are mostly fertilized by night-flying moths, and those which reserve their odours for the evening probably escape the visits of diurnal insects which would consume their nectar without effecting fertilization. The absence of odour in showy flowers and its preponderance among those that are white, may be shown to be a fact by an examination of the lists in Mr. Mongredien's work on hardy trees and shrubs. He gives a list of about one hundred and sixty species with showy flowers, and another list of sixty species with fragrant flowers; but only twenty of these latter are included among the showy species, and these are almost all white flowered. Of the sixty species with fragrant flowers, more than forty are white, and a number of others have greenish, yellowish, or dusky and inconspicuous flowers. The relation of

white flowers to nocturnal insects is also well shown by those which, like the evening primroses, only open their large white blossoms after sunset. The red Martagon lily has been observed by Mr. Herman Müller to be fertilized by the humming - bird hawk moth, which flies in the morning and afternoon when the colours of this flower, exposed to the nearly horizontal rays of the sun, glow with brilliancy, and when it also becomes very sweetscented.

To the same need of conspicuousness the combination of so many individually small flowers into heads and bunches is probably due, producing such broad masses as those of the elder, the gueldre-rose, and most of the Umbelliferæ, or such elegant bunches as those of the lilac, laburnum, horse-chestnut, and wistaria. In other cases minute flowers are gathered into dense heads, as with Globularia, Jasione, clover, and all the Composite; and among the latter the outer flowers are often developed into a ray, as in the sunflowers, the daisies, and the asters, forming a starlike compound flower, which is

1 Trees and Shrubs for English Plantations, by Augustus Mongredien. Murray, 1870.

itself often produced in immense profusion.

The beauty of alpine flowers is almost proverbial. It consists either in the increased size of the individual flowers as compared with the whole plant, in increased intensity of colour, or in the massing of small flowers into dense cushions of bright colour; and it is only in the higher Alps, above the limit of forests and upwards towards the perpetual snow-line that these characteristics are fully exhibited. This effort at conspicuousness under adverse circumstances may be traced to the comparative scarcity of winged insects in the higher regions, and to the necessity for attracting them from a distance. Amid the vast slopes of debris and the huge masses of rock so prevalent in higher mountain regions, patches of intense colour can alone make themselves visible and serve to attract the wandering butterfly from the valleys. Mr. Herman Müller's careful observations have shown, that in the higher Alps bees and most other groups of winged insects are almost wanting, while butterflies are tolerably abundant; and he has discovered, that in a number of cases where a lowland flower is adapted to be fertilized by bees, its alpine ally has had its structure so modified as to be adapted for fertilization only by butterflies. But bees are always (in the temperate zone) far more abundant than butterflies, and this will be another reason why flowers specially adapted to be fertilized by the latter should be rendered unusually conspicuous. We find, accordingly, the yellow primrose of the plains replaced by pink and magenta-coloured alpine species; the straggling wild pinks of the lowlands by the masses of large flowers in such mountain species as Dianthus alpinus and D. glacialis; the saxifrages of the high Alps with bunches of flowers a foot long, as in Saxifraga longifolia and S. cotyledon, or forming spreading masses of flowers, as in S. oppositifolia; while the soap2 Nature, vol. xi. pp. 32, 110.

worts, silenes, and louseworts are equally superior to the allied species of the plains.

Again, Dr. Müller has discovered that when there are showy and inconspicuous species in the same genus of plants, there is often a corresponding difference of structure, those with large and showy flowers being quite incapable of self-fertilization, and thus depending for their very existence on the visits of insects; while the others are able to fertilize themselves should insects fail to visit them. We have examples of this difference in Malva sylvestris, Epilobium angustifolium, Polygonum bistorta, and Geranium pratense— which have all large or showy flowers and must be fertilized by insects,-as compared with Malva rotundifolia, Epilobium parviflorum, Polygonum aviculare, and Geranium pusillum, which have small or inconspicuous flowers, and are so constructed that if insects should not visit them they are able to fertilize themselves.1

