12 
PACIFIC SCIENCE, Vol. IX, January, 1955 
ficed, especially as the slit in the shell becomes 
very narrow, at times almost obsolete, and is 
always occluded by the overgrowth of sponge 
tissue and spicules across it. 
Adaptations for the protection of the em- 
bryos must be briefly mentioned for the Sili- 
quariidae. In Stephopoma , the embryos with 
their trumpet-mouthed shells, enclosed in a 
thin egg capsule, are simply held freely in the 
mantle cavity of the parent, as is the case in 
the more primitive members of the Verme- 
tidae. In Pyxipoma there is a spacious brood 
pouch excavated in the cavity of the head, 
immediately below the buccal mass and the 
oesophagus. It opens to the exterior by a small 
circular aperture on the right side at the ter- 
mination of the ciliated oviducal groove run- 
ning forwards from the female aperture, and 
close to the termination of the food groove 
below the right tentacle. 
No species of the genus Vermicularia ap- 
pears to have been examined from living 
material. The present writer (1951 b) suggested 
reasons for its transfer to a position alongside 
the siliquariids, pointing out that in its struc- 
ture it was certainly unlike any of the gastro- 
pods properly belonging to the Vermetidae. 
Following access to a better supply of pre- 
served material of Vermicularia spirata , the 
close relationship of this genus with both the 
Siliquariidae and the Turritellidae is fully con- 
firmed. It forms an interesting example of a 
form intermediate in structure, and presum- 
ably in habits, between the freely moving, 
ciliary feeding turritellids and the fixed or 
embedded siliquariids as has been indicated 
in a more recent discussion (Morton, 1953): 
Vermicularia then has close relationships with Tur- 
ritella in the pallial tentacles, and the structure of the 
gill, the food groove and the foot; as well as in the 
radula, the embryo shell, the operculum and the onto- 
geny and sculpture of the adult shell. On the other 
hand, it is close to Stephopoma in many features, such 
as the uncoiled shell, the sessile "vermetid” habit, and 
the greater elongation of the gill filaments . . . . In 
choosing with which family the genus Vermicularia 
should most properly be placed, we should probably 
select the Turritellidae, which would enable the Sili- 
quariidae to be reserved for those sessile forms with 
reduced, trumpet-mouthed apices. But the relationship 
of Vermicularia to both is very close and there can be 
little reason to doubt that the genus is near the point 
at which a deposit feeding prosobranch with a tur- 
ritellid organisation gave rise to one or more lines of 
sessile, uncoiled derivatives. 
DISCUSSION 
To "define” a group of organisms at the 
family level in terms useful to the systematist 
is a problem rather different from "character- 
ising” it in such a way as to bring out its 
interest to the evolutionist. In this case we 
are concerned less with extreme cases or with 
anomalies in marginal form, and more with 
the central pattern of organisation that unifies 
the family, and with the evolutionary trends 
being followed within it. These trends would 
appear in many cases to be of two rather differ- 
ent kinds: there is first the phenomenon of 
"adaptive radiation” between different genera, 
and this must account for many of the charac- 
teristics of the Vermetidae and Siliquariidae, 
especially the modifications of the foot and 
pallial cavity; but there are, further, a number 
of trends running through each family, often 
apparently developed convergently in differ- 
ent genera, which seem to be much less 
closely, if at all, related to adaptation. 
It must be obvious that in the vermetid 
and siliquariid stocks we are dealing with two 
fundamentally different groups of Mesogas- 
tropoda. The writer has already (1951 b) out- 
lined at length the differences between the 
two families, and has suggested that the Sili- 
quariidae approach rather closely to the Turri- 
tellidae in their affinities. It is much less clear 
from which group the Vermetidae may have 
originated, and there seem to be no living 
mesogastropods which at all clearly suggest 
an ancestral form. In the development of 
ciliary feeding within the Siliquariidae and the 
more primitive section of the Vermetidae, we 
would seem to have no more than an instance 
of parallel adaptation. There are moreover 
several other apparently unrelated groups of 
prosobranch gastropods which have devel- 
