242 OBELISCUS. 



that the necessity has been felt of recognizing several genera 

 in place of the one. Stenogyra in its original limits cor- 

 responds now to a subfamily rather than a genus. As sec- 

 tions of Stenoyyra, Slmttlewort.h ranked the following: 

 Opeas Albers (1850), with the species subula, octonoidcs, 

 margaritacea, alabastrina, gompharium, goodaUi. 



Pseudobalea Shuttlw. (1854), species dominicensis Pfr. 



Obeliscus Beck (1837), species swiftiaita, turb raster. 



Subulina Beck (1837), species octona, acicularis. 



No type was selected for Stenogyra; and if the ordinary 

 process of elimination be applied to the group, the name 

 Stenogyra must replace the so-called Obeliscus of. Shuttle- 

 worth's list, which are not typical Obeliscus as now restricted. 

 This view I took in 1899 when I selected 8. terebraster as 

 type of Stenogyra. By another method of elimination, Steno- 

 gyra would take the place of Pseudobalea Shuttl. Mr. E. A. 

 Smith suggests that Stenogyra be dropped entirely, on the 

 ground that it was proposed without a type, to cover a num- 

 ber of groups already instituted. In a case like this where 

 the author's evident intention was to form a new genus by 

 the union of several prior groups, and yet by the inclusion 

 of new species renders it not exactly equivalent to them 

 singly or collectively, no two subsequent students can be ex- 

 pected to arrive at the same conclusion in its restriction. It 

 becomes a matter of convenience or personal predilection. I 

 can see no serious objection to the use of Stenogyra as a sec- 

 tion of Obeliscus for the 0. terebraster group, which differs 

 sufficiently from typical Obeliscus to require a special desig- 

 nation. 



Obeliscus is closely related 'to Rhodea and Neobeliscus, both 

 evidently branches from the Obeliscus stem. Rhodea diverges 

 from the parent stock in its later stages of growth, Neobelis- 

 cus in its embryonic stage. The forms remaining in Obe- 

 liscus 'constitute several subordinate groups, some of which 

 may hereafter be raised to genera. The rank of the several 

 groups in this classification depends largely upon the prom- 

 inence of their special modifications, and does not necessarily 

 involve the comparative antiquity of the group. Thus Rlio- 



