BY H. L. KESTEVEN. 73 



which, being moulded inside the horny Nectoconch, are devoid 

 of the adult characters ; for these latter I used the term 

 Pseudo-X ectoconch {aiite p. 62) ; in such instances the Anune- 

 anoconch of the former type is also present. 



The P&tudo-S ectoconch will generally have smooth and 

 rounded whorls ; since such is the character of the inside of 

 the majority of the horny Nectoconchs, they, in many in- 

 stances, present a complete model of the inside of the Necto- 

 conch; in others, they fall short of this, owing to the inter- 

 position of a septum between the upper and lower whorls ; 

 this deficiency reaches its maximum in those cases where the 

 whole N ectoconch is shed, and the apical whorl closed by a 

 shelly septum (^caducous ajJex). 



Ananeanoconchs of the former group often present interest- 

 ing features. In the generality of cases, the Ananeanoconch 

 presents the contour and sculpture of the adult shell ; such is 

 the case throughout the genus Cyniatiuin, including most of 

 those segregated by Dall under the name Se-pta, and with 

 only one or two exceptions. In a few cases, however, we find 

 the Ananeanoconch presents characters differing in some 

 marked way from those of the adult, presenting, in fact, ex- 

 amples of adolescent jmecilogony , when we compare allied 

 species. The most striking instance I know of among the 

 Mollusca, is one which I myself recorded in "A Note on Two 

 Species of AstraUum from Port Jackson" (26). In this case, 

 two species, so similar to one another that they had not 

 previously been recognised as distinct, were shown to have 

 Ananeanoconchs presenting different types of coiling ; the 

 one being truly trochiform, with columella and no trace of 

 umbilicus ; the other, at the same stage, is Z,(o^/rt-like, with a 

 wide umbilicus. 



At that time I offered the opinion that : "These two species 

 would seem to present an exception to the rule that distinc- 

 tive characters are inherited earlier in successive generations, 

 for, by that theory, .4. finihriatuni should be descended from 



