140 JOUkNAL OF CONCHOI.OGY, VOL. I4, NO. 5, JANUARY, 1914. 



the Urocoptidce, and the land operculates. In the HeUcinid<B it about 

 equals the length of the shell, and in the Nerltidce exceeds it ; while 

 in the marine LittorinidcE it reaches three or even seven times that 

 length, and has to be curled up like a watch-spring. The "teeth," 

 again, are simple structures, not to be compared with the teeth of 

 higher animals. Most of them have an organic basis of chitin, super- 

 ficially hardened by mineral deposits ; only in the brown teeth of the 

 Docoglossa the chief constituent is silica hydrate.^ In the Ptdmonata 

 they are of a beautiful silvery whiteness, standing out on a dark 

 ground like spun glass. So white are they that the radula of a Pupa 

 or a Laomn lying in water can easily be distinguished from a particle 

 of dust. The chief exception (among the Pulmoiiata) is that the 

 posterior part tends to a brownish colour in most of the Agnatha and 

 also in Circinaria. 



As a general rule there are many rows of teeth across the mem- 

 brane, all pointing backward like the roughnesses on the cat's tongue, 

 and serving a similar purpose of scraping, though the aculeate teeth 

 of the carnivorous Agnatha may be used to seize their prey. In the 

 Pulmonata we never find a suctorial mouth, or an aborted radula like 

 that of Harpa^ or such slender and delicate teeth as those of Chenopiis 

 or Cerithiopsis, which can only be used on the softest substances. I 

 cannot name even a manifestly degraded radula like that of Buccin- 

 opsis or Cohunbella. The teeth are always very numerous, ranging 

 down from the Mexican Lydnoe ghiesbreghti, which has nearly 40,000; 

 Limax viaximus has about 29,000; Helix aspersa 14,000; and even 

 so small a species as Pyra/nidula rupestris can muster 5,500. But the 

 number is commonly smaller when the teeth have unusual forms, and 

 in Brachypodella it may come far below a thousand. It is worth not- 

 ing that the English species of Vitrea fall into two very distinct 

 sections. In nitiduia, pura, etc., the radula resembles that of Limax, 

 and the teeth range from 2,500 to 4,500 ; while in aiiiaria, cellaria, 

 and their allies, where the teeth are considerably modified, the number 

 is only about a thousand. There is a similar distinction in the foreign 

 species of Vitrea. 



Allowing for incomplete growth at the hinder end of the radula, 

 the teeth are quite uniform from front to back. Every malformation 

 runs down the whole file of teeth from back to front. This, indeed, 

 is what we should expect, for every defect in the parent cell must be 

 propagated the whole length of the radula. The transverse rows from 

 side to side of the radula are symmetrical, in the sense that the two 

 sides of the median line correspond tooth to tooth, or very nearly so; 



jNIiss Sollas in Quartcr/y Journal of Microscopical Science, vol. 51 (Felj. 1897), pp. 115 — 



