these cases there is often a doubt as to whether the divergent specimens are really
identical, but it would seem that in Doris (especially in the sections Staurodoris and
Archidoris) a slight labial armature may be present or absent, and specimens of
Acanthodoris from New Zealand identified by Bergh with A. pilosa have no armature of
hooks on the verge. With regard to the radula no general statement can be made. In
some groups, e. g. the Polyceridse, it is remarkably constant, and it may be said that as a
rule narrow radulse with differentiated teeth show less variation than those which are
wide and contain many teeth of much the same pattern. The shape of the separate teeth
and the general proportions of the arrangement are usually preserved within the same
species, e.g. a short, broad radula does not become a long and narrow one. But the
number and size of the teeth may vary greatly in individuals of different age and size, or
sometimes without special reason. Thus in Tritonia liombergii the formula varies from
45 x 120.1.120 to 100 X 265.1.265. In JEolids the number of denticles on the teeth is
often variable, particularly when the teeth are broad as in JEolidia.
Apart from other causes, age may affect not only the size but the shape of
Nudibranchs. As a rule the older animals are larger, deeper in colour, and, if the species
bears tubercles, pits, or processes, show these in a more developed form. But sometimes
the opposite happens. The appendages do not keep pace with the general growth of the
animal, and their shape becomes obscured. Thus in the young Plocamopherus the dorsal
appendages are relatively larger and more branched than in the adult. According to
Trinchese the young Lomanotus eisigii is seolidiform in appearance, but the skin grows
up between the cerata, which finally become a row of points on an undulating lateral
membrane. In the young Grosslandia (allied to Scyllsea) the wings are bifid and bear
finger-like processes. In the adult the bifurcation and the processes both disappear, and
the wing becomes a roughly triangular flap.
But it is in colour that the variations of Nudibranchs are most striking. The
common Doris tuberculata is protean in this respect. Pure yellow specimens are
sometimes found, but usually there are mottlings of one or more tints on a light ground.
The range of colours comprises red, yellow, pink, brown, and grey of many shades,
purple, slatey blue, sage green, and perhaps others. In a series of specimens received
from the Isle of Man slate-colour and greyish-blue predominate. At Plymouth red and
yellow, though not universal, are very common. I have occasionally seen there bright
yellow specimens marbled with bright red which might have competed with any tropical
Dorid in brilliancy. It is worth noticing that this variation in colour and pattern, though
luxuriant, is not unlimited. As far as I know, the markings of D. tuberculata never form
stellate patterns on the back, as they do in the equally variable D. testudinaria, and there
are never any spots on the under side of the mantle.
These variations are probably due in part to climate. At Plymouth many marine
animals as well as seaweeds are reddish. Tropical Nudibranchs, as well as the Ascidians
and sponges among which they live, are more deeply and gorgeously coloured than those
of temperate and northern waters. These last are often of a semi-pellucid white, and
when the coloration appears brilliant it will generally be found to consist of bright spots
distributed over a colourless surface, and not of large pigmented areas. Colour also
depends to some extent on food, especially among the Solids. Though some of their
markings may be due to pigment in the integuments, yet the predominant colour is
generally determined by the hepatic diverticula as seen through the transparent cerata,
and these vary with the food. Thus Fiona marina is of a greyish-blue when it feeds on
Velella, but of a pale brown when it eats young barnacles. Coryphella gracilis (orange-
brown) and G. smaragdina (green) are probably similar variations produced in the same
animal by different diets. So, too, in Favorinus albus the hepatic diverticula may be
white, greyish-brown, coffee-coloured, red, or green, and white spots may be present or
absent on the integuments. But though Trinchese noticed and figured all these variations
at Genoa, Vayssi&re found that they did not exist at Marseilles.
The common JEolidia papillosa is almost as variable in colour as D. tuberculata..
Hecht mentions a case of its becoming violet after eating Actinia eguina, and apart
from such special and temporary modifications the species comprises several races,
as described by Alder and Hancock, distinguished by their size and the length of the
cerata as well as by the colour, which ranges through many shades of grey, buff, yellow,
brown, rose, and green. As in D. tuberculata, the specimens found on our southern
coasts show a tendency towards a rosy or pinkish coloration. The colour of Dorids is
to some extent affected by their food, though less than that of HSolids. The brightly
coloured species often frequent and feed on similarly bright sponges1 or Ascidians, and
when thev do not obtain their usual food in confinement they lose their colour. Hecht
mentions that Elysia viridis (which usually feeds on Godium) became much larger when
fed by him on Cladophora, and developed red cells in the epithelium.
This great variability naturally makes the definition of species a difficult business.
In particular it is very hard to say whether preserved Nudibranchs which are similar
but come from widely distant localities (such as Great Britain and the Falkland Islands)
are specifically identical or not. Even when the characters of a living specimen are
certain, their value as specific characters can be determined only after comparison with
a series of other specimens. Hence species which are based on one (or even on two or
three) specimens are nearly always open to suspicion, unless their peculiarities are so
decided as to make them also representatives of separate genera. Alder and Hancock
doubtless erred in creating too many species, particularly in the JEolididas, but the error
is on the right side, for if certain types only are selected as species and others are
dismissed as varieties, there is danger that the varieties will be neglected and the real
multiplicity of forms forgotten. Yet morphologically a variety may be as important as
a species. The only difference is that if two forms are connected by intermediate
gradations they are called varieties, whereas if the connecting links are absent they are
called species. Thus the forms described below as Doris veirucosa and D. maculata are as
distinct as any accepted species. They can easily be distinguished in appearance, and
present real structural differences. Yet it is highly probable that they pass into one
another through the medium of Do'ris pseudoverrucosa, and also pass into many other
forms which have received specific rank such as D. januani, ocelligera, atypica, and fallc-
landica. But it is clearly safer to maintain the two species as separate until it is proved
that they pass into one another. And even then (provided our conception of species is
1 Thus the red British Dorids Rostanga cocdned and D. fiammea eat red sponges such as Micro-
cion a dtrasanguinea.