species of small fruit-pigeons (Ptilopus) each usually found in a single island or small group of islands in the Pacific. After describing the many peculiarities of colour which characterise these insular species, he says:

Consequently it appears certain that most of these species were developed singly, each on its own island. If this be the case, the colours which now distinguish the different species cannot be recognition-marks, because there is no other species in each island with which they could be confounded."

The fallacy of this argument, which was repeated by the late Dr. Mivart in the case of the beautifully coloured lories of the Papuan and Pacific islands, is obvious on a little consideration. For if these writers had asked themselves the question-Developed from what? they would have seen that each of these distinct species must have been developed from some other species which had accidentally reached the island. During the process of modification which was required to bring the old species into harmony with the new conditions of its new home there would necessarily be a time when there were two forms living side by side-the immigrant species and the partially modified form or incipient new species which was to replace it; and in order that the process of modification and adaptation to new conditions should be carried on and completed it is universally admitted that the intercrossing of the two forms must in some way be checked. Now some distinctive marking enabling each kind to know its fellows is one of the best ways of effecting this, and therefore all suitable variations of colour and marking, which were not otherwise injurious, would be preserved and gradually intensified in the very process of development of the new species. When this was completed, after perhaps hundreds or thousands of years, the original species, being less adapted to the conditions of existence, would be crowded out, and the modified form alone remain, preserving those peculiarities of colour or marking which now seem useless, but which were really one of the essentials to its development.

It thus appears that insular species must have been formed in exactly the same way as all other species, and that the distinctive characters developed were essential to the complete differentiation of all species. Here too we find a rational explanation of those diversities of ornament in the males of so many birds and butterflies, which, beginning as mere recognition-marks, became sometimes greatly developed through agencies referred to in our tenth chapter,

CHAPTER IX

WARNING COLORATION AND MIMICRY

The skunk as an example of warning coloration-Warning colours among insects—Butterflies-Caterpillars-Mimicry-How mimicry has been produced--Heliconida-Perfection of the imitation-Other cases of mimicry among Lepidoptera-Mimicry among protected groups-Its explanation-Extension of the principle—Mimicry in other orders of insects-Mimicry among the vertebrata-Snakes-The rattlesnake and the cobra-Mimicry among birds-Objections to the theory of mimicry-Concluding remarks on warning colours and mimicry.

WE have now to deal with a class of colours which are the very opposite of those we have hitherto considered, since, instead of serving to conceal the animals that possess them or as recognition marks to their associates, they are developed for the express purpose of rendering the species conspicuous. The reason of this is that the animals in question are either the possessors of some deadly weapons, as stings or poison fangs, or they are uneatable, and are thus so disagreeable to the usual enemies of their kind that they are never attacked when their peculiar powers or properties are known. It is, therefore, important that they should not be mistaken for defenceless or eatable species of the same class or order, since in that case they might suffer injury, or even death, before their enemies discovered the danger or the uselessness of the attack. They require some signal or danger-flag which shall serve as a warning to would-be enemies not to attack them, and they have usually obtained this in the form of conspicuous or brilliant coloration, very distinct from the protective tints of the defenceless animals allied to them.

The Skunk as illustrating Warning Coloration.

While staying a few days, in July 1887, at the Summit. Hotel on the Central Pacific Railway, I strolled out one evening after dinner, and on the road, not fifty yards from the house, I saw a pretty little white and black animal with a bushy tail coming towards me. As it came on at a slow pace and without any fear, although it evidently saw me, I thought at first that it must be some tame creature, when it suddenly occurred to me that it was a skunk. It came on till within five or six yards of me, then quietly climbed over a dwarf wall and disappeared under a small outhouse, in search of chickens, as the landlord afterwards told me. This animal possesses, as is well known, a most offensive secretion, which it has the power of ejecting over its enemies, and which effectually protects it from attack. The odour of this substance is so penetrating that it taints, and renders useless, everything it touches, or in its vicinity. Provisions near it become uneatable, and clothes saturated with it will retain the smell for several weeks, even though they are repeatedly washed and dried. A drop of the liquid in the eyes will cause blindness, and Indians are said not unfrequently to lose their sight from this cause. Owing to this remarkable power of offence the skunk is rarely attacked by other animals, and its black and white. fur, and the bushy white tail carried erect when disturbed, form the danger-signals by which it is easily distinguished in the twilight or moonlight from unprotected animals. consciousness that it needs only to be seen to be avoided gives it that slowness of motion and fearlessness of aspect which are, as we shall see, characteristic of most creatures so protected.

