STATION IN FRESH-WATER MUSSELS. 277 



of West Virginia (below Clarksville) is polluted, and its fauna is 

 destroyed. I have older material, however, from one locality in 

 the Monongahela proper; farther up, I collected some near the 

 mouth of Cheat River, just at the state line; and at one locality in 

 the headwaters (West Fork Riz^er) . The measurements are as 

 follows. 



Loc. No. Max. Min. Av. 



Monongahela, Charleroi 13 61 48 53 suhrot. (kirtl.) 



Cheat, Cheat Haven 2 52 40 46 kirtl. (subrot.) 



West Fork, Lynch Mines 2 44 41 42.5 kirtl. 



This material is rather scanty ; nevertheless the decrease in 

 obesity in the upstream direction is clearly seen. The locality 

 Charleroi has a higher percentage (63) than the next localities below 

 in the Ohio (47 and 48 at Industry and Coraopolis), but agrees with 

 that of Cooks Ferry. This, probably, is again due to the age of the 

 shells. In the Ohio below Pittsburgh, giant specimens, with the 

 length over 100 mm., are not uncommon, while among the shells 

 from Charleroi there is not a single one over 100 mm. (largest 

 91 mm.). 



Very probably, this law holds good also in other tributaries of 

 the Ohio. I have not enough material to demonstrate this. In one 

 case, that of the Ttiscaraivas River in Ohio, I have rather abundant 

 material from the Holland collection, but, unfortunately, no exact 

 localities are given ; both forms, however, are represented in this 

 river. .From the Levisa Fork Big Sandy River, Prestonsburg, 

 Floyd Co., Ky., I have a single specimen, the diameter of which is 

 50 per cent., that is to say, it is a subrotunda standing just at the 

 lower limit, and close to kirtlandiana, and this would correspond 

 to the station well up the river. From Little Kanazvha River, 

 Grantsville, Calhoun Co., W. Va., I have one specimen, the diameter 

 of which is 43 per cent, this is the typical kirtlandiana, correspond- 

 ing to the small size of this river. 



All this serves to show that our contention is upheld, that F. 

 subrotunda and kirtlandiana are forms of the same species, differ- 

 ing only in obesity. The former is the swollen form of the large 

 rivers ; the latter is the compressed form of the small streams. Both 



PROC. AMER. PHIL. SOC, VOL. T IX, R, AUGUST 20, I92O. 



