in. COTTOIDS OF NORTH AMERICA. 35 



ought at least to find both sexes in the same hydrographical basin. But, along 

 the southern and eastern shores of Lake Superior, we find the G. FranJdinii, which 

 inhabits the same basin as G. Richardsonii, and, even if both of these species were 

 provided with the same number of rays to the ventrals, nobody would ever think 

 of uniting them in one, so much do they differ in other respects. 



If we take up the species with three soft rays only to the ventrals, we witness 

 similar phenomena. Without speaking of G. Fahricii, which we have not had 

 under actual examination, we find, in the same latitude, G. gohioides in the waters 

 running west of the Green Mountains, and G. holeoides east of the same orographic 

 range. Should the streams in which they live have a direct communication, a 

 zoologist could not reasonably identify them. Finally, G. viscosus which inhabits 

 eastern Pennsylvania, and G. gracilis, Connecticut, New York and Massachusetts, 

 are widely distinct. The two species which resemble each other most, G. viscosus 

 and G. Franl-Hnii, are geographically the most remote. 



It is not improbable that some one may hereafter propose to unite in a separate 

 genus the species provided with four soft rays to the ventrals. Our impression, 

 however, is, that such a generic subdivision would be useless, inasmuch as it would 

 interrupt the philosophic idea to which we have been led by our investigations. 

 Indeed, a genus, in our mind, is a group varying, it is true, as to the number of 

 species which it may contain, but having, at the same time, a physical and a 

 metaphysical signification. A genus involves a progressive idea whose realization 

 is materially canned out in the species. Now we are at loss to find what progress 

 is involved in the fact that some species have one ray more or less to the A^entral 

 fins. These two facts are cotemporaneous, and their value is entirely in the dis- 

 crimination of the species, and, indeed, in this respect they have an actual signifi- 

 cation in the manner in which they are distributed among them. 



The same peculiarity is observed amongst Acanthocotti, and those also would 

 have to be likewise subdivided. If the characters of three or four soft ventral rays 

 were of a generic value, either the species with three or those with four should 

 have appeared first in geological times. 



The color in Cotti has not yet afforded any safe distinctive mark between the 

 various species. The ground is generally brownish-yellow, sometimes blackish- 

 brown, maculated and dotted with a deeper black or brown. The upper edge of 

 the anterior dorsal in G. viscosus is orange, Avhilst in G. gracilis it is red. Whether 

 that liue is specific needs still to be investigated, as well as its ^^articular appear- 

 ance in the other species of the genus. 



The following synoptical table will exhibit the most prominent differences 

 between the species of the genus Cottus : 



