254 PROCEEDINGS OF THE MALACOLOGICAL SOCIETY. 



either side of the median are distinguishable ; hut of these the first 

 and second, and again the fourth and fifth, counting from the centre, 

 show less differentiation from each other than from the third, so that 

 three tracts on either side are really all that practically strike the eye. 

 Now Troschel and others have laid considerable stress on the breaking 

 up of the rhipidoglossate radula into three zones on either side and on 

 the occasional replacement of a group of marginal laterals by a single 

 large tooth, which generally retains sufficient traces of the individuals 

 it replaces to suggest that it represents the fusion of a series. This 

 is noticeable in Addisonia, Cocculina, the Neritidse and Helicidse 

 among the Rhipidoglossa, and also in certain archi-tsenioglossate 

 Cyclophoridae, the Solariidae, and even in OviiJa} Hence it may be 

 legitimately inferred th it the tsenioglossate radula, whose formula is 

 1:1:1:1:1:1:1, is derived from the rhipidoglossate by a fusion 

 of the elements of the three original zones. 



The docoglossate radula, judged from its modern representatives, the 

 Patellidse, is explicable on similar lines, for if the aborted median 

 tooth, sometimes represented by a rudimentary plate, be allowed for, 

 the remaining teeth form three series of pairs on either side. 



A further fusion, or, what is more probable in this case, the 

 suppression of the outer laterals, in the Rhachiglossa and more 

 primitive of the Toxoglossa gives rise to the 1:1:1, the typical 

 radula of these forms ; and a further abortion of the laterals to the 

 single median tooth left in Ilarpa, Marginella, and most of the 

 Volutidae. In the Cones, on the other hand, it is the median tooth 

 that is abolished, leaving the double row of barbed laterals. 



What, then, is to be said of such exceptions as occur in lanthina 

 and Scala and others? Simply that either individual requirements 

 have necessitated a return to the more primitive form of radula, or 

 that their ancestry goes further back in geological time than suspected. 

 Fossils doubtfully referred to the latter occur iu the Silurian and 

 Devonian, but lanthina extends, so far as at present known, only 

 back to the Middle Tertiaries. 



The Opisthobraochia offer every variety of radula in their ranks, 

 from the uniform multiserial to the single row of median teeth in 

 Elysia, and much further research will be necessary ere a solution of 

 this diversity is found. 



The Pulmonata present considerable variation in the composition of 

 the radula. The Auriculidse, pronounced by Pelseneer^ to be the 

 most primitive of the group, have also a primitive type of radula, but 

 geologically date only from the Cretaceous period. The Limnseidse, 

 which appear in the Jurassic, have a similar primitive radula. So, too, 

 have the Siphonariidse, but here, if Hercynella be correctly referred to 

 this family, which seems doubtful, we have an ancestry dating from 

 the Devonian. 



1 In Ovtda ovum, indeed, fusion has proceeded so far that the formula is 1:1:1:1:1. 



2 " Recherches sur divers Opisthobranches " : Mem. Cour. 4° Acad. Sci. Belg., 



tome liii, p. 114. 



