134 HUNTING EXTINCT ANIMALS 



our animals lived is another much discussed problem 

 with almost as many conclusions as there are writers on 

 the subject. There are no faunas which arose independ- 

 ently and have not received some contribution from other 

 continents. The mammals arose apparently on the north- 

 ern hemisphere or were at least spread over it from their 

 earlier beginnings. Their presence in South America indi- 

 cates that they came from some other region, though their 

 peculiarities may be due to isolation. The fresh water 

 fishes of South America have more affinities with those of 

 Africa than with other continents. The fresh water mussels 

 and snails also show a striking resemblance between the 

 two continents. This is true of some of the fossils of the 

 Cretaceous, like the dinosaurs. In general there is an 

 agreement that Africa and South America were united 

 across the South Atlantic in the early Cretaceous, but the 

 usual opinion has been that these two continents were 

 separated early in this period. This was too early to ac- 

 count for the mammals, etc. By several authors there has 

 been a tendency to continue this connection to later times, 

 in which case we should expect to find resemblances be- 

 tween the land animals of the Eocene at least. Such 

 resemblances have in many cases been asserted but mostly 

 on incomplete material. For instance, Ameghino consid- 

 ered that the Pyrotherium belonged in the elephant family 

 and was closely related to remains of elephants found 

 in the Fayum desert in northern Africa. The nearly com- 

 plete skull which we found, however, shows no resemblance 

 to the early elephants except in the shape of the grinding 

 teeth, but in the important structural points it approaches 

 to Toxodonts, which are a purely South American group. 

 Of the evidence for an African migration very little if 

 anything remains. There are forms which Ameghino 

 interpreted as Primates and which would point to a con- 

 nection with North America, but these with fuller material 

 prove to be marsupials and that connection fades. The 



