is left behind, which will be yellow. Color is a partial extinction of light—not of light as a whole, but a suppression of one of its constituents. If you take a yellow glass and allow the light to fall through it, you will find it transparent; in the same way a blue glass is transparent; but if these glasses are the complementary blue and yellow color, and placed one on top of the other, no light comes through them. The yellow glass sifts out all the blue rays, and the blue glass sifts out all the other rays, and no light can get through. If the colors are not pure, it is usually because the yellow has some green in it, and so has the blue. Neither the yellow glass nor the blue is competent to sift out these rays, so we see green come through them both. This is the case in mixing blue and yellow in paints: the resulting green does not come from the mixture, but is the sediment—you might almost call it—left after the pure blue and pure yellow have neutralized each other.
It is clear that, if two waves can be made to set into vibration the same medium at the same time, and from almost exactly the same center, one of them being a half-wave or several half-waves' length behind the other, we shall have, as in the case of water and sound, no movement, or darkness. If there is not exactly a half-wave's distance between them, some color-waves will neutralize each other and be extinguished, and we shall get the complementary color—the resultant of all that is left unneutralized.
This is the cause of all the flitting and changing colors in soap-bubbles, mother-of-pearl, peacocks' plumage, opals, and iridescent glass. By some means certain vibrations have been extinguished by interference, and we see the resultant of the rest. Whenever light goes from one medium into another, even when both media seem perfectly transparent, there is a partial reflection from the surface where the media meet. Hold a pin against the surface of a piece of glass (unsilvered plate glass is the best): you will see two faint reflections of the pin, one from the front surface of the glass and one from the back, and yet the main part of the light reflected from the pin goes through, as you can easily tell by looking through the glass at the pin. So it is with a soap film: when light falls on it, most of it goes through, but there is a slight reflection from the outer surface of the thin lamina of soap-suds and another slight reflection from the back of it. The two sets of reflected waves start from points so very near each other that they both act on the medium in different directions at the same time and in the same place, and we have color.
If light went forward like a regiment of soldiers in line, there might be just as much interference from the plate of glass as there is from the film of soap-suds; but it does not—it goes out in
