220 



Dll. A. GtlfTHEE ON A COILEOTIOK OF 



[Feb. 4, 



proved by the series of C. gariepinus in the British Museum. 

 Peters (Eeise n. Mossambique) was of a different opinion and has 

 attempted to prove the variability of this character, but, in my 

 view, he has confounded two or even more species under the name 

 of 0. mosmmbieus. 



Two small specimens of C'larlas collected on the Shebeli E. 

 (1 Febr., 1895) are not in sufficiently good condition to be deter- 

 mined. The form of the vomerine band is very different fi-om 

 that of the fish described as C. smithii. A collection of a large 

 series of specimens of all ages of any species of Clarias from the 

 same locality is very much needed ; but until this is done, it seems 

 to be safer to utilize all characters observable in apparently mature 

 or nearly mature specimens. 



. 6. EuTEOPiTJS dephessiuostiiis, Ptrs. — Shebeli R. 



7. Stnodontis geledbnsis, sp. n. 

 Allied to S. serratus ' . 



General form of the body somewhat elongate; snout rather pro- 

 duced, subconical ; diameter of the orbit two sevenths of the length 



Rg 2. 



of the snout. The gill-opening extends downward to before the 

 root of the pectoral fin. Mandibulary teeth in moderate number, 



' I must again draw attention to an unfortunate clerical error in Cat. Fisli. 

 T. p. 212, where the line " B. Mandibulary teeth not longer than the eye," 

 oue;ht to have been placed above " Synodontis serratus." 



