154 MEMOIRS OF THE CARNEGIE MUSEUM. 



breeding season, and are earlier than Conner's records. Some of the specimens 

 collected on April 24 were found discharging. This would establish the breeding 

 season as lasting from August to May, a condition, which is much like that seen in 

 A. grandis. It seems, that the glochidia are discharged rather early, for I never 

 found any charged females in May. 



Remarks: There still might be some lingering doubt, as to whether A. catarada 

 is specifically distinct from A. grandis. However, I have never seen any specimens, 

 which might be called intergrades, and wherever I have examined the beak-sculp- 

 ture, it has always been possible to distinguish these two species by this character. 

 Nevertheless they are closely allied, and there is no question that A. catarada is 

 the eastern representative of the western A. grandis. The two forms are geo- 

 graphically well separated. It is true that in New York the ranges do overlap, 

 but I think this is due to a secondary, postglacial expansion of the two species 

 into the glacial area. During glacial times, they undoubtedly were completely 

 separated (Ortmann, 1913, pp. 325, 363-365). Walker's opinion, expressed 

 incidentally, that catarada is closely allied to the European A. cygnea, and might 

 be a co-immigrant with Margaritana margaritifera from Europe, cannot be main- 

 tained. Compare also Latchford (1914, p. 10). 



A rather strong argument for the specific distinctness of A . grandis and catarada 

 is, as I believe, their different behavior or reaction to environment. Both possess a 

 creek- and a pond-form, but while the reaction to the pond-environment in A. 

 grandis is to develop a large, high and comparatively short shell (A. grandis gi- 

 gantea), in A. catarada a more elongated shell is the result, with a "rostrate" 

 posterior end. This latter characteristic shape of the shell is never found distinctly 

 in A. grandis. On the other hand, the creek-form of A. grandis is subovate, with 

 the anterior end broad, and the posterior narrowed, while the creek-form of A. 

 catarada is rather subelliptical, with the anterior and posterior ends generally 

 rather equally narrowed. 



The short and high type of shell, so common in A. grandis, is much rarer in 

 A. catarada, but it is present. Lea's "species" williamsi, virgulata, and trxjoni 

 represent this. They may be described as follows. 



A. williamsi. {Type locality Lower Potomac.) High and short, with better de- 

 veloped posterior wing, so that the upper and lower margins of the shell become 

 subparallel. Sometimes the posterior wing may be so high that the posterior 

 part of the shell appears higher than the anterior (ovate, with anterior end nar- 

 rower) . 



A. tryoni. Closely resembling A. williamsi, but with the posterior wing not 



