Page:Mammalia (Beddard).djvu/268

 has a well-marked coracoid process. The radius and ulna are separate; so too are the tibia and fibula. Eohippus, belonging to the same sub-family, is slightly more primitive; for the hind-feet have a rudiment of digit I. Orohippus is a little nearer to the Horses in that the molar teeth have acquired a little further advance towards the equine type. Instead of the tubercles of the teeth remaining for the most part separate, they have fused into a set of ridges, of which, however, the pattern is less complex than in the modern Horses. In other respects Orohippus is much like Hyracotherium. Pachynolophus seems to be but a synonym.

The next stage is shown by Mesohippus, a Lower Miocene form, usually referred to the neighbourhood of Palaeotherium. It has nearly lost one of the toes of the fore-foot, a rudiment only remaining; the metapodials, at any rate of the fore-feet, seem to be slightly increased in length. The orbit is not encircled by bone, but there is a strong process from the frontal, which nearly meets the zygomatic arch.

Anchitherium, from the Upper Miocene, is not far removed in structure from the last-mentioned form; it is a trifle nearer the existing Horse in several points. The ulna is further reduced and fused with the radius below: the rudiment of digit V is still more rudimentary; the two lateral digits are smaller in proportion to the central one than they are in Mesohippus; the fibula is fused below with the tibia. From this form to Equus is a small series of steps, characterised by the still further reduction of all the digits except III, by the still further reduction of the already rudimentary ulna and fibula, and by the increasing depth of the molar teeth, which are of course, in Equus, hypselodont.

Another interesting conclusion may seem to follow when we consider the geographical range of the ancestral Horses. Hyracotherium and Pachynolophus occurred both in the Old and New World. From them may have arisen the Horses of both hemispheres. After that point there is a division. Mesohippus is American, and we get at Equus in that continent through Desmatippus and Protohippus. On the other hand there are no remains known of Mesohippus in Europe; and unless subsequent researches prove the existence of Mesohippus, we have to rely upon forms which are placed with Anchitherium and Hipparion.

It seems that in America the next genus in the direct line of equine descent to Mesohippus is Miohippus. It is smaller in