474 Haronp HEATH, 
proved or disproved it does not necessarily throw any light on the 
point at issue. 
About the shell there is nothing new to be said. It is absent 
in the present day species, and the imbricating chiton-like plates 
on the dorsal side of the larva may in reality be modified spines. 
In this connection it may be said that there is very little foundation 
so far as I can see, for the belief that the chitons (and the soleno- 
gastres if they are in reality of a common stock) have been derived 
from ancestors with a distinct head provided with tentacles, a highly 
arched visceral mass covered with a simple shell and posteriorly 
placed mantle cavity containing a small number of gills; and that 
in adapting themselves to a littoral habitat they have become 
transformed into the modern types. Comparative anatomy speaks 
in no certain tone regarding the subject, and embryology gives very 
little information indeed in support of such a belief. The very ill 
defined head the chiton possesses is even less evident in the larva; 
there is no indication of a visceral hump at any time; and in the 
formation of the mantle cavity there is no evidence whatever that 
it was originally located posteriorly, nor that it contained in the 
adults of a former period a small number of gills. 
It has been asserted repeatedly that the branchial folds of the 
Neomeniidae are merely modified walls of the cloacal chamber, and 
that the plume-like gills of the Ohaetodermatidae are but a further 
development and not therefore ctenidia. It is true that there are 
lamellae in both types of branchiae, and they are both placed in a 
cavity (that I believe to be a true mantle cavity at least in the 
Chaetodermatidae) at the posterior end of the body, but here the 
resemblance stops. The COhaetoderma type of gill is so remarkably 
similar to the generally recognized ctenidia of the prosobranchs or 
chitons for example that I cannot force myself to the belief that 
they have had a different phylogenetic origin. Whether the neo- 
menian branchiae are a late development or not is far from being 
demonstrated, but to me their resemblance to ctenidia is not as 
close as some authors would have us believe. 
As is well known it is exactly these external features that 
are the most readily modified, and are the ones most difficult to 
deal with. On the other hand the internal systems of organs, 
especially those comprising the nervous and coelom, are comparativly 
conservative and probably will afford the most fruitful field for tlıe 
student of molluscan relationships. In fairly closely related groups 
