248 EAST AFRICAN FLORA AND FAUNA part hi 



As has been already suggested, a great change in the 

 character of the flora of a country must inevitably influence 

 the distribution of its animals by altering the food supply. 

 Thus insects will vary with the flowers, and birds with the 

 insects ; small mammals will disappear as the vegetation 

 becomes sparser ; earthworms will be restricted to the more 

 fertile regions, and termites or " white ants " replace them, as 

 the soil becomes harder and more barren. In the case of the 

 fresh-water fauna, the range of possible variation in the environ- 

 ment is more limited. The current of a river may become 

 slower, or the quantity of water in a pool may lessen ; but the 

 conditions do not vary fundamentally, unless the water dries up 

 entirely, and such a change as this is at once fatal to animal 

 life. Nevertheless, the fresh-water fish of Equatorial Africa 

 present as many anomalies in distribution as the land animals 

 and plants. When Giinther ^ in 1869 described the collection 

 made by Petherick in the Upper Nile, he pointed out that the 

 fauna of which it is a part has a very remarkable distribution, 

 being more nearly related to the faunas of the Palestine and West 

 African rivers than to those of Lake Nyasa and the Zambesi. 

 Later collections have fully confirmed the truth of Dr. Gunther's 

 view. Though the fish of the Zambesi, the Upper Congo, and 

 their connected lakes, such as Nyasa and Tanganyika, are in 

 the main identical in genera with those of the Upper Nile, they 

 differ from them in species ; while the Upper Nile contains 

 some of the very same species as the Jordan and the Sea of 

 Galilee. My collections prove that the rivers of British East 

 Africa, such as the Tana, Athi, and Sabaki, and the rivers and 

 lakes of the Rift Valley have the same generic fauna. The 

 Tana, however, is intermediate between the Nile and the 

 Zambesi, for it has two species from each, viz. Clarias lazera, 

 C. v., and Clarotes laticeps, Rlipp., from the former, and Eutropius 

 depr'essirostris, Ptrs., and Synodontis zambezensis, Ptrs., from the 

 latter, as well as one {Barbus intermedius, Riipp.) from Abyssinia.^ 



The fact that renders the evidence of fresh-water fish so in- 

 structive is that the distribution of the species is often extremely 

 local. Thus many of the small British and Irish lakes have 



^ In J. and B. H. Petherick's Travels in Central Africa, Appendix C, vol. ii. pp. 

 197-268. 



^ A. Giinther, " Report on the Collection of Reptiles and Fishes made by Dr. J. W. 

 Gregory during his Expedition to Mount Kenya," Proc. Zool. Soc. (1894), pp. 84-91. 



