290 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxv. 



trate the supposed evolution, the lines of a Drepanella have been 

 drawn in black over a figure of B. tuberculata. This shows that by 

 merely dividing the sickle-shaped marginal ridge of Drepanella into 

 a series of three or more node-like parts, the result is in essential 

 accord with the structure characterizing the group of B. tahereulata. 

 Above the submarginal ridge the valves of Drepanella have two per- 

 sistent lobes corresponding to the median and anterior lobes of most 

 Beyrichise. They may be dissected into subsidiary nodes (as, for 

 instance. /-*. crassinoda and D. nitida) and are sometimes connected 

 vent rally (D. bigeneris), as is commonly the case in the groups of B. 

 buchiana and B. klo&deni. The subsidiary nodes of the dissected 

 anterior lobe in Drepanella maera, D. crassinoda, and D. nitida can 

 be matched exactly in respectively Beyrichia tuberculata, B. noetlingi, 

 and B. baueri. But the tendency to dissection of the median lobe 

 exhibited by the oldest species of Drepanella is never observed in 

 species of Beyrichia. Indeed, this lobe soon became the most con- 

 stant feature for the whole family. Aside from this occasional dif- 

 ference, the greatest distinction between 

 Drepanella and the tuberculata section 

 of Beyrichia is that, whereas in the 

 former the outer sickle-shaped ridge is 

 the most constant feature, in the latter it 

 became through dissection the least stable. 

 The youngest unquestionable Drepa- 

 nella known is the D. richardsoni of the 



FlG. 10. A LEFT VALVE OF BEY- T -> • l 1 • / \i • x ,1 • , 1 



richia tuberculata (Klce- Richmond m Ohio. In this the anterior 



den), x 15, with the lines op e nd of the sickle-sha pod submarginal 

 a Drepanella drawn over it. ...... n 1 



ridge is thick and tends to connect with 



the basal part of the expanded and prominent, though still partially 

 dissected anterior lobe. Continuing this line of departure from the 

 older, typical species of the genus, a stage might be expected in which 

 the anterior lobe would be swollen to such a degree that the component 

 nodes or tubercles of the earlier dissected stages would be entirely 

 obscured. In fact, we have such a stage in a late Richmond species 

 described by LJlrich as Ctcnobolb'ma tumida. As stated in the dis- 

 cussion of that genus, the species is not a Ctenobolbina, the bulbous 

 part of the carapace which was thought to correspond to the simi- 

 larly bulbous posterior end of ('. ciliata being, in fact, anterior. The 

 original specimens of the species were not in condition to permit 

 working out all its characters exactly, nor had any reason occurred 

 at that time to lead the author to suspect that the swollen end of the 

 carapace is anterior and not posterior. Such a suspicion, ending 

 finally in conviction, arose only during the course of the present re- 

 vision of the family. Recognizing the median lobe of Beyrichia in 

 the small vertical node or ridge located well to one side of the middle 



