20 AFRICAN GAME ANIMALS 



effect to this field of theory. Notwithstanding the present 

 apparent futihty of such speculation, participation in it is 

 irresistible. We may justify such indulgence of the fancy 

 as a groping toward the light or a legitimate attempt to 

 illuminate the obscurity surrounding the remote past. In 

 the northern hemisphere or holarctic region the paleo- 

 mammal records are at least partially complete in some 

 few groups, but in Africa the paleontological records are so 

 absurdly scanty that we are left with hardly a straw to 

 support our speculation. Africa is so dark paleontologic- 

 ally that we may almost consider ourselves engaged in 

 mere guesswork. 



Localities in which mammal remains have been found 

 in Ethiopia, or Africa south of the Sahara Desert, are limited 

 to a few small deposits. In the Cape region are two de- 

 posits of Pleistocene age containing species allied to or iden- 

 tical with those existing to-day in the same district. The 

 most important of these localities is Barkley West, situated 

 on the Vaal River, near the boundary of the Orange River 

 Colony. Fraas described, in 1907, from this locality, a 

 species of Damaliscus, a hippopotamus, and a mastodon. 

 From the Zululand coast, in Natal, Scott* has described a 

 buffalo, a black rhinoceros, a hippopotamus, and an ele- 

 phant; all forms closely related to living African species. 

 Broom has recently described from the Orange River a 

 Pleistocene species of hartebeest aUied to Coke's, but ex- 

 ceeding it in size, and a large horse from the shore deposits 

 near Cape Town. More recently part of a mandible of a 

 Dinotherium, a primitive elephant bearing mandibular 

 tusks, together with remains of various other animals, have 



* "Geol. Surv. Natal and Zululand," 1907, p. 253. 



