[Illustration] FIG. 151.--Anterior surface of axis of Red Deer, x 2/3. _o_, Odontoid process; _pz_, posterior zygapophysis; _sn_, foramen for second spinal nerve. (From Flower's _Osteology_.) Later in historical sequence than _Poebrotherium_, and structurally intermediate between it and _Protolabis_, is the Miocene genus _Gomphotherium_. It shows an advance in structure upon _Poebrotherium_, in that the orbit is completely encircled by bone, though the posterior wall is thin; the lower canines instead of being incisiform are curved back as in later Camels, and separated by a wide diastema from the preceding and the succeeding teeth. Later in age than _Poebrotherium_ is _Protolabis_, a _Tylopod_ in which the full number of teeth is still retained; its skull presents no particular changes from the Poebrotherine type; the nasals, however, are somewhat shortened. Later still in point of time is _Procamelus_. In this form we have apparently an ancestral stock, whence both Camels and Lamas were derived. The upper incisors are as in existing forms, but the first and second persist for a somewhat longer time. The skull shows two well-marked types of structure; in _P. occidentalis_ {290} there are more points of likeness to the Lama, in _P. angustidens_ to the Camel. In both, the orbits are completely encircled by bone. The nasals are much shortened. The odontoid process of the axis is still more concave than in _Poebrotherium_, but not spout-like as in existing forms. This fact shows that the spout-like character of the Camels' odontoid process is not a point of affinity to other Artiodactyles--in fact the occurrence of the same form of odontoid process in Perissodactyles is enough proof of this. We must come to the conclusion that the form is adaptive in all cases. If we were not obliged on palaeontological evidence to come to this conclusion, the structure in question is just one which would be fastened upon as evidence of genetic affinity; for it is a resemblance in a small though distinctive point of structure having no obvious relation to utility. The metacarpals and metatarsals have coalesced to form the cannon bones, though a rudiment of one metacarpal seems to remain. The genera referred to appear to be on the direct line of descent of the modern representatives of the family. But there are other forms which are offshoots of the main stem. Such are _Homocamelus_, _Eschatia_, and _Holomeniscus_. The last two are Pliocene and American; the teeth are much reduced. C. PECORA. The Pecora are a group which possess so many characters in common that it is not an easy task further to subdivide them.