HYPSELOSTOMA-BOYS1DIA GROUP. 173 



lostoma and Boysidia group, like Odontocyclas, Fauxulus, and 

 some other members of the same subfamily, had independent 

 angular and parietal lamellae. In one series this primitive 

 condition has persisted ; in another these lamellaa have united. 

 The primary division of the group should therefore be into 

 two series : 



(1) Forms which have retained the angular and parietal 

 separate and parallel, and (2) those in which these two lamella 

 have become more or less completely concrescent, as they have 

 in Oastrocopta. 



A secondary division of both of these series may then be 

 based upon the character of the last whorl, whether adnate to 

 the preceding, as in normal Pupaa (a primitive condition), or 

 becoming free, extended or upturned (a highly evolved con- 

 dition). 



This gives us the following arrangement: 



Parietal and Parietal and angular 



angular separate concrescent 



Last whorl free Gyliauchen Hypselostoma 



Last whorl adnate Paraboysidia Boysidia 



If my interpretation of the facts of structure is sound, it 

 appears that the free last whorl in Hypselostoma and Gyli- 

 auchen is a secondary character assumed independently in 

 terminal members of two phyletic series. These two genera 

 are not directly related, but are allied through genera of nor- 

 mal form. 



The two groups Bensonella and Anauchen are side lines, 

 differentiated in tooth structure. 



The most primitive Boysidia, B. strophostoma, has retained 

 an early structure in the separate angular and parietal lamel- 

 lae, and in this single character (but not in form of the colu- 

 mellar lamella), it approaches Paraboysidia. In the latter 

 group, the species rol)usta and kelantanensis have evolved 

 parallel to the Boysidia; in the single character of united an- 

 gular and parietal lamellae, retaining the columellar structure 



