134 THE NEOG^EIC REALM. [CHAP. 



Africa and South America in Tertiary times was only in high 

 latitudes, a warm epoch in the southern hemisphere must have 

 been necessary for the passage of such tropical forms. It might 

 be urged that as Ceratodus dates from the Trias, the other two 

 genera might have reached their present habitats at a very 

 distant epoch ; but their specialisation is against their antiquity. 

 Another family which is essentially southern is that of the Osteo- 

 glossidce, represented by Arapaima of the Brazilian rivers, and 

 Osteoglossum, with one species from Brazil and the Guianas, a 

 second from Sumatra, and two others from Australia. But the 

 northern origin of the family is indicated by the occurrence of the 

 extinct genus Dapedoglossus in the Eocene of Wyoming. Here 

 there is a presumption that Osteoglossum originated in Asia, from 

 which it passed in one direction by way of Malaysia to Australia, 

 and in another through Africa to South America. Two other 

 families of freshwater fishes have a somewhat similar distribution ; 

 the first being the Chromidid<z y which includes spiny fishes mainly 

 characteristic of tropical America and Africa, but extending 

 eastwards into Syria, and sparingly represented in Southern India 

 and Ceylon. In a fossil state they occur in the Cretaceous of 

 Syria ; and, although none of the genera are common to the two 

 continents, they are highly suggestive of a connection between 

 Africa and South America. The second family is that of the 

 Characiniidce, comprising fish more nearly allied to the carps, and 

 now exclusively confined to tropical America and Africa. Although 

 the palaeontological record is a blank, this can scarcely be taken 

 as a sufficient indication that the family has always been a 

 southern one. 



From the foregoing facts it may be considered that the as- 

 sumption of an Antarctic continent is unnecessary 

 to explain the origin of the many forms of vertebrate 

 life which are now exclusively or mainly southern ; nearly all of 

 these, with the exception of the edentates and penguins, being of 

 northern derivation, and thus apparently showing a southern 

 migration of the older forms of life. The Cretaceous and Tertiary 

 break between North and South America appears, however, to 

 have prevented the occurrence of such migration in the western 

 hemisphere till the close of the Miocene : and it is accordingly 



