224 ON ACCIDENTAL OR SUBJECTIVE COLORS. 



the complementary tint is replaced by a slight shade of the same color with that 

 space. Thus, in the experiment of Prieur de la Cote d'Or, when the small strip 

 of white pasteboard has a certain breadth, ten or twelve millimetres, for exam- 

 ple, it occurs, for certain positions of the paper, that the edges only of the latter 

 appear to take the complementary tint, and the interior is lightly tinged with 

 the same color as the paper : for example, on a red paper, the small strip will 

 have the edges green and the interior rose- colored ; that is to say, at a certain 

 distance from the outline of the red space, the green complementary tint is re- 

 placed by a shade of red. 



It has also been seen that the appearance of the images of the first class suc- 

 ceeds the phenomenon of the persistence of the primitive impressions. Now, 

 we find, in regard to the accidental colors of the second class, a fact which is 

 the analogue of the former. In effect, physicists recognize that the impression ■ 

 produced on the retina extends a little beyond the space directly excited by the 

 light ; so that if we observe a red object, for instance, on a black ground, the 

 red image perceived is a little more extended than it would be if the sensation 

 were limited to the space on which the red light directly strikes. From this it 

 results, that if we consider the appearances which are manifested beyond the 

 colored space, beginning at its real outline and proceeding outwardly, the im- 

 pression of its color extends to a short distance, beyond which the complemen- 

 tary color appears ; while this latter, as we have seen, is replaced at a still 

 greater distance by a new development of the first color. 



Conclusions. — Recapitulation. — Thus, on comparing the two preceding classes 

 of subjective colors, we see, that in the first, as soon a-s the retina ceases to 

 be directly excited by the presence of the colored object, there occur : 1st. A 

 ■persistence, generally very short, of the primitive impression; 2d. The appari- 

 tion of the accidental image; 3d. Usually successive disappearances and reap- 

 pearances, more or less numerous, of that accidental image; and, in certain 

 cases, alternate apparitions of the primitive impression and of the accidental 

 image. 



In the second class, beyond the outline which the colored space or object 

 would directly present, we find : 1st. A prolongation, to a certain distance, of 

 the real impression ; 2d. Beyond that prolongation and to a distance usually 

 considerable, a development of the accidental color ; 3d. Under certain circixm- 

 stances, beyond the space colored by the accidental color, a space which in this 

 case is of small extent, we observe a new development of the real color of the 

 object. 



It will hence be seen that the phenomena of the second class are really, in 

 relation to space, what the phenomena of the first class are in relation to time. 

 There are other appearances which depend alike on these two sorts of accidental 

 colors. 



Neio class of phenomena. — "We will cite only a few examples : 1st. When 

 instead of placing the object which we are observing on a white or black ground, 

 we place it on a colored ground of sufiicient extent, the color of the image after- 

 wards perceived on casting the eyes on a white surface is not simply the com- 

 plementary of the color of the object ; it is found to be combined with the color 

 of the ground on which the object is placed. Thus a small piece of orange- 

 colored paper placed on a sheet of yelloio paper, will produce a green accidental 

 image on a violet ground ; that is to say, the image will be composed of blue, 

 which is the complementary of the orange, and of yellow, which is the color of 

 the paper itself on which the object was placed. Thus, too, a small piece of 

 violet paper placed on a sheet of red paper will produce an orange-colored image 

 on a green ground. These effects were remarked by Darwin, and depend on 

 the two classes of subjective colors which we have been considering.. In effect, 

 while we are observing an orange-colored object placed on a yellow ground, a 

 light tint of violet produced by the neighborhood of the yellow will become 



