FAUNA OF THE AFRICAN LAKES. 



517 



when it was actually obtained from a neighbouring swamp ? 

 I have dealt with the difficulty in what seems the most common- 

 sense way, by definitely including all forms recorded from the 

 drainage area of a lake as belonging to it. Nevertheless it must 

 be pointed out that even this decision does not avoid the trouble 

 altogethei-, for it is not an easy matter to determine what con- 

 stitutes the drainage area of a lake in a country where the 

 boundaries of river basins are quite indefinite in the rainy 

 season. Again, the possibility is not excluded, that oiganisms 

 from adjoining fresh-waters may not find physical conditions 

 suitable for them in the big lake in their vicinity, in which case 

 they should not be included in its fauna. Evidence of tbe 

 existence of this very phenomenon is to be found as regards 

 certain plankton organisms which are absent from both Tangan- 

 yika and Kivu. Throughout this paper I Lave placed a § against 

 the names of all species which have not definitely been recorded 

 from the waters of the lake in question. 



Subject to these difficulties in compiling exact lists, the 

 following approximate figures can be given : — Tanganyika contains 

 402 species of which no less than 293 are endemic, Nyasa with 

 a total of 361 and 86 endemics coming second. Victoria Nyanza 

 has a somewhat smaller total, viz. 289, but a larger number of 

 endemic forms (110). The tliree smaller lakes show a great 

 falling off in numbers, Albert Nyanza mustering ouly 67 species 

 of which 9 are endemic and Edward Nyanza 54 with 11 endemic 

 species. Kivu, finally, only contains 23 types in all, 4 of which 

 are peculiar to the lake. Thus, as regards its fauna, Tanganyika 

 is distinguished from the other lakes not only by containing a 

 larger number of forms, but especially by displaying a remarkable 

 proportion of endemic species. 



While it is not necessary to discuss here the origin of fresh- 

 water organisms, a few observations on one aspect of the question 

 are needed at this stage. By movements of the earth's crust, 

 portions of the sea have, from time to time, been cut off' with 

 their inhabitants, and by gradual freshening have been sub- 

 sequently converted into fresh-water lakes. During the lattei* 

 process, there is little doubt that certain of the organisms would 

 be unable to withstand the change and would perish. Inland 

 lakes which appear to be the modified remainders of isolated 

 portions of the ocean are sometimes spoken of as *' relict lakes" 

 (Reliktenseen). It is, of course, precisely this view of the origin 

 of Tanganyika which was put forward by Giinther and afterwards 

 elaborated by Moore. On this hypothesis, the peculiar fauna 

 was oi'iginally that of a former sea, then that of a salt lake, and 

 finally that of a fresh-water lake. 



The marine-like appearance of certain of the Gasteropod shells 

 of Tanganyika was expressed by the conchologist Bourguignat 

 by employing the term " thalassoid " with reference to them — a 

 term which did not commit its author to any view as to origin 

 (42, p. 9). In his paper " On the Zoological Evidence for the 



