228 R. I. POCOCK [makch 



sufficient importance to warrant the belief that they have been 

 isolated during the enormous length of time that has elapsed since the 

 Jurassic epoch. The species of Thelyphonidae, Solifugae, and most of 

 the scorpions, for example, are only generically different from their old 

 world allies ; and, although the scorpions of the family Bothriuridae are, 

 with the exception of the Australian genus Cercophonius, peculiar to the 

 region, they are, in my opinion, merely a specialised group descended 

 from the American members of the Vejovidae (Iuridae), a family 

 which at the present time is represented by genera in the Mediter- 

 ranean and Oriental regions, in California, Mexico? Bolivia, and Chili. 

 It is mainly to the south of the area occupied by the Vejovidae in 

 South America that the Bothriuridae occur in force, and their structure 

 and distribution suggest that they became gradually differentiated as 

 the Vejovidae spread southwards into South America. Fortunately 

 the survival of the Vejovidae has left a clue to the origin of the 

 Bothriuridae. 



So, too, in the case of the tailless Pedipalpi of the family Admetidae, 

 which are confined to the region. The structural peculiarities of the 

 genera seem hardly important enough to justify the view that they 

 have been isolated since the early portion of the Secondary epoch ; and 

 the "supposition that the eastern and western representatives of the 

 Amblypygi were probably cut off from intercommunication with each 

 other at an earlier date than were the scorpions, whip-scorpions, and 

 Solifugae, is supported by the fact that at the present time they do not 

 extend to the north of the Tropic of Cancer in the Old World, and 

 scarcely surpass it in the New. This would lead to an earlier differen- 

 tiation of the genera on the two sides of the Pacific, and the differences 

 would go on increasing during the gradual refrigeration of the northern 

 hemisphere, which culminated in the glacial period, when the Pedi- 

 palps in question were probably exterminated as far south as a line 

 represented by the northern tropic, and the northern ancestors of the 

 Admetidae were blotted out. 



If, however, the view that the Neotropical scorpions, Pedipalpi, etc., 

 did not enter South America in pre-Jurassic times be rejected, it still 

 remains to be seen whether any other country than North America 

 can be looked upon as the source that has supplied South America 

 with its fauna. 



On palaeontological and geological grounds it is believed by Mr. 

 Lydekker that the ancestors of the mammalian ungulate fauna of South 

 America migrated from South Africa by means of a land connection in 

 Tertiary times, a small contingent of marsupials crossing in the same 

 way and during the same epoch from Australia. This migration of 

 African and Australian forms took place before the northern irruption 

 of the higher mammals, such as cats, bears, llamas, etc., into South America 

 in post-Miocene times ; and since, as is believed, the ancestors of the 

 -existing population of African eutherian mammals and of Australian 



