BOYCOTT : THE RADULA OF HYAI.INIA. 3OI 



ectocone is represented by a well marked fold and notch but the tip 

 of the cusp thus formed has no refractile cutting point. The con- 

 dition is represented in figure 2 and is evidently different from the 

 more usual shoulder (figure i) or small ectocone with a cutting point 

 (figure 3). The common abnormalities of Hyalinia radula — bifid 

 cones — occurred in five of the whole 160 specimens. 



Some mention is required of the characters of the snails and their 

 shells from which these radulse came. In no case from an examina- 

 tion of the living animal have I had the slightest doubt that I was 

 dealing with the species commonly known as helvetica. The external 

 appearance of this species when alive and active seems to me to be 

 highly characteristic and exceptionally uniform, compared with e.g.^ 

 ceUaria, and to a less degree alliaria. Careful observation of live 

 specimens crawling about is, I think, of special importance in Hyalmia. 

 Generally speaking the four series present no differences which seem 

 of much importance. In the shell the Tremadoc and Bicknor speci- 

 mens are thinner than those from Banstead and Marple, especially 

 the former which were difficult to collect and deal with without 

 breaking a good number. They came from a slaty rock, the Bicknor 

 specimens from limestone. The Tremadoc group is further dis- 

 tinguished by a narrower umbilicus^; the mouth is also relatively 

 higher and the columellar margin more vertical than in Marple and 

 Bicknor specimens in which the mouth is narrower and set more 

 obliquely. The Banstead series agrees more with those from Tremadoc. 

 The white about the umbilicus is best developed in the Banstead 

 series and is practically absent in those from Tremadoc. The spire 

 in the Tremadoc group is flatter than in the others. Such are the 

 differences which by our present methods of examination are qualitative 

 only in character. Definite measurements give the impression of 

 greater uniformity. The average dimensions which are susceptible to 

 fairly easy and tolerably accurate evaluation are given in Table VI. in 

 terms of the major diameter. The shells are those in the 7 and 8 mm. 

 groups in each series. - 



Locality 

 Banstead 

 Marple 

 Bicknor 

 Tremadoc 



The differences of conoidity, closeness of whorls, relative size of 

 last whorl, etc., are evidently not very great, though the method of 



1 Large specimens everywhere seem to have a relatively wider umbilicus. 



2 There is no obvious growth change in these figures. The 6 mm. group of Banstead 

 helvetica gives loo, 90, 53, 35 ; the 5 mm. group igo, 8g, 51, 34 : the 4 mm. group 100, 89, 52, 35. 



