to have acquired the power of feeding on corals and medusæ ; ; and the beautiful bands and spots and bright colours with which they are frequently adorned, may be either protective when feeding in the submarine coral groves, or may, in some cases, be warning colours to show that they themselves are poisonous and uneatable.

A remarkable illustration of the wide extension of warning colours, and their very definite purpose in nature, is afforded by what may now be termed "Mr. Belt's frog." Frogs in all parts of the world are, usually, protectively coloured with greens or browns; and the little tree-frogs are either green like the leaves they rest upon, or curiously mottled to imitate bark or dead leaves. But there are a certain number of very gaily coloured frogs, and these do not conceal themselves as frogs usually do. Such was the small toad found by Darwin at Bahia Blanca, which was intense black and bright vermilion, and crawled about in the sunshine over dry sand-hills and arid plains. And in Nicaragua, Mr. Belt found a little frog gorgeously dressed in a livery of red and blue, which did not attempt concealment and was very abundant, a combination of characters which convinced him that it was uneatable. He, therefore, took a few specimens home with him and gave them to his fowls and ducks, but none would touch them. At last, by throwing down pieces of meat, for which there was a great competition among the poultry, he managed to entice a young duck into snatching up one of the little frogs. Instead of swallowing it, however, the duck instantly threw it out of its mouth, and went about jerking its head as if trying to get rid of some unpleasant taste.1

The power of predicting what will happen in a given case is always considered to be a crucial test of a true theory, and if so, the theory of warning colours, and with it that of mimicry, must be held to be well established. Among the creatures which probably have warning colours as a sign of inedibility are, the brilliantly coloured nudibranchiate molluscs, those curious annelids the Nereis and the Aphrodite or seamouse, and many other marine animals. The brilliant colours of the scallops (Pecten) and some other bivalve shells are perhaps

1 The Naturalist in Nicaragua, p. 321.

1

an indication of their hardness and consequent inedibility, as in the case of the hard beetles; and it is not improbable that some of the phosphorescent fishes and other marine organisms may, like the glow-worm, hold out their lamp as a warning to enemies. In Queensland there is an exceedingly poisonous spider (Latrodectus sp.), whose bite will kill a dog, and cause severe illness with excruciating pain in man. It is black, with a bright vermilion patch on the middle of the body; and it is so well recognised by this conspicuous coloration that even the spider-hunting wasps avoid it.2

Locusts and grasshoppers are generally of green protective tints, but there are many tropical species most gaudily decorated with red, blue, and black colours, and these are often inedible. Mr. Charles Horne states that one of the bright coloured Indian locusts was invariably rejected when offered to birds and lizards.3

The Rev. James Sibree informs me that while most of the Malagasy locusts are protectively coloured, the largest of them all is very brilliant-green, gold, bright blue and scarlet wings, all give it a gorgeous appearance; but it is so vilesmelling that it cannot be handled or kept in a cabinet, and is therefore avoided by all birds.

The examples now given lead us to the conclusion that colours acquired for the purpose of serving as a danger-signal to enemies are very widespread in nature, and, with the corresponding colours of the species which mimic them, furnish us with a rational explanation of a considerable portion of the coloration of animals which is outside the limits of those colours that have been acquired for either protection or recognition. There remains, however, another set of colours, chiefly among the higher animals, which, being connected with some of the most interesting problems in natural history, must be discussed in a separate chapter.

1 Mr. Belt first suggested this use of the light of the Lampyrida (fireflies and glow-worms)-Naturalist in Nicaragua, p. 320. Mr. Verrill and Professor Meldola made the same suggestion in the case of medusa and other phosphorescent marine organisms (Nature, vol. xxx. pp. 281, 289).

2 W. E. Armit, in Nature, vol. xviii. p. 642. Professor Herdman has also made numerous experiments showing that conspicuously coloured Nudibranchs are usually rejected by fishes (Trans. Biol. Soc. of Liverpool, vol. iv. p. 150). 3 Proc. Ent. Soc., 1869, p. xiii.

T

CHAPTER X

COLOURS AND ORNAMENTS CHARACTERISTIC OF SEX

Sex colours in the mollusca and crustacea-In insects-In butterflies and moths-Probable causes of these colours-Sexual selection as a supposed cause-Sexual coloration of birds-Cause of dull colours of female birds-Relation of sex colour to nesting habits-Sexual colours of other vertebrates-Sexual selection by the struggles of males— Sexual characters due to natural selection-Decorative plumage of males and its effect on the females-Display of decorative plumage by the males-A theory of animal coloration-The origin of accessory plumes-Development of accessory plumes and their display-The effect of female preference will be neutralised by natural selectionGeneral laws of animal coloration-Concluding remarks.

