# 



:;i 



PROCEEDINGS OF THE ACADEMY OF [Vol. LXXIV 



single cusp, although more commonly one or two cusps will be con- 

 siderably reduced. However, the cusps of the centrals, laterals, 



i inner marginals are quite often broken or worn, so as to give 

 an apparent variation, but an examination of all parts of the 

 radula will easily remove this source of error. In specimens of 

 similar size, the maximum variation in the number of marginals 

 was •"• teeth, and part of this may have been due to the difficulty 

 in ascertaining to which row the divergent outer teeth belonged. 

 However, a dwarfed form of Helicina adspersa (H. a.marmorata 

 d'Orb. +tenuilabris Pfr.) had only 80 marginals on each side, 

 while the larger form had 106. 



On the other hand, big variations are fairly common. In one 

 of the radulae of Lucidella aureola and in one of Oligyra orbiculata, 

 two rows of A-centrals were present on each side, so that there 

 were 9 teeth in the central group; this, if a constant character, 

 would exclude them from the Helicinidae, as at present defined. 

 In these cases, the outer A-central was a little smaller than the 

 inner, and, especially in the latter species, somewhat approached 

 the B-central in shape. In another radula of H. orbiculata, all but 

 one of the cusps on one row of A-centrals were absent or vestigial, 

 although the other side was perfectly normal. In a radula of Hender- 

 sonia occulta, the seventh marginal on one side was vestigial, 

 while an entire longitudinal row of marginals on the other side was 

 branched, and bore from 2 to 4 quite perfect tips, each with its 

 full complement of cusps. In a specimen of Helicina adspersa 

 marmorata, most of the cusps were represented by mere knobs on 

 one row of comb-laterals, although those of the other side were nor- 

 mal. However, this type of variation comes under the heading of 

 monst rosil ies or mutations and is very easily separated from fluctuat- 

 ing or intergrading variation. It does not affect the use of the radula 

 as an index of relationship, any more than the presence of gill- 

 slits, or a fuller complement of aortic arches in an adult human 

 being, affeel the use of these characters in the classification of 

 vertebrates. 



In the following arrangement of the groups of species from the 

 mainland of North America, the radular characteristics are listed 

 in some detail. Opercular and other shell characters are omitted 

 3 used for the separation of groups; they are treated in detail 

 by A. .). Wagner I 1907-1911). In each case the name of the group 

 is given hist., followed by a list of synonyms. In the West Indian 



