IN THE ALLEGHENY REGION OF SOUTH-WESTERN VIRGINIA. 200 



The following pages furnish the results of an examination into the fish-faunte of 

 the head waters of four rivers, with reference to the above propositions. These rivers 

 are the James, the Roanoke, the Kanawha, and the Tennessee. The first two rise 

 in Montgomery and Giles counties, Virginia; the James, by its important tributary, 

 Craig's Creek. Both of these approach usually on opposite side of mountain range.s, 

 the tributaries of the Kanawha; the Roanoke to Strouble's and Tom's Creeks, and 

 the James to Sinking and Stony Creeks. Those that head in the same valley 

 descend to ihe north-east and south-west, being separated by well-marked elevated 

 tracts, in some cases by exceedingly high ones. None of the tributaries examined 

 had their rise in a common swamp. 



The Tennessee rises by ijs longest tributary, the north fork of the Holston, in 

 Bland county, flowing from the most elevated portion of the Rich valley on the north 

 side of Walker's (or Gap) mountain ; while a strong tributary of the Kanawha, 

 Walker's Creek, descends from the same point to the north-east. A dry, elevated 

 portion of the valley separates their sources. A straight line connecting the sources 

 of Craig's Creek and the north fork of tiie Holston would not exceed fifty-five miles 

 in length. 



These streams were fished many times by the writer during a residence of nearly 

 four months in that region. It was done with a fine seine, and for some miles at a 

 time, and it is believed the results form a tolerably close approximation to the true 

 character of the fish fauna of the streams in question. The means employed 

 certainly furnished as reliable data for the relative abundance of species as could 

 well be, since at each draw of the net the number taken was large, often reaching one 

 to two hundred specimens. When the same results wei^e thus expressed over some 

 extent of space and time, and in every variety of situation, they might be accepted as 

 reliable for the stream in general. The fishes of the Roanoke were taken in the 

 seventh montli, those of the James and Kanawha in the eighth and ninth, and those 

 of the Holston* in the tenth, 1867. 



The observation of fishes in their native haunts brings much of beauty, as well as 

 of interest, to the eye of the naturalist In the spring of the yeai-, after the 

 Catostomi of the head of the Roanoke had deposited their eggs beneath their 

 accustomed piles of stones, the smaller species of Cyprinidaa gathered themselves to 

 try and rob the nests, so fir as lay in their power. Refulgent shoals of Clinostomus 

 affinis and Hi/psilepis ardens would lie close Jto the heaps down the stream, and when 

 startled b}^ the approach of a stranger, would disappear like the passage of a stream 



* I must here record the obligations under which I have been placed by the kindness of ray friend William 

 Alex. Stuart, of Saltville, Va., one of the proprietors of the salt wells and furnaces at that place, Through 

 his aid I was enabled to procure a remarkably good series of the fishes of the north fork of the Ilolslon. 



