GWATKIN : SOMK ^[OLLUSCA^• RADUI..*:. I43 



licking movement of the front end ; but I doubt if this is a full 

 account of it. I commend to your study the little Margarita helicina, 

 found in England as far south as Scarborough. It is of a convenient 

 size for observation in a zoophyte trough ; and I think we shall have 

 made some way when we have made out the. whirlpool motion of the 

 radula. It is many years since I saw it, but I believe it gives a 

 problem not even yet fully solved. 



The value of the radula in classification is variable. For the genus 

 it is always important, and commonly decisive ; for the species it is 

 often decisive, sometimes useless. Thus, in the Biiccinida the indi- 

 vidual variation seems greater than the specific, and in huge genera, 

 like Clansilia and AchatiiieUa, the species cannot generally be dis- 

 tinguished by the radula. In the former, Mr. Davy Dean's researches 

 would seem to shew that the clausium is a more distinctive character. 

 Yet even in these, the radula is not quite useless. Achatinella separ- 

 ates at a glance from Amastra, and if the material could be obtained, 

 I believe a clear distinction might be found between the American 

 Claiisilias and those of the Old World. But the difificulty of classifi- 

 cation is always the same, whether we are classifying religions, or 

 governments, or beasts — the difficulty of finding a character whose 

 witness is never overborne by the convergent witness of other char- 

 acters. Thus, there is no enormous difference of shell between Helix 

 and JVatalina, or of anatomy between Alurex and Ranella \ yet the 

 radulfe are very unlike. Conversely, we find a few curious likenesses 

 of radula in genera which must stand very far apart, like Omphalotropis 

 and Ovnla, Brachypodella and what I shall presently call the Coosa 

 group of Ancylus. 



Upon the whole, however, the indications of the radula are more 

 often confirmed by the rest of the anatomy than those of the shell. 

 Thus, the huge radula of Natalifia, with its aculeate teeth and enor- 

 mous retractor muscles, implies a considerable difference of anatomy 

 from that of Helix with its quadrate teeth and moderate buccal mass. 

 Patella and AcmcBa have very similar shells ; there is much more 

 difference in the anatomy and the radula. So, too, the Laomas of 

 Southern Europe and the Tesseraria of New Zealand used to pass as 

 Fupas, and many another fraudulent pretender has been detected by 

 the radula. 



Let me give you now an illustration more in detail. For some 

 years past I have studied the genus Ancylus, so far as my material 

 permitted, in alliance with Mr. Bryant Walker, of Detroit. He has 

 described the shell, while mine has been the more modest part of pre- 

 paring the radula. So it is only of the radula that I can speak, though 



