1898.] NATURAL SCIENCES OF PHILADELPHIA. 265 



and Malakozoqlogische Blatter represent it ; and Arango, in his 

 Contribution a la Fauna Malacologica Cabana, notes the internal 

 structure of the Cuban species. Most of these observations, how- 

 ever, are not sufficiently detailed or exact to meet the requirements 

 of the case, now that a classification is based largely upon internal 

 structure ; and our own work is therefore founded wholly upon the 

 study of a series of sections including nearly every species in the 

 collection of the Academy. 



Without entering into any elaborate expose or criticism of the 

 work of former authors upon the Cylindrellas, attention should be 

 directed to the contention of Crosse & Fischer 1 and later of Fischer 2 

 that the series should be distributed between two family groups, the 

 CyUndrellidce and the Pupidce, a conclusion based wholly upon the 

 structure of the teeth and jaws ; the first family having greatly mod- 

 ified teeth and plaited jaw, the second having normal dentition and 

 solid jaw. 



That this splitting of the Cylindrellas into two is an unnatural 

 division, seems to us to be proven by the following considerations : 

 (1) The discovery of completely Cylindrelloid shells (Epirobia) 

 with the " normal " type of teeth. (2) The presence of transition 

 stages in the teeth in the genus Holospira, and (3) the recent dem- 

 onstration by the senior author of this paper 3 of the rapid changes 

 undergone by the teeth of some gene"ra under the stress of changed 

 habits, without corresponding changes in the rest of the anatomy, 

 as seen in Papuina, Polymita, etc. (4) The general law of change 

 in the structure of the jaw, as illustrated in the families Endodon- 

 tidce and Helicidce, must now be recognized as largely discounting 

 the old value placed upon that organ as a factor in systematic mal- 

 acology ; and in any family of snails we may expect to find both 

 the more primitive plaited and the later solid type of jaw. The 

 final reason for rejecting the idea that any of the group under con- 

 sideration are Pupidce, is that none of them, so far as known, possess 

 the extremely characteristic complication of the male genital organs 

 found in Pupa, Buliminus, Clausilia, and their immediate allies, and 

 which constitute one of the most important characters of the 

 family Pupidce. 



1 Journ. de Conchyl., 1870. 



2 Manuel de Conchyliologie. 



3 Manual of Conchology (2), IX, introduction and portions relating to ar- 

 boreal Helices. 



18 



