BOYCOTT : THE RADULA OF HTfALINlA. ±1^ 



of Hj. lucida, while a third lateral tooth of a more marginal type was 

 found in Hy. alUaria and some Hy. ^^glabra"'^ and the same tendency 

 for the first tooth with marginal characters "to move outwards in the 

 series glabra, allia?ia, cellaria, draparnaldi" is evident in the growth 

 of the radula in a single species. 



Something in the way of giving numerical expression to these 

 changes may be done. The'endocone of the third lateral may be 

 absent (type E), small (D), or present in good development (A, B, 

 and C). The ectocone may be present (A), represented by a 

 "shoulder," which bears no cutting point (B, C), or absent (D, E). 

 The endocone of the first marginal may be present, though always 

 small (A), or absent (B-E). Table IV. shows how frequently these 

 various forms occurred in the present series ; it is obvious that it 

 must sometimes be a matter of opinion whether a cone is " small," 

 or whether a corner or "shoulder" bears a minute cutting point. 

 When an item occurs on one half of the radula and not on the other 

 the fact may be brought into due numerical order by reckoning it as 

 half an occurrence. Thus an ectocone on the third lateral tooth 

 occurs in four of the sixty-six specimens in group V. on both sides, 

 and in four others on one side only of the radula ; the total frequency 

 of its occurrence is then reckoned as 6, and (8) is added to denote 

 that it was found in 8 different specimens. 



Table IV. 



The other features concerning the shape of the teeth I propose to 

 leave for further discussion. It is not easy to reduce the facts to 

 any absolute and objective form of demonstration ; the method of 

 multiple measurements, of which E. W. Bowell has given some 

 examples," will doubtless prove an important advance. No examples 

 of the truncated central mesocone occurred in the present series. 



1 Journal of Conch., vol. xii., p. 159. 



2 Proc. Maine. Soc, vol. viii. (1909), p. 380. 



