This contrast theory should, however, not be interpreted too narrowly. There are pairs of so-called complementaries which make a very crude, harsh, even painful impression. The theory is happily supplemented by showing<1> that the ideal combination involves all three contrast factors, hue, saturation, and brightness. Contrast of saturation or brightness within the same hue is also pleasant. For any two qualities of the color circle, in fact, there can be found degrees of saturation and brightness in which they will form an agreeable combination, and this pleasing effect will be based on some form of contrast. But the absolute and relative extension and the space-form of the components have also a great influence on the pleasurableness of combinations. <1> A. Kirschmann, "Die psychol.-aesthet. Bedeutung des Licht und Farbencontrastes," _Philos. Studien_, vol. vii. Further rules can hardly be given; but the results of various observers<1> seem to show that the best combinations lie, as already said, among the complementaries, or among those pairs nearer together in the color circle than complementaries, which are "warmer." The reason for this last is that, in Chevreul's phraseology, combinations of cold colors change each other's peculiar hue the most, and of warm colors the least; because the complementaries of these cold colors are "warm," i.e. bright, and each, appearing on the field of the neighboring cold color, seems to fade it out; while the complementaries of the juxtaposed warm colors are not bright, and do not have sufficient strength to affect their neighbors at all. With a combination of blue and green for instance, a yellow shade would appear in the green and a red in the blue. Such a result fails to satisfy the demand, already touched on, for purity and homogeneity of color,--that is, for unimpeded seeing of color. <1> Chevreul, _De la Loi du Contraste Simultane des Couleurs_. E.S. Banker, op. Cit. What significance have these abstract principles of beauty in the combination of colors for representative art? In the choice of objects with a definite local color, of course, these laws will be found operative. A scheme of blues and yellows is likely to be more effective than one of reds and violets. If we analyze the masterpiece of coloring, we shall find that what we at first supposed to be the wonderful single effects of color is really the result of juxtapositions which bring out each color to its highest power. III