D.— ZOOLOGY. 89 



discussion and diagrams that the tubular shell and elongated body of 

 Dentalium have exercised an undue influence upon Prof. Naef's mind as 

 furnishing a kind of connecting link between Gastropods and Cephalopods. 

 They are, of course, indubitably secondary features. Prof. Naef's com- 

 parison between the apical slit in the shell of Dentalium and the marginal 

 slit of the lower Gastropods will be dealt with at a later stage. (2) Having 

 concluded on these grounds that the common ancestor of Gastropods, 

 Scaphopods and Bivalves possessed a flat shell and no narrow waist 

 capable of rotation between Kopffuss and visceral dome, it is easy to see 

 that the phyletic linkage of this group to still lower forms of Mollusca 

 must be with forms of the Placophoran rather than the Cephalopod type. 

 I do not mean that the Conchifera do not form a natural assemblage, but 

 simply that the first Conchifera must have been essentially Chiton-like 

 in organisation except for the simplicity of their shell : the conversion of 

 the discoidal shell of the earliest Conchifera into cones, tubes, and spires 

 has taken place independently in Cephalopods, Scaphopods, and 

 Gastropods, in relation to very different habits of life. 



If these considerations are well based, there can be no presumption in 

 favour of a pelagic ancestry of Gastropods, and the attempt I have made 

 to explain the evolution of the Veliger larva without regard to such 

 ' phylogenetic reminiscences ' can be submitted without anxiety as to 

 objections on that account. The only pelagic feature of a Veliger is its 

 velum, and that, as we have seen, comes down from a Trochosphere, not 

 from a pelagic Mollusk. The bilobed origin of the foot in Patella and 

 Trochus admits of various alternative explanations. 



It will be noted that Prof. Naef's argument in support of the sudden 

 and muscular character of the original process of torsion remains un- 

 affected. He constructed his case from the observation of larval behaviour, 

 but applied it to the behaviour of hypothetical primaeval adults, which 

 could not possibly have behaved like larvae if they had existed. On the 

 other hand, if you prolong backwards into the Cambrian the larval 

 sequence which is demonstrable to-day, and project into it a con- 

 tinuation of the train of modifications in the mode of development of the 

 torsion which we have also seen to be operating, step by step, and 

 sub-order by sub-order, there appears to be neither speculation nor 

 hypothesis in the conclusion that torsion in Gastropods arose as a larval 

 mutation : the logic is that of simple mathematical extrapolation, or of 

 projecting a curve the equation of which is known. 



Of course I cannot tell whether you consider my proposition reasonable, 

 or not, on the evidence I have put before you, and I have sought to base 

 it entirely on positive grounds which are open to verification. Let me, 

 however, now draw your attention to the secondary or corroborative 

 evidence. At the outset of my discussion I remarked on the sharpness of 

 the gap which separates Gastropoda from all other groups of Mollusca. 

 The one thing, i.e. the only thing of importance, that distinguishes 

 Gastropoda from other Conchifera is their torsion, and that is complete 

 from the start. Torsion makes the Gastropod, and it appears in the 

 systematic sequence as a true saltation. Now if torsion arose in the 

 first instance by gradual modifications of adult form, each step fitted to 

 some particular combination of external conditions or internal functionings, 



