III. COTTOIDS OF NORTH AMERICA. 59 



backwards, their tips do not reach the posterior margin of the first dorsal, and 

 consequently not the anterior margin of the anal. 



Br. 6. D viii. 16. A 11. C 2. 1. 5. 4. 1. 1. V I. 3. P 12. 



The lateral line runs straight, and near the back. It passes into the series of 

 minute pores before it reaches the end of the second dorsal, and its fall on the 

 peduncle of the tail is indicated by a slight undulation. The cephalic channels 

 appear to be very much developed, perhaps more so than in any other species of 

 the genus, if we judge of them by the wide duct which passes through the pre- 

 opercular. 



The vent is nearer the base of the caudal than the end of the snout. 



Nothing can be said about the color, on account of the circumstances under 

 which the specimen was found. 



The species inhabits Lake Ontario, in the vicinity of Oswego (N. Y.), in company 

 with Triylopsis Thompsonii, constituting a part of the food of Lota maculosa. 



The only specimen hitherto known of this species was collected by Prof. Baird, 

 and is now preserved at the Smithsonian Institution. 



XIII. COTTUS FABRICII, GIRARD. - 



Syn. Cottus Fabricii, GIRARD, Proc. Amer. Assoc. Adv. Sc. II., 1850, p. 411; and, Proc. Bost. Soe. Nat. 

 Hist. III., 1850, p. 189. 



Cottus (/olio, FABR. Faun. Groonl. 1780, p. 159. 

 Cottus tricuspis (Mus. Reg.). GRAAH, Reise Ostk. Grb'nl. 1832, p. 194. 



In the Fauna of Greenland, we find one species of true Cottus, identified by 0. 

 Fabricius with G. yobio of the authors of Northern Europe. But there would be 

 a very remarkable peculiarity in the habits of this species, if it be true that it lives 

 in company with Acanthocottiis scorpiiis and scorpioides, and consequently in salt 

 water. We are not prepared to tell how far fresh water Cotti may be accommodated 

 in a saline medium, but it seems to us much more probable that the Greenland 

 Acanthocotti themselves, like those of New England, leave the sea at a certain 

 season, ascend the fresh water inlets, and then may be found in company with the 

 Greenland Miller's Thumb. On the other hand, if we take into account the reduced 

 saltness of the Arctic Sea, we can easily conceive how a fresh water fish may occur 

 at the entrance of streams or rivulets emptying into the bays. If in the Tem- 

 perate Zone, the Miller's Thumbs keep far away from the sea-shore, in the Arctic 

 Zone, they may reach the sea without extending into its depth. 



At any rate we have here a true Cottus, since Fabricius, who could not be mis- 

 taken respecting G. gobio, identifies it with the latter. Besides, his diagnosis leaves 

 no doubts on the subject, " smooth with two spines on the head." Now this may 

 be said of all the species of the genus without exception. Then, without giving its 

 size, he tells us that it is much smaller than the Cottus (Acanthocottus) scorpius and 

 scorpioides. Although very similar to G. gobio, we consider it as a distinct species. 

 And if Fabricius be correct in the formula of the fins, we notice a fact which 

 excludes at once the idea of identity with either C. cognatus or C. Eichardsonii ; 