As supplementing these curious facts showing the relation of colour in flowers to the need of the visits of insects to fertilize them, we have the remarkable, and on any other theory utterly inexplicable circumstance, that in all the numerous cases in which plants are fertilized by the agency of the wind they never have specially coloured floral envelopes. Such are our pines, oaks, poplars, willows, beeches, and hazel; our nettles, grasses, sedges, and many others. In some of these the male flowers are, it is true, conspicuous, as in the catkins of the willows and the hazel, but this arises incidentally from the masses of pollen necessary to secure fertilization, as shown by the entire absence of a corolla or of those coloured bracts which so often add to the beauty and conspicuousness of true flowers.

The adaptation of flowers to be fertilized by insects-often to such an extent that the very existence of the species depends upon it-has had widespread influence on the distribution of plants and the general aspects of vege

1 Nature, vol. ix. p. 164.

2

tation. The seeds of a particular species may be carried to another country, may find there a suitable soil and climate. may grow and produce flowers, but if the insect which alone can fertilize it should not inhabit that country, the plant cannot maintain itself, however frequently it may be introduced or however vigorously it may grow. Thus may probably be explained the poverty in flowering-plants and the great preponderance of ferns that distinguishes many oceanic islands, as well as the deficiency of gaily-coloured flowers in others. This branch of the subject is discussed at some length in my Address to the Biological Section of the British Association, but I may here just allude to two of the most striking cases. New Zealand is, in proportion to its total number of flowering plants, exceedingly poor in handsome flowers, and it is correspondingly poor in insects, especially in bees and butterflies, the two groups which so greatly aid in fertilization. In both these aspects it contrasts strongly with Southern Australia and Tasmania in the same latitudes, where there is a profusion of gaily-coloured flowers and an exceedingly rich insect-fauna. The other case is presented by the Galapagos islands, which, though situated on the equator off the west coast of South America, and with a tolerably luxuriant vegetation in the damp mountain zone, yet produce hardly a single conspicuously-coloured flower; and this is correlated with, and no doubt dependent on, an extreme poverty of insect life, not one bee and only a single butterfly having been found there.

Again, there is reason to believe that some portion of the large size and corresponding showiness of tropical flowers is due to their being fertilized by very large insects and even by birds. Tropical sphinx-moths often have their probosces nine or ten inches long, and we find flowers whose tubes or spurs reach about the same length; while the giant bees, and the 2 See Nature, September 6th, 1876.

numerous flower-sucking birds, aid in the fertilization of flowers whose corollas or stamens are proportion ately large.

I have now concluded this sketch of the general phenomena of colour in the organic world. I have shown reasons for believing that its presence, in some of its infinitely-varied hues, is more probable than its absence, and that variation of colour is an almost necessary concomitant of variation of structure, of development, and of growth. It has also been shown how colour has been appropriated and modified both in the animal and vegetable world, for the advantage of the species in a great variety of ways, and that there is no need to call in the aid of

any other laws than those of organic development and "natural selection" to explain its countless modifications. From the point of view here taken it seems at once improbable and unnecessary that the lower animals should have the same delicate appreciation of the infinite variety and beauty-of the delicate contrasts and subtle harmonies of colour-which are possessed by the more intellectual races of mankind, since even the lower human races do not possess it. All that seems required in the case of animals, is a perception of distinctness or contrast of colours; and the dislike of so many creatures to scarlet may perhaps be due to the rarity of that colour in nature, and to the glaring contrast it offers to the sober greens and browns which form the general clothing of the earth's surface.

