294 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxv. 



far as essentials go, the result is not unlike the Ordovician genus 

 Drepanella, with its great, sickle-shaped marginal ridge. 



Much similarity is traceable also between B. clavata and the syn- 

 thetic B. interrupta group. The posterior ridge is not so well de- 

 veloped in that group, but in B. v-scripta and B. granulifera it is 

 clearly suggested to where it is lost in the low anterior lobe; and just 

 over its ventral part is the antero-median isthmus. Whether either 

 of these resemblances are indicative of close genetic alliances can 

 not be decided with the evidence now available. The youngest un- 

 questionable Drepanella, D. richardsoni (Miller), of the Richmond 

 group, analyzes more in accord with Beyrichia tuberculata than with 

 B. cldvata. As stated in another note, the union of the posterior lobe 

 of B. tuberculata with the lower of the three parts of the anterior 

 lobe (see figure) would give every essential of Drepanella. While 

 the importance of the difference is recognized, and it is a fact that 

 links establishing the relation are unknown, the writers, nevertheless, 



Figs. 31—33. — 31. Left valve of Beyrichia clavata Kolmodin. :vz. Left valve of 

 Beyrichia granulifera, new name (Bollia granulosa Krause), x 15. (After 

 Krause.) '■'>'■'>. Right valve of Beyrichia tumida (Ulrich), x t». The figures 

 illustrate the resemblance of the beyrichia clavata group to the 15. inter- 

 rupta group and show the similar antero-ventral prolongation of the poste- 

 rior lobe. 



are convinced that the B. tuberculata section was evolved out of 

 Drepanella. 



Despite the unbroken antero-ventral continuation of the posterior 

 ridge in B. clavata, the direct derivation of this species from Drepa- 

 nella seems unlikely, except it be through B. tumida (Ctenobolbina 

 tumida I'lrich.)" Derivation from something like B. granulifera 

 and />'. v-scripta is at least equally plausible. However, neither of 

 these possible solutions is entirely satisfactory, so that for the present 

 the origin of B. clavata must be left as undecided. 



As for II. umbonata, which is somewhat doubtfully referred to this 

 group, the alliance with Drepanella seems much more natural. Ex- 

 cept that the valves are. on the whole, more convex, and the lobes 

 thicker and less sharply defined, every other essential feature may be 

 duplicated in typical Drepanella like D. crassinoda and I>. macra. 



Another drepanelloid Beyrichian and possible member orderivation 

 of this group is the Devonian B. kolmodini Jones. This species has 



" See notes on Drepanella richardsoni and Beyrichia tumida on page 290. 



