simultaneously or at very close intervals, produces the impression of white. Starting from a preconceived analogy with the notes of the gamut, Newton divided the solar spectrum, or the image obtained by decomposing white light with a refracting prism, into seven different colors. This division is really arbitrary, for the colors pass from one to another by insensible transitions, and each of them may be characterized either by the degree of its refraction in the prism, or by the length of the undulations to which it corresponds.
When we collect a part of the spectral rays in one point, we obtain either one of the primitive colors in a greater or less state of purity, or a new tint. If we divide the spectrum arbitrarily into two parts and collect the rays of these parts separately, we obtain two distinct colors, the superposition of one of which on the other gives white light. The experiment can be performed with an ordinary spectrum divided arbitrarily into two parts; and it is effected, we might say naturally, in the phenomena of rotatory polarization in which the most brilliant hues are shown.
We mention these properties to deduce two conclusions from them. We remark, first, that the mixture of simple or homogeneous lights in any proportion always produces upon the eye a single impression, that of one color. While the ear can distinguish all the notes that go to make up a harmony, the eye can grasp only one color, without being able to distinguish whether it is really simple, or is formed of different lights.
In the second place, the mixture of colors provokes only one new impression, that of purple, for example, which we may obtain by mixing red and violet, while the varieties of rose are nothing but mixtures of purple and white. White may be produced by two simple colors alone, as by red and green; more generally, if we isolate three suitably chosen colors in the spectrum, such as particular shades of red, green, and violet, we may, by mixing them in different proportions, imitate the impressions produced by all the colors. Artificial colors, formed, for example, of rays selected from the spectrum, may be simple or compound, without the eye being informed of the difference, except, perhaps, when they have a shade of purple or of rose, for we know then that those colors do not exist in a simple state.
The same is the case with the colors of nature, or of industry. An object appears colored to us because it sends us only a part of the light it borrows from the general illumination. The sorting out is made either by transmission, as in colored glasses, or by reflection, as in the case of the metals, or by diffraction, as in the wings of some butterflies, or in the coronas which we sometimes perceive around the moon; the portion of the light that does not reach the eye having been absorbed or sent off in a different direction. Leaving out of the account the effects of fluorescence, we perceive that objects do not have colors of themselves, but simply borrow from the general light-
