[ 399 ] 



V. On a Collection of Fishes from the Lake N garni Basin, Bechuanaiand. 

 By G. A. BouLENGER, F.B.S., V.P.Z.S. 



[Received December 17, 1910 ; Eead February 21, 1911.] 



[Plates XXXVIII.-XLIII. and Text-figures 85-87.] 



In his little book ' Memoire sur les Poissons de I'Afrique Australe ' (Paris, 1861), 

 Count Francis de Castelnau gave the descriptions of a number of new fishes from Lake 

 Ngami, then a lake of some importance discovered by Livingstone and Oswell. He 

 had sent there one of his " preparateurs," Frederic Daviaud, who brought back to 

 Cape Town, where Castelnau Avas Consul, a number of dried specimens, probably 

 accompanied by notes on the coloration, from which the descriptions were drawn up. 

 The types of these are all lost, and as the definitions are inadequate, it has been 

 impossible hitherto to allocate a position in the system to most of the species 

 described by Castelnau. 



Over forty years have elapsed since the publication of the ' Memoire' quoted, and in 

 the meanwhile no one seems to have collected fishes in the lake, which is rapidly drying 

 up. I was therefore happy to hear, two years ago, through my colleague Mr. Ogilvie- 

 Grant, that Mr. R. B. Woosnam was preparing an expedition to Bechuanaiand, and 

 that it might be possible to get at the lake for the purpose of obtaining a series of its 

 fishes. An application having been made by Dr. P. L. Sclater to the Eoyal Society's 

 Government Grant Committee, Mr. Woosnam was provided with the necessary means 

 to extend his collecting-trip in that direction. Although unforeseen circumstances 

 have prevented his reaching the lake itself, he has nevertheless been able to form 

 a considerable collection of fishes from the Okovango river that flows into it, a collec- 

 tion by means of which I have been able to identify, with some approach to certainty, 

 most of the species described by Castelnau, and thus remove a stumbling-block 

 in African systematic ichthyology. 



xls Castelnau's little book is not easily procurable, I have reproduced most of the 

 original descriptions, in order to enable others to judge of the degree of probability 

 of my identifications, which, needless to say, in view of the unsatisfactory nature of 

 Castelnau's Avork, are in some cases little more than guesses. On the other hand, a few 

 of the species in Mr. Woosnam's collection could not be referred to any previously 

 named, and are here described as new. 



The Ngami Fish-fauna shows no feature differentiating it from that of the Zambesi, 

 and not a few species are common to both. Although no striking discoveries have 



VOL. XVIII. — PART V. No. 1. — Mai/, 1911. 3 i 



