736 PERISSODACTYLS xxx. 9- 



lines' are so numerous that they immediately throw doubt on the idea 

 that there has been any single uniform 'trend' in horse evolution. At 

 least twelve types sufficiently marked to be classified as genera are 

 known, in addition to those directly on the line leading to Equus\ 

 of course there is a much larger number of shorter, independent, 

 evolutionary lines within these genera. We have enough evidence to 

 glimpse the extraordinary complexity that would be revealed by the 

 complete evolutionary 'bush', even in this single family. A further 

 complication is produced by migrations. It is at present believed that 

 the main course of horse evolution went on in North America, with 

 migration at various times to the Old World and South America. 

 Certainly a more continuous series of forms has been revealed in 

 North America than elsewhere, but it must be remembered that they 

 have been looked for intensively, and brilliantly studied. It is not 

 impossible that further study of Old World horses will produce still 

 greater complications by revealing sequences of evolution within that 

 area. 



Throughout the Eocene epoch the horses all possessed four toes 

 in each limb. The fossils classed as *Orohippns and *Epihippus from 

 the Middle and Upper North American Eocene are little different 

 from *Hyracotherium, except for molarization of the hinder premolars. 

 The size remains small. 



The Oligocene horses, *Mesohippus and *Miohippns, walked with 

 three toes on the ground, and all the premolars were molarized. The 

 ectoloph was well formed but the inner cusps were still separate, and 

 the teeth low-crowned. Some horses of this type (*Anchitherium and 

 its descendant *Hypohippus) persisted into the Miocene, presumably 

 surviving as browsers in forests, while other descendants took to the 

 plains. These browsing horses migrated to the Old World in the 

 Miocene, then died out there, as they did also in North America. 



*Parahippus of the American Lower Miocene shows the beginning 

 of the adaptation for life on the plains. The lateral digits II and IV 

 still carried hoofs, but since the central proximal phalanx was much 

 the longest and strongest, it is probable that the lateral ones touched 

 the ground only to maintain balance over uneven surfaces, or in soft 

 conditions. The teeth were still rather low, but were beginning to be 

 elongated and to show cement on the crowns. The protoloph and 

 metaloph were connected by a narrow bridge. There was a partial 

 post-orbital bar. 



*Merychippus comes from later Miocene beds and could have been 

 directly derived from *Parahippus by increase in the depth of the 



