334 MEMOIRS OF THE CARNEGIE MUSEUM. 



(only three specimens taken), and in the Allegheny River from Armstrong County 

 up to Warren County. It is also found in French and Connewango Creeks, 

 tributaries of the Allegheny. In the Allegheny itself it is found rather regularly, 

 but not in great numbers; it is most abundant in French Creek and at Cochranton 

 it was a common species. 



In the Monongahela-drainage I never found it alive, but its former occurrence 

 there is indicated by shells from the Indian garbage heap at Point Marion. It 

 also occurs farther up in the headwaters in West Virginia, but is very rare. 



It is remarkable that this species has not turned up in the Ohio below Pitts- 

 burgh, and that there are no localities for it positively ascertained in the Ohio, 

 until we reach Cincinnati. 



I alwaj^s found this species on riffles, perferably upon a bottom composed of 

 firmly packed and rather fine gravel, in swiftly flowing, shallow water. In the 

 Allegheny in Armstrong Countj^ I have also seen it in coarse gravel. 



General distribution: Type locality, Ohio River, Cincinnati (Lea). 



Aside from western Pennsj'lvania and the upper Monongahela in West Vir- 

 ginia, this species is known from the state of Ohio. In the Ohio proper only the 

 type locality is known; but according to Sterki (1907«) it is found in the Mahoning, 

 Tuscarawas, and Scioto Rivers. In Indiana it is found in the Ohio and Wabash 

 (Call, 1896a), and in White River (Carnegie Museum) and in Illinois only in the 

 southern portion (Wabash, see Baker, 1906). In addition it occurs in south- 

 eastern Michigan, in the Lake Erie-drainage (Detroit and Raisin Rivers, see Walker, 

 1892 & 1898), and it is also given (Walker, 1913) in the Lake Erie list, but has 

 never been found on the Pennsylvanian shores of the lake. 



According to these records, this species is rather restricted in its range, and 

 belongs to the Ohio River and its tributaries, chiefly the northern ones, and seems 

 to have its metropolis in the smaller tributaries of the upper Ohio River in the 

 states of Ohio and Pennsylvania. It also has crossed over into the lake-drainage, 

 but particulars are lacking. 



An allied species is T. torulosa (= perplexa), which seems to inhabit the 

 lower Ohio and the Tennessee drainage. It is interesting to note that the latter 

 develops in the upper Tennessee a compressed form without tubercles, which should 

 be called gubernaculum Reeve, and which is a form parallel to rangiana, but has 

 the dark green postbasal expansion of torulosa. This case should be kept in mind, 

 for it is an interesting instance of the reactions of the Naiades to environmental 

 conditions. The form most closely related to T. rangiana undoubtedly is T. 

 sampsoni (Lea) of the lower Wabash, which differs only in having the sheU more 