IN the preceding chapters we have dealt chiefly with the coloration of animals as distinctive of the several species; and we have seen that, in an enormous number of cases, the colours can be shown to have a definite purpose, and to be useful either as a means of protection or concealment, of warning to enemies, or of recognition by their own kind. We have now to consider a subordinate but very widespread phenomenon the differences of colour or of ornamental appendages in the two sexes. These differences are found to have special relations with the three classes of coloration above referred to, in many cases confirming the explanation already given of their purport and use, and furnishing us with important aid in formulating a general theory of animal coloration.

In comparing the colours of the two sexes we find a perfect gradation, from absolute identity of colour up to such extreme difference that it is difficult to believe that the two forms can belong to the same species; and this diversity in the

colours of the sexes does not bear any constant relation to affinity or systematic position. In both insects and birds we find examples of complete identity and extreme diversity of the sexes; and these differences occur sometimes in the same tribe or family, and sometimes even in the same genus.

It is only among the higher and more active animals that sexual differences of colour acquire any prominence. In the mollusca the two sexes, when separated, are always alike in colour, and only very rarely present slight differences in the form of the shell. In the extensive group of crustacea the two sexes as a rule are identical in colour, though there are often differences in the form of the prehensile organs; but in a very few cases there are differences of colour also. Thus, in a Brazilian species of shore-crab (Gelasimus) the female is grayish-brown, while in the male the posterior part of the cephalo-thorax is pure white, with the anterior part of a rich green. This colour is only acquired by the males when they become mature, and is liable to rapid change in a few minutes to dusky tints.1 In some of the fresh-water fleas (Daphnoida) the males are ornamented with red and blue spots, while in others similar colours occur in both sexes. spiders also, though as a rule the two sexes are alike in colour, there are a few exceptions, the males being ornamented with brilliant colours on the abdomen, while the female is dull coloured.

Sexual Coloration in Insects.

In

It is only when we come to the winged insects that we find any large amount of peculiarity in sexual coloration, and even here it is only developed in certain orders. Flies (Diptera), field-bugs (Hemiptera), cicadas (Homoptera), and the grasshoppers, locusts, and crickets (Orthoptera) present very few and unimportant sexual differences of colour; but the last two groups have special musical organs very fully developed in the males of some of the species, and these no doubt enable the sexes to discover and recognise each other. In some cases, however, when the female is protectively coloured, as in the well-known leaf-insects already referred to (p. 207), the male

1 Darwin's Descent of Man, p. 271.

is smaller and much less protectively formed and coloured. In the bees and wasps (Hymenoptera) it is also the rule that the sexes are alike in colour, though there are several cases among solitary bees where they differ; the female being black, and the male brown in Anthophora retusa, while in Andræna fulva the female is more brightly coloured than the male. Of the great order of beetles (Coleoptera) the same thing may be said. Though often so rich and varied in their colours the sexes are usually alike, and Mr. Darwin was only able to find about a dozen cases in which there was any conspicuous difference between them. They exhibit, however, numerous sexual characters, in the length of the antennæ, and in horns, legs, or jaws remarkably enlarged or curiously modified in the male sex.

It is in the family of dragonflies (order Neuroptera) that we first meet with numerous cases of distinctive sexual coloration. In some of the Agrionidæ the males have the bodies rich blue and the wings black, while the females have the bodies green and the wings transparent. In the North American genus Hetærina the males alone have a carmine spot at the base of each wing; but in some other genera the sexes hardly differ at all.

The great order of Lepidoptera, including the butterflies and moths, affords us the most numerous and striking examples of diversity of sexual colouring. Among the moths the difference is usually but slight, being manifested in a greater intensity of the colour of the smaller winged male; but in a few cases there is a decided difference, as in the ghost-moth (Hepialus humuli), in which the male is pure white, while the female is yellow with darker markings. This may be a recognition colour, enabling the female more readily to discover her mate; and this view receives some support from the fact that in the Shetland Islands the male is almost as yellow as the female, since it has been suggested that at midsummer, when this moth appears, there is in that high latitude sufficient twilight all night to render any special coloration unnecessary.2

Butterflies present us with a wonderful amount of sexual

1 Darwin's Descent of Man, p. 294, and footnote.

2 Nature, 1871, p. 489.