ON ACCIDENTAL OR SUBJECTIVE COLORS. 223 



were so hidden by clouds that the shadow extended over all objects. On the 

 contrary, the effect produced by the colored windows of Gothic churches pro- 

 ceeds, in great part, from the circumstance of their being environed by dark 

 walls. 



The facts which relate to this class of colors are susceptible of numerous ap- 

 plications in all the arts where the assortment of tints is involved. M. Chevreul 

 mentions a great number relating to the art of the upholsterer, figured impres- 

 sions on paper or cloth, &c. For example, he remarks that it is never proper, 

 in assorting furniture, to combine upholstery of a bright red color with mahogany, 

 for the greenish complementary tint which is developed by such accessories 

 causes the reddish color which is prized in mahogany to disappeai-, and the latter 

 then resembles oak or walnut, &c. M. Chevreul also applies these considera- 

 tions to the arrangement of flowers in a parterre. The principal rule to be ob- 

 served in this case is to place blue flowers beside orange-colored; violet flowers 

 beside yellow. As regards red and rose-colored flowers, they never appear with 

 so much advantage as when surroxinded by verdure and white flowers. These 

 should also be interposed between groups formed of blue and orange-colored 

 flowers, violet and yellow flowers. We see that this arrangement is founded on 

 the consideration that two complementary colors vivify one another by being 

 associated, and that white heightens the tone of the colors which it is near. 

 Considerations of the same kind are applicable to the female toilet. For exam- 

 ple, the head-dress which would seem most advantageous for a fair complexion 

 is a green hat, lined with rose; because the latter color is reflected on the visage, 

 and the green which surrounds it further tends to add to this the effect of its 

 complementary color. 



Advantage has also been derived from the effects which result from a difference 

 of brightness in adjacent objects. Painters, in order to judge well of the effect 

 of a painting, observe it through a black tube. The colors in this way acquire 

 considerable vivacity and brightness. The magic effect of the diorama depends 

 in part on the circumstance that the picture alone is well illuminated, while all 

 that surrounds it, as well as the hall containing the spectators, remains in a cer- 

 tain degree of obscurity. When it is desirable to produce a strong effect at the 

 theatre, it is necessary to diminish as much as possible the light in the body of 

 the hall. 



This second class of colors may be made to afford curious and amusing effects 

 in the following manner : Let colored papers having a certain transparency be 

 spread on a rectangular frame of wood or pasteboard, in such a way that half 

 the rectangle shall present a red paper and the other half a paper of some other 

 color; then let a flower, of which all the parts have little width, two or three 

 millimetres at most, be cut out of white pasteboard; for this purpose the petals 

 and leaves may be cut in open-work. This flower is now to be placed on the 

 colored papers in such manner that the stem and leaves shall be relieved on the 

 red paper, and the flower on the other colored paper. If the frame be then held 

 between the window and eye in a certain position, the stem and leaves will 

 appear green, and the flower will seem colored with the tint complementary to 

 that of the ground. If we have several such frames, in which one-half is 

 always red and the other half of different colors, the same white flower, applied 

 in succession on different grounds, will each time take a new color. It will be 

 blue on an orange colored ground, rose on a green ground, while the parts 

 which would naturally be green continue of that color. 



Intrinsic nature of the phenomenon of subjective colors of the second order. — 

 It has been seen that the subjective colors of the first class are sometimes fol- 

 lowed by one or more reappeai-ances of the primitive impression. The colors 

 which we are now examining sometimes present a phenomenon which is, in re- 

 lation to space, what the former is in relation lo time. We observe, under some 

 circumstances, that at a certain distance from the outline of the colored space, 



