36 MUSSELS OF CUMBERLAND EIVER AND TRIBUTARIES. 



except a few deep holes; the bottom is gravel and mud. The mus- 

 sels were collected with a rake and by wading, and were mostly near 

 the bank in the mud, only a few being found in the gravel. Sixteen 

 species were obtained in all, two of which, S. costata and S. compla- 

 nata, were rare in the main river. The mussels were said to have 

 been formerly abundant, but they had been nearly cleared out by 

 pearlers, and not enough marketable species were left to make fishing 

 profitable. 



Haynes Lake lies several miles below Clarksville, on the north side 

 of the river, and is apparently a part of the old river channel. It is 

 about a mile long and surrounded by woods; the bottom is soft mud 

 and the watei- is about 3 feet deep, with a temperature of 89°. Very 

 large specimens of Anodonta grandis gigantea were obtained, 2 of 

 which contained sporocysts of some distomid, while 2 others were 

 gravid (Sept. 4). The nacre of 8 was purplish, that of the remain- 

 ing 17 a beautiful creamy white. Of the 2 specimens of A. imhecillis 

 1 was gravid. 



Elk Creek Shoals, 13 miles above Dover, had a current of 3 miles 

 an hour in water 10 feet deep over a bottom of gravel mixed with 

 some sand. Nine of the pigtoes obtained here were gravid (May 30), 

 and on the land bar above the shoals was found one dead TrundUa 

 sulcata, a species which is exceedingly rare in the Cumberland. 



Walter's shelling camp was about a mile below these shoals, and 

 Mr. Walter very kindly conveyed us up and down the river in his 

 launch, giving much valuable information. He had a pile of shells 

 containing about 150 tons, of which the most important button shells, 

 in the order of their abundance, were the Ohio River pigtoe, the 

 Cumberland pigtoe, the monkey -face, the yellow sand-shell, the but- 

 terfly, the niggerhead, and the southern mucket. 



At Glasgow Landing, 2 miles above Dover, on May 29 the current 

 was about 4 miles an hour, the water high and muddy but rapidly 

 falling, and the bottom gravel mLxed with clay. About one-third of 

 the pigtoes were gravid, the glochidia being usually in the lower half 

 of the outer gills. The niggerheads were also in the early stages of 

 gravidity, all four gills being red and padlike; one elephant-ear was 

 gravid. At the foot of Dover Island the conditions are the same as 

 just recorded except that the water was 20 feet in depth. A small 

 species of Atax, with broad white marks on the back, was found on 

 several of the mussels obtained here. Marginal distomid cysts were 

 fairly common, especially in P. secuns (the butterfly). This same 

 butterfly was frequently gravid, the pigtoe was less often gravid, and 

 a single specimen of Q. fragosa had glochidia in all four gills. 



A noteworthy feature of the lower river, somewhat marked at 

 Clarksville, but decidedly more so at Dover and below, is the land- 



