2. CTEN'OPOMA. 'A7'A 



1. SPIROBRANCHUS. 



Spirobranchus, Ciiv. S,- Val vii. p. 392. 



Body compressed, oblong ; none of the bones of the head siTrated ; 

 cleft of the mouth extending below the orbit. Small teeth in tlie 

 jaws, on the vomer and the palatine bones. Dorsal spines twelve 

 or thirteen, anal spines seven. Yentral fins well developed. None 

 of the fins elongate. Lateral line interi'ui)ted. Air-bladder none. 

 Pyloric appendages two ; intestines of moderate length. 



Fresh waters of the Cape of Good Hope. 



1. Spirobranchus capensis. 

 Cuv. 8f Val. vii. p. 392. pi. 200; Cuv. Rii/itc Anim. 111. Poiss. pi. 75. fig. 1. 

 B. 4. D. '-g?. A. |. L. hit. 30. L. transv. 10. 



BroAvnish, with three oblique blackish streaks running from the 

 orbit towards the praeoperculum. 

 Cape of Good Hope. 



a. Half-groAvn. 



h-c. Adult: stuffed. Cape of Good Hope. Prcscntedby Sir A. Smitli. 



2. CTENOPOMA. 

 Ctenopoma, Peters in Miill. Arch. 1840, p. 481. 

 Body compressed, oblong ; operculum, inter- and sub-operculum 

 serrated ; cleft of the mouth extending to below the orbit. Small 

 teeth in the jaws, on the vomer and the palatine bones. Dorsal 

 spines seventeen, anal spines ten. Ventral fins composed of one spine 

 and five rays. Lateral line interrupted. Air-bladder simple. Pyloric 

 appendage one. 



Fresh waters of Quellimane. 



1. Ctenopoma multispine. 



Ctenopoma multispinis, Peters, I. c. iiib. 10. fig. lO-lo. 

 ? Anabas scandens, Bimtconi, Spec. Zool. Mas. p. 270. 

 Spirobranchus smithii, Bianconi, I. c. p. 279. tab. 10. 



B. 6. D. ^. A. J. V. 1/5. Caic. pylor. I. 



Greenish, spotted with darker. 

 Pools near Quellimane. 



o. Fine specnroen. From Professor Peters's Collection. 



Bianconi states that there are only six soft rays in the anal fin, 

 which is probably an error. We can scarcely bcheve that Annbas 

 scandens occurs in Mozam])ique : Bianconi's specimens having been 

 purchased, some Indian fishes may have been mixed with them ; or, 

 which is more probable, the same specimens which Bianconi formerly 

 called A. scandens, he may afterwards have determined as Sj^iro- 

 hranchns smWiii, — the last error with regard to generic denomina- 

 tion being greater than the first. T may here also mention that 

 Megalepis Alessandrini. of the same author is Upeneoides vitt((ta.'i, in 

 Avhich the brtrbcls have been overlooked ! 



