96 SECTIONAL ADDRESSES. 



adult recapitulation naturally falls with it. On the Haeckelian hypothesis 

 the inheritance runs like this : — 



Emarginula : Egg — Larva — Adult 



I X (with slit) 



Fissurella : Egg — Larva — Neanic stage — New Adult 



(with slit) (with hole) 



Fissurella is regarded as inheriting all that Emarginula has to give, and 

 as then adding a new stage to the series. This differs from the previous 

 adult stage, but is built up out of a Neanic stage, which is claimed to 

 represent the previous adult. 



But we have seen that the complete adult ancestral stage is not 

 inherited : the whole of the post-Neanic ontogeny which includes a ' slit- 

 band ' is absent from the ontogeny of Fissurella. We must therefore 

 distinguish like from unlike, and represent the two ontogenies differently : 



Emarginula : Egg — Larva — Young Adult — Old Adult 



X -i- (with slit) (with slit-band) 



Fissurella : Egg — Larva — Young Adult — Old Adult 



(with slit) (with hole) 



From this analysis it results, so far as this one particular character 

 is concerned, that the ontogeny of Fissurella repeats the ontogeny of 

 Emarginula up to the Neanic stage of the latter, but no further, and then 

 deviates. There is thus no ' compression ' of the whole adult stage of 

 Emarginula into the Neanic stage of Fissurella. It is much the same with 

 regard to all other characters, with this qualification, that the point at 

 which divergence takes place may be quite different for different organs 

 (e.g. character of gills, kidneys, &c.). One ontogeny is derivable from 

 another, but when modification is introduced, it is not by the addition 

 of a new total ' stage ' at the end of the previous life-history, but by 

 interstitial changes, so to say, either in individual organs or in parts of 

 organs, and usually in quite early stages of growth and differentiation. 



I have claimed that torsion arose in the free-swimming larval stage of 

 Gastropod ancestors, and that by so arising it created the order Gastropoda; 

 also that the marginal slit arose in an early post-larval creeping stage 

 to meet respiratory difficulties then first encountered as a result of the 

 mutation. I now claim that the hole of Fissurella arose by a modification 

 of the marginal slit at a stage of development scarcely later than that of 

 the slit itself, but at a later period in the phyletic history. By this I mean 

 that the immediate adult ancestor of Fissurella was to all intents and 

 purposes an Emarginula with marginal slit and long slit-band ; and that 

 the slit was transformed into a hole very much as it is transformed to-day 

 and at the same stage of the life-history. 



We have an irresistible tendency when considering the evolution of 

 living things to look for gradual changes — ' By Nature's gradual processes 

 be taught ! ' to requote Wordsworth — , but there is no getting over the fact 

 that the conversion of a slit into a hole sooner or later involves an act of 

 discontinuity, — a mutation. At the critical period in the evolutional 

 process one generation had a migrating slit, and the next generation, or 

 some individuals in it, changed the slit into a permanent hole. Now in an 



