228 MEMOIRS OF THE CARNEGIE MUSEUM. 



Monongahela River, Charleroi, Washington Co. (G. A. Ehrmann)."^ 



Other localities represented in the Carnegie Museum: 



Lake-drainage: 

 St. Joseph River, Fort Wayne, Allen Co., Indiana (C. Goodrich)- 



Ohio-drainage: 

 Tuscarawas River, Ohio (Holland collection). 



West Fork River, Lynch Mines, Harrison Co.; Lightburn and Weston, Lewis Co., West Virginia. 

 Little Kanawha River, Burnsville, Braxton Co., West Virginia. 

 North Fork Hughes River, Cornwallis, Ritchie Co., West Virginia. 

 Elk River, Sutton, Braxton Co., West Virginia. 

 Pocatalico River, Raymond City, Putnam Co., West Virginia. 

 Coal River, Sproul, Kanawha Co., West Virginia. 

 Mud River, Milton, Cabell Co., West Virginia. 

 Levisa Fork Big Sandy River, Prestonsburg, Floyd Co., Kentucky. 

 Licking River, Farmer, Rowan Co., Kentucky. 



Tennessee-drainage : 

 Bear Creek, Burleson, Franklin Co., Alabama (H. H. Smith). 

 Elk River, Estill Springs, Franklin Co., Tennessee (H. H. Smith). 

 Hurricane Creek, Gurley, Madison Co., Alabama (G. H. Clapp, donor). 

 Flint River, Maysville, Madison Co., Alabama (H. H. Smith). 



Paint Rock River, Paint Rock, Holly Tree, Trenton, and Princeton, Jackson Co., Alabama (H. H. Smith). 

 South Chickamauga Creek, Ringgold, Catoosa Co., Georgia. 

 Holston River, Holston Station, Grainger Co., Tennessee. 



Distribution and Ecology (See fig. 23) : Type locality, Kentucky River (Rafin- 

 esque) . 



As has been mentioned above, it is hard to trace the distribution of this form 

 in its relation to 0. subrotunda. Simpson (1900) gives only: "Ohio-drainage 

 and southern Michigan." 



According to my experience this is the small-stream-form of 0. subrotunda, 

 and wherever the latter tends to ascend into the tributaries of the larger rivers, 

 it turns into levigata. This is true in the Beaver-drainage, in the Allegheny and 

 Monongahela in Pennsylvania. But this form is not found in all the tributaries. 

 Ui)on the whole it is one of the rarer shells in our state. Rather compressed 

 shells are found down the Ohio for a certain distance, but they become scarcer and 

 scarcer. I have traced them down as far as Portland, Meigs Co., Ohio. But since 

 there are only isolated specimens among large numbers of typical subrotunda, I 

 have not taken the trouble of picking them out. 



"^ This form must have once existed in the Monongahela near Point IVLarion, Faj'ette Co., for 

 specimens found in an Indian garbage heap are this (Ortmann, 1909c, p. 13). It is abundant in the 

 upper Monongahela system (see below). 



