104 THE THEORY OF EVOLUTION 



thickness, but is complete and separate from the 

 radius, and in the lower hind leg, the fibula is separate 

 and has a very slender shaft, but it is entire and 

 uninterrupted. The feet are still three-toed, but 

 the lateral toes are longer and more useful in carry- 

 ing weight. 



Once more omitting many intermediate forms, 

 we may examine the most ancient known horses, 

 those of the lower Eocene, which were little creatures, 

 not exceeding a fox in size. The grinding teeth have 

 a very simple and primitive pattern, though the 

 beginnings of the complicated horse-pattern may be 

 discerned in them; the premolars are all smaller and 

 simpler than the molars. The neck and limbs and 

 especially the feet are relatively short; there are four 

 functional toes in the front foot and three in the 

 hind, but the latter also has splint-like rudiments 

 of the first and fifth, plainly indicating derivation 

 from a five-toed foot. 



Here the line breaks off and cannot at present be 

 traced farther back, the preceding Paleocene of 

 North America having yielded nothing that can be 

 regarded as ancestral to this family. The lower 

 Eocene horses in Europe are of the same type and 

 there also they appeared unheralded by any pre- 

 decessors in the Paleocene. The obvious conclusion 

 from these facts is that the family arose in some 

 region, as yet unidentified, but which not improbably 

 was central Asia, and thence migrated into Europe 

 and North America; the latter continent was made 



