200 ME. G. A. BOULEXGER ON THE [Feb. 2, 



4. On the Type Specimen of a West African Fish, Clarias 

 Iceviceps Gill. By G. A. Boulenger, F.R.S., V.P.Z.S. 



[Received January 9, 1904.J 



The name Clarias Iceviceps was introduced by Gill in Proc. Ac. 

 Philad. 1862, p. 139, for a small fish in the Collection of the 

 Smithsonian Institution, bearing no indication of locaHty, but 

 regarded as probably from Liberia. The definition is given in the 

 following words : — "' Height at anus a tenth of length ; head 

 (latei'ally) a sixth ; its breadth an eighth ; the sui'face smooth ; 

 maxillary barbels twice as long as head. D. 86. A. 61." 



In a genus represented in West Africa by so many closely allied 

 species, so brief a definition could hardly answer its purpose, and 

 Clarias kevicej^s has proved a stumbling-block to workers at 

 African ichthyology. Sauvage recognised the species in a specimen 

 from the Gold Coast; Gunther suggested its identity with G. hyihi- 

 pogon of tSauvage ; and I myself proposed to regard it as a synonym 

 of C. sake, described by Hubi'echt from Liberian specimens. 

 But all this was mere guesswork, and lately, when describing 

 another close ally, under the name of C. ]xichynema, I felt it 

 imperative to settle the matter by an appeal to the type specimen 

 preserved in the U.S. National Museum. 



I am glad to say that my application was granted, in a most 

 courteous manner, by the Assistant Secretaiy of that Institution, 

 notwithstanding the rule that iinique type specimens are not to 

 be sent on loan. I must add that the exception made in my 

 favour is greatly due to the kind advocacy of Dr. Theo. Gill, the 

 original describei- of the specimen. I beg to tender my sincere 

 thanks to the authorities of the Smithsonian Institution and to 

 Dr. Gill for their great kindness. 



I hei'C give a description of the type of C. Iceviceps, followed by a 

 comparative description of three further specimens, viz. the types 

 of C. Mngsleyce Gthr., and the specimen from Assini, Gold Coast, 

 correctly identified by Sauvage. 



Clarias l.eviceps. 



Gill, Proc. Ac. Philad. 1862, p. 139. 



Depth of body 8| times in total length, length of head 5 times. 

 Head once and ^ as long as broad, smooth ; occipital process 

 pointed ; frontal fontanelle sole-shaped, not quite twice as long as 

 broad ; occipital fontanelle smallei-, in advance of occipital process ; 

 eye very small, 3 times in length of snout, 6 times in interorbital 

 width, which equals half length of head and a little exceeds width 

 of mouth ; band of prjcmaxillary teeth 5 times as long as broad ; 

 vomerine teeth conical, forming a curved band which is a little 

 nari'ower than the pi'wmaxillary band ; barbels rather thin, nasal 

 nearly half length of head, maxillary once and -S, outer mandibular 

 once, inner mandibular not quite |. Gill-rakers few, about 20 on 

 first arch. Clavicles concealed under the skin. Dorsal fin 87. its 



