives, one referable to the raccoon family and one to the dogs, while a hare and a small member of the Artiodactyla, of indeterminate family, complete the list of northern forms, though this list will doubtless be extended by future discovery. The peccaries, deer, antelopes, tapirs, horses, †mastodons, cats, weasels, otters, squirrels, mice, etc. had not reached the southern continent, or were still so rare that remains of them have not been found. This rarity and relative insignificance of the northern forms gave a very different aspect to the fauna.

On the other hand, the indigenous South American groups were very fully represented. Many kinds of opossums and a few large carnivorous types, much like the so-called Tasmanian Wolf (Thylacynus), were the remnants of a much larger assemblage of marsupials which inhabited South America in the Miocene. Of the Edentata, there were great abundance and variety, many large †glyptodonts and some gigantic armadillos, as well as numerous examples of normal size; the †ground-sloths, though somewhat smaller than those of the Pleistocene, were mostly of gigantic size, and true or arboreal sloths (Tardigrada) have been reported. The very numerous rodents, with the exception of the intrusive hare, all belonged to typically South American families. Some of the rodents were gigantic and one (†Megamys), a member of the Chinchilla family, was equal to a rhinoceros in size and the largest known representative of the order. Especially characteristic was the abundance of the cavy family (Caviidæ).

The hoofed animals, with the single known exception of the immigrant artiodactyl, all belonged to the autochthonous orders, all of which are extinct at the present time. Forerunners of the extraordinary genus †Macrauchenia, which was one of the most conspicuous elements of Pleistocene life, were quite common in the Pliocene and differed from the Pampean genus chiefly in their smaller size and less advanced specialization. We find here also the last survivors of another family of the †Litopterna, the †proterotheres (†Proterotheriidæ), which imitated the horses in such a surprising manner that some authorities believe them to have been actually related to those perissodactyls. The Monte Hermoso genus (†Epitherium) had feet which were wonderfully, though but superficially, like those of the three-toed horses. The †Toxodonta were numerous and most of them were large, ponderous animals; one genus (†Trigodon) had the interesting peculiarity of a single median horn on the forehead, much like that of a rhinoceros. Horned species were always rare among the indigenous groups of South American ungulates, and all that have been discovered so far belonged to the †toxodonts. The remaining group, that of the †Typotheria, was also well represented, both by larger and by very small forms, some no larger than a rabbit (†Pachyrukhos).

The presumably lower Pliocene (perhaps upper Miocene) fauna of the Paraná formation is as yet known only from very fragmentary material. Representatives of the dogs, raccoons and bears have been reported, but the identifications are doubtful; at all events, these would seem to have been the most ancient of the northern immigrants. A considerable number of marsupials, both opossums and large predaceous types, have been found. The rodents were very numerous, all belonging to South American families and some of them very large. The edentates were gigantic †ground-sloths and †glyptodonts, with numerous armadillos of ordinary size. The hoofed animals all belonged to the indigenous South American orders, the predominant place being taken by the †toxodonts, some of which were large. There were many †typotheres, both of the larger and smaller kinds. The †Litopterna were represented both by the horse-like †proterotheres and the long-necked †macrauchenids, the latter smaller and less specialized than those of the Pampean.

2. Miocene

North America.—Upper Miocene beds cover extensive areas of the Great Plains region and are scattered from Montana far into Mexico. The rich fauna is an outgrowth and development of that of the middle Miocene, with but few immigrant additions and, on the other hand, passes so gradually into that of the lower Pliocene, that any line of separation between them is very difficult to draw. The rodents, numerous as they are among the fossils, are almost certainly very incompletely represented in the collections; the families are almost all still in existence, but nearly every genus is extinct, and thus the vernacular names used to designate them must be understood in a broad sense. Hares, mice, pocket-gophers, squirrels, marmots, beavers and the extraordinary †mylagaulids were all abundant.

In even more strongly marked sense must the broad meaning for the vernacular names of the other mammals be emphasized, for we have to deal almost exclusively with extinct genera, which differed much from their modern descendants. Many of the Carnivora have been obtained; there were numerous dogs, some rivalling the largest of existing bears in size, true felines and †sabre-tooth tigers, which were smaller and lighter animals than the great beasts of the Pleistocene; weasels, martens, otters and raccoons, but no bears. The bears, a family of Old World origin, are not certainly known in America before the Pleistocene, but had probably reached this continent in the Pliocene.

