556 PALEONTOLOGY 



day. In other words, the coast-line seems to have been continuous 

 at that time, a neck of land uniting Asia and North America where 

 now there exists the Bering Strait. The Pliocene land-animals of the 

 northern hemisphere agree in suggesting the same connection. Hence, 

 the ultimate separation of the so-called Old and New Worlds is shown 

 by fossils to be quite a modern event in geological history. 



Again, it has been proved by recent researches 1 that the mollusks, 

 brachiopods, and trilobites found in the Devonian rocks of South 

 Africa, agree much more closely with those occurring in the corre- 

 sponding formations of South and North America than with those of 

 Europe. The South African sea in the Devonian period seems there- 

 fore to have extended directly into the American region, but to have 

 been separated by a barrier from the European region. Similarly, 

 there is evidence of circumscribed seas separated by land-barriers in 

 the Triassic, Jurassic, and other epochs; and when the fossils from all 

 parts are sufficiently well known, it will be possible to determine even 

 some of the minor geographical features of each successive period. 



To restore the old continents and to discover their varying connec- 

 tions and disintegrations is an especially fascinating problem. A 

 means of solution is provided by the various terrestrial vertebrates, 

 which, under ordinary circumstances, are unable to cross seas. When 

 a new race suddenly appears in any land, it obviously implies the 

 removal of the barrier which previously prevented that race from 

 spreading. The primitive elephants, for example, suddenly invaded 

 Europe at the beginning of the Miocene period. Recent discoveries 

 in the Egyptian desert have proved that their ancestors lived and 

 evolved in the Eocene and Oligocene periods in northern Africa. 2 

 Therefore, during this earlier time, the European and African regions 

 were separated by some barrier, doubtless the sea; at the dawn of 

 the Miocene period earth-movements of some kind resulted in a land 

 connection over which mammals could migrate. 



The use of terrestrial vertebrates in deciphering the past history 

 of continents is, however, less simple than it may at first sight appear; 

 and the case of South America may be quoted as an interesting illus- 

 tration. With reference to the latest phases in the development of 

 this land, only two main conclusions are well founded. The first fact 

 to notice is that, of the jaguars, pumas, wolves, bears, tapirs, deer and 

 llamas, which now characterize South America, and of the mastodons 

 and horses which lived there in the Pleistocene period, there are no 

 remains in the geological formations of that country below the top 

 of the Pliocene. Hence, as representatives of all these quadrupeds 

 lived at an earlier date in North America, there must have been some 



1 F. R. C. Reed and P. Lake, Ann. S. African Museum, vol. iv, pts. 3, 4, 6 

 (1903-04). 



3 C. W. Andrews, "On the Evolution of the Proboscidea," Phil Trans. 1903, 

 on. B. 217. 



