different. This is especially the case with yellow and blue, which on the screen combine to produce white, generally with a pink tint, but cannot be made to give green. The reason of this difference in the two results is that in the former case we do not get a true combination of the colours at all. When the mixed pigments are illuminated by white light, the yellow particles absorb the red and blue rays, but reflect the yellow along with a good deal of the neighbouring green and orange. The blue particles, on the other hand, absorb the red, orange and yellow, but reflect the blue and a good deal of green and violet. As much of the light is affected by several particles, most of the rays are absorbed except green, which is reflected by both pigments. Thus, the colour of the mixture is not a mixture of the colours yellow and blue, but the remainder of white light after the yellow and blue pigments have absorbed all they can. The effect can also be seen in coloured solutions. If two equal beams of white light are transmitted respectively through a yellow solution of potassium bichromate and a blue solution of copper sulphate in proper thicknesses, they can be compounded on a screen to an approximately white colour; but a single beam transmitted through both solutions appears green. Blue and yellow pigments would produce the effect of white only if very sparsely distributed. This fact is made use of in laundries, where cobalt blue is used to correct the yellow colour of linen after washing.
Thomas Young suggested red, green and violet as the primary colours, but the subsequent experiments of J. Clerk Maxwell appear to show that they should be red, green and blue. Sir William Abney, however, assigns somewhat different places in the spectrum to the primary colours, and, like Young, considers that they should be red, green and violet. All other hues can be obtained by combining the three primaries in proper proportions. Yellow is derived from red and green. This can be done by superposition on a screen or by making a solution which will transmit only red and green rays. For this purpose Lord Rayleigh recommends a mixture of solutions of blue litmus and yellow potassium chromate. The litmus stops the yellow and orange light, while the potassium chromate stops the blue and violet. Thus only red and green are transmitted, and the result is a full compound yellow which resembles the simple yellow of the spectrum in appearance, but is resolved into red and green by a prism. The brightest yellow pigments are those which give both the pure and compound yellow. Since red and green produce yellow, and yellow and blue produce white, it follows that red, green and blue can be compounded into white. H. von Helmholtz has shown that the only pair of simple spectral colours capable of compounding to white are a greenish-yellow and blue.
 |
| Fig. 1. |
Just as musical sounds differ in pitch, loudness and quality, so may colours differ in three respects, which Maxwell calls hue, shade and tint. All hues can be produced by combining every pair of primaries in every proportion. The addition of white alters the tint without affecting the hue. If the colour be darkened by adding black or by diminishing the illumination, a variation in shade is produced. Thus the hue red includes every variation in tint from red to white, and every variation in shade from red to black, and similarly for other hues. We can represent every hue and tint on a diagram in a manner proposed by Young, following a very similar suggestion of Newton’s. Let RGB (fig. 1) be an equilateral triangle, and let the angular points be coloured red, green and blue of such intensities as to produce white if equally combined; and let the colour of every point of the triangle be determined by combining such proportions of the three primaries, that three weights in the same proportion would have their centre of gravity at the point. Then the centre of the triangle will be a neutral tint, white or grey; and the middle points of the sides Y, S, P will be yellow, greenish-blue and purple. The hue varies all round the perimeter. The tint varies along any straight line through W. To vary the shade, the whole triangle must be uniformly darkened.
 |
| Fig. 2. |
The simplest way of compounding colours is by means of Maxwell’s colour top, which is a broad spinning-top over the spindle of which coloured disks can be slipped (fig. 2). The disks are slit radially so that they can be slipped partially over each other and the surfaces exposed in any desired ratio. Three disks are used together, and a match is obtained between these and a pair of smaller ones mounted on the same spindle. If any five colours are taken, two of which may be black and white, a match can be got between them by suitable adjustment. This shows that a relation exists between any four colours (the black being only needed to obtain the proper intensity) and that consequently the number of independent colours is three. A still better instrument for combining colours is Maxwell’s colour box, in which the colours of the spectrum are combined by means of prisms. Sir W. Abney has also invented an apparatus for the same purpose, which is much the same in principle as Maxwell’s colour box. Several methods of colour photography depend on the fact that all varieties of colour can be compounded from red, green and blue in proper proportions.
 |
| (After Müller-Pouillet’s Lehrbuch der Physik, 1897.) |
| Fig. 3. |
Any two colours which together give white are called complementary colours. Greenish-yellow and blue are a pair of complementaries, as already mentioned. Any number of pairs may be obtained by a simple device due to Helmholtz and represented in fig. 3. A beam of white light, decomposed by the prism P, is recompounded into white light by the lens l and focussed on a screen at f. If the thin prism p is inserted near the lens, any set of colours may be deflected to another point n, thus producing two coloured and complementary images of the source of light.
Nature of White Light.—The question as to whether white light actually consists of trains of waves of regular frequency has been discussed in recent years by A. Schuster, Lord Rayleigh and others, and it has been shown that even if it consisted of a succession of somewhat irregular impulses, it would still be resolved, by the dispersive property of a prism or grating, into trains of regular frequency. We may still, however, speak of white light as compounded of the rays of the spectrum, provided we mean only that the two systems are mathematically equivalent, and not that the homogeneous trains exist as such in the original light.
See also Newton’s Opticks, bk. i. pt. ii.; Maxwell’s Scientific Papers; Helmholtz’s papers in Poggendorf’s Annalen; Sir G. G. Stokes, Burnett Lectures for 1884–5–6; Abney’s Colour Vision (1895). (J. R. C.)
COLOURS, MILITARY, the flags carried by infantry regiments and battalions, sometimes also by troops of other arms. Cavalry regiments and other units have as a rule standards and guidons (see Flag). Colours are generally embroidered with mottoes, symbols, and above all with the names of battles.
From the earliest time at which men fought in organized bodies of troops, the latter have possessed some sort of insignia visible over all the field of battle, and serving as a rallying-point for the men of the corps and an indication of position for the higher leaders and the men of other formed bodies. In the Roman army the eagle, the vexillum, &c. had all the moral and sentimental importance of the colours of to-day. During the dark and the middle ages, however, the basis of military force being the individual knight or lord, the banner, or other flag bearing his arms, replaced the regimental colour which had signified the corporate body and claimed the devotion of each individual soldier in the ranks, though the original meaning of the
