On Fishes from Gullaland. 557 



LXXI. — On a Collection of Fishes from Gallahinil. 

 By G. A. BOULENGER, F.R.S. 



Thanks to tlie p^enerosit}' of Mr. W. N. M'^Milhin, the 

 British ^luseum lias recently received a large coll-ction of 

 lishes made by Mr. P, C. Zaphiro during a trip to Kaffa and 

 Lake Rudolf. The great care bestowed by Mr. Zaphiro on 

 the preservation and labelling of the specimens, about 400 in 

 number, most of which are accompanied by notes on the 

 coloration in the fresh state, and the great immber of new 

 or rare forms from comparatively little-known waters *, 

 render this collection a very valuable one, by which our 

 knowledge of African ichthyology is further advanced. The 

 collection was made in live distinct hydrographic systems : — 

 1. The Blue Nile; 2, The Omo with Lake Rudolf ; 3. Lakes 

 Abaya and Ganjule with the Sagan River, which connects 

 them with Lake Stephanie f; ^' Likes Zwai and Suksuki ; 

 5. The Hawash. According to Mr. Zaphiro, Lake Stephanie 

 contains no fishes. The species represented in the collection 

 are first enumerated in hydrographical order +. 



The collection made by Mr. Zaphiro has added to the list 

 of the species of Barhus of the B. Bynni group, the variety 

 of which in Southern Ethiopia and East Africa constitutes so 

 striking a feature. Undoubtedly fast-running mountain- 

 streams are more favourable to the existence of Cyprinids than 

 of fishes of a more strictly tropical character. But it, never- 

 theless, remains a ])uzzling fact that these /?a/*5ws should have 

 split up into such a numljer of allied forms, whilst the Laheo, 

 seemingly still better adapted for waters of a torrential 

 nature, have remained unaltered, only four species, three 

 of them of wide distribution in Africa, being known from the 

 district. 



• Cf. Gunther, P. Z. S. 1896, p. 217; Vinciguerra, Anu. Mus. Genova, 

 (2) xvii. 1897, p. 34-3, and xix. 1898, p. 210; Boulenger, Ann. & Mag. 

 Nat. Hist. (7) x. 1902, p. 421, and P. Z. S. 1903, ii. p. 328; Pellegrin, 

 Bull. Mus. Paris, 1905, n. 290. 



t A communication between these lakes appears to exist only at 

 certain seasons, according to Oscar Neumann (Geogr. Journ. xx. 1902, 

 p. 384). "The sources of the Sagan lie east of the south end of Lake 

 Abaya [Pngade]. But there is a broad channel connecting Lake Ganjule 

 [Margherita] with the Sagan. Tlie bed of this channel was dry at the 

 time, but tliere were some large and small watiTpools scattered over it. 

 When the water rises in L. Ganjule for about o inches, which will 

 probably tike place every year at the beginning of the rainy season, a 

 large river will run from Lake Ganjule to the Sagun." 



X Cf. maps in l?otte:.'o's ' L' Omo ' (Milan, 1899) and in O. Xeumann'd 

 paper in Geogr. Journ. xx. (19():j). 



