Colour is not a material existence; it is a sensation. The colour of an object varies slightly with the brilliancy of the light emanating from it to the eye; and where the eye is abnormal, as in the 'colour-blind,' the apparent colours of objects may differ widely from their colour as perceived by normal eyes. Light is due to waves—or other periodic disturbances whose recurrence resembles that of waves—in the ether of space; and just as air-waves of a certain definite frequency of recurrence will induce in the ear the sensation of a sound of a particular pitch, so will the impact of 'ether-waves' of a certain particular frequency induce in the eye a sensation of light of a particular colour. We are experimentally acquainted with ether-waves whose frequencies range between 20,000,000,000,000 and 40,000,000,000,000,000 per second; but the eye is blind to all except a comparatively narrow range of these—viz. from 392 billion to 757 billion per second. Within that comparatively small range, however, we have a large choice of fractional and integral numbers; and each number, each frequency, has its own colour. When we look at the spectrum or rainbow we have marshalled before us a series of colours, of which the extreme visible red is produced by about 392 billion, the extreme visible violet by about 757 billion vibrations per second. Between these the eye may rest upon certain distinctive colours, such as yellow, blue, and so on; and the frequencies corresponding to these respective colours are, taking the centre of each distinctive colour as displayed in the spectrum—red, 492.4; orange-red, 484.1; orange, 503.3; orange-yellow, 511.2; yellow, 517.5; green, 570; blue-green, 591.4; cyan-blue, 606; blue, 635.2; violet- blue, 685.8; puce-violet, 740.5—all in billions (1,000,000,000,000)—per second. Light due to wave-motion of one simple frequency would be 'homogeneous' or 'monochromatic' light; it would produce the simplest colour-sensation; but no such thing is experimentally attainable. The light from burning sodium is a compound of two yellow lights, very near one another in the spectrum, and corresponding to the respective wave-frequencies of 508.9 and 510.6 billions per second; and this is the nearest attainable approximation to monochromatic light.
The eye takes up, singularly enough, any congeries of several monochromatic lights impinging simultaneously upon the same spot in the retina, and the resulting sensation is always that of a single colour, not necessarily resembling any of the components. The retina is composed of numerous ultimate nerve-elements, each of which is capable of perceiving one of three physiologically primary colours. These colours are red, green, and violet (Young and Helmholtz); vermilion, emerald green, and ultramarine blue (Clerk Maxwell); or red, green, and blue (Fick). Simultaneous affection of the elements sensitive to red and of those sensitive to green produces, according to the ratio between the respective irritations, any colour of the spectrum from red through orange and yellow up to green; similarly, green and violet lights blended in different proportions produce all the intermediate blues; and when the whole three sets of nerve-elements are irritated, the sensation is still that of a simple colour, or, it may be by due adjustment, of white light. Coloured lights may be mixed so as to show this, either by causing coloured lights from different sources to coincide in the eye or on a screen, or else, as in the colour-top, by causing ocular impressions of different colours to succeed one another in the eye with such rapidity that the eye or the brain blends them. But nearly every example of what we call coloured light is in reality an admixture of several monochromatic lights—e.g. the light passing through a piece of green glass is composed—as will be found on trying to make a spectrum of it by means of a prism—not only of green light, but also of blue and yellow, and, often enough, of red and violet; but the resultant sensation is that of green, a simple colour.
White light is in general due to a simultaneous impact of wave-motions of all visible frequencies. But the sensation of whiteness may also be produced by the simultaneous impact on the retina of two suitable spectral colours, such as yellow and ultramarine blue; and such a pair of colours are said to be complementary to one another; other examples are red and greenish-blue, greenish-yellow and violet, orange and cyan-blue (a rather greenish blue). It will be observed that it is said that yellow and ultramarine blue lights make white light, while it is well known that yellow and blue pigments make a green pigment. The explanation of this is that the light from the yellow pigment is not pure; it contains green light; similarly that from the blue pigment contains green light; when the pigments are mixed, the eye receives a simultaneous impression of blue, yellow, and green; but the blue and the yellow destroy one another, being, to the eye, complementary colours; jointly they produce white light; and thus the green alone remains, diluted with white. Complementary coloured lights may both or either be of any degree of complexity; if an aggregate impression of blue light and an aggregate impression of yellow be superimposed on the same part of the retina, whatever be the mode of their production, the result may be the same—an impression of white light.
