382 HISTORICAL GEOLOGY. 



been found. As time went on, these separated more and 

 more widely, and gave off sub-branches, which again 

 divided, and so on. In general terms, it may be said that 

 some of the existing orders may be traced back to the 



Eocene. Many of the exist- 

 ing families commenced in 

 the Miocene ; existing genera 

 in the Pliocene ; but existing 

 species only in the Quater- 

 nary. This is well illustrated 

 by one great branch, the Un- 

 gulates, or hoofed animals. 

 These consist now of many 

 widely separated sub- 

 branches ; but in the earliest 

 FIG. 344.-Head of Dinotherium Tertiary they seem to unite 



giganteum, greatly reduced. . J J . 



into one, a primal ungulate. 



As we go up, this branch separates, even in the Upper 

 Eocene, into odd-toed (perissodactyls) and even-toed 

 (artiodactyls) ungulates. In the Miocene, each of these 

 again separates, the former into the elephant family 

 (Proboscidians) with five toes, the tapir and rhinoceros 

 families with three toes, and the horse family, with three 

 toes passing into one; the latter into the hog and hippo- 

 potamus families with four toes, and the ruminant family 

 (horned animals) with two toes. 



Genesis of the Horse. Let us trace one of these 

 branches throughout. We select for this purpose the 

 horse. A most wonderful series representing this family, 

 about forty species in all, has been furnished by the 

 American Tertiaries, and the successive steps traced by 

 Professor Marsh. First of all, in the early Eocene Wah- 

 satch beds, we find the Eohippus (dawn-horse). This 

 little animal (the size of a fox) had three toes on the 

 hind-foot, and four perfect toes and a fifth splint, and 

 perhaps dew-claw, on the fore-foot. Next, in the Middle 



