ORTMANN: monograph of the naiades of PENNSYLVANIA. 23 



feature noticeable in almost all species found in Lake Erie. In other respects 

 this variety is rather indifferent; it inclines most toward the var. trigona, but 

 there is a tendency in some individuals to become more triangular (with higher 

 beaks), like F. fiava undata. . With regard to obesity, it is very variable, and some 

 specimens are as flat as F. flava; in fact, young specimens are very often found, 

 which are indistinguishable from young F. flava. 

 Localities represented in the Carnegie Museum: 



Lake Erie, Presque Isle Bay, Erie, Erie Co., Pa.; and also in Horseshoe Pond on Presque Isle. 



Lake Erie, Crystal Beach, Welland Co., Ontario, Canada (F. Behrlc); Port Rowan, Norfolk Co., On- 

 tario, Canada (C. Goodrich); Sandusky Bay, Cedar Point, Erie Co., Ohio (O. E. Jennings); La 

 Plaisance Bay, Monroe Co., Michigan (C. Goodrich). 



Winona Lake, Kosciusko Co., Indiana (E. B. Williamson).'' 



Distribution and Ecology (See fig. 3): Type locality, Lake Erie, Presque Isle 

 Bay, Erie, Erie Co., Pennsylvania. Type set: Carnegie Museum, Cat. No. 6L4513. 



Aside from the range indicated by the above localities, no details are known. 

 Walker (19106, p. 22) mentions that Quadnda undata has invaded the St. Lawrence 

 sj^stem, and that it is found in the lake-drainage in Wisconsin, Illinois, and southern 

 Michigan, and that it goes eastwards as far as Buffalo, New York, and Port Dover, 

 Ontario (in Lake Erie); and in 1913, Walker cites both ruhiginosa and undata 

 from Lake Erie, but the form from Lake Erie has not been discussed in detail. 

 We do not know whether the var. parvula is restricted to the lake, or also found 

 elsewhere. Marshall (1895, pp. 89 and 93) calls the Lake Erie form from New 

 York both ruhiginosa and trigona. 



As far as my material goes (I have seen over fifty specimens from Lake Erie) 

 F. flava parvida is a well-marked local form, easily distinguished from the other 

 phases of the species. However, it would not be astonishing if it should pass else- 

 where into the other forms, in fact, the shells from Winona Lake represent to a 

 degree transitions both toward F. flava and F. flava undata. 



Possibly the dwarf form of Quadrula ruhiginosa from Tippecanoe and Kuntz 

 Lakes in northern Indiana, mentioned by Wilson & Clark (1912a, p. 43), difficult 

 to distinguish from trigona, belongs here. 



The mutual connection of the three forms described above has hitherto been 

 misunderstood, and some authors have been quite emphatic in the assertion that 

 ruhiginosa ( = flava) should not be united specifically with undata or trigona (Call, 

 1895; Walker, 19106). But upon the basis of my own studies of the conditions 

 in western Pennsylvania, I am as emphatic in maintaining that flava is onlj^ the 



" These were received as " coccinea," but undoubtedly correspond to " nibiginosa " of Norris' 

 list (1902) from Eagle ( = Winona) Lake. 



