FAUNA OP THE AFRICAN LAKES. 607 



A veiy similar result is ai'rived at on comparing the number of 

 groups represented in the dift'ereiit lakes. From A^'ictoria Nya.nza 

 <lo\vn\var(ls the decrease in size is acconipn,nied by a decrease in 

 the number of groups present. Tlie figures (including Tangan- 

 yika for com[)arison) are found to l)e as follows : — Tiuiganyika 

 26 (groups), Victoria 25, Nyasa 24, Albert 15, Edward 14, and 

 Kiwi 6. It is possible, however, that the low totals for the 

 smaller lakes are, in part, a result of less complete investigation. 



Enough has been said to show that Mooi-e's contention is 

 substantially correct as far as these African lakes are concerned. 

 It would be interesting to discover whether a similar relation 

 between size and number of specific forms can be made out for 

 other groups of lakes or even if it is a principle of general appli- 

 catiou. Moore makes a. comparison, which would seem to be 

 justified, between this phenomenon and that exhibited by the 

 llora, of oceanic islands, whei'e the smaller the island, — although 

 it may be as thickly covered with vegetation as any other area, — 

 the fewer the species of plants which inhabit it. 



The last matter to be considered is the undoubted affinity 

 which exists between certain African atid Indian fresh-water 

 types — an affinity which has been noticed already, when re- 

 viewing the groups in which it is most pronounced. This inter- 

 relationship is exhibited in many groups of animals, and extends 

 not merely to forms from the Indian Peninsula itself, but from 

 the whole of that pai't of Asia, including the Malay Archipelago. 

 An interesting account of these affinities is given by Annandale 

 in a paper entitled " The African Element in the Freshwater 

 Fauna of British India " (10). So far as the present tre<ati8e is 

 concerned, consideration is limited to those animals which are 

 known to occur in one or other of the big African lakes. Briefly 

 enumerating the ca.ses, the Cyprinidpe and Mastacembelida; among 

 the fishes indicate this affinity in a marked degree. The Polyzoa 

 afford striking evidence, since the genus Arachnoidea is known 

 from Tanganyika and East Indian seas, while in the case of 

 Plmnatella ( Afriiidella) tanganyikce the actual species has been 

 foimd in an Indian lake as well as in Tanganyika. Oaridina 

 nilotica with its varietal forms occui's in several of the African 

 lakes, while it is widely distributed in Indian and Malay regions 

 and extends still further east into Australia. Among the 

 Coelenterata a medusa has now been found in India which is 

 generically identical with that from Tanga.nyika and Victoria 

 Nyanza, while finally among the sponges SjJongiUa earteri from 

 Lake Victoria is known in India a.nd Java. 



This is not the place to discuss the geological evidence for a 

 former land-connection between these areas, but the views 

 commonly held can be stated in a few words. During the 

 Carboniferous period, and persisting subseijuently through the 

 Permian and Triassic, there appears to have existed a vast 

 tropical continent which extended from Brazil to Australia, 

 embracing of course Africa and India. This continent is known 



