When colours not complementary are used, the results depend on the proportions. Red and yellowish green give all the variations through reddish orange, orange-red, orange, orange-yellow, yellowish orange, yellow, and greenish yellow. These colours lie on the straight portion at the left of the curve in Fig. 86. Red and purple give all the intermediate purples; these colours lie along the straight line at the bottom of the same curve. Red and green give all the variations through orange and yellow, but they are all whitish; they lie along a straight line drawn from red to green, which passes nearer to white than the line from red to yellowish green. To find what are the results of combining any two colours a line is drawn between their positions in Fig. 86 and the character of the resulting colours is noted.

Three colours can also be combined by disks or spectrum colours. The result is the same as if two had been first combined and then the third added. The position in the colour diagram of Fig. 86 is worked out in the same way.

By using three spectrum colours, red, green, and purple, we can produce nearly all the hues of the spectrum with fair approximation. This suggests that the system of colours may be reducible to three elementary colour-sensations. These sensations cannot be colours of the spectrum because the results of spectrum mixture are generally somewhat whitish.

The study of abnormal forms of colour vision known as colour blindness (see below) indicates that our three fundamental sensations are far more saturated colours than those we find in the spectrum; the green in par

ticular is a far purer green than we ever actually see, the green of the spectrum being very whitish. The red is not very different from the end-red of the spectrum but the other colour is a deep indigo-blue and not a purple. These fundamental sensations find places in the colour diagram at the corners of the large triangle in Fig. 94.

The combinations of pigments, such as paints, often give very different results from the combinations of the colours. If we mix the paints with which two paper disks have been coloured, a paper coloured by the mixture will never be of the same colour as the resultant from a direct mixture of the colours of the two disks by means of the colour-top.

This can be illustrated by a disk prepared as in Fig. 95. The shaded portions are to be painted with blue, the light portions with yellow, and the central portion with a green formed by a mixture of half blue and half yellow. When the disk is rotated, the blue and yellow directly mixed never produce green but a greyish colour with a blue or yellow cast.

When increasing quantities of yellow paint are mixed with the blue paint, the colour passes through various shades of bluish green, green, and yellowish green. When yellow a and blue colours are mixed, the resulting colour passes through greyish blue, grey, and greyish yellow. With some blues the grey has a very slight

FIG. 95.-Mixing Yellow greenish tinge.

and Blue.

The reason why blue and yellow pigments give green can be illustrated by using blue and yellow glass. When two such pieces of glass are placed together, all light passing through both of them is green. Blue glass is blue because the glass absorbs the red, orange, and yellow light and allows the blue light to pass. Yellow glass absorbs the blue and allows the red, orange, and yellow to pass. Each of them allows a portion of the green to pass. When both of them are together, the blue keeps out the red, orange, and yellow, while the yellow keeps out the blue. Consequently only the green gets through.

Blue paints are blue because the minute particles of which they are composed send back to the eye mainly colours from the blue end of the series of hues. Yellow paints send back mainly those from the red end. Both send back some green. When they are mixed, the blue paint absorbs all the red end and the yellow absorbs all the blue end, leaving only green to be sent back. Similar results are obtained from the other paints; their mixtures are matters depending on their particular composition and not on their colours.

These accidents of the action of paints formerly led

people to suppose that colours followed the same laws. Thus red, yellow, and blue were formerly called the fundamental colours. The artist often speaks of his paints as his "colours;" his laws of combination of the fundamental "colours" are quite correct, if by "colours" we understand paints. To avoid confusion with the other use of the word colour, it is preferable not to use it to mean paint or pigment. Red, yellow, and blue are the fundamental pigments, and red, green, and blue are the fundamental colours.

9

WE

CHAPTER XIII

COLOUR SENSITIVENESS

E are frequently called upon to distinguish small differences in colour; how accurately can we do it?

The colour-top furnishes one method of answering the question. Suppose we take as a definite question: How accurately can we judge the mixture of small portions of blue with a large mass of red? The little red disk is placed in the centre of the top; it remains unchanged during the experiment. The large red and blue disks are placed together so that a minute por-. tion of the blue appears. The top is spun; no difference is detected. A little more blue is added and the top is again spun. This is repeated till the difference is noticed. The amount of blue can be measured by the graduated disk (Fig 92). of the whole circle. The red times as much as the blue. one part of blue to nine of red before the difference is detected.

Suppose it covers

90

must cover 10%, or nine Therefore we can add

The result depends upon the sensitiveness of the person. A dyer will detect minute differences that escape ordinary individuals; persons who have paid little attention to art are often incapable of detecting large differences.

130