Yet Tritonia is more primitive than the Dorids, so that these latter and the Pleurobranchids
must represent parallel, not successive stages of development. In external features at any
rate the possibility of such convergence is certain, for not only Dorids and Pleurobranchids
but also the Lamellariidge and Oncidiidge have the same general appearance.
But the resemblances between Tectibranchs and Nudibranchs are not confined to the
Pleurobranchids and Lophocercids among the former, and many features characteristic
of particular groups of Nudibranchs do not occur in those two families but in other Tectibranchs.
In fact most of the peculiarities of the Cladohepatica may be found scattered
among the Tectibranchiata, though not united in one form, and though the majority of the
Cladohepatica can be explained as successive specializations of a type resembling Tritonia,
the points in which they differ from Tritonia often prove not to be novelties but features
occurring in the Tectibranchs, which as a group are less specialized than Tritonia. It
may be useful to collect some instances where features characteristic of the Nudibranchs
as a whole (e. g. nervous system, loss of shell, internal vas deferens), or of particular
groups (e. g. divided liver, narrow radula, armature in genitalia) appear in Tectibranchs.
Taking the external characters first,, the shell is entirely lost not only in Pleuro-
branchaea but in Pelta and in Acclesia, a genus of the Aplysiidge. The perfoliate rhinophores
of many Nudibranchs may be compared to the same organs in Tylodina which are
lamellated internally, and also to the lamellated sense-organs of several Bullacea (most
highly developed in Haminaea and Aplustrum), which though not tentacular seem to
perform the same functions as rhinophores.
In the alimentary system the following points may be noted. The division of the
liver is not entirely confined to the cladohepatic Nudibranchs, and (though unknown in
the Pleurobranchids) occurs in other Tectibranchs. In some Gdvoliniae this organ is
formed of two separate lobes, each with its own duct. In Gastropteron it consists of a
number of small branched glands (about fifteen) which enter the stomach by separate
ducts. The ramifications of these glands interlace with one another and with the lobes of
the hermaphrodite gland. The liver also consists of three separate portions entering the
stomach by separate ducts in the Oncidiidse, an aberrant family of Pulmonates with
affinities to the Tecti- and Nudibranchiata.
The broad radula of the Pleurobranchids resembles that of many Dorids, but narrow
radulas occur in other families of Nudibranchs and can be paralleled among the Tectibranchs.
Thus a triseriate radula (1.1.1) occurs in Pelta, Diaphana, Thecosomata, and many Cladohepatica.
The raduls of Philine and Scaphander, in which there are only two large
laterals with or without a small median tooth (1.1.1 or 1.0.1), resemble that of the Gonio-
doridids, and the radula of Gylichna (a central tooth with one large lateral followed by two
to five small ones) shows some analogies to that of Adalaria. The Bullid Newnesia has a
uniseriate radula. Teeth are entirely absent in Doridopsidge and Phyllidiids (Holohepatica),
Tethymelibids (Cladohepatica), Doridium and Retusa (Bullacea). Most remarkable is the
occurrence of an upper unpaired jaw similar to that of the Pulmonata in Ogives, a phanero-
branchiate Dorid, and Bullastra.1 In the G-ymnosomata, jaws when present are fused to1
This very obscure form is regarded by Bergh (Mai. Unters., iv, 3, 1901, p. 254) as belonging to
the Bullacea. It seems to me as if it might be equally well regarded as an aberrant Pulmonate. But
gether below. Stomach-plates or spines which do not occur in either the Pleurobranchidge
or Lophocercidge are found in most Bullacea, the Gymnosomata, Aplysiidge, and among
Nudibranchs in the Tethymelibidge, Scyllgeidge, Bornella, and Bathydoris hodgsoni. In
some species of Aplustrum too (A. physis, A. velum) the mandibles form a ring or band
analogous to the labial armature of many Dorids.
The central nervous system is concentrated in Notarclms and Acclesia, the visceral
ganglia 'being drawn up between the pleural ganglia and the visceral commissure being
shortened. Concentration (varying in degree in the different genera) also occurs in
Gastropteron, Aplustrum, Umbrella, Tylodina, and the Pleurobranchidge. A blood-gland is
found in these last and in most Holohepatica; also in Bulla, Aplustrum, Scaphander, Philine,
and Alcera.
In the genitalia, an external spermatic groove is usual among the Tectibranchiata^
but a curious combination of genera replace it by an internal vas deferens: Acteon, Aplustrum,
Lophocercidge, Umbrella, Tylodina, and Pleurobranchidge. In these last and in Aplustrum
the orifices are contiguous, as in most Nudibranchs. In Acteon, the Lophocercidge, Fiona,
Galma, and the ascoglossan Nudibranchs, they are separate: Most Tectibranchs have two
spermatothecas, but Acteon and Gastropteron, like the Cladohepatica (except some Asco-
glossa), have one. The Thecosomata and. Gymnosomata have. also only one or none at
all, like Phylliroe. A chitinous armature of the male genitalia, such as is common in Nudibranchs,
is not recorded in the Pleurobranchidge but is frequent in the Aplysiidge, where
it appears as numerous simple spines or chitinous hooks set upon softer prominences.
In Diaphana and Bullacta the verge is armed with a single spine or hook as in some
^Solids, Kent/rodoris, etc. In Scaphander the configuration of the hermaphrodite gland
is almost the same as in some Dorids. I t is spread over the surface of the liver and sends
prolongations into it: The structure in Alcera appears to be similar.
Thus many important characters common to the Dorids and Pleurobranchidge are
also found in other Tectibranchs. Such are the absence of a shell, concentration of the
nervous system round the oesophagus, an internal vas deferens, and a blood-gland. Other
characters are found in several Nudibranchs and Tectibranchs but not in the Pleurobranchidge.
Such are an armature of the stomach, an armature of the genitalia, very
narrow radulge, lamellated sense organs on the head, a liver in several divisions. It
would seem that when one group of animals (here the Nudibranchs) is derived from
another (here the Tectibranchs), and some families of the parent group approach so near
to some families of the derived group that they may be regarded as annectent, then the
derived group may exhibit not only the characters of the annectent families and others
that may be supposed to have grown out of them, but also characters not found in the
annectent forms at all but distributed among other members of the parent group. This
may be explicable by the disappearance of extinct forms, which leaves us with a very
limited and inadequate notion of the extent to which characters may have been combined
in the past. It is also possible that an annectent form may transmit not only its own
characters but the potentiality of other characters existing in distant members of the
group to which it belongs. Among the Tectibranchs not only the Pleurobranchidge and
whatever the affinities of Bullastra may be, the fact that Mgives, alone of all Nudibranchs, has a
jaw like a pulmonate, remains.