THE TANGANYIKA PROBLEM. 329 



regarded as belonging to the halolimnic group, although they 

 have not the strikingly marine attributes characteristic of 

 the other members of this series. At the same time, and 

 in Chapter VII., it was emphasized that although the fishes 

 which now inhabit Tanganyika all belong to what have 

 become exclusively fresh-water forms, the fish-fauna of the 

 lake differs entirely in character from the fish-fauna of any 

 others of the great lakes of Central Africa, and it was 

 further shown that the majority of a large section of this 

 fish-fauna, the Cichlidae, is wholly restricted to the con- 

 fines of Tanganyika. Actually about half the Old World 

 species of this family being endemic in the lake even 

 now. In this way it was rendered evident that one 

 of the chief peculiarities of the more striking members 

 of the halolimnic group, namely, their geographical isola- 

 tion, is shared equally by a section of the fishes, and 

 consequently that these fishes were in one sense as 

 distinctive of the fauna of the lake as the more striking 

 invertebrate halolimnic animals themselves. Or in other 

 words, it was found that Lake Tanganyika possesses a 

 fish-fauna which is distinctive and characteristic of that 

 lake, and is wholly unlike the fish-fauna of any other 

 of the great African lakes. From these considerations 

 tions I was inclined, as I pointed out in Chapter VII., to 

 regard at any rate the endemic Cichlidse, and probably the 

 ancient Ganoids, together with some of the Caricinidse as 

 the remaining and somewhat scattered piscine portion of 

 the halolimnic fauna ; the wide dispersal of these animals 

 being due to the obvious migratory capacities of fishes 

 when compared with the invertebrate section of the 

 halolimnic group. 



Turning now to the consideration of the nature of 

 this group of animals, so many of which are obviously 



