16 



OBSERVATIONS ON THE LOCAL VARIATION OF 

 CLAUSILIA BIDENTATA. 



By a. E. boycott. 



Presidential Address delivered at the Annual Meeting, October 12th, 1918. 



The stone walls in the neighbourhood of Portmadoc in Carnarvon- 

 shire are the haunt of great numbers of Claiisilia bidefitata, which in 

 suitable conditions of weather may be collected with a mininmm of 

 time and trouble. Their insistence made me wonder on an idle 

 summer holiday whether the specimens from different places in the 

 same neighbourhood were perceptibly different in any obvious char- 

 acter, and this curious interrogatory led to the inquiry of which I 

 here give some account. 



The number of characters presented by a snail shell is infinitely 

 great j it is impossible to study them all. Taking as axiomatic the 

 superiority of quantitative over purely qualitative information, one 

 naturally explored first those properties which are capable of more or 

 less exact numerical evaluation. The occasion encouraged a natural 

 inclination to believe that quantities which are easily ascertained are 

 apt to be determined more correctly than those which are difficult to 

 measure, and the only data with which I propose to deal here are 

 the length and breadth of the shell. The problem then in short is 

 whether bidentata from one locality are bigger or smaller than biden- 

 tata from another locality — a clear preliminary to the question why 

 different places are associated with different sizes. 



The definition of a '■'■locality" is not altogether easy. The word is 

 used in a pretty elastic way ; a district may be a locality, so may two 

 or three half rotten boards and a couple of bricks. Whether it be 

 really an advantage or not, I should prefer the expression locus as 

 denoting a smaller, more definitely circumscribed area, such as those 

 with which I am at present concerned. It would be tempting to say 

 that a locus is an area so small that within its limits one or many 

 species show no topographical variation, that is, all the individuals 

 form a homogeneous family, and if I were repeating this investigation 

 a delimitation of loci along this line would be one of its objects. For 

 the present it must suffice to indicate that what I mean by a locus 

 in this paper is a length of stone wall never more than loo yards 

 long nor less than about 20, free from any discontinuity, and so 

 situated with regard to trees, aspect, and such like that there is no 

 plain lack of similarity between any of its parts. 



To ascertain the approximate size of a snail shell such as bidentata 

 presents no particular difficulty. Only adult shells with well-formed 



