110 MEMOIRS OF THE CARNEGIE MUSEUM. 



tendency to develop many different phases, which may turn up anywhere under 

 proper conditions, but which do not lead to the development of geographical races, 

 at least in our territory. 



General distribution: Type locality, North America (Spengler). 



According to Simpson (1900, p. 725), E. violaceus is found on the Atlantic 

 side from the St. Lawrence to Georgia. The southern boundary is extremely un- 

 certain, and is obscured by a great development of local forms, the standing of 

 which is very doubtful." Northward, it goes to Maine (Jackson, 1908, Lermond, 

 1909, Nylander, 1914), and, according to Call, to Nova Scotia. From the lower 

 St. Lawrence-drainage, it has been reported from many localities in Vermont, 

 New York, and Canada, up to and into Lake Ontario. It has never been found 

 in Lake Erie, but turns up again in the upper lakes district, in western Ontario 

 (see our localities in Sandy and Pigeon Lakes, and in Muskoka and Parry Sound 

 Cos.). It has been discovered in various places in northern Michigan, as far as 

 Lake Superior (See Walker, 1898, p. 5; 1913, map on p. 30, and Winslow, 1917, 

 p. 11). An isolated lotaUty is the one reported by Sterki (1907a, p. 393) from a 

 mill-race at New Philadelphia, Tuscarawas Co., Ohio, and recently I found a single 

 specimen in Grand River, Eagleville, Ashtabula Co., Ohio. 



This westward extension of the range apparcntl.y was not by way of Lake 

 Erie, but from the lower St. Lawrence, in the region about Ottawa, in the direct 

 line toward Lake Huron, so that in Michigan it is a recent immigrant from the 

 North. Walker (1913) has treated this question in detail: the migration was by 

 waj^ of the Trent or Nipissing route (l. c, pp. 44, 45, fig. 4). 



The established range of this species is unique, and larger than that of any 

 other form of the Atlantic slope. Most peculiar is the westward extension in the 

 north. There is only one species, which might be compared with this, Eurynia 

 nasuta, but the range of this from west to east goes by way of Lake Erie, and is 

 undoubtedly governed by different laws, as I have shown (Ortmann, 1913a, pp. 

 378 ff.). 



Elliptic cupreus (Rafinesque) (1820). 



Unio productus Conrad. Simpson, 1914, p. 690." 



Plate VIII, fig. 6. 

 Records from Pennsylvania: 



Ortmann, 1910, p. 117 (Fulton Co.). 



'1 Conrad (1834, p. 8) says that the shells of the type of U. purpureus (= E. violaceus) are in the 

 ■ Savannah, Oconee, and Ocmulgee Rivers, and also in Flint River, Georgia, but that to the west of this 

 he did not find them. 



"- L". fisherianus Reeve (1865, PI. 24, fig. 11.3) is not U. fisherianus of Lea, but clearly the present 

 species. 



