146 MEMOIRS OF THE CARNEGIE MUSEUM. 



elsewhere, and its presence in such ponds can be explained only by the assumption, 

 that exceptional means of dispersal are available, presumably transport by aquatic 

 birds. This is also suggested by its distribution across the northwestern section 

 of our state, where it seems as if it had gone across countrj^ not following the 

 streams. This, however, requires further investigation. It should also be pointed 

 out that it is very remarkable that these factors, if acting at all, did not act in an 

 eastward direction across the Alleghenian divide; and that this species is entirely 

 missing in the Atlantic-drainage, although it approaches the divide very closely 

 in Indiana and Westmoreland Cos. But this holds good only for Pennsylvania, 

 and is different in New York (see below) . 



General distribution: Type locality, Fox River of Wabash River, Indiana (Say). 



According to Simpson (1900, p. 644),^"- the range covers the "entire INIissis- 

 sippi system; the up])er St. Lawrence-drainage; Red River of the North; Lake 

 Winnipeg; Manitoba; southwest Texas" (he also with doubt mentions. south- 

 eastern Pennsylvania, but this should be cancelled). This is certainly a correct 

 general statement. But the exact boundaries of this species are yet largely ill- 

 defined. In western Pennsylvania we apparently have part of the northeastern 

 boundary of the range, and it here coincides practically with the divide between the 

 interior basin and the Atlantic-drainage. However in New York state this is 

 different, for here this species passes over into the Atlantic-drainage of the Hudson 

 as well as the St. Lawrence River (see DeKay, 1843; Dewey, 1856; Marshall, 

 1895; Baker, 18986). In part in its eastward advance, it seems to have used the 

 route of the Erie canal, reaching Albany County, and has here invaded the terri- 

 torj^ of the eastern A. cataracta. Its relation to this latter species in this region 

 (western and central New York) has never been closely studied, and on account 

 of the great resemblance of these two species it is very likely that they have often 

 been mistaken for each other. In this region A. grandis (generally kno-wai under 

 the name of A. leivisi) is found chiefly in streams and in some smaller lakes, and 

 it does not go into the large lakes. The same seems to be true all along the northern 

 boundary of its range, and, as soon as the great lakes are reached, it turns into the 

 var. footiana. In Pennsylvania, it goes up to the divide, and it is even in Conneaut 

 Creek, a tributary of Lake Erie; but it is not in the lake. 



Generally in the region of the divide, intergrades are found between the two 

 varieties. Farther south grandis prevails, and finally is the only form in existence. 



"- To Simpson belongs the credit of having reduced the number of species of Anodonta to reasonable 

 limits, and he has to a large extent worked out the sj-nonymy of this species; but I hardly think that he 

 has gone far enough, and believe, that some " species " admitted by him, will finally fall as synonyms 

 under A. grandis. 



