system usually consists of seven principal ganglia. The liver, the kidney, and the glands
of the genitalia are all ramified, sometimes .in a very complicated manner.1 There are
often two spermatothecas. As a rule there are three genital orifices and the genital ducts
are completely or incompletely triaulic.1 2 * The follicles of the hermaphrodite gland are
themselves hermaphrodite, and are not divided into male and female acini. The vent is
usually dorsal; more rarely (Phyllobranchus) lateral. As a rule there is only one pair of
tentacles (the rhinophores) or none at all.
I t may be mentioned that the anatomy of all the Ascoglossa is exceedingly complicated
and often varies in apparently allied species. The following characteristics of
families are therefore given with all reserve.
The Hermaeidae are seolidiform animals bearing not very numerous cerata. ‘ Stiligw,
in which the rhinophores are smooth and ungrooved, is externally indistinguishable from
the true iEolids except by the absence of oral tentacles (which are also wanting in the
iEolid Embletonia). In Ercolania the rhinophores are slightly furrowed. In Hermsea,
Placida, etc., they are auriform. The anal papilla is usually dorsal and anterior.8 The
teeth are smooth or serrulate; there is no ingluvies buccalis. The stomach is small
and lies transversely across the body-cavity, Into it enter two liver-canals,4 * one on each
side, from which rise branches that sometimes form a layer of tubes in the integuments.
They enter the cerata and ramify there, or more rarely (Placida) remain simple. There
are two or three genital orifices and sometimes the male orifice is separated from the
others by a considerable interval. The hermaphrodite gland fills the hinder part of the
body-cavity. In some cases, at any rate, there are two receptacula seminis, and the
genitalia have accessory ramified glands. The verge is armed with a spine.
In a formal classification it is perhaps simplest to_ include Alderia amongst the
Hermaeidje, but it is a connecting link between them and the Limapontiidse. In many
points it is allied to the latter, but is dissociated from them by its cerata and the
compact structure of its hermaphrodite gland. Lobiancoia Trinchese is described as
having the general structure of Hermsea, but the diverticula of the liver do not enter the
cerata. The Phyllobranchidae are characterized externally by the possession of flat leaflike
cerata of unique appearance. They usually have grooved and bifid rhinophores as
well as grooved tentacles. To the buccal mass is attached an ingluvies buccalis which is
sometimes very large. The teeth are usually denticulate, and only two or three are in
use at one time. There is a dilatation (proventriculus) before the stomach. Both the
digestive and reproductive organs are extremely complicated; the latter are provided
1 Something analogous to this ramification of the genitalia may be seen in the prostate of
Ploeamopherus (a phanerobranchiate Dorid) which is dendritic and surrounds the spermatotheca.
2 But there is some divergence in the statements of various authors on these points. According
to Pelseneer Cy&rce, Elysia and Limapontia have one spermatotheca. In Hermsea bifida there are
only two orifices; the prostate is ovoid and the albumen gland not ramified in the papillae.
8 In the somewhat doubtful Hermseopsis variopicta it is said to be lateral, but this has been
contradicted.
4 Both as to liver-canals and as to the genitalia there is considerable discrepancy in the
statements of Bergh, Trinchese and Pelseneer.
with numerous accessory ramified glands, and the genital system is either completely
or incompletely triaulic. The genera, though externally similar, present considerable
differences of structure; thus in Gyerce the leaf-like appendages are merely tegumentary,
whereas in Phyllobranchus and Galiphylla they contain branches of the liver. The family
is recorded from the Mediterranean and Atlantic but not from British waters.
In the two remaining families, Elysiidse and Limapontiidae, there are no cerata and
no appendages whatever except the rhinophores, which are sometimes much reduced. In
the Elysiidse the sides of the body are expanded into thin wings within which the subdivisions
of the liver are profusely ramified, so that the arrangement is functionally
similar to the cerata of other families. Also the afferent blood-vessels appear as threads
or ridges on the dorsal surface. I t does not appear to me that Placobranchus is entitled
to rank as a separate family. Like Elysia it has wings, but they are folded more stiffly on
its back, and it has an ingluvies buccalis. The Limapontiidse are small slug-like animals
without any special arrangements for respiration, which must be performed by the integuments
generally. They frequent tide-pools and brackish water. They sometimes leave
the water, and the abnormal development of Genia, which issues from the egg in the
adult form, is probably connected with this habit. Despite the difference in their external
appearance the two families seem to be nearly related. A buccal crop is in both (almost
invariably) absent, and the hermaphrodite gland is not a compact mass but a layer of
diffused follicles. In the Elysiidse the ramifications of the liver and the genitalia are more
considerable, and are no doubt correlated with the greater space afforded by the lateral
expansion of the body. In the Limapontiidae the external symmetry is greater, for the
vent is posterior and median, whereas in Elysia it lies in front of the pericardium and to
the right. But in Thundilla, which has in other respects the appearance of Elysia, it is
terminal. This form has also a buccal crop. Limapontia- depressa is flatter and wider
than other members of the genus, and if the lateral expansions of the body were somewhat
increased it would show much the same external configuration as Tharidilla. Elysiella
and the Bosellia of Trinchese also show a transition towards the shape of Limapontia.
Alderia is another annectent form, uniting some of the characters of Limapontia, Elysia,
and Hermsea. I t resembles the first inasmuch as the anal papilla is terminal and the
rhinophores are reduced to rounded prominences. As in Elysia the sides of the body are
developed into flaps, but these flaps bear cerata as in Heimsea, and, as in H. bifida, the
genitalia open by only two orifices.
All the Ascoglossa show signs of specialization and none of them can be considered
as archaic forms. The apparent simplicity of the Limapontiidse is accompanied by such
features as the posterior position of the aual papilla and two widely separated female
orifices» so that their want of appendages is probably retrograde, not primitive. Genia
has even acquired a special larval history. On the other hand the Hermseidse, especially
H. bifida, present few features which cannot be paralleled among the JEolididse and allied
families. The radula of Myrrhine approximates to the ascoglossan type, as does also that
of Favonnus. The jaws are very weak in Doto and absent altogether in Eedyle iveberi,
which is a cladohepatic Nudibranch, though of doubtful affinities. Anterior tentacles are
absent in Embletonia and reduced in several other forms. The anal papilla is antero-
dorsal in Dotonidse. The male and female orifices are separated in Fiona. The kidney