set free by the wound. This operation requires only a few teeth which are employed in
an unusual way, and not exposed to the wear and tear of breaking up food which may be
spiculous. I t is an intelligible (though certainly not a necessary) consequence of this
mechanism that the teeth do not fall off, but that those in front are gradually covered up
by the anterior tissues of the mouth-parts, and one can understand that this arrangement
might be reproduced in molluscs not otherwise closely allied. Still these considerations
should not blind us to the remarkable resemblance between the mouth-parts of Lobiger
and the Phyllobranchidae. But the fact that this resemblance is strongest in the family
of the Ascoglossa least like the Lophocercidse in other respects and most highly
specialized in both external and internal structure, is an objection to giving it a phylogenetic
significance, for I do not see how it is possible to derive the one family directly
from the other. I am inclined to see in this resemblance an instance of a phenomenon to
which I shall recur in the next chapter, namely that, in the Opisthobranchiata, forms
whose general structure is very different may be found to exhibit the same structure in
particular organs.
For taxinomic purposes it seems to me impossible to unite the Lophocercidge and
nudibranchiate Ascoglossa in one group and separate them from the Bullacea, and
Cladohepatica respectively, for they have more links with the groups from which we
should separate them than they have with one another. And if we were to form them
into one group, a fortiori we ought to unite Pleurobrancheea with the Dorididas,1 for their
respective diagnoses1 * 3 show hardly any difference, except the presence or absence of a
ctenidium. But undoubtedly Plev/robranchasa is closely allied to Plenrobrcmchus which
has a shell, and through Pleurobranchus to other Tectibranchs. Nothing, it seems to
me, is gained by thus breaking down the distinction between Tectibranchs and Nudi-
branchs. I t would, indeed, be most misleading to represent the latter as an independent
group standing apart. On the contrary, all Nudibranchs are clearly derived from
Tectibranchs, and all groups of Tectibranchs show members which are, so to speak, trying
to become Nudibranchs. But for taxinomic purposes the line of division seems clear,
viz. that Nudibranchs are Opisthobranchs which have neither a ctenidium nor a shell nor
an external spermatic groove, and, as far as our present knowledge goes, this dividing
line does not dissociate allied forms so much as would the removal of the Lophocercidge
from the Bullacea.
1 As is proposed by Guiart, Mollusques Tectibranches in Causeries Scient. de la Soc. Zool. de
la France, No. 4, 1900.
3 I use this expression intentionally, for there are many details of structure which are not
mentioned in the diagnosis, but indicate divergence.
VIII.
AFFINITIES AND EELATIONSHIPS.
The Opisthobranchiata, and indeed the whole subclass Euthyneura of which they
form part, show a strong tendency towards reduction or loss of the shell, hermaphroditism,.
and detorsion, accompanied by concentration of the nervous system. I t is generally
agreed that the visceral torsion of the streptoneurous Gastropods is not caused by the
coiling of the shell but that both are the result of early changes in the embryo. Detorsion
however is very commonly associated with the reduction or disappearance of the shell.
The Pterotrachseidge, which belong to the normally streptoneurous group Heteropoda, are
without shell and mantle and have also undergone detorsion. The Palmonatahave as a
rule lighter shells than the Streptoneura and contain a much larger proportion of forms
in which the shell is small, internal, or absent. I t is clear that in its more advanced stage
the acquisition of symmetry and of concentration in the nervous system is greatly assisted
by the disappearance of the shell and the freedom thereby given to nearly all the organs
to re-arrange themselves symmetrically.
Diminution or disappearance of the shell occurs in nearly all the great groups of the
Mollusca. It would seem that these heavy coverings are felt to be a burden as well as a
protection. Active animals can move more easily in lighter armour, while parasitic or
sedentary animals which are sufficiently protected by their habitat often become obese
and grow over the shell, enclosing it in. their integuments. Among living Cephalopoda
only Nautilus has a true external shell. In all the Decapoda it is internal and reduced;
in the Octopoda it has vanished, leaving slight vestiges. The Aplacophora have no
shell, though the mantle bears calcified spicules, and in several Chitons (Cryptochiton
and Ci'yptoplttx) the dorsal plates have become more or less covered, by the integuments.
It is hardly possible to imagine a shell-less Lamellibranch, for the structure
of these animals depends largely on their being bivalves, but a few forms (Chlamy-
doconcha, Scioberetia, Entovalva) have overgrown and completely enclosed the shell,
and in Teredo the long vermiform animal extends far beyond the small shell-valves,
which only cover an insignificant portion of its length. In the Gastropods the
shell sometimes becomes very small compared to the whole animal, without being
enclosed. This happens in several pelagic Heteropods, in some active carnivorous Pul-
monates (Testacella, Daudebardia), and in a few Opisthobranchs (Cryptophthalmus, Smarag-
dinella.). More commonly the shell is overgrown by the mantle and becomes partially or
wholly internal. This is very common among the Opisthobranchs and the Pulmonata
Stylommatophora but hardly occurs among the Basommatophora. Among the Streptoneura
partial covering of the shell is not infrequent, and it becomes wholly
internal in several families, e.g. Lamellariidas, Cyprasidas (Pustulana)i Fissurellidas
(Pupilia). Finally the shell may disappear altogether as in many Opisthobranchs and
Stylommatophora, and among Streptoneura in Titiscania, Entoconcka, and the Ptero-
trachaeidaa. Sometimes the shell seems to be merely absorbed and the structure of the
11