422 Mr. G. A. Boulcnger on 



great Central African lakes. The study of these fishes is a 

 bewildering one from the very close affinity which connects 

 them, as I have already pointed out in a recently published 

 paper dealing with the collection made by Mr. Ilinde in the 

 Kenia district * ; and nothing- short of the large series (about 

 350 specimens) got together by Mr. Uegen could have enabled 

 me to estimate correctly, as I believe, the value of the cha- 

 racters which have been regarded by some authors as generic, 

 whilst by others they have been refused even specific importance. 

 ThuSj the character on which Riippell founded his genus 

 Laheoharhiis, though insuflScient for generic separation, is of 

 the greatest value for the distinction of species, as I have 

 ascertained that it is not sexual nor dependent on age in tjie 

 Abyssinian fishes, since the lips are as much produced in a 

 Barhus [Laheoharhus] ne.dgin of 6|- inches as in one of 20. 

 The idea expressed by Giinther a few years ago f, that adult 

 specimens of Barhus intermedlus have the lips more developed, 

 is not supported by the evidence available at present, and is 

 clearly the result of an error of determination, since the largest 

 specimen named " Barhus intermedlus^ Riippell," from East 

 Africa J, on M'hich this view is founded, measures only 

 1'^ inches, whilst the type of B. intermedlus from Lake Tsana§, 

 for the loan of which I am indebted to the kindness of the 

 Directors of the Senckenberg Museum^ measures 13 and lacks 

 the mental lobe. 



It is extremely surprising how fishes, agreeing completely in 

 the f( rm and scaling of the body and in the position and struc- 

 tureof the fins, in the pharyngeal teeth, and alsoin the skeleton, 

 so far as I have been able to ascertain, may differ very con- 

 siderably in the proportions of the various parts of the head, 



* P. Z. S. ]902, ii. p. 221. t P. Z. S. 1894, p. 91. 



t I propose the name Barhus Gregorii for the species referred by 

 Giinther to B. intermedins, from which it differs in the much-developed 

 lower lip with a rounded median lobe, and in the longer barbels, the 

 anterior measuring If to If, the posterior If to 2 diameters of the eye. 

 D. IV 8-9 ; A. Ill 5 ; Sc. 31-33 ^-Ht. In its smaller scales, of which 

 three series intervene between the lateral line and the ventrals, it differs 

 from the other allied species of the Tana sj-stem {B. tcaiensis, Gthr., 

 Hindii, Blgr., perplexic<ms, Blgr., labiatus, Blgr.) ; in these the number of 

 scales is 25-30 ^^% 2 between the lateral line and the ventrals, and the 

 number of branched dorsal rays is 9 or 10. A specimen from the Gota 

 Eiver, in the Degen collection, appears to be referable to B. Grerjorii. 



§ Often written " Tana " on modern maps. I think it desirable to 

 retain the old orthography, in order to avoid confusion with the system 

 of the Tana River further south in East Africa, the fishes of which have 

 much in common with tho.ve of the great Abyssinian lake which is the 

 source of the Blue Nile. 



