It is impossible to doubt that colour has been gained by many fishes as a protection: no one can examine the speckled upper surface of a flounder, and overlook its resemblance to the sandy bed of the sea on which it lives. Certain fishes, moreover, can through the action of the nervous system change their colours in adaptation to surrounding objects, and that within a short time. (32. G. Pouchet, 'L'Institut.' Nov. 1, 1871, p. 134.) One of the most striking instances ever recorded of an animal being protected by its colour (as far as it can be judged of in preserved specimens), as well as by its form, is that given by Dr. Gunther (33. 'Proc. Zoolog. Soc.' 1865, p. 327, pl. xiv. and xv.) of a pipe-fish, which, with its reddish streaming filaments, is hardly distinguishable from the sea-weed to which it clings with its prehensile tail.
But the question now under consideration is whether the females alone have been modified for this object. We can see that one sex will not be modified through natural selection for the sake of protection more than the other, supposing both to vary, unless one sex is exposed for a longer period to danger, or has less power of escaping from such danger than the other; and it does not appear that with fishes the sexes differ in these respects. As far as there is any difference, the males, from being generally smaller and from wandering more about, are exposed to greater danger than the females; and yet, when the sexes differ, the males are almost always the more conspicuously coloured.
The ova are fertilised immediately after being deposited; and when this process lasts for several days, as in the case of the salmon (34. Yarrell, 'British Fishes,' vol. ii. p. 11.), the female, during the whole time, is attended by the male. After the ova are fertilised they are, in most cases, left unprotected by both parents, so that the males and females, as far as oviposition is concerned, are equally exposed to danger, and both are equally important for the production of fertile ova; consequently the more or less brightly-coloured individuals of either sex would be equally liable to be destroyed or preserved, and both would have an equal influence on the colours of their offspring.
Certain fishes, belonging to several families, make nests, and some of them take care of their young when hatched. Both sexes of the bright coloured Crenilabrus massa and melops work together in building their nests with sea-weed, shells, etc. (35. According to the observations of M. Gerbe; see Gunther's 'Record of Zoolog. Literature,' 1865, p. 194.) But the males of certain fishes do all the work, and afterwards take exclusive charge of the young. This is the case with the dull-coloured gobies (36. Cuvier, 'Regne Animal,' vol. ii. 1829, p. 242.), in which the sexes are not known to differ in colour, and likewise with the sticklebacks (Gasterosteus), in which the males become brilliantly coloured during the spawning season.
The male of the smooth-tailed stickleback (G. leiurus) performs the duties of a nurse with exemplary care and vigilance during a long time, and is continually employed in gently leading back the young to the nest, when they stray too far. He courageously drives away all enemies including the females of his own species. It would indeed be no small relief to the male, if the female, after depositing her eggs, were immediately devoured by some enemy, for he is forced incessantly to drive her from the nest. (37. See Mr. Warington's most interesting description of the habits of the Gasterosteus leiurus in 'Annals and Magazine of Nat. History,' November 1855.)