when looking at the setting sun, or a flaring gas light. If the light at which we look is coloured, we shall see the complementary colour in the impression left on the retina. Sir David Brewster was one of the first to notice and experiment upon these very interesting facts.
If we cut out any simple figure, a small cross for instance, in scarlet paper, place it upon a white background and look at it fixedly for a minute or two, we shall find that its tint will gradually become duller. If we now suddenly look at a piece of white paper, we shall see the cross depicted upon it in green, which is the complementary colour to red. It should be explained, that the complementary of any colour is that which is necessary to make white light. Thus, blue, yellow, and red (as we shall find out when we come to speak of the prismatic spectrum), mixed in certain proportions, form white light; consequently the complementary of orange, which is composed of red and yellow, will be blue; of green, which is yellow and blue, red; of purple, which is blue and red, yellow, and vice versâ. The complementary of black is white, and of white, black as a rule; but if the white object be very brilliant, the black spectrum will speedily become coloured. The impression left by the setting sun is of this character. At first, while the eye is open, the image is black, then brownish red, with a light blue border; but if the eye be shut suddenly, it becomes green, with a red border, the brilliancy of colour being apparently in proportion to the strength of the impression. These spectra may be perceived for a long time, if the eye is gently rubbed with the finger now and then. Some eyes are more impressionable in this respect than others, and Beyle gives an instance of an individual who saw the spectrum of the sun for years, whenever he looked at a bright object. A modern instance of