As is so very generally true, the commonest and best-preserved of the fossils are those of the hoofed animals. The †mastodons were of the four-tusked kind (†Gomphotherium or †Trilophodon), the skull and teeth of which differed so markedly from those of the true elephants. The relatively small, low-crowned and simple grinding teeth were common to all the †mastodons, but the tusks were different from those of the larger members of the group. The upper tusks were comparatively short and nearly straight and retained a band of enamel, while the lower tusks were still shorter, chisel-shaped and so worn as to prove that they were regularly used, no doubt in cropping leaves; the shortness of these lower tusks was compensated for by the great elongation of the lower jaw. The head was proportionately broad and low and, for Proboscidea, these were small animals, not more than five or six feet high at the shoulder. The body, limbs and feet had already attained substantially their modern grade of structure, advance among the Proboscidea being chiefly restricted to the teeth and skull.

Four families of Perissodactyla were represented in the upper Miocene. The rhinoceroses, which were very abundant, were present in considerable variety; some were hornless, others had a single small horn on the end of the nose. Among these rhinoceroses there was much difference in bodily proportions, some being extremely heavy, with very short legs and feet, and these were the commonest, while others had longer legs and less massive bodies. Tapirs, on the contrary, would seem to have been scantily represented; at least, they are rare among the fossils. The extraordinary, aberrant †chalicotheres, perissodactyls with claws instead of hoofs, still persisted, but are far better known from the lower Miocene, in connection with which they will be described. The dominant perissodactyl family was that of the horses, of which no less than five genera are already known. There were some with very low-crowned teeth, which must have fed principally by browsing upon leaves and such soft diet; but the grazing kinds, which had high-crowned, cement-covered and very complex grinding teeth, had come to the fore. Still retaining three toes in each foot, with the middle toe so enlarged as to bear nearly the entire weight, save in snow or soft ground, these eminently cursorial animals, which had the slender limbs of a deer, must have roamed the plains in great herds.

Still commoner were the Artiodactyla. Many species of grazing camels, which were the predominant artiodactyl family in North America during upper Miocene times, were the ancestors both of the true camels of the Old World and the South American llamas. †Giraffe-camels have not yet been found and no doubt they were much less abundant than in the middle Miocene, but that they had not completely disappeared is shown by their recurrence in the Pliocene. As compared with earlier ages, the †oreodonts had begun a rapid decline and had lost notably both in numbers and variety, but one most curious beast (†Pronomotherium, Fig. 197, p. 375) marked the final step in the development of the short-faced, proboscis-bearing series, which may be traced back to its beginnings in the Oligocene. In this wonderful creature the skull was so short and deep as to suggest that of a gorilla or some other great ape. No other artiodactyls even approximate these later proboscis-bearing †oreodonts in the altogether exceptional form of the skull. Grazing †oreodonts (†Merychyus), of moderate and small size with high-crowned teeth, were evidently quite common on the upper Miocene plains. The †hornless deer and “†deer-antelopes” differed but little from those of the lower Pliocene. Peccaries were fairly abundant.

The upper Miocene fauna was especially characterized by the large number of mammals, belonging to several different orders, which had acquired the high-crowned, persistently growing pattern of grinding teeth. Many of the horses, camels, ruminants and rodents displayed this structure, and, as was first pointed out by Kowalevsky, the explanation is probably to be found in the spread of grassy plains at the expense of the forests. On account of the silica which they contain, the grasses are very abrasive and rapidly wear the teeth down. In adaptation to this new source of abundant and nutritious food, many kinds of mammals developed a form of tooth which was fitted to compensate by growth for the loss through abrasion.

The middle Miocene, small areas of which occur in Montana, eastern Oregon and northeastern Colorado, has received various local names, the typical one being the Deep River of Montana. Very probably, these scattered areas are not strictly contemporaneous, but form a closely connected series. That a land-connection with the eastern hemisphere existed, is made clear by the appearance of several unmistakably Old World types of animals and the beginnings of migration from South America are perhaps also to be noted, though this cannot be positively stated. The evidence for the South American connection is the finding in the middle Miocene of Oregon of what are believed to be the earliest remains of †ground-sloths yet discovered in North America, but the material is too scanty for altogether certain determination.

The smaller animals are not very well represented in the middle Miocene faunas, as conditions appear to have been unfavourable to their preservation; something is known of them, nevertheless. The very curious extinct family of rodents known as the †Mylagaulidæ, the presence of which was noted in the upper Miocene and lower Pliocene, first appeared here. These †mylagaulids, which were distantly related to the modern Sewellel (Aplodontia rufa), were characterized by the great enlargement and complication of one of the grinding teeth in each jaw and the consequent reduction of the others. One genus of this family, as in the Pliocene, had the peculiarity, unique among rodents, of developing a large horn upon the nose, like a miniature rhinoceros. Among the Carnivora, we find a great variety of dogs, large and small, all belonging to extinct genera, as indeed is true of the other carnivores also. True felines have been found, but as yet, none of the †sabre-tooth series; the abundance of the latter, however, in both preceding and succeeding formations, is sufficient proof that the discovery of them in the middle Miocene is merely a question of time. Mustelines were present, and especially noteworthy is the appearance of the first American otters, immigrants from the Old World.

