objects to which they are attached; for instance, it was shown that the pupa of the small tortoise-shell butterfly, when placed on a dark background, became itself very dark; but when placed on a white background it became light-coloured. This susceptibility to change of colour is greatest at the stage when the larva first fixes itself before changing to the pupa. The effect is produced through the skin generally, and not through the eyes. Again, trout change colour according to that of the bottom of the stream they inhabit; so also do minnows; hence the interior of a minnow-can is painted white, so that the bait may be light-coloured, and more conspicuous to a pike or perch. Changes of colour depend on the eye in the case of fish, and a blind trout remains dark, the pigment cells relaxing and becoming flattened, thereby exposing their maximum amount of surface area. Cave animals, on the other hand, become pale, because the pigment which is now useless degenerates and disappears. The effect of cold in causing change of colour indirectly through the nervous system, has been demonstrated by suddenly exposing animals to cold which had previously been protected from it for some time, the result being distinct blanching. Melanism, or dark coloration, is common on oceanic islands, and humidity of the atmosphere is as a rule associated with the darkening of colours. So it is with increased elevation, and it is possible that the object of this is to increase the absorption of heat. Brilliant colours are not dependent on or proportionate to the amount of light. The effect of environment in causing changes of colour is well shown in cases of what is known as seasonal dimorphism, where animals produce two broods in each year, each of different appearance with regard to colouring, and each capable of producing the other. A good example of this is found in the two continental butterflies Vanessa prorsa and Vanessa levana. Vanessa levana, the spring form, has a red ground colour, with black spots and dashes, and a row of blue spots round the margin of the hind wings; Vanessa prorsa, the summer form, is deep black, with a broad yellowish-white band across both wings, and with no blue spots. These were formerly called distinct species, but have recently been shown to be varieties of one and the same species. That these differences are due to the direct action of cold and heat has been shown by keeping the pupæ of levana at a low temperature. These, which would ordinarily have produced the summer form prorsa, hatched, under the altered conditions of temperature, partly as levana and partly as an intermediate form. Again, very young Canaries change their colour to orange when given Cayenne pepper, and certain parrots have been shown to change their colour when fed on the fat of a particular fish. It is very difficult to draw the line between the direct action of environment through the nervous system, and the action of Natural Selection; for to which can we attribute the whitening of Arctic animals ? We have now to consider the great mass of cases illustrating the preservation and accentuation of colour through the agency of Natural Selectionone of the most striking of the later developments of the theory. SIGNIFICANT COLOURS. These are colours which are of direct advantage to their possessor as colour, and not merely because they are associated with other properties which are useful, as in the case of hæmoglobin and chlorophyll. The classification of these colours is a matter of some difficulty, for cross-relations occur which are difficult to express. There are three chief classes or groups :1. Apatetic; the purpose of which, or rather the object gained by which, is to hinder recognition by other animals. 2. Sematic, or signalling colours; the purpose of which is to facilitate or aid recognition by animals of the same or of other kinds. 3. Epigamic; which include those cases in which differences occur between the male and female sex, as in the peacock and pea-hen, the duck and drake, &c. This is a special and important group. APATETIC COLOURS are again divided into : a. Protective resemblances, aiding escape from enemies, as in those cases where animals resemble sticks or plants, and so escape notice. b. Aggressive resemblances; the purpose of which is to aid the approach to prey; for example the resemblance of the colour of the lion to that of the desert. c. Alluring resemblances; constituting a small group of cases, in which an animal acts as a bait by taking on the form of something attractive to its prey. SEMATIC COLOURS have two subdivisions : a. Warning colours. These constitute a curious group of cases, in which animals have bright conspicuous colours, for the purpose of warning other animals off them, and which are signs of inedibility or of the possession of dangerous powers of attack. b. Recognition colours. These are for the purpose of easy recognition by animals of the same kind; and are best seen in the cases of gregarious animals, such as deer, whose safety largely depends on association and mutual defence. PROTECTIVE RESEMBLANCES. Such forms of protective colouring as aid the escape from enemies by hindering recognition may be one of two kinds : (1) General; in which the colouring is such as to assimilate the animal to its environment, and so render it less conspicuous, as in the case of the whiteness of Arctic animals, such as the Polar bear; and the sandy colour of desert animals, or the transparent blueness of pelagic forms. (2) Special; where the resemblance is to some particular object, and where the animal escapes, not through being concealed from view and so over looked, but through being mistaken for something else. Of these cases some extraordinary instances are known. The resemblance may be to another animal or to a plant, flower, or leaf, or to inorganic sub stances. Furthermore, the protective colouring may be either constant or variable; a good example of variable protective colouring being shown by the Octopus and Chameleon. Again, in animals such as insects, which undergo metamorphosis, and in which the form, structure, and habits are widely different in the larval and adult stages respectively, both these stages may be protectively coloured, but the resemblance will be to entirely different objects. Let us take examples from the different groups of animals, and we shall see that the reality of protective colouring is impossible to doubt. MAMMALS.-The whiteness of Arctic animals has already been referred to. The American polar bear is white all the year round; the ermine or stoat changes to white in the winter, and the Arctic fox usually does this also. The Alpine hare always becomes white in the winter in Scandinavia, and usually in Scotland, although rarely so in Ireland. This change consists in an actual blanching of the hairs from the tips inwards, with a new growth of additional white hairs. The general tawny colour of deer is also protective; the protection afforded by spots is seen by their resemblance to the circular spots of light caused by sunlight passing through the leaves of a wood, while stripes facilitate escape in long grass or reeds. |