248 



•Annals of the Carnegie Museum. 



This form is very likely the southern representative of the foregoing. 

 It agrees with it in every respect. In one of the specimens from Louisi- 

 ana and three from Arkansas the supra-anal was separated from the 

 anal, in the others this separation was absent. The inner edge of 

 the anal is finely crenulated. The posterior margins of the palpi are 

 connected for about one-third of their length, and the inner lamina of 

 the inner gills is free from the abdominal sac. 



The gill-structure of the females is identical with that of C. undulata. 

 The gravid female had only a few glochidia in the outer gills, while the 

 inner ones were yet partly charged, and both suprabranchial canals, 

 as well as the cloacal chamber, were filled with masses of loose glo- 

 chidia, partly sticking together, but not in the shape of placentae. 



Glochidia like those of C. undulata. Length 0.20; height 0.21 mm. 



Crenodonta heros (Say). 

 According to the description and figure given by Lea (as multi- 

 plicutits, Obs., VII, i860, p. 222, pi. 30, fig. 105), this species without 

 doubt belongs here. 



Crenodonta trapezoides (Lea). 

 I have received, from L. S. Frierson, one male and two females from 

 Bayou Pierre, De Soto Parish, Louisiana (collected Aug. 6, 1910), 



-> an 



\br 



Fig. 5. Crenodonta trapezoides (Lea). Male, from Bayou Pierre, De Soto 

 Parish, La. (Cam. Mus., No. 61, 4,586.) 



from A. A. Hinkley two females from Pearl River, Jackson, Hinds Co., 

 Mississippi (collected Nov. 5, 1910), and from H. E. Wheeler a male 