Of the hoofed animals, the most interesting are the Proboscidea, the most ancient of which that are definitely determinable in America occur in this horizon. The place of origin and ancestry of these animals were long exasperating puzzles. Appearing suddenly in the Miocene of Europe and North America, in which regions nothing was known that could, with any plausibility, be regarded as ancestral to them, they might as well have dropped from the moon, for all that could be told concerning their history. The exploration of the Eocene and Oligocene beds of Egypt has dispelled the mystery and shown that Africa was the original home of the group, whence they gradually spread to every continent except Australia. Little is known of these earliest American proboscideans, but they were doubtless small †mastodons of the four-tusked type.

Among the Perissodactyla, the rhinoceroses were perhaps the most conspicuous; the native American stocks of this family appear to have mostly died out and to have been replaced by two or more phyla of immigrants from the Old World, some of which were hornless, others had a small horn on the tip of the nose and others again had a second and smaller horn on the forehead. Tapirs, though unquestionably present, are rare as fossils and not well known. Several distinct phyla of horses may be distinguished, which were like small ponies in size, but of more slender form; they were all three-toed, but there were marked differences among them with regard to the degree to which the middle toe (the third of the original five) had been enlarged to carry the whole weight and the lateral toes (second and fourth) reduced to mere “dew-claws.” While browsing horses, with low-crowned teeth, still persisted in large numbers, we find also the extremely interesting beginnings of the highly complex, cement-covered and high-crowned teeth of the grazing kinds. The clawed †chalicotheres were present, though very little is known about them because of the fragmentary character of the remains.

The Artiodactyla were much more varied and abundant, though they did not rival the great assemblage of these animals found in the European Miocene. Of the peccaries little more can be said than that they were present in these faunas. The †oreodonts were very numerous, both individually and generically; two stages of the proboscis-bearing kind are found here together, the older, long-faced genus (†Promerycochœrus) surviving from the Oligocene, while the newer Miocene type was short-faced and had a moderate proboscis (see Fig. 196, p. 373). Others had more the proportions of peccaries and still others were very small and presumably aquatic in habits. Camels abounded, both the grazing kinds which were ancestral to the modern forms of South America and Asia, and the great, browsing †giraffe-camels. The †hornless deer and the antlered †deer-antelopes were much like those of the Upper Miocene, slender and graceful little creatures, and there were also considerably larger ruminants (†Dromomeryx) with straight, simple and non-deciduous horns, which may be called antelopes.

The line of division between the lower Miocene and the uppermost Oligocene is a very obscure and difficult one to draw. Personally, I prefer to begin the Miocene with the widespread formation of the Great Plains, which has been variously named Arikaree, Harrison, Rosebud, etc., but this is a moot point. Concerning the lower part of these beds Osborn says: “They may be either: (1) Upper Oligocene or (2) transitional from Oligocene to Miocene, or (3) of pure Lower Miocene age.” The upper division is referred to the Miocene without question by any one, but for the purposes of this rapid sketch it will be best to treat the two faunas together. This many-named formation, for which the term Arikaree is here employed, as having priority, is found over extensive areas of South Dakota, northern Nebraska and central Wyoming. The fauna was almost entirely a development from that of the North American Oligocene, with very little admixture of foreign elements, so that the land communication with the eastern hemisphere must have been difficult. In this, as in most of the Miocene formations, the smaller mammals are not fairly represented, and it is evident that much remains to be learned with regard to them; this is especially true of the upper division of this stage.

The rodents, which were fairly numerous, were directly continuous with those of the upper Oligocene and included forms which were more or less distantly connected with the modern hares, squirrels, beavers, sewellels, pocket-gophers and kangaroo-rats. A few Insectivora of doubtful reference have been found. Among the Carnivora there was also considerable variety: dogs, large and small, were abundant, but all of them were decidedly primitive from the modern standpoint; the cats were represented both by the true felines, which were probably immigrants, and by the †sabre-tooth series. There were several large and powerful mustelines, or members of the weasel family, which were likewise immigrants, one of which resembles in many ways the modern Wolverene (Gulo). Very interesting is the beginning of the raccoon family (Procyonidæ) or, at least, what is believed to be such, which arose from a branch of the dogs; this most ancient of the raccoons was †Phlaocyon, a small and slender animal.

