Vol. XXII, pp. 205-208 December 31 , 1909 



PROCEEDINGS 



OF THE 



BIOLOGICAL SOCIETY OF WASHINGTON 





THE SCALES OF THE COBITID AND HOMALOPTERID 



FISHES. 



BY T. D. A. COCKERELL. 



According to Boulenger, the Cobitids and Homalopterids form 

 subfamilies (Cobitidinte and Homalopterinas) of the Cyprinidse. 

 Gill, as far back as 1861, proposed to treat both groups as dis- 

 tinct families, Cobitidte and Homalopteridse, a course in which 

 he has been followed by several writers. Having regard for all 

 the characters, it seems that we may recognize a distinct family 

 Cobitidffi, but the Homalopterids may be regarded as very aber- 

 rant Cobitids. The arrangement will then be as follows (using 

 the characters cited by Gill and Boulenger): 



CoBiTiD.^. Maxillaries not bordering the mouth, barbels three 

 to six pairs ; pharyngeal teeth in one row ; air-bladder 

 with a long capsule; scale small or absent. 



(1) Cohiiinse. Air-bladder variably modified or reduced, but 



always distinct. 



(2) Homalopterinse. Air-bladder rudimentary. Curiously flat- 



tened fishes, adapted to life in mountain streams, 

 where they are able to adhere to the rocks in a strong 

 current. 



Gyrinochilus Vaillant, regarded by Gill as the type of a family 

 Gyrinochilidse, is certainly very different from these, and has 

 large scales. I regret that I have never seen a specimen. 



The scales of the Cobitidaj, as here defined, appear to be of a 

 sufficiently uniform type. In the genus Misgurnus of the Cobit- 

 inse they are relatively large and well developed, and have an 

 exceedingly beautiful pattern. They are fairly large, however, 

 in Homaloptera. 



33— PRoc. Biol. Soc. Wash., Vol. XXII. 1909. (205) 



