of the Fishery Board for Scotland. 
243 
ing animals — that of smell — I am disposed to think that this lateral sheen 
and non -sexual coloration is useful in keeping shoals of the same species 
together. It is well known that shoals of sprats and young herrings keep 
well defined and separate, unless in rough weather, which may drive them 
together, but they afterwards separate. 
In regard to sexual colours proper, it is noteworthy that they are 
absent in almost all forms producing pelagic ova ; and indeed, as might 
be anticipated, in all those whose reproduction is mixogamous (where the 
ova of each female are normally impregnated by the spermatozoa of many 
different males), as in the Gadidae, Pleuronectidae, herring, smelt, sand-eel, - 
&c. There is a striking exception in the skulpin or dragonet (Calliony- 
mus lyra), which produces isolated pelagic ova, and in which the sexes 
are markedly dissimilar, and this puzzled me much until I came across an 
observation of Saville Kent's, who had watched the process of impregna- 
tion. A male and female at the spawning time were observed to rush up- 
wards together in the water with their ventral surfaces closely apposed; and 
reproduction is hence not mixogamous. Sexual colour seems to be associated 
with the male, only when the eggs of a certain female can be impregnated 
by one individual male, and it is characteristic of many fish with demersal 
ova (lumpsucker, Cottus, Stickleback, Salmonidae, &c), and the region, or 
mode, in which this form of coloration is manifested is usually such as to 
interfere as little as possible with the protective coloration, e.g., lump- 
sucker. In Cottus scorpius, the brilliant markings in the male, resem- 
bling the wing of a tiger moth, are confined to the region posterior to the 
pectoral fin, and to the under surface of that fin, and can be concealed by 
it. There is one point to be noted, that this form of coloration is not 
limited to the surface in some forms. In the male lumpsucker, where the 
abdomen is coloured at the spawning time, the rosy hue gradually per- 
vades every tissue of the body — bones, cartilage, muscle, &c, and even 
tints the spermatic fluid. In Cottus bubalis, where the female is also 
brilliantly tinted, the internal parts are greenish. 
It is further noteworthy that in many of those forms in which the 
impregnation of the demersal ova may be accomplished usually by a single 
male, the ova themselves are often brightly tinted — (Agonus, Cyelopterus, 
Liparis, Salmo, Syngnathus, Gastrosteus, &c.) That this is not due 
merely to participation in a general colour change, such as occurs in the 
male lumpsucker, is shown by the fact that the female of the same species 
is quite devoid of bright coloration, and yet produces a mass of beautiful 
rosy ova. This coloration can scarcely be protective.* The pigmentation 
of the ova in such cases is probably connected with the sexual function of 
the male — either of impregnation or of guardianship. 
* In several cases, at least, the mass of pinkish ova of the lumpsucker begins to' 
lose this colour on the surface after deposition, and passes to greenish or brownish, a 
tint which may serve for protection, by simulation of the colour of seaweeds, &c. 
This change appears to be due to the action of light ; when a mass, which is green 
on the surface, is broken up, the interior may be found rosy ; and by exposing one 
side of a portion to the light, and keeping the other side in darkness, I found that 
the exposed side lost its rosy hue, which was retained by the darkened portion ; and 
even single ova showed the change on the opposite sides. 
