18 MEMOIRS OF THE CARNEGIE MUSEUM. 



Clarion Rivers the fauna is entirely destroyed, and the few survivals we have in 

 the headwaters of the Kiskiminetas do not include this species. 



F. flava prefers fine gravel and sand, and avoids rough bottom and rocks. Its 

 favorite stations are on bars of fine, firmly packed gravel, just below riffles. 



General Distribution: Type locality, Small tributaries of Kentucky, Salt River, 

 and Green River (Rafinesque). 



Outside of Pennsylvania this form has a wide range. It has been reported 

 from western New York (see below) westward as far as Kansas and southeastern 

 Nebraska. Northward it passes into Canada and the drainage of the Red River 

 of the North (Winnipeg, compare also our specimens from North Dakota). In 

 Michigan, it is all over the southern half of the lower peninsula (Walker's map, 

 1898, PI. 1). It probably reached the lake-drainage by several ways, but not 

 through Pennsylvania, since it is missing in the Beaver and upper Allegheny basins. 



From the Ohio River southward its range becomes obscure. As I have dis- 

 covered, it is i)resent in the Little and Big Kanawha-drainages in West Virginia, 

 and also in the Big Sandy and Licking Rivers in Kentucky. The type locality is 

 in central Kentucky. Records from Tennessee are missing, except that given by 

 Wilson & Clark (1914) Stones River, Tennessee, tributary to the Cumberland. 

 The localities quoted from Mississippi, Alabama, and Texas are more than doubtful. 

 But, as our material shows, forms representing the species are certainly found as 

 far west as Arkansas and Oklahoma, and although certain authors might call, 

 and have called, these by different names, I am unable to distinguish them from 

 the northern form, except by their more shining epidermis. (See Wheeler, 1918, 

 p. 123.) 



It would be interesting to know whether outside of Pennsylvania the rule 

 likewise holds good, that this form prefers smaller creeks. This is certainly the 

 case, wherever I have collected it in West Virginia and Kentucky.^* The localities 

 in Arkansas, and partly also in Oklahoma, are small creeks, while in the larger 

 rivers (Ouachita) F. flava undata and trigona are found. Call (1900, p. 506) says 

 that ruhiginosus {flava) is found in Indiana, in streams both large and small, while 

 Wilson & Clark (1912a, pp. 42, 43) report /am from the headwaters of the Kankakee 

 system in Indiana, while they cite trigona from the lower Kankakee and Iroquois 

 Rivers in Illinois. In the uppermost Wabash and in the Maumee C. Goodrich 

 collected only typical flava, and here undoubtedly is one of the places, where it 



" However, in Elk River in West Virginia this is not perfectly clear. Here is found a form which 

 may be called a dwarfed F. flava trigona. But this is in keeping with the general character of the Elk 

 River fauna, w'hich should be designated as a dwarfed big-river-fauna. It is not the place here to give 

 details of these remarkable conditions. 



