14 
PACIFIC SCIENCE, Vol. IX, January, 1955 
the reduction of the gill, the shell in Ver- 
metidae becomes elongated and very irregu- 
larly coiled: with the loss of the operculum 
the animal becomes able to retreat deeply into 
the shell, and the mantle in the female be- 
comes medianly slit as described above, as an 
adaptation for fixing the egg capsules directly 
to the shell This modification seems always 
to be associated with the loss of the oper- 
culum and the consequent need for deep 
protective retreats into the shell. The reduc- 
tion or loss of the operculum may itself be 
primarily a consequence of the mucous feed- 
ing habit, enabling the animal to put out its 
mucous traps unimpeded by the edge of the 
operculum, as suggested by Yonge and lies 
(1938). 
In the group including the Siliquariidae, 
the Vermiculariidae, and the Turritellidae, we 
may imagine an ancestral form with a poten- 
tiality to develop both elaborate opercular 
setae and also pallial tentacles— these are the 
chief diagnostic features of this assemblage, 
which has been shown to share a large num- 
ber of structural characters in common. In the 
genera so far investigated, Turritella retains 
both the opercular setae and the pallial ten- 
tacles, Vermicularia loses the setae, and in the 
Siliquariidae the pallial tentacles are not rep- 
resented. We should not perhaps regard the 
loss of tentacles or setae in one or other of 
these groups as being too literally a phylo- 
genetic event; or suppose that an ideal com- 
mon ancestor possessing each of them ever 
existed. All that the facts entitle us to suggest 
is the existence of a common stock which 
had a tendency to produce evolutionary forms, 
either of the siliquariid or the vermiculariid 
type. The Turritellidae perhaps comes closest, 
among surviving families, to such an an- 
cestral stock but this family is in itself un- 
doubtedly specialised upon a distinct line, 
and almost certainly gave rise directly to 
neither. 
In the evolution of a radiating group such 
as the prosobranch gastropods, we are likely 
to find different adaptive patterns almost from 
family to family; and the two groups of 
vermiform mesogastropods illustrate this 
principle very well. At the level, however, at 
which families are diversified into genera, it 
would appear that evolution due obviously 
to adaptive changes has relatively much less 
effect. Such a conclusion is best illustrated by 
the details of shell sculpture and dentition, 
and, in the present case, by the characters of 
the nuclear shell and the operculum; and it 
is in part because of their nonadaptive char- 
acter that these structures have come to be 
the most useful ones in a reliable classifica- 
tion. The evolution of other molluscan groups 
seems often to tell a similar story. We find 
first a series of adaptive characters setting off 
a family or a group of families from neigh- 
bouring ones, such as, in the vermetids, 
adaptations to ciliary or mucous feeding. 
Then, most often at the generic or inter- 
specific level, evolutionary diversity of a dif- 
ferent kind seems to enter the picture. In two 
others groups of gastropods examined by the 
writer in some detail, the Strombacea (1951 d) 
and the Ellobiidae (in press), a similar situa- 
tion has appeared, and further examples 
might easily be cited from other groups. 
Mollusca evolving at the generic level are 
uneasy ground for the dogmatic selectionist. 
In some cases there appear to be features of 
an "orthogenetic” kind (though this is a 
word that has acquired some objectionable 
shades of meaning) running in a parallel way 
through different, but related series of genera. 
Many of these trends may force themselves 
into expression in the history of a group so 
as to dominate much of its evolution. Thus 
it happens that many molluscan groups pre- 
sent beautiful examples of adaptive evolution 
at the family level, and may then become per- 
meated with what appears to be functionally 
meaningless minor trends, often, as in the 
case of shell structure, aesthetically pleasing 
and always interesting to the taxonomist. The 
selectionist’s plea will be that most of these 
characters, at present inexplicable on adaptive 