The general view of the subject now given must convince us that, so far from colour being-as it has sometimes been thought to be- unimportant, it is intimately connected with the very existence of a large proportion of the species of the animal and vegetable worlds. The gay colours of the butterfly and of the alpine flower which it unconsciously fertilises while seeking for its secreted honey, are each beneficial to its possessor, and have

been shown to be dependent on the same class of general laws as those which have determined the form, the structure, and the habits of every living thing. The complex laws and unexpected relations which we have seen to be involved in the production of the special colours of flower, bird, and insect, must give them an additional interest for every thoughtful mind; while the knowledge that, in all probability, each style of coloration, and sometimes the smallest details, have a meaning and a use, must add a new charm to the study of nature.

Throughout the preceding discussion we have accepted the subjective phenomena of colour--that is, our perception of varied hues, and the mental emotions excited by them-as ultimate facts needing no explanation. Yet they present certain features well worthy of attention, a brief consideration of which will form a fitting sequel to the present essay.

The perception of colour seems, to the present writer, the most wonderful and the most mysterious of our sensations. Its extreme diversities and exquisite beauties seem out of proportion to the causes that are supposed to have produced them, or the physical needs to which they minister. If we look at pure tints of red, green, blue, and yellow, they appear so absolutely contrasted and unlike each other, that it is almost impossible to believe (what we nevertheless know to be the fact) that the rays of light producing these very distinct sensations differ only in wave-length and rate of vibration; and that there is from one to the other a continuous series and gradation of such vibrating waves. The positive diversity we see in them must then depend upon special adaptations in ourselves; and the question arises for what purpose have our visual organs and mental perceptions become so highly specialised in this respect When the sense of sight was first developed in the animal kingdom, we can hardly doubt that

what was perceived was light only, and its more or less complete withdrawal. As the sense became perfected, more delicate gradations of light and shade would be perceived; and there seems no reason why a visual capacity might not have been developed as perfect as our own, or even more so, in respect of light and shade, but entirely insensible to differences of colour, except in so far as these implied a difference in the quantity of light. The world would in that case appear somewhat as we see it in good stereoscopic photographs; and we all know how exquisitely beautiful such pictures are, and how completely they give us all requisite information as to form, surface-texture, solidity, and distance, and even to some extent as to colour; for almost all colours are distinguishable in a photograph by some differences of tint, and it is quite conceivable that visual organs might exist which would differentiate what we term colour by delicate gradations of some one cha

racteristic neutral tint. Now such a capacity of vision would be simple as compared with that which we actually possess; which, besides distinguishing infinite gradations of the quantity of light, distinguishes also, by a totally distinct set of sensations, gradations of quality, as determined by differences of wave lengths or rate of vibration. At what grade in animal development this new and more complex sense first began to appear we have no means of determining. The fact that the higher vertebrates, and even some insects, distinguish what are to us diversities of colour, by no means proves that their sensations of colour bear any resemblance whatever to ours. An insect's capacity to distinguish red from blue or yellow may be (and probably is) due to perceptions of a totally distinct nature, and quite unaccompanied by any of that sense of enjoyment or even of radical distinctness which pure colours excite in us. Mammalia and birds, whose structure and emotions are so similar to our

own, do probably receive somewhat similar impressions of colour; but we have no evidence to show that they experience pleasurable emotions from colour itself when not associated with the satisfaction of their wants or the gratification of their passions.

The primary necessity which led to the development of the sense of colour, was probably the need of distinguishing objects much alike in form and size, but differing in important properties;such as ripe and unripe, or eatable and poisonous fruits; flowers with honey or without; the sexes of the same or of closely-allied species. In most cases the strongest contrast would be the most useful, especially as the colours of the objects to be distinguished would form but minute spots or points when compared with the broad masses of tint of sky, earth, or foliage against which they would be set. Throughout the long epochs in which the sense of sight was being gradually developed in the higher animals, their visual organs would be mainly subjected to two groups of rays-the green from vegetation, and the blue from the sky. The immense preponderance of these over all other groups of rays would naturally lead the eye to become specially adapted for their perception; and it is quite possible that at first these were the only kinds of lightvibrations which could be perceived at all. When the need for differentiation of colour arose, rays of greater and of smaller wave-lengths would necessarily be made use of to excite the new sensations required; and we can thus understand why green and blue form the central portion of the visible spectrum, and are the colours which are most agreeable to us in large surfaces; while at its two extremities we find yellow, red, and violet, colours which we best appreciate in smaller masses, and when contrasted with the other two or with light neutral tints. We have here probably the foundations of a natural theory of harmoni ous colouring, derived from the order in which our colour-sensations have