Warning Colours among Insects.

Its

It is among insects that warning colours are best developed, and most abundant. We all know how well marked and conspicuous are the colours and forms of the stinging wasps and bees, no one of which in any part of the world is known to be protectively coloured like the majority of defenceless insects. Most of the great tribe of Malacoderms among beetles are distasteful to insect-eating animals. Our red and

black Telephorida, commonly called "soldiers and sailors," were found, by Mr. Jenner Weir, to be refused by small birds. These and the allied Lampyrida (the fire-flies and glow-worms) in Nicaragua, were rejected by Mr. Belt's tame monkey and by his fowls, though most other insects were greedily eaten by them. The Coccinellidae or lady-birds are another uneatable group, and their conspicuous and singularly spotted bodies serve to distinguish them at a glance from all other beetles.

These uneatable insects are probably more numerous than is supposed, although we already know immense numbers that are so protected. The most remarkable are the three families of butterflies-Heliconidæ, Danaida, and Acræidacomprising more than a thousand species, and characteristic respectively of the three great tropical regions-South America, Southern Asia, and Africa. All these butterflies have peculiarities which serve to distinguish them from every other group in their respective regions. They all have ample but rather weak wings, and fly slowly; they are always very abundant; and they all have conspicuous colours or markings, so distinct from those of other families that, in conjunction with their peculiar outline and mode of flight, they can usually be recognised at a glance. Other distinctive features are, that their colours are always nearly the same on the under surface of their wings as on the upper; they never try to conceal themselves, but rest on the upper surfaces of leaves or flowers; and, lastly, they all have juices which exhale a powerful scent, so that when one kills them by pinching the body, the liquid that exudes stains the fingers yellow, and leaves an odour that can only be removed by repeated washings.

Now, there is much direct evidence to show that this odour, though not very offensive to us, is so to most insecteating creatures. Mr. Bates observed that, when set out to dry, specimens of Heliconidae were less subject to the attacks of vermin; while both he and I noticed that they were not attacked by insect-eating birds or dragonflies, and that their wings were not found in the forest paths among the numerous wings of other butterflies whose bodies had been devoured. Mr. Belt once observed a pair of birds capturing insects for

their young; and although the Heliconidæ swarmed in the vicinity, and from their slow flight could have been easily caught, not one was ever pursued, although other butterflies did not escape. His tame monkey also, which would greedily munch up other butterflies, would never eat the Heliconidæ. It would sometimes smell them, but always rolled them up in its hand and then dropped them.

We have also some corresponding evidence as to the distastefulness of the Eastern Danaidæ. The Hon. Mr. Justice Newton, who assiduously collected and took notes upon the Lepidoptera of Bombay, informed Mr. Butler of the British Museum that the large and swift-flying butterfly Charaxes psaphon, was continually persecuted by the bulbul, so that he rarely caught a specimen of this species which had not a piece snipped out of the hind wings. He offered one to a bulbul which he had in a cage, and it was greedily devoured, whilst it was only by repeated persecution that he succeeded in inducing the bird to touch a Danais.1

Besides these three families of butterflies, there are certain groups of the great genus Papilio-the true swallow-tailed butterflies-which have all the characteristics of uneatable insects. They have a special coloration, usually red and black (at least in the females), they fly slowly, they are very abundant, and they possess a peculiar odour somewhat like that of the Heliconidæ. One of these groups is common in tropical America, another in tropical Asia, and it is curious that, although not very closely allied, they have each the same red and black colours, and are very distinct from all the other butterflies of their respective countries. There is reason to believe also that many of the brilliantly coloured and weakflying diurnal moths, like the fine tropical Agaristidæ and burnet-moths, are similarly protected, and that their conspicuous colours serve as a warning of inedibility. common burnet-moth (Anthrocera filipendula) and the equally conspicuous ragwort-moth (Euchelia jacobea) have been proved to be distasteful to insect-eating creatures.

1 Nature, vol. iii. p. 165.

The

Professor Meldola observed that specimens of Danais and Euplæa in collections were less subject to the attacks of mites (Proc. Ent. Soc., 1877, p. xii.); and this was corroborated by Mr. Jenner Weir. Entomologist, 1882, vol. xv. p. 160.

R