178 Annals of the Carnegie Museum. 



to take from Table III. Correlations and observations on Paraptera 

 are apt to be misleading for the reason stated. For this, however, a 

 great deal one way or another can be inferred from the even number of 

 cases mentioned. It follows that greater rigidity of the posterior 

 portion of the shell is required in Lake Erie; vice versa in the Ohio. 



{e) On the Thickness of the Shell. 



While Lake Erie contains more lime, it is evident that the great 

 majority of shells are thinner there. It is known that the presence of 

 CO2 (44) in water tends to erode shells, yet far more of the finer shell 

 architecture, such as beak-sculpture, seems to be preserved in specimens 

 from Lake Erie than in those from the L^pper Ohio Drainage. March 

 states (30) that shells from districts highly charged with CaCOs have 

 thin shells and are not eroded at the beaks. This may be attributed 

 to the lack of humic acid, which does not occur where limestone is. 

 Greater thickness of shell seems to be connected with presence of 

 humic acid, or absence of chalk. Dr. Ortmann noted the presence of 

 humic acid in the Shenango, the water supply of which is largely derived 

 from swamps containing peat and decaying vegetable matter. This 

 explanation of March's does not correlate with my measurements in the 

 case of Anodo7!ta, Pleiirohema, EUiptio in the Shenango, although it 

 does in the case of other species. With regard to marine Pelecypoda, 

 Baker (4) observes that the thinness of shells in the Caspian Sea 

 appears to be directly correlated with the degree of saltiness of the 

 water. It is safest to assume that each shell takes what lime it needs 

 under the circumstances. Rich (41) tells of some shells from a soft 

 water lake in New York, which were almost free from lime. 



It is thought that perfect uniformity in these results has not been 

 obtained because of: 



1. Lack of sufficient material in species, the variability of which 

 may be unusually great. 



2. Ecological habits of species, which may play a larger part than 

 we now know. 



Sources of Error. 



With a reasonable amount of care the rapid measurement of these 

 dimensions was not inconsistent with accuracy, with the sole exception 

 of that of the length of the hinge. This seemed to be too important a 

 character to neglect, yet often it had to be estimated by the eye where 