arisen, and the nature of the emotions with which the several tints have been always associated.1 The agreeable and soothing influence of green light may be in part due to the green rays having little heating power; but this can hardly be the chief cause, for the blue and violet, though they contain less heat, are not generally felt to be so cool and sedative. But when we consider how dependent are all the higher

1 There is reason to believe that our capacity of distinguishing colours has increased even in historical times. The subject has attracted the attention of German philologists, and I have been furnished by a friend with some notes from a work of the late Lazarus Geiger, entitled, Zur Entwickelungsgeschichte der Menschheit (Stuttgart, 1871). According to this writer it appears that the colour of grass and foliage is never alluded to as a beauty in the Vedas or the Zendavesta, though these productions are continually extolled for other properties. Blue is described by terms denoting sometimes green, sometimes black, showing that it was hardly recognised as a distinct colour. The colour of the sky is never mentioned in the Bible, the Vedas, the Homeric poems, or even in the Koran. The first distinct allusion to it known to Geiger is in an Arabic work of the ninth century. "Hyacinthine locks' are black locks, and Homer calls iron "violet-coloured." Yellow was often confounded with green, but, along with red, it was one of the earliest colours to receive a distinct name. Aristotle names three colours in the rainbow-red, yellow, and green. Two centuries earlier Xenophanes had described the rainbow as purple, reddish, and yellow. The Pythagoreans admitted four primary colours-white, black, red, and yellow; the Chinese the same, with the addition of green. If these statements fairly represent the early condition of colour-sensation they well accord with the view here maintained, that green and blue were first alone perceived, and that the other colours were successively separated from them.

These

latter would be the first to receive names; hence we find purple, reddish, and yellow, first noticed in the rainbow as the tints to be separated from the widespread blue and green of the visible world which required no distinctive colour-appellation. If the capacity of distinguishing colours has increased in historic times, we may perhaps look upon colour-blindness as a survival of a condition once almost universal; while the fact that it is still so prevalent is in harmony with the view that our present high perception and appreciation of colour is a comparatively recent acquisition, and may be correlated with a general advance in mental activity.

animals on vegetation, and that man himself has been developed in the closest relation to it, we shall find, probably, a sufficient explanation. The green mantle with which the earth is overspread caused this one colour to predominate over all others that meet our sight, and to be almost always associated with the satisfaction of human wants. Where the grass is greenest, and vegetation most abundant and varied, there has man always found his most suitable dwelling-place. In such spots hunger and thirst are unknown, and the choicest productions of nature gratify the appetite and please the eye. In the greatest heats of summer, coolness, shade, and mois-. ture are found in the green forest glades; and we can thus understand how our visual apparatus has become especially adapted to receive pleasur able and soothing sensations from this class of rays.

The preceding considerations enable us to comprehend, both why a perception of difference of colour has become developed in the higher animals, and also why colours require to be presented or combined in varying proportions in order to be agreeable to us. But they hardly seem to afford a sufficient explanation, either of the wonderful contrasts and total unlikeness of the sensations produced in us by the chief primary colours, or of the exquisite charm and pleasure we derive from colour itself, as distinguished from variously coloured objects, in the case of which association of ideas comes into play. It is hardly conceivable that the material uses of colour to animals and to ourselves required such very distinct and powerfullycontrasted sensations; and it is still less conceivable that a sense of delight in colour per se should have been necessary for our utilization of it.

The emotions excited by colour and by music, alike, seem to rise above the level of a world developed on purely utilitarian principles.

ALFRED R. WALLACE.