220 ANNALS NEW YORK ACADEMY OF SCIENCES 



mens upon which Swallow bases his description of Terebratula hrerilo- 

 hata. In the smaller examples, of course, the convexity is still lower, and 

 no trace of lobation is to be seen. 



While recognizing the relation of these faintly plicated shells to the 

 two examples which more closely agree with Swallow's description, it has 

 seemed to me from the evidence at hand that it would be well to recog- 

 nize them as representing a distinct variety. 



Harttina anna var. graciliformis var. nov. 



Under this title, I am including two specimens which I at first, though 

 really against the evidence in hand, provisionally identified with Dielasma 

 gracile. Their size is much smaller, but their shape is almost exactly 

 that of the larger shell, elongate and subpentagonal, with the greatest 

 width near the middle or a little below. No fold or sinus appears to 

 have been developed. The ventral valve has the usual dental plates and 

 the dorsal a median septum. , 



Harttina indianensis var. exporrecta var. nov. 



Associated with examples which have been referred to Harttina hrevi- 

 lohata var. marginalis and to H. indianensis, I have in several instances 

 found shells more or less closelv resembling them, Irat distinguished bv 

 having a broader, rounder shape. Neither valve has a distinct sinus, but 

 traces of a sinus can sometimes be noticed in both valves. 



So far as observed, the specimens referred here have only the median 

 septum of Harttina without the lateral plates and platform of Dielasma. 



Dielasma formosum var. whitfieldi var. nov. 



This species is abundant at two or three localities almost to the ex- 

 clusion of other types of Terebratula. It is one of the poorly character- 

 ized forms, having a broadly ovate shape and nearly obsolete fold and 

 sinus. The greatest width is usually below the middle, and the outline 

 in front is often somewhat flattened. The sinus, when present, is de- 

 veloped onty toward the front, where it is shallow and undefined. Ven- 

 tral beak small. The convexity varies from rather low to rather high. 

 There is also variation in the width, some specimens being wider than 

 others. 



Dielasma formosum var. seminuloides var. nov. 



This form resembles D. formosum var. ivliitficldi, except tbat it is 

 much more spreading. Though the variety whitfieldi manifests a tend- 

 ency to pass into these rotund forms, some of them could not, I tliink, 