The earliest traces of the Proboscidea in America have been reported from this formation, but the fragmentary specimens are inconclusive. The Perissodactyla are among the commonest fossils. The rhinoceroses belonged to native stocks, including both the horned and hornless forms. The horned genus (†Diceratherium) differed from all other rhinoceroses in having a transverse pair of horns on the nose, and the species of the lower Miocene were quite small and light; the hornless genus (†Cænopus) was a larger and heavier animal. Tapirs are rare as fossils and consequently not well known. While there were several kinds of horses, they all agreed in having short-crowned and relatively simple grinding teeth and three-toed feet; they were smaller and of lighter, more slender build than those of the middle Miocene. The wonderful aberrant perissodactyls with clawed feet, the †chalicotheres (suborder †Ancylopoda), appear to have been more abundant in the Arikaree than at any other time in North America, though their history in this continent extends from the middle Eocene to the lower Pliocene. †Moropus, the lower Miocene genus, was as grotesque a creature as could well be imagined and, in advance of experience, no one ever did imagine such a beast. With rather small and somewhat horse-like head, long neck, long fore limbs and shorter hind limbs, these extraordinary animals united short, three-toed feet, which were armed with enormous claws. The long persistence (to the Pleistocene of Asia) and wide geographical range of the †chalicotheres are sufficient evidence that their very unusual structure must have been advantageous to them, but the problem of their habits and mode of life is still unsolved. From the character of the teeth, the long neck and fore limbs, it may, however, be inferred that they fed chiefly upon the leaves of trees.

Even more numerous and varied were the Artiodactyla. Peccaries of a primitive sort were common, and we find the last of the series of “†giant pigs,” which had been a very conspicuous group throughout the Oligocene. The lower Miocene genus, †Dinohyus, was a monstrous beast, six feet or more in height, with formidable canine tusks and a very long head made grotesque by bony excrescences upon the skull and jaws. For a pig, the legs were very long and the feet slender, having but two toes. The †oreodonts were present in great numbers, both small and large forms; except for bodily stature and modifications of the head, they all looked very much alike; †Merycochœrus, with its incipient proboscis, here made its first appearance. The last representatives of a family (†Hypertragulidæ) of small and graceful artiodactyls are found in this formation. One of these (†Syndyoceras, see Fig. 215, p. 403), an animal considerably larger than the existing Musk-Deer, was in its way even more bizarre-looking than the †chalicotheres; with an antelope-like head, it had four horns, one pair over the eyes, curving inward, and a shorter pair, with outward curvature, on the muzzle. Another genus (†Hypertragulus) was very much smaller and very slender.

The camels were beginning to diversify and give rise to several phyla. One of the genera (†Protomeryx), which did not much exceed a sheep in size, probably represented the main stock, which led to the camels and llamas of to-day. A second (†Stenomylus) was a still smaller animal, with remarkably long and slender legs, and might be called a “gazelle-camel,” while a third (†Oxydactylus, see Fig. 209, p. 392), which was larger and apparently the beginning of the †giraffe-camels, was noteworthy for its long neck. All of these lower Miocene camels had deer-like hoofs, the characteristic pad or cushion which gives such an exceptional appearance to the feet of modern llamas and camels not being fully developed till a later period. A very important new element in the North American fauna was the appearance of the first deer (†Blastomeryx), which came in the latter part of the Arikaree stage and were the forerunners of a renewed immigration from the Old World, which had been broken off during the upper Oligocene. This, however, is a disputed point; Professor Osborn and Dr. Matthew believe that these animals were truly indigenous and derived from a long line of American ancestry. The same genus continued through the middle Miocene, as we have already seen, and therefore no further description of it is called for.

The limits of the South American Miocene are very doubtful. The Paraná formation, here regarded as lower Pliocene, may prove to be more properly referable to the upper Miocene. No other upper Miocene is known.

To the earlier, probably middle, Miocene may be referred the wonderful Santa Cruz fauna of Patagonia. It is extremely difficult to convey to the reader any adequate conception of this great assemblage of mammals, because most of them belonged to orders which have altogether vanished from the earth and are only remotely like the forms with which we are familiar in the northern hemisphere. To one who knows only these northern animals, it seems like entering another world when he begins the study of the Santa Cruz fossils. If any North American mammals had then entered South America, which is not probable, they had not extended their range as far as Patagonia. Marvellously rich and varied as the Santa Cruz fauna was, it did not contain everything that we should expect to find in it; several recent families of undoubtedly indigenous South American origin have left no ancestors in the early Miocene formations. For this, there are several obvious reasons. In part, these gaps in the history are merely due to the accidents of collecting and some of them will almost certainly be filled by future exploration. Other absentees will probably never be found, because the Santa Cruz beds are known only in the very far south, and the Miocene climate of the region, though much milder and more genial than the present one, must have been unsuitable for many tropical animals. Again, the Patagonia of that time appears to have been a country of open plains, with few trees, and hence arboreal forms were rare.