III. COTTOIDS OF NORTH AMERICA. 53 



Baird, about Carlisle (Pa.), in Mountain and Yellow Breeches Creeks, and Letart 

 Spring, the largest of which were three and six-eighths of an inch. They all 

 agree with the above description, excepting the color of an individual from Mountain 

 Creek, which we found of a dark and uniform black. 



We owe to the kindness of Prof. S. S. Haldeman, an authentical specimen, three 

 inches long, so that there can be no doubts left with regard to the species which is 

 here described. 



The specimen figured was caught in the vicinity of Carlisle, and is preserved at 

 the Smithsonian Institution, together with several others of the same vicinities. 

 Specimens were also obtained from the Schuylkill at Reading (Pa.), and from the 

 tributaries of the Potomac, at Rohrersville (Md.), and in Rock Creek, Washing- 

 ton (D. C). So that the range of G. viscosus is Eastern Pennsylvania and Mary- 

 land. 



IX. COTTUS FRA]\KLII\II, Aoass. 



Plate II. Figs. 5 and G. 



Syn. Col/its FranMinii, Agass. Lake Sup. 1850, p. 303. — Girard, Proc. Amer. Assoc. Adv. Sc. XL, 1850, 

 p. 411; and, Proc. Bost. Soc. Nat. Hist. III., 1850, p. 189. 



Here is a species which we might easily have identified with the preceding, so 

 much it resembles it by its general appearance, had we not looked into the ana- 

 tomical as well as zoological peculiarities of both species. 



Before we were prepared to formulate distinctively the characters by which this 

 species differs from its congeners, the form of the skull had already satisfied us that 

 it was distinct. In the first place, and to speak only here of this species and of C 

 viscosus, which appear so much alike, the conformation of the skull has something 

 so peculiar that, when once well understood, it will be easy to tell at first sight to 

 which of these two species such and such skulls may belong (Compare Figs. 8 and 

 12 of Plate III.). 



If the differences exhibited in those profiles are not specific, comparative osteo- 

 logy can no longer be a sure guide in the study of species, nor can anatomy be 

 of any help to zoology. But to this conclusion we have not yet arrived; we 

 know what compai'ative osteology of the skull is worth, and, confident in the future 

 of that science, we should have established the two species as distinct on those 

 characters alone. 



Zoologically speaking, the general form is short and stout. The greatest depth 

 is contained five times and a half in the total length, and is proportionally greater 

 than in G. viscosus. The least depth is one-nineteenth of the length. The body 

 tapers rapidly away, as in G. meridionalis and G. Alcordil. The peduncle of the 

 tail is more slender, and the back more arched than in G. viscosus. The thickness 

 is greater than the depth for a considerable length; towards the tail, however, the 

 depth becomes greater. The body, as a whole, has rather a cylindrico-conical 

 shape. 



The head, proportionally shorter than that of G. viscosus, is contained two times 

 and a half in the lemrth of the bodv, the caudal fin excluded. The snout, also, is 



