| R + G + V | = | 250·0 |
| (R + G) + V | = | 250·5 |
| (R + V) + G | = | 252·5 |
| (G + V) + R | = | 248·0 |
| (R + G + V) | = | 250·0 |
The mean of the first four is 250·25, which is only 1/10% different from the value of 250 obtained from the measurement of (R + G + V) combined. Other measures fully bore out the fact that the luminosity of the mixed light is equal to the sum of the luminosities of its components. It is true that we have here only been dealing with spectrum colours, but we shall see when we come to deal with the mixture of colours reflected from pigments that the same law is universally true.
It will be proved by and by that a mixture of three colours, and sometimes of only two colours, be they of the spectrum or of pigments, can produce the impression of white light. If then we measure all the components but one, and also the white light produced by all, then the luminosity of the remaining component can be obtained by deducting the first measures from the last. For instance, red, green and violet were mixed to form white light. The luminosity of the white being taken as 100, the red and violet were measured and found to have a luminosity of 44·5 and 3 respectively. This should give the green as having a luminosity of 52·5. The green was measured and found to be 53, whilst a measurement of the red and green together gave a luminosity of 96·5 instead of 97.
CHAPTER VIII.
Methods of Measuring the Intensity of the Different Colours of the Spectrum, reflected from Pigmented Surfaces—Templates for the Spectrum.
Fig. 14.—Measurement of the Intensity of Rays reflected from white and coloured surfaces.
We will now proceed to demonstrate how we can measure the amount of spectral light reflected by different pigments. Let us take a strip of card painted with a paste of vermilion, leaving half the breadth white; and similarly one with emerald green. If we place the first in the spectrum so that half its breadth falls on the red, and the other half on the white card, we shall see that apparently the red and orange rays are undiminished in intensity by reflection from the vermilion, but that in the green and beyond but very little of the spectrum is reflected. With the emerald green placed similarly in the spectrum, the red rays will be found to be absorbed, but in the green rays the full intensity of colour is found, fading off in the blue. What we now have to do is to find a method of comparing the intensities of the different rays reflected from the pigments, with those from the white surface. We will commence with the second of the two methods which the writer devised with this object, and then describe the first, which is more complex. Suppose we have, say a card disc three inches in diameter, painted with the pigment whose reflective power has to be measured, and place it on a rotating apparatus with black and white sectors of say five inches diameter, and capable of overlapping so as to show different proportions of black to white (see [Fig. 42]). If we throw a colour patch (shown in [Fig. 14] as the area inside the dotted square) on the combination of black and white, and at the same time on the pigmented disc, it is probable that either one or other will be the brighter. By moving the slit along the spectrum it is evident, however, that a colour can be found which is equally reflected from them both whilst rotating. Take as an example the sectors as set at two parts white, to one part black, the centre disc being vermilion, the slit is moved along the spectrum until such a point is reached that the colour reflected from the ring and the disc appears of the same brightness, for it must be recollected that they cannot differ in hue, as the light is monochromatic. It will be found that the place where they match in brightness is in the red, the exact position being fixed by the scale at the back of the slide. Taking the proportion of black to white as three to one, the match will be found to take place in the orange. Increasing the proportion of black more and more, a point will be reached where the reflection takes place uniformly along the blue end of the spectrum, this will be from the green to the end of the violet. By sufficiently increasing the number of matches made, a curve of reflection can be made showing the exact proportion of each ray of the spectrum that is reflected. The uniform reflection along the blue end of the spectrum shows that a certain amount of white light is reflected from the pigment.