BOYCOTT : THE RADULA OF HYALINIA. 299 



Interpreting the results along these lines we arrive at the following 

 characteristics by which each local series may be differentiated from 

 each of the others. Cases which are queried are probably rather than 

 certainly distinct. Cases which are left blank show no significant 

 difference, but it should be remembered that the degree of difference 

 necessary for significance varies with the number of specimens 

 examined and it by no means follows that the measurement of longer 

 series would not reveal distinctions where, on the present evidence, 

 they cannot be said to exist. ^ 



It is evident that there are one or more particulars in which each 

 series differs from the others. On the whole Banstead is most and 

 Marple least divergent from the general run. Tremadoc and Marple 

 are nearest one another. Banstead and especially Bicknor have 

 narrow radulse, and Bicknor a generally small organ. Banstead has 

 the greatest number of rows with many marginals and hence a large 

 number of teeth, which, as the radula is not conspicuously large, are 

 rather small. 



The configuration of the variable teeth {i.e., the third lateral and 

 first marginal) also shows some local variation. The cutting points 

 may in a general way be regarded as falling into three categories of 

 development and to each may be assigned a numerical value. They 

 may be absent (=o), small or imperfect (=0-5) or fully developed 

 (^i). For example, in the eleven specimens in the 8 mm. group 

 from Bicknor the endocone of the third lateral tooth is absent in one, 

 small in 7, and well-developed in 3, giving a total of o + 3'5 + 3=^6'5 

 marks out of a possible total of 11, or 60 per cent. This figure of 60 

 may be taken as representing approximately the degree of development 

 of that cutting point in that particular series of specimens. The 

 calculation is obviously based partly on fact and partly on one's per- 

 sonal interpretation of the significance of " small." The " figures of 

 merit " for the present series are : 



Table IV. 



Eudocone Ectocone Eudocone 



3rd lateral 3rd lateral 1st marginal 



Tremadoc 7-S '9 mm 52 31 o 



9-9 '9 mm 70 60 o 



Bicknor 7-8-9 mm 57 47 o 



9-9'9 mm 75 65 o 



Marple 7-S'9 mm 55 ...... 27 i 



9-9-9 mm. 75 45 ..... o 



Banstead 7-8-9 mm. 



I In the present state of our knowledge one has to restrict each series to those collected at 

 one time over quite a small area of ground in order to exclude time and locality factors which 

 have not yet been examined. Hence it is njt always easy to secure adequate series of the 

 appropriate sizes. 



