534 DR. W. A. CUNNINGTON ON THE 



Lakes Edward and Kivu and nowhere else. Tliis is a matter of 

 some interest, since there is evidence for believing that these two 

 lakes, between which there is now no connection whatever, were 

 ill I'elatively recent times in communication with one another 

 {cf. Moore, 137, p. 89). The geniis Faraiilapia is another repre- 

 sented in the list by a large number of species, these being 

 associated in groups which in most cases are peculiar to one of 

 the three bigger lakes. Felmatochromis, which is widely dis- 

 tributed in West Africa, occurs in Tanganyika and in Victoria 

 ISTyanza. but otherwise only in a single instance (Lake Edward), 

 in the Inkes with which this paper is concerned. 



There is no occasion to refer individually to the numerous- 

 Cichlid genera vv^hich are endemic in the African lakes, the 

 important point being that so many of them are found in Tan- 

 ganyika. Attention may, however, be called to Chainpsochromis: 

 from ISTyasa with three species, and Bathi/baies from Tanganyika 

 with six, as these exhibit the greatest number of species among 

 the different endemic genera. The genus Lamproloyus is of 

 considerable interest and importance, being represented by a 

 total of 27 species from the Congo and Lake Tanganyika. It is. 

 nevertheless essentially characteristic of the latter, for 24 endemic 

 species are recorded from tlie lake, while the remaining three 

 species are peculiar to various parts of the Congo river sj'stem. 

 The distribution of the three forms of Hemitilapia is quite 

 different from anything else met with in the list, since the genus 

 is only known from Lakes Nyasa and Victoria. A single endemic 

 species occurs in the former lake and two endemic species in the 

 latter. 



Since the Cichlids are so pre-eminent among the fishes of 

 Tanganyika,, and indeed constitute a most striking feature of the 

 lake fauna as a whole, it becomes necessary to refer to certain 

 conclusions which have been drawn from a study of the anatomy 

 of these particular forms. From various considerations, such as 

 the number of rays, the midtiple lateral lines, and the number of 

 vertebrae, Boulenger (20, p. 2 : 25, p. 393) at one time regarded 

 certain Tanganyika genera as showing features of generalisation. 

 At a later date, however, after a further investigation into the 

 anatomical evidence bearing on the ])hylogeny of the group, 

 he arrived (28, p. 422 : 38, p. 43) at a different conclusion, 

 and considers the Tanganj'ika genera and species as essentially 

 specialised. This is an important pronouncement, since it brings 

 the case of these fishes into line with most of the other animal 

 groups under discus.sion, in which cei'tain marks of specialisation 

 are found to be characteristic of the endemic Tanganyika forms. 



Passing now to the family Anabantidfe, a solitary representative 

 (Anabas mtiriei) has been obtained from Victoria iSTyanza. It is 

 a Nilotic type, and its existence here calls for no particular com- 

 ment. — The last group in the list, that of the Mastacembelidae, is 

 represented in Africa only by the genus Mastacevibelus itself. 

 Many species are known fiom the fresh-waters of the continent, 



