120 MEMOIRS OF THE CARNEGIE MUSEUM. 



noquenessing), Cowanshannock Creek (to middle Allegheny), and Potato Creek 

 (flowing into the uppermost Allegheny), is it outside of the Terminal Moraine, 

 but never far from it. The explanation of this peculiar fact will be attempted 

 elsewhere; here it suffices to direct attention to it, in order that the case may be 

 kept in mind. 



As regards its ecology, L. viridis is in Pennsylvania distinctly a form of the 

 smaller streams, goi^g up far into the headwaters, into streams which are indeed 

 the smallest in Pennsylvania, which contain mussels. Also Wilson & Clark 

 (1912a) have observed that it belongs to the small creeks in the upper Kankakee- 

 drainage in Indiana. Wherever I found this species, it seems to be averse to very 

 strongly flowing waters (in riffles) ; its usual station is near (below) riffles, where 

 there is a moderate flow of water over coarser or finer gravel, packed firmly by 

 fine sand or mud. It is generally not an abundant species, but locally it has been 

 taken in some numbers. In the beach-pools of Lake Erie I found it in the char- 

 acteristic sand of the lake-shores, with a scanty vegetation of algae. 



General distribution: Type localily, rare in the Ohio, more common in the 

 Kentucky and small tributaries (Rafinesque). 



Simpson (1900) condenses the range as follows: "Ohio and St. Lawrence 

 drainages areas; west to Arkansas, north through Nebraska to Wisconsin; Hudson 

 River." This, however, does not bring out the chief feature of the distribution. 

 This species is distinctly more northern, and its main range is north of the Ohio in 

 Iflinois, Indian^, Michigan, Ohio, and Pennsylvania; it crosses over into the lake- 

 drainage (Walker, 1913), and has in the St. Lawrence system a great extension 

 eastwards, crossing over in New York into the Atlantic-drainage (Mohawk and 

 Hudson Rivers). In Canada, it goes as far as the vicinity of Ottawa and Montreal 

 (Marshall, 1895), and extends to Vermont (CaU, 1878, and Gray, 1883). It also 

 has been reported from Connecticut." 



In this eastern extension of its range it invades in part that of the next species 

 (L. subviridis), but it is not known whether it is found in any one locality associated 

 with the latter. Probably it went here (as did Fusconaia flava) by way of the 

 Erie Canal, although CaU suggests the possibility of distribution by birds. 



In the southward direction, in southwestern Pennsylvania, as indicated above, 

 and generally to the South of the Ohio River, it is missing or rare. I have never 

 found it in the Ohio proper from Pittsburg down to Cincinnati, although it is 

 mentioned in the Cincinnati list (Harper, 1896). I have found a single individual 



" New Haven Canal, according to Linsley (1S4.5), but not mentioned by Perkins (1869). Adams 

 (1842) has distinguished this eastern form as var. plcbcia (See Simpson, 1014, p. 483, and Johnson, 191.5, 

 p. 25). 



