D.— ZOOLOGY. 97 



organism provided with a free mantle wliicli goes on growing through life, 

 and preserves a slit margin as long as it grows, the mutation to form a 

 hole must be justthe same in later as in earlier stages of the migration, 

 and no greater if it occurs at the beginning than if it occurs at the end. 

 We never can be present throughout an act of evolution, for the simple 

 reason that, until countless generations have passed, the mutation is an 

 individual peculiarity, or a local variety, or something not yet sufl&ciently 

 widespread to ensure our recognition of its significance. Nature's ' gradual 

 processes ' of evolution lie not so much in the absence of mutation as in 

 the spreading of a mutation through the community. The nearest approach 

 to witnessing such a process in our time has been the observation by 

 entomologists of the spreading of melanism in moths. 



In this address I have sought to keep morphological facts clearly 

 distinguishable from interpretations, but I have also attempted to show 

 that morphological facts require bionomical facts to elucidate their 

 significance. On purely morphological grounds I attempted to show that 

 we are under no intellectual necessity of concluding that everything new 

 must arise late in the life-history, and the development of Fissurella shows 

 us that to-day at any rate the marginal slit is converted into a hole at 

 the very outset of the adult life. I am well aware of the fact that there 

 are many other stable conditions of Zygobrancliiate holes, and that in 

 Rimula, for example, the hole, instead of being apical, is halfway between 

 the apex and the margin. It therefore furnishes to superficial appearances 

 a halfway house between Emarginula and Fissurella, and renders it 

 perfectly possible, some would say probable, that the immediate ancestor 

 of Fissurella was a Rimula, with a short slit-band, and not Emarginula 

 with a long one. I submit that the existence of Rimula makes no difference 

 to the problem of recapitulatory development as evidenced by Fissurella. 

 There is a stage in the development of Fissurella when its hole is also in 

 the middle, and it is commonly claimed that Fissurella on that account 

 goes through a Rimula stage after its Emarginula stage. But it is equally 

 true of Rimula as of Emarginula, that it possesses something which 

 Fissurella at the corresponding stage does not possess, viz. a ' slit-band,' 

 BO that any representation of the definitely adult stage of Rimula is absent 

 from the life-history of Fissurella as completely as is that of Emarginula 

 itself. All that these three genera possess in common is a short transitory 

 post-larval stage with a slit and no band, and it is at this stage that the 

 slit is converted into a hole in Fissurella. 



Under heredity we cover a multitude of things, and it seems to become 

 increasingly clear that half the things which constantly occur in a given 

 ontogeny, i.e. half the links in the necessitarian chain, are not pre- 

 determined by intrinsic structure so much as dependent on the operation 

 of influences from surrounding or adjacent parts of the developing organism. 

 At an earlier stage I suggested that the marginal slit itself may have been 

 determined originally— and, I now add, may still be determined— by the 

 pouring out of a horizontal stream of deoxygenatcd and poisonous water 

 against the growing mantle-edge. Suppose now that in the series of 

 generations between Emarginula and Fissurella the changing conformation 

 of the body, associated with perpetual downgrowth of the mantle-edge 

 and elevation of the visceral cone, should have gradually involved a 



1928 



