The light of D had to be reduced to 14° before it was extinguished; therefore to extinguish the original light of this colour in the spectrum would require 180/14, or 12·9 times the intensity of the white light of the reflected beam. With the E light it would take 180/22, or 8·2 times the white light to extinguish it, and so on. If we tabulate the results in this manner, and take the white light necessary to extinguish the D light empirically as 98·5, which is its percentage luminosity in the spectrum of the electric light, we can then compare the extinguishing factor with the luminosity in each case.
| Colour. | White required to extinguish the Spectrum. | White required to extinguish the Spectrum, with 50 as That required at E. | Luminosity of Spectrum. |
| near line B | ·6 | 3·9 | 4·9 |
| C | 3·2 | 19·5 | 20·6 |
| D | 12·9 | 78 | 98·5 |
| E | 8·2 | 50 | 50 |
| F | 1·2 | 7·5 | 7·5 |
| G | ·087 | ·56 | ·6 |
The very close resemblance between the last two columns indicates that the same luminosity of white light is necessary to extinguish the same luminosity of most colours, within the limits of observation that is to say. Indeed the method of extinction was a plan which Draper and Vierordt essayed, but the results, tabulated from experiments made by them with the apparatus they employed, give a curve of intensity very unlike that given in Chapter VII. In these experiments the luminosity of the orange light corresponding to the D line coming through the slit was measured, and it was found to be 37·5/180 of the white light. Now according to the last table but one 14/180 of this light was extinguished by the full white light, consequently 37·5/180 × 14/180, or 1/62 of the orange light was extinguished by the white light. In other words, if white light be sixty-two times latter when the two are mixed will be invisible. The extinction of all colours requires somewhat more light than this, and a calculation shows that the extinction of every colour is effected by white light, which is seventy-five times brighter than the colour. Artists are well aware that a pale wash of a pigment may be washed over drawing paper, and when dry is invisible to the eye. The above experiments fully account for it.
The other experiment which was to be tried was to see how much white light could be extinguished by a colour. There are several ways by which this can be effected. For instance we may superpose a white dot on the colour patch by placing a card, in which a circular hole is cut, in the reflected beam near the prism, from which the reflection takes place; or by putting a black circular disc of small dimensions pasted on a glass in the same position, by which means the white light is superposed over the whole of the colour patch, with the exception of what, when the colour is cut off, is a black spot; or again by placing a rod to shade half the patch from the white light, but leaving the whole of it exposed to the coloured beam. All these methods have been tried, and it appears that the size of the piece of the patch over which the white light is thrown may have some effect on the resulting curve, but of one thing there is evidence, viz. that a great deal more white light can be mixed unperceived with orange light, than can be with the green, blue, or violet. From one experiment it was found that 1/36 part of white light of the same luminosity as the orange could be mixed with the orange and not be perceived; but that with the green light at E 1/90 would just be visible, whilst at F in the blue-green the 1/120 could be distinguished. Looking at these results, and applying them in elucidating the experiments in which it was attempted, but without success, to match the intermediate colours between violet and green (of which the light at F is a case in point), by mixing them together, unless white light were added to the simple colour; and the success of the other experiment, in which orange light could be obtained of the same hue as that at D by a mixture of the red and green, it will be noticed that 3·3 times more white light can be added to the orange than to the green light at F, without its perception. The white light produced by the mixture in the first case might well show when mixed with the green, but might pass wholly unperceived when mixed with the orange.
CHAPTER XI.
Primary Colours—Molecular Swings—Colour Sensations—Sensations absent in the Colour-blind.
For some purposes it is advantageous to show experiments before indicating the deductions from them which may lead to a theory. Those described in Chapter IX. will enable us to treat the theory of colour perception from a standpoint of some advantage. How is it that the combination of three colours suffices to form white, or to match any colours we wish, be they spectrum colours to which a little white is added, or the colours of pigments? The most plausible theory that can be advanced is that it is only necessary for the eye to be furnished with a three-colour-perceiving apparatus to give the impression of every colour, and yet this would be somewhat difficult to believe had we not had the experiments narrated in that chapter before us. We should have almost expected some machinery in the eye to exist, which would answer to the rhythmic swing of the rays of every wave-length which together make up white light. But now we have to stand face to face with the results of experiment, and we find that at the most only three colours are necessary to make up white light, and that from these three spectrum colours we can form any others, with the limitation already mentioned, when some simple colours are in question.
We must here digress for a moment, and notice the fact that from our experiments we have derived the three primary colours as they are called, viz. red, violet, and green; the definition of a primary colour being that it cannot be formed by the mixture of any other colours. We have ascertained that yellow and blue make white. It is therefore evident that blue, yellow, and red cannot be primary colours, since two of them form white; and we have moreover shown that yellow can be made from green and red; hence it might be fair to assume that the three primary colours are red, green, and blue. But blue, when mixed with a very small percentage of white light, can be made by green and violet. Hence, in the white light formed by the two colours yellow and blue, we have the first made by green and red, and the second by green and violet; hence the three colours which really make the white light are red, green, and violet. The approximate positions of these three colours in the spectrum are those already indicated; though, as we shall presently see, it is highly improbable that any person whose eyes are what are called normal, has ever experienced the fundamental green sensation.