The Amynodonts were short, heavy types, probably marsh-haunting in habit, and possibly with a proboscis like that of the Tapir. The orbit is higher than it is in the purely terrestrial
Hyracodonts, and it is suggested that when swimming it was raised above the surface as with the Hippopotamus. "This feature," observes Professor Osborn, "with the long curved tusks, undoubtedly used in uprooting, suggests the resemblance between the habits of these animals and those of the hippopotami." There were no horns in the Amynodonts. The face is shorter than in the Hyracodonts, and the mastoid is covered as in recent Rhinoceroses. The canines are very strongly developed into tusks, but the incisors show signs of disappearance. We know of the genera Amynodon, Metamynodon, and Cadurcotherium. All except the last, which is European, are American in range.
Fam. 4. Titanotheriidae.—These Oligocene Ungulates, often attaining to large dimensions, are nearly peculiar, so far as is at present known, to the North American Continent, and are at least most abundant in it.[[167]] Many generic names, such as Titanotherium, Brontotherium, Brontops, Titanops, and Menodus, have been given to them; but a recent study of the entire material accessible for description or already described has led Professor Osborn to the opinion that there was but a single genus, to which the name Titanotherium must be applied. Of this genus there are some thirty well-characterised species, of which the gradual evolution can be traced from the lowest strata of the White River beds where their remains occur. An entire skeleton of T. robustum enables us to understand the osteology of these forms and to compare them with other Perissodactyles. This animal was more than 13 feet long, standing some 7 feet 7 inches in height. It seems to have presented during life the aspect of a Rhinoceros with perhaps a touch of Elephant. The skull is not unlike that of a Rhinoceros in general dimensions and shape; but there are a pair of apparent horn cores anteriorly, which are smaller in the more ancient forms and acquire a large size, a forward direction with a divergence of the two in the later forms. A glance at the accompanying figures of skulls (Fig. 137) of early and later Titanotheres will exhibit the changes in this particular which the skulls underwent in the lapse of time occupied by the deposition of these Oligocene beds. The nasals are short in the later, longer in the more early species, such as T. heloceras and T. coloradense. The zygomatic arch projects much, and is "shelf-like" in the later forms, the skull thus getting an immense breadth, which,
together with the long and divergent horn cores, must have given to the living animal a most bizarre appearance. It is an interesting fact that this animal, though a Perissodactyle, agrees with the Artiodactyla in the nineteen dorso-lumbar vertebrae, of which seventeen bear ribs.
Fig. 137.—Three figures showing the cranial evolution of Titanotherium. Upper figure, T. trigonoceras; middle figure, T. elatum; lower figure, T. platyceras. (After Osborn.)
The genus further agrees with the Artiodactyles in the structure of the carpus. The toes of the fore-limb are four, those of the hind-limb three; but while the hind-limb is undoubtedly Perissodactyle in the arrangement of its component parts, the fore-limb shows a hint of an Artiodactyle mode of structure. This limb is paraxonic, the axis of the limb passing between the two middle digits. It may be that this genus represents more nearly than any other Perissodactyle or Artiodactyle the primitive stem from which both have diverged, though, of course, it is not old enough to be very near to the actual ancestor. The molar dentition is the typical one; the incisors seem to vary as to their presence or absence, and, if present, in their numbers. In comparing the older with the more recent forms it is noteworthy that there has been an increase of size exactly as there has been during the evolution of the Camels and some other groups of Ungulates. As already mentioned, the size of the horn cores also increases until it culminates in the extraordinary species, T. platyceras and T. ramosum, in which these are half as long as the skull, flattened in form, and connected at their bases by a "web" of bone. Arrived at this amount of specialisation the genus Titanotherium apparently exhausted its capacities for modification and ceased to be. The many generic names may be explained by sexual differences on the one hand and an incomplete knowledge of connecting links on the other.[[168]]
Palaeosyops is somewhat like a Tapir in build, the skull especially resembling that of the Tapir. As in Titanotherium the molar teeth, instead of having an outer wall formed by fused cusps, have a
-shaped outer wall on one side and two or one cusps on the opposite side. It is, moreover, an Eocene form, and in correspondence with its greater age is more primitive in some points of structure, for example, in the absence of horns and in the full dental formula. The fore-limbs are four-toed, the hind