292 ORTMANN—CORRELATION OF SHAPE AND 
The true metanevra is restricted mainly to larger rivers. But 
occasionally it goes into smaller rivers, but hardly ever into the 
headwaters. It has been reported repeatedly from rather small 
streams, but then it was generally represented by the var. ward. 
The original localities of the latter are Walhonding River, Ohio 
(tributary to Tuscarawas); Wapsipinicon River (Wassepinicon), 
Iowa; and Coal River, Logan Co., Va. (Coal River is now in 
Kanawha and Boone Cos., W. Va.). Three specimens from the 
latter locality were in the Hartman collection, and are now in the 
Carnegie Museum, 
Sterki'® reports wardi from Sugar Creek, another tributary of 
Tuscarawas River. 
Aside from the types of wardi from Coal River, the Carnegie 
Museum possesses this form from the Ohio, Monongahela and Alle- 
gheny Rivers in the vicinity of Pittsburgh. But these specimens 
are not very typical, are rare, and intergrade with the normal 
metanevra. They are found, however, at and near the upstream 
limit of the distribution of the species. A rather good specimen of 
qwardi comes from the Little Kanawha River at Burnsville. This 
again is a smaller stream. 
Thus the tendency is observed, when the species enters smaller 
streams, to develop a compressed variety. But, in addition, this 
variety inclines to obliterate a peculiar sculpture, very character- 
istic for the main species, that of large knobs upon the surface of 
the shell. This should be kept in mind. 
QUADRULA CYLINDRICA (Say). Simpson, ’14, p. 832. 
This widely distributed species ordinarily is strongly nodulous, 
with great knobs upon the posterior ridge, and the shell is generally 
much swollen. Its characters are rather uniform over the range. 
However in the headwaters of the Ohio River, in Ohio and 
Pennsylvania, and in the upper Tennessee-region, two peculiar small- 
stream-races have developed. In the Tuscarawas River, in Beaver 
River, and French Creek, there is a remarkably compressed form, 
which, in addition, is practically smooth, having lost not only the 
‘6 Proc. Ohio Acad. Sci., 4, 1907. 
