(>i<t\i\n\- Families wo Genera oi Najades. 855 



(i) those with the entire post-basal area occupied by the marsupial 

 expansion; (2) those with the marsupial expansion restricted to tin- 

 posterior ridge; (3) those with the marsupial expansion in front of the 

 posterior ridge. These divisions undoubtedly are natural, and 

 Walker thinks that the first represents the most primitive condition, 55 

 truin which the other two are to be derived. 



Having regard to the shell only, this view is quite plausible, but in 

 studying the structure of the soft parts, it becomes evident, that it is 

 scarcely tenable. Although I have seen comparatively few species, 

 it is certain that the simplest structure is found in T. triquetra, 

 which represents the second group of Walker. In this species the 

 typical features of the genus are barely indicated. From this form 

 we can imagine that the other two have been derived, and have 

 descended in apparently two parallel lines; in the one (represented by 

 T. haysiana in our material), the marsupial swelling advances forward 

 from the posterior ridge, in the other (represented by rangiana, 

 florentina and capsajormis) it becomes greatly enlarged, and often 

 corneous. In the latter forms, the inner compartment of the soft 

 parts is most capacious, and developed to its greatest extent, and con- 

 sequently these must be the most advanced types within the genus. 

 (Possibly, however, some of the third type may be as highly specialized, 

 but in another direction.) 



Truncilla triquetra Rafinesque. 



Twelve males, two sterile, and seven gravid females are at hand, 

 from the Ohio drainage in western Pennsylvania. 



The gravid females were found in September and October, but 

 further details as to the breeding season are not known. 



The soft parts have been described by Lea (Obs., X, 1863, p. 420). 

 (See also Ortmann 191 lb, p. 321.) 



Anal and supra-anal separated by a mantle-connection of medium 

 length, but shorter than the anal. Anal with fine crenulations, bran- 

 chial with papilla,', which stand somewhat remote from the outer 

 edge. The latter is, corresponding to the teeth on the margin of the 

 shell, toothed or scalloped. In the female, the inner edge of the 



55 He compares the general shape of the shell with the female of Lampsilis, and 

 thinks that it is closely allied to it. I rather believe, that the roots of Truncilla 

 are to be sought in forms which stood between Amygdalonajas and Eurynia. 

 Some characters of the shell resemble strongly those of Amygdalonajas. 



