286 SILURIDiE. 



the head, its distance from the caudal fin measuring one to two diameters of eye. 

 Anal fin with 50 to 62 rays, separated from the caudal by a distinct interspace. 

 Pectoral fin two-fifths to one-half the length of the head, the spine serrated on the 

 outer border and three-fifths to three-fourths the length of the fin. Ventral fin 

 shorter, equally distant from the end of the snout and from the root of the caudal fin, or 

 a little nearer the former. Caudal fin rounded, two-fifths to half the length of the head. 



Olive to blackish above, white beneath ; a more or less distinct dark band on each 

 side of the lower surface of the head. 



The largest specimen examined by me measures 650 millimetres. 



This species has a wide distribution in East Africa, being known from Abyssinia and 

 Lake Victoria to German East Africa and Lake Tanganyika. The type specimens 

 described by Vinciguerra came from the Webi Shebeli. 



List of specimens examined : — 



2 Lake Victoria.— Sir H. Johnston, 4.12.04. 



1 Lake Victoria. — Col. Define Radcliffe, 1903. 



3 Entebbe.— Degen, 31.8.05-5.9.05. 



1 Unfras E., L. Tsana.— Degen, 9.5.02. 



2 Warrar, Hawash R.— Degen, 6.2.02. 



1 Hawash R. — O. Neumann, 8.01. 



2 Marmasa, Grota R. (Hawash system), 4-5000 feet. — Degen, 21.1.02. 

 2 Elgo R., L. Abaia, 3000 feet.— Zaphiro, 3.9.05. 



1 Webi Shebeli.— Dr. Donaldson Smith, 1894. (Type of C. smithii.) 



1 Voi R., Taita.— C. S. Betton, 1897. 



1 Ngatana, Tana R.— Prof. Gregory, 1893. 



2 Bubu R., Urangi. — 0. Neumann, 7.03. 



3 Kandoa, Usagara. — Capt. Bloyet, 1882. 

 5 Zanzibar. — Sir L. Playfair, 1865-67. 



2 Zanzibar. — Sir J. Kirk, 1867. 



1 Lake Tanganyika. — Moore, 1896. 



1 Niamkolo, L. Tanganyika. — Dr. Cunnington, 1904. 



Clarices robecchii is intermediate between C. anguillaris and C. lazera, with both of 

 which it has been confounded by me and others, but on the whole nearer to C. lazera, 

 from which some specimens are difficult to distinguish. It has also been confounded 

 with C. gariepinus, Burchell, in which the vomerine teeth are nearly all conical. 



G. vinciguerrce, Blgr., based on a single specimen from the Kassam River, Shoa 

 (altitude about 4000 ieet), is very closely related to C. robecchii and may prove to be an 

 abnormal individual of that species. It has been separated on the ground of its 

 having a longer head, a little more than once and two-thirds as long as broad, and 

 eight times as long as its distance from the dorsal fin, and a very high number (90) of 

 gill-rakers, considering the size (410 millimetres) of the specimen. 



