242 
Part III. — Tenth Annual Report 
material produced by a given proportion of the males among fish with 
demersal ova be relatively so much greater than among fish with pelagic 
ova % The duration of the spawning period in a given individual in the 
two cases will not explain it. Demersal ova may be emitted in successive 
crops, and the number emitted at one time is probably as great with 
pelagic ova. The maturation of the spermatozoa is, no doubt, much more 
rapid than the maturation of the ova ; but this argument would apply to 
both cases. It might be supposed that the mode or conditions of repro- 
duction would furnish a clue ; fish with demersal ova having generally a 
closer individual sexual relation than fish with pelagic ova. Neverthe- 
less, considerations of monogamy, mixogamy, &c, do not explain the 
matter ; for the herring and sand-eel — in which the preponderance of the 
male element is marked— are quite as mixogamous as the plaice or the 
gurnard. The difference is probably due to the spermatic fluid having a 
specific gravity less than that of the sea-water. With , pelagic ova pos- 
sessing an inferior micropyle, and passing through many fathoms of water 
to the surface, fertilisation by the upwardly-moving sperms will probably 
entail less loss of the latter than in the fertilisation of demersal ova lying 
on the bottom, and sometimes aggregated in masses. In the one case 
the action of gravity goes hand in hand with the intrinsic movement of 
the sperms. But in the other case each sperm must, so to speak, fight its 
way to the micropyle of the demersal ovum against physical opposition. 
5. Sexual Coloration. 
Some observations have been made on this subject which may be briefly 
referred to here. The coloration of many fishes may be, in a wide manner, 
classified according to three regions — the ventral, dorsal, and lateral. 
The dorsal colours are nearly always protective — a fact noticeable as much 
in pelagic, and freely roaming forms, as among flat fishes and shore forms. 
The ventral colour, or rather, absence of colour, is also protective in the 
majority of cases, in function, if not in origin, as may be observed in 
looking up at, say, a whiting or a herring poised in the water. The 
absence of colour may be due to the absence of light ; and Cunningham's 
important observations on the production of pigment on the lower surface 
of young flat fishes by lighting them from below, are of much interest in 
this connection. The contrary does not hold good, for the young Zoarces 
extracted from the mother, or at the period of birth, have. well-marked 
characteristic pigmentation dorsally and laterally, although they were pre- 
viously cut off from the influence of light. Ventral coloration may, how- 
ever, in some cases be sexual or other than protective. Lateral coloration 
may be sexual, although in several cases such coloration invades both 
the dorsal and the ventral surfaces. On the other hand, lateral colora- 
tion appears in some cases, at least, to have relation to the 'instinct' of 
shoaling, as among herring and mackerel. For instance, if a herring is 
held by the head and tail, and looked down upon, the coloration is dull 
and protective, as it is gradually raised, opposite the light, and the late- 
ral surface comes into view, the characteristic iridiscence catches the eye 
at a certain angle, and the silvery sheen ; but as the fish is moved upwards 
the iridiscence is lost, and, when viewed from below, the dullish white pro- 
tective appearance is only visible. Examination of males and females 
revealed no distinction in coloration. So with the mackerel, where the 
lateral banding and sheen are distinct ; with the sprat, the sand-eel, the 
smelt, the salmon, &c. From the fact that fish are, in this connection, 
largely cut off from the guidance of the sense which is so useful to herd- 
