III. COTTOIDS OF NORTH AMERICA. 9 



from the fresh waters of Europe. Following Artedi's method, he describes first the 

 fresh water and next the salt water species. Nevertheless, when writing the his- 

 tory of this genus, he did not find it necessary to separate generically these two 

 groups, although he had already pointed out their principal difference. Two 

 species only were known to him, one of them (C. minutus, Pall.), very imperfectly. 



On ascertaining the actual number of species to be so large, and each group 

 still maintaining constant generic differences, we concluded to give the name of 

 ACANTHOCOTTUS to the marine species, which, generally speaking, are of a larger 

 size than those living in fresh water. They are characterized by having many 

 spines to the opercular apparatus ; the preopercular itself having several of them 

 always strongly developed. The surface of the head, and often the circumference 

 of the orbits also, are either armed with spines, or else serrated or notched in different 

 ways. The nasal and turbinal bones are in most of the species surmounted by a spine 

 or ridge. The head itself is rather higher than broad ; sometimes much deformed, 

 with proportionally large eyes, and a deep occipital depression, as for instance, in 

 A. ventralis and claviger. The mouth is always more deeply cleft than in the fresh 

 water species (Coitus proper), but the dentition, as a whole, is nearly the same. 

 The nostrils open exteriorly by two apertures quite distant from each other, the 

 anterior tubular, and much larger than the posterior, which is even with the 

 surface of the head, and situated close to the orbits. The body is scaleless; the 

 back is often arched, and the fir.st dorsal fin almost as high as the second. Some 

 species have three, others four, soft rays to the ventral fins. The lateral line runs 

 uninterrupted and distinct from the head to the base of the caudal fin. 



The name of Cottus is retained for the fresh water species, the subject of the 

 present monograph, and as the generic characters will be given hereafter, it is 

 unnecessary to present them in this place. 



Not long since, the generic name of Uranidea was given to a species of our 

 genus Cottus, and had the name of Cottus belonged to the marine species of the group, 

 instead of being founded on the fresh water C. gobio, the new name Acanthocottus 

 would have been unnecessary. In that case we might have called the fresh water 

 species Uranidea, and the marine ones, Cottus, with equal propriety, although the 

 original idea of the genus Uranidea was a mistake of its author. 



We have formed another genus under the name of COTTOPSIS, to include one 

 single species, the Cottus asper, originally placed in the genus Cottus by Sir John 

 Richardson, afterwards in the genus Tradiidermis, by Mr. Heckel, and again by 

 Sir John Richardson, in that of Centridermichthys. The genera Trachidermis and 

 Centridermichthys have proved to be identical, 1 and in addition, to have no imme- 

 diate affinities with C. asper. The history of the genus Cottopsis will be found in 

 Chapter IV. Finally, we have established the genus TmGLOPSis 3 for an entirely 

 new fish discovered by Prof. Baird in Lake Ontario. It comprehends as yet, only 

 one species, the description of which will be found in Chapter V. 



The manner in which the characters are shared by the genera of Cottoids, either 



1 Proc. Bost. Soc. Nat. Hist., iii., 1850, p. 302. 2 Proc. Boat. Soc. Nat. Hist., iv., 1851, p. 19. 



2 



