THE TANGANYIKA PROBLEM. 341 



interesting ; the Cichlidae, as we have seen, abound in Tan- 

 ganyika, they are distributed more sparsely over the rest of 

 Africa and in South-eastern Asia, but they abound also in 

 the rivers of South America. Of the other families of fish 

 which occur in Tanganyika, the Siluridae are apparently 

 cosmopolitan ; and many types belonging to the group 

 are at the present time marine. The Cyprinodontidffi are 

 peculiar to Europe, Asia and America. 



The Cyprinida^ to Europe, Asia and North America. 



The Characinida? to Africa and Central and South 

 America. 



The Polypteridse only to certain parts of Africa, including 

 Tanganyika. 



The Lepidocirenidae to Africa and South America. 



The Mormyridse are peculiar to Africa. 



The Mastacembelidae to Africa and Southern Asia. 



The chief feature of the distribution of these families 

 which occur in Tanganyika now, is thus seen to be that 

 they belong to groups which are found to be specially 

 related to both sides of the Atlantic as it now exists. I see 

 no reason to dispute the view originally put forward by 

 Dr. Gimther, that the fresh-water fish-fauna of Central 

 Africa has radiated from the region of the great lakes. 

 For this view seems to be strongly supported by the 

 evidence produced by Mr. Boulenger, who has shown 

 that the Cichlidae of Tanganyika are among the more 

 primitive of that group, and when we consider the 

 above features of the distribution of the fishes, together 

 with the facts, that there are marine animals in Tan- 

 ganyika, and that the early representatives of these 

 fishes were marine, there appears to be a very strong 

 indication, indeed, that the fish-fauna of Tanganyika, as 

 well as some portion of the fish-fauna of the American 



