EXTINCT CAMELS.--The earliest cameloid type is the genus _Protylopus_,[194] of which we are acquainted with an imperfect skull {288} and the greater part of a radius and ulna belonging to one individual, and most portions of the hind-limbs in other specimens. The one species, _P. petersoni_, was about the size of a "jack rabbit," and is late Eocene (Uinta formation) and American in range. The teeth of this mammal are the typical forty-four, and the canines are not pronounced, being incisiform in shape. In the skull the nasals overhang, as in the genus _Poebrotherium_. The orbit is not closed by bone. There is in this ancient Camel a trace of the supra-orbital notch so characteristic of the Camel tribe. "The vertebrae resemble those of the modern Lamas closely in their general proportions." The lumbars have the usually Cameloid formula of 7. This genus has but two functional toes on the hind-feet, the second and fifth being reduced to vestiges. It is interesting to note that the radius and ulna appear to remain distinct, except in very old animals, in which they come to be co-ossified in the middle only, thus foreshadowing their complete union in the next genus, _Poebrotherium_. The present genus, moreover, as well as _Poebrotherium_, was distinctly unguligrade; it has not acquired the characteristic phalangigrade mode of progression of the modern types of Camels. [Illustration] FIG. 150.--Skull of _Poebrotherium wilsoni_. _i_^1, _i_^2, _i_^3, Incisors 1-3. x 1/2. (After Wortman.) The American and Oligocene _Poebrotherium_ has been recently and exhaustively studied by Professor Scott.[195] It was considerably smaller than a Lama. Its neck was long as compared with other Artiodactyles, but still shorter than that of the Lama. It was a lightly-built, graceful creature, with apparently some external likeness to a Lama. It is an important fact to notice that at this {289} period, and for a long time after, there were no types referable to the Camelidae in the Old World. Though a Camel in many features of its organisation, _Poebrotherium_ was "generalised" in many ways. Thus the metacarpals and metatarsals were not fused to form a cannon bone, and the two lateral digits were represented by splint rudiments of metacarpals and metatarsals. The dentition was complete. The skull though distinctly Tylopodan, also shows more generalised characters. Thus the orbit is not quite, though nearly, completed by bone. In the Camel it is quite closed. The nasal bones are much longer, reaching nearly to the end of the snout. The odontoid process of the axis vertebra is not spout-like as in existing forms, but cylindrical, though slightly flattened upon the upper surface. The scapula is described as being more like that of the Lama than of the Camel, though variations occur which approximate to the Camel. The brain, judging of course from casts, has those sulci "which are common to the whole series of Ungulates, and closely resemble those of a foetal Sheep."