Vol. XXII, pp. 121-124 June 25, 1909 



PROCEEDINGS 



OF THE 



BIOLOGICAL SOCIETY OF WASHINGTON 



NOTES ON THE SCALES OF FISHES. THE HERBI- 

 VOROUS CYPRINID^. 



BY T. D. A. COCKERELL AND OTIS CALLAWAY. 



It has been remarked by .lordan and Evermann (Fishes of 

 North and INIiddle America, p. 202) that the subfamily Chondros- 

 tomime, as recognized in America, is extremely heterogeneous . 

 A study of the scales of most of the genera clearly indicates that 

 the group is even less uniform than has been supposed, and 

 makes it practically certain that it should be divided into three 

 sul)families, which are not especiallj' related. Chrosovms stands 

 entirely apart in the structure of its scales, which resemble in 

 sculpture those of Catostomidse. The Catostomids are evidently 

 an ancient, and in some respects, primitive group, which prol)a- 

 ])l3' inliabited North America before there were any true Cyprini- 

 da^ on tliis continent. Chrosomus may not l)e supposed to have 

 been derived from the Catostomids, l)ut must apparently be 

 regarded as an ancient offshoot from the stem which gave rise 

 to Cyprinida? and Catostoniidje alike. 



Contrasting with Chrosomus in nearly every way, but related 

 to it in the small scales and the traces of basal radii, is the 

 Chondrostomine group proper, including fishes of the Pacific 

 coast region, closely related to those of the Old World.* These 

 must be regarded as comparatively recent (iNIiocene ?) immi- 

 grants, which liave never crossed the Rocky Mountains. The 

 third series typified by Pimephales, consists of large-scaled fishes 

 with no traces of basal radii. They are entirely difi'erent from 

 the two small-scaled groups, and must have had a quite distinct 



* Basal radii are common in Old World Cyprinidse, but exist in very few North 

 America genera. Rhinichlhyg and Agosia have them, but they are absent in the great 

 mass of the carnivorous forms. 



18— Proc. Bioi.. Soc. Wash., Vol. XXII, lyO'J. (121) 



