246 ON THE DISTRIBUTION OF FRESH-WATER FISHES 



It is known that the Rocky Mountains, elevated at the commencement of Ctenozoic 

 time, divide fish faunae which are in many respects very different, though numerous 

 genera exist on both sides the range. 



It would appear, from the previous considerations, that the distribution of fresh- 

 water fishes is governed by laws similar to those controlling terrestrial vertebrates, 

 and other animals, in spite of the seemingly confined nature of their habitat. With 

 this general principle in view, we may revert briefly to this distribution over the 

 eastern district of the nearctic region. 



That this is a single though large zoological district, is indicated by the distribu- 

 tion of several species over its entire area, or nearly so, so far as yet examined : those 

 oi Semotilus, Geratichthys, HijpsiJepis, Catostomics, etc., or by the universal recurrence 

 of the same in suitable situations ; and by the representation of these and other 

 genera, by nearly allied species, in its difTerent portions. The fauna of the tributa- 

 ries of the Mississippi constitutes, it might* be said, that of our district; while the 

 slight variations presented by the Atlantic coast streams might be regarded as 

 exceptional. The fauna of the great lakes combines the peculiarities of both,, 

 possessing as a special peculiarity (I) which belongs to the lake region, which, in the 

 district, commences at latitude 42°, and extends to the arctic regions, the range of 

 the genus Goregonu-s. 



The peculiarity of the Atlantic subdistrict (II) may be said to be the abundance of 

 Esox, Sabno and Anguilla, and the absence of Haploidonotihs. The first two are 

 abundant in the lake region, while Anguilla has but a partial distribution there, and 

 Haploidonotus does not occur. In (III) the Mississippi basin Esox is represented by 

 but few species, and remarkably few individuals. Salmo occurs abundantly in the 

 upper parts of the Missouri tributaries, exists in the western mountain streams of the 

 AUeghenies, becoming rare in those of the Kanawha, and only occurring near the 

 highest summits in those of the Tennessee, south to the line of North Carolina. The 

 gradation from the Mississippi grouping of species to the Atlantic is very gradual, and 

 takes place in successional order from those emptying into the Gulf of Mexico to- 

 wards the east and north-east, until we reach the rivers of Massachusetts and Maine, 

 where the greatest modification of the fauna exists. The latter fact has been 

 pointed out by Agassiz, who calls this region a " zoological island," and enumerates* 

 the characteristic nearctic genera which are wanting there. I give now a list, show- 

 ing the points at which Mississippi genera cease as we follow the rivers of the Gulf 

 and AtUmtic coasts, so far as our present knowledge extends : 



Gulf rivers. Haploidonotibs has not yet been indicated from eastward of these. 

 Roanoke. Gampostoma ceases here. 



* Amer.'Joarn. Sci. ArtS; xvii, 363, 1854. 



