CHAPTER X.

ON THE MIXTURE OF COLOURS.

THOSE Who watch a painter at work are astonished at the vast number and variety of tints which can be made by mixing in varying proportions a very few pigments: from red and yellow there is produced a great series of orange tints; yellow and blue furnish a multitude of green hues ; blue and red, a series of purples. The results seem almost magical, and we justly admire the skill and knowledge which enable him in a few seconds accurately to match any colour which is within the compass of his palette. As we continue our observations we soon find that the matter is more complicated than it appeared at first sight, each pigment having a particular set of properties which it carries into its mixtures, and these properties being by no means. fully indicated by its mere colour. Thus, some blue pigments furnish fine sets of greens, while others, as beautiful and intense, yield only dull olive-greens; some reds give glowing purples, while from others not less bright it is possible to obtain only dull, slaty purples. Before touching on these complicated cases it will be well to study this subject under its simplest aspects, and to content ourselves for the present with making an examination of the effects which are produced by mixing light of different colours. This can not be brought about by mixing pigments, as was for a long time supposed. In some cases the mixture of pigments gives results more or less like those produced by

the mixture of coloured light, but as a general thing they differ, and in some cases the difference is enormous. In the previous chapter, for instance, it was shown that while the mixture of yellow with blue pigments produced invariably a green hue of varying intensity, the mixture of blue with yellow light gave a more or less pure white, but under no circumstances anything approaching green. The mixture of two masses of coloured light can easily be effected in a simple manner so as to be exhibited readily to a large audience. Two magic lanterns are to be employed, as shown in Fig. 45, the usual slides being replaced by plates

FIG. 45.-Two Magic Lanterns casting Yellow and Blue Light on the same Screen, where it forms White.

of coloured glass, as indicated.

Each lantern then will fur

nish a large bright circle of coloured light, which can be projected on a white screen, the room in which the experiments are made being first darkened. In this way it will be found that violet-blue light when mixed with green light gives blue or greenish-blue light, according to the proportions of the two constituents; green with red gives various hues of orange or whitish yellow, instead of a set of dull, indescribable shades of greenish, reddish, or brownish grey, as is the case with pigments. These and many

other beautiful experiments on the mixture of coloured light can easily be made; and even the effects produced by varying the intensity of either of the masses of coloured light can readily be studied by gradually diminishing the brightness of one of the coloured circles, the other remaining constant.

To all these experiments it may be objected that we are not using coloured light of sufficient purity; that our yellow glass transmits, as was shown in Chapter VII., not only yellow but red, orange, and green light; and that the other stained glasses are not much better off in this respect. Hence, to meet all such objections, physicists have been finally obliged to conduct their researches in this direction on the pure coloured rays of the spectrum. The difficulties encountered in the use of this method are much greater, but the results so obtained are far more precious. Very beautiful investigations of this character have been made by Helmholtz, Maxwell, and J. J. Müller. The general character of their results is something like this: By mixing two kinds of pure coloured light they obtained, as a general thing, light having a colour different from either of the original ingredients; for example, red and yellowish green gave an orange hue which looked in all respects like the pure orange of the spectrum; also, in this new orange it was impossible by the eye to detect the presence of either red or yellowish-green light. This was true of all mixtures; in no case could the original ingredients be detected by the eye. In this respect the eye differs from the ear; for by practice it is possible with the unaided ear to analyze up compound sounds into the elements which compose them, at least to some extent. It was also ascertained that the same colour could be produced in several different ways— that is, by combining together different pairs of spectral colours. Thus, violet with cyan-blue gave an ultramarine hue, but violet gave the same colour when mixed with bluish-green, or even with green; in the last case the tint

A

a

D

Red.

Red-orange.
Orange.
Orange-yellow.
Yellow.

Greenish-yellow and Yellowish-green.

was somewhat whitish. By mixing certain colours of the spectrum, it was found that one new colour or colour-sensation could be produced which originally was not furnished by the B plain pure spectrum itself; we re- c fer to purple, or rather to the whole class of purples, ranging from violet-purple to red-purple. These are produced by mixing the end colours of the spectrum, red and violet, in varying proportions. Furthermore, mixtures of certain colours of the spectrum gave rise E to white; this was true, for example, of red and bluish-green, and of yellow and ultramarineblue. The white in these two I cases, though so different in origin, had exactly the same appearance to the eye. Again, by mixing three or more spectral colours no new hues were produced, but simply varieties of those which could be obtained from two colours.

Green and
Blue-green.

Cyan-blue.

Blue and
Blue-violet.

Violet.

FIG. 46.-Prismatic Spectrum.

of Helmholtz, that all the colours of the spectrum, Fig. 46, from red to yellowish-green, gave by mixture resultant hues which were always identical with some of the colours situated between red and yellowish-green, thus:

The effect of the mixture in these cases was to produce colours which were, to all appearance, as pure as the corresponding colours of the spectrum itself.

Furthermore, all colours of the spectrum from violet to bluish-green furnished mixtures corresponding to the colours contained between these limits, thus:

TABLE II.

Bluish-green and ultramarine-blue gave... Cyan-blue.

Bluish-green and violet gave...

Violet and cyan-blue gave...

Cyan-blue or ultramarine-blue.*
Ultramarine-blue.

În these cases also the resulting tints could not be distinguished from the corresponding spectral colours. The results thus far are simple in character, and easily remembered by any one who recollects the arrangement of the colours of the spectrum.

On the other hand, green, when mixed with any colour of the spectrum, gave a resultant colour, which was less saturated or intense, and appeared more whitish, than the corresponding spectral tint, thus :

Green and red gave..

TABLE III.

Orange,
Yellow,
Yellowish-green,

whitish.

* According to the proportions.