white caused by the combination; yellow, for instance, will be as bright as the brightness of red and green added together; white will be as bright as the brightness of the three primaries added together.

The different vibrations of which white light consists are capable of producing every possible color sensation, yet we cannot say that all color is due to them alone. Were it not for the power which natural bodies have of selecting their own particular color, or rather colorproducing vibrations, the phenomena would be incomplete. No natural body creates color; it simply absorbs a part of the light thrown upon it, and rejects the remainder. It is the portion so rejected, and not that retained, which determines the color.

Intercepting the white beam of light from the lantern by red glass, we prevent the vibrations which produce the sensations of violet, blue, green and greenish-yellow from passing through. The red glass absorbs all these rays; consequently the red rays alone are left, and we have a red disc on the screen. We will now from our second lantern project a greenish-blue light, this color being as nearly as possible the sum of the colors absorbed by the red medium. By moving the lanterns in order that this color may be, as it were, superposed, or added, to the red, we produce white, thus proving that the red glass had absorbed all those rays of which the bluish-green consists, and these rays had only to be given back to reproduce the original white disc. Repeating this experiment, but with blue and yellow lights, white is also produced, and not green, as might be expected from our experience of the result obtained by a mixture of blue and yellow pigments. By passing a beam of light through the blue and yellow glass together, placing both colors in the one lantern, green is produced, because the blue and yellow glass are not only transparent to their respective

colors, but also to the green, which is contiguous to both; but the blue glass absorbs the yellow and less refrangible rays, and the yellow glass absorbs the blue and more refrangible rays, consequently green is the only color left to pass through both, and therefore when a beam of white light goes through these colors together, green is the only possible result. This is exactly what occurs in the mixture of pigments, and it is through the confounding of the mixture of lights and the mixture of pigments that many erroneous conclusions are arrived at.

So far absorption has been considered in connection with transmitted light; we have now, in order to show its connection with the colors of pigments, to consider it when combined with reflected light. The coloration of pigments, of flowers, of green leaves, indeed of all natural bodies, "is due to the combination of reflection with the phenomenon of absorption."* I have here a colorless glass vessel, containing a red solution, and of course it exhibits the same phenomena as colored glass; it requires light either transmitted or reflected to be seen, for when placed in front of the black screen it appears dark and colorless. By placing a white object in the solution, we notice that part of the light directed upon it undergoes absorption whilst passing through the fluid, and the portion left is reflected by the object, and passes back again to the eye. This light, therefore, has itself gone twice through the fluid, or double the distance of the object from the front surface of the glass vessel. It is evident that our red color is due to what is called "selective absorption," red being the only color not devoured, as it were, by the fluid. We will now add to the solution some powdered chalk, each particle of which may be regarded as a minute white object

Stokes.

reflecting a small quantity of the light not absorbed; all I have said of the one white object applies also to these minute particles. All color pigments are composed of such minute objects, mixed with a vehicle which has the power of sifting the white light shed upon them, and selecting certain rays, whilst the particles reflect the remainder. The condition of the petals of flowers and of green leaves is just the same. A white lily, for instance, if it were quite smooth, would have the appearance of thin glass; it is, however, composed of a vast quantity of minute cells, so that it resembles finely-powdered glass, from each little particle of which light is reflected backwards and forwards, and there being nothing in the lily to cause the selection of one ray over another, the light is white. A scarlet geranium is similarly composed, but its particles, or cells, are infused with colored matter, which absorbs the green and blue rays, and the unabsorbed scarlet is reflected. All green leaves. owe their color to the same cause; light passing backwards and forwards through a coloring matter which extinguishes the red rays, is received upon the retina as green.

A property of absorption, not only interesting, but of some importance to colorists, is the change of hue colors undergo when deepened or lightened. Each additional layer of a colored medium naturally absorbs more of the light not actually destroyed, and therefore deepens the color, until at last it, too, is altogether extinguished. Light, however, has been shown to consist of vibrations of different degrees of strength; this being the case, the stronger vibrations will continue to act after the weaker are quenched; or as the color deepens its hue alters. To explain this more clearly, I will quote from a lecture by Professor Stokes, F.R.S., who, for the sake of simplicity, presupposes two kinds of light, blue and red, having intensities respectively of 100 and 10. "There is, of course, a great predominance of blue over red.

Now suppose in passing through a stratum of a certain thickness half the blue light is lost, and only half transmitted, and that ninety per cent. of the red light is transmitted. Then after passing through the first stratum the intensities will be respectively 50 and 9; after passing the second stratum of the same thickness, the intensities will be 25 and 8.1; after the third, 12.5 and 7·8; after the fourth, 6-2 and 66, or about equal; but after passing through the next stratum they will be 3.1 and 5.8; so that although the quantity of red light was so much smaller to begin with, the red is more lasting, and in light which has passed through five of these strata the red now predominates over the blue." I have here a glass vessel containing water; by adding a little red fluid you will perceive it has a blue hue, but by adding still more of the color, it assumes a scarlet hue. The cause is evident; at first the weaker blue vibrations were able to pass through, but when the absorbing liquid became more dense, the stronger red vibrations could only make their way. Conversely with this effect is the result obtained by the addition of white to a color pigment, for in this case the effect should be towards a hue of the more refrangible rays of blue and violet. The diagrams are colored with crimson lake, dark Brunswick green, and Prussian blue; these colors all show a decidedly bluer hue when mixed with white than would have been expected from the pure color.

I have shown that the primary sensations of color, red, green and violet, cannot be produced by the combination of any of the other color rays of white light, but that conversely all the other sensations of color are obtained by these three colors in different proportions. The same may be said. of the primary colors of pigments. Red, yellow and blue cannot be produced by the mixture of other color pigments, but these colors are capable of producing by admixture

almost all other colors. It will be at once evident that the colors of each set are different, the primary green and violet sensations being substituted by yellow and blue as primary color pigments; and this substitution causes one to enquire whether the remaining primary of light, red, is of the same hue as the primary pigment of that name? It is here desirable to point out the very general idea the names of colors convey to the mind. When we say red, what do we mean? The term red is often applied to a color that may be almost orange on the one side or violet on the other, green, commonly speaking, may be almost yellow or blue, violet may mean almost pink or blue. I consider that there is as much difference in the primary red of the spectrum and the primary red pigment as there is between the green light and the yellow pigment, or between the violet light and the blue pigment.

Now, the primary red of light is a scarlet red; indeed, in some works the term scarlet has been applied to it. The green light is more inclined to yellow than blue, and the violet light is a decided blue violet, and has been named blue by some writers. In pigments we find that the primary colors are the reverse of this. The best red pigment for the mixture with the greatest number of other colors is a crimson red; the best yellow is more inclined to orange than blue; the best blue is of a green hue. When these pigment colors are lightened by the addition of white, there is obtained, perhaps as near as it is possible to obtain with. pigments, the secondary colors of light. It would thus appear that the colors which were formerly considered primary colors, are really the secondary colors of light, and therefore, though red, yellow and blue may be the primary colors of pigments, they are certainly not primary color sensations caused by the vibrations of light.