The teeth on the front part of the radula are often much worn (Fig. 112), and probably fall away by degrees, their place being taken by others successively pushed up from behind. At the extreme hinder end of the radula the teeth are in a nascent condition, and there are often as many as a dozen or more scarcely developed rows. Here, too, lie the cells from which the teeth are originally formed. The length and breadth of the radula vary greatly in different genera. In _Littorina_ it is very narrow, and several times the length of the whole animal. It is kept coiled away like a watch-spring at the back of the throat, only a small proportion of the whole being in use. I have counted as many as 480 rows in the common _Littorina littorea_. In _Patella_ it is often longer than the shell itself, and if the radula of a large specimen be freshly extracted and drawn across the hand, the action of the hooks can be plainly felt. In _Aerope_, the Turbinidae generally, and _Haliotis_ it is very large. In _Turritella_, _Aporrhais_, _Cylichna_, _Struthiolaria_, and the Cephalopoda it is small in proportion to the size of the animal. In the Pulmonata generally it is very broad, the length not exceeding, as a rule, thrice the breadth; in most other groups the breadth is inconsiderable, as compared to the length. The Radula is wanting in two families of Prosobranchiata, the Eulimidae and Pyramidellidae, which are consequently grouped together as the section Gymnoglossa. It is probable that in these cases the radula has aborted through disuse, the animals having taken to a food which does not require trituration. Thus several genera contained in both these families are known to live parasitically upon various animals--Holothurians, Echinoderms, etc.--nourishing themselves on the juices of their host. In some cases, the development of a special suctorial proboscis compensates for the loss of radula (see pp. 76–77). In _Harpa_ there is no radula in the adult, though it is present in the young form. No explanation of this fact has yet been given. It is also absent in the Coralliophilidae, a family closely akin to _Purpura_, but invariably parasitic on corals, and probably nourished by their exudations. There is no radula in _Entoconcha_, an obscure form parasitic on the blood-vessels of _Synapta_, or in _Neomenia_, a genus of very low organisation, or in the Tethyidae, or sea-hares, or in one or two other genera of Nudibranchiata. [Illustration: FIG. 112.--Example of a front portion of a radula (_Cantharus ringens_ Reeve, Panama), much worn by use. × 70.]