Colours vary in hue, in purity, and in luminosity. The hue determines the name of the colour—e.g. vermilion, scarlet; the purity or absence of admixture with white light determines its richness—vermilion reflects 80 per cent. of red light mixed with 20 of white; the luminosity or brightness determines the shade or tone of colour. Interfere with any of these and the impression produced upon the eye is modified. Take for example a definite red light, saturated or unadulterated with white light, such a red as may be found in the spectrum; progressively lower its purity by mixing it with white light—it becomes brighter, but passes through light red and pink to pinkish-white tints; lower its luminosity by mixing with black—it passes through terra-cotta tones to brown, which is a dull red; interfere with both purity and luminosity by mixing with various greys—it goes through russets and maroons, the so-called tertiary colours. Any colour in nature can be matched either by a spectral colour or by a purple, treated in this way; and for each such modification of the original colour there will be a different complementary, which when mixed with it forms not a white but a grey.
The colour of transparent objects is due to Selective Absorption. A red object seen through greenish-blue glass appears black; greenish-blue glass is opaque to the light from the red object. Hold the same piece of glass up to the sky, and the red lights, which are components of the white light of day, are cut off; what passes through produces a sensation of greenish-blue. The red, which is cut off by absorption, and the greenish-blue, which passes through, are complementary to one another—both being really complex, not monochromatic. The colour of a transparent body will also apparently depend upon the thickness of the layer examined: a thin layer of iodine-vapour absorbs all the constituents of visible white light except blue and red; it therefore appears in daylight to be purple; a thicker layer effects the complete absorption of the red but not that of the blue, and a thicker layer of iodine-vapour therefore appears blue. If looked at in red light, a thin layer of iodine-vapour appears red, while a thick layer will present the blackness of opacity.
Before a non-luminous object can be seen otherwise than by transmitted light it must reflect light; if it reflect none it will appear black; a dustless pool in a mountain-hollow, a liquid in a deep black vessel, may reflect no light to the eye of the observer, and will appear black. (Black is the negation of colour, because it implies that there is no sensation of light; gray, produced by mixing white and black, is white deficient in luminosity). Let the pool become turbid, and there will be some light reflected towards the observer. A coloured liquid in a deep black vessel will have its colour revealed by sprinkling a white powder into it. White light (daylight) enters the liquid; it is reflected in all directions by the white powder; but it is in part absorbed by the liquid, which accordingly appears coloured. Of precisely the same kind is the reflection of light by a solid object. Bodies allow light to traverse them to a very small depth, and then the light is, by internal reflection, turned back in all directions; absorption, meanwhile, comes into play, and the result is that the object appears to have a definite colour, the purity of which is marred by surface-reflection. The white light reflected from the surface of a metal masks its true colour, which is brought out by repeated reflection. Gold is deep orange; copper, scarlet; silver, yellowish-bronze; brass, a rich golden red.
If the light supplied to an object do not contain those kinds of light which it can reflect, the object will appear black or colourless; a bunch of flowers looked at by the yellow light of burning sodium or of a spirit-lamp with common salt in the wick, will all appear black or colourless except those which are yellow. Surface-reflection modifies the result.
There is one class of cases in which colour is not due to absorption. A haze is blue if its particles be fine enough: if it be composed of coarse particles it at once reflects white light in all directions; but fine particles cause repeated reflection, and at each reflection the reflected light becomes bluer; because those rays which would have been most refracted (the blue and violet) are in fact most largely reflected. The colour of the sky is that of a haze, reflecting light downwards; if there were no dust-haze or water-haze above us, the sky would be black. The light which is not reflected from such a haze is either transmitted through it, yellower or redder in colour, or else it is entirely absorbed. The sun thus appears yellower than it would do if our atmosphere did not intervene. See also DICHOISM, INTERFERENCE, IRIDESCENCE, LIGHT, OPTICS, PHOSPHORESCENCE, PLEIOCHROISM, RAINBOW, SPECTRUM, DYEING, and the articles on the several colours: and for further information, consult Colour by A. H. Church (Lond. 1887), and Colour by C. T. Whitnell (Cardiff: 1888). For Heraldic Colours, see HERALDRY; for the Ecclesiastical Colours, VESTMENTS; and for Colour in Animals, PIGMENT.