BY H. L. KESTEVKN. 61 



leave a lasting impress. If this were not so, our classifications 

 of tne members of the different phyla would be far more 

 perfect, more natural than they are. As an illustration of 

 this, we can confidently say that the species of Onchidium 

 are recently evolved from some marine ancestor, but though 

 an Opisthobranch origin has been suggested, there is not 

 sufficient evidence to decide whether the ancestor was Opis- 

 thobranch or Prosobranch ; we continue to class them with 

 the Pulmonata, though by no means satisfied that they are 

 derived from the same stock as other Pulmonates. This group 

 (Pulmonata) then, is probably at least diphyletic, and may 

 be polyphyletic. 



If we imagine a feature "A" to be acquired by an adult 

 organism, say a gastropod, during a given period "a," that 

 character appearing late in life on the ultimate whorl of the 

 adult shell, then, in the succeeding period "b," this charac- 

 ter will have been, by tachygenetic tendencies carried back 

 on to the penultimate whorl, and, in the period "c," appears 

 on the antepenultimate whorl, and so on. Finally, however, 

 this minor tachygenetic wave will break against a aock, and 

 be eliminated from the ontogenetic record. The rock against 

 which it breaks is the hereditary resistance of the larval 

 stages. It is this resistance of the larval stages which has pre- 

 served with, but minor modifications, the Trochosphere stage 

 in the ontogeny of the Rotifera, the Annulata, and the Mol- 

 lusca, and the Veliger in that of all Gastropoda. In some 

 cases, the adult organism has proven equally resistant, and 

 there has resulted the "persistent type," e.g., Lmgula. 



The morphological features of the gastropod, like those of 

 most other larvae, are stable, and have been beyond the reach 

 of those minor fluctuations which have resulted in the differ- 

 entiation of the various subdivisions of the group. (It may 

 be suggested that the extreme reduction of the velum in 

 some of the Pulmonata, shadowing forth, as it does, its 

 approaching elimination from the ontogenetic record of these 

 forms, indicates that the Veliger ancestor was a completely 



