ON ACCIDENTAL OR SUBJECTIVE COLORS. 227 



potliesis of a moral cause ; the image should then always appear lo us of the 

 same magnitude with the object. Analogy, moreover, would lead us to regard 

 the colors of the second class as owing, like the first, to a physical modification 

 of the organ. 



If, now, we compare the accidental impressions with the corresponding real 

 impressions, we shall soon see that the first should be regarded as of a nature 

 opposed to that of the second : 1st, the opposition is evident for the case of a 

 white object on a black ground ; 2dly, it may be said, in general, that two com- 

 plementary colors are opposed, since their tints neutralize one another by pro- 

 ducing white ; but, as has been seen, the opposition goes further in the case of 

 a real impression and the corresponding accidental impression ; it then destroys 

 the perception of brightness by producing black ; 3dly, while two complement- 

 ary real colors produce ichite by their combination, two accidental complement- 

 ary colors, on the contrary, produce black. 



The foregoing being admitted, let us recur to the collective phenomena which 

 succeed the contemplation of colored objects, or which accompany the contem- 

 plation. When a portion of the retina, after having been excited by the pres- 

 ence of a colored object, is suddenly withdrawn from that action, it regains by 

 degrees the normal state, producing for us the phenomena of the persistence of 

 the primitive impression and of the accidental colors of the first class. 



On the other hand, if, while the retina is subjected to the action of the colored 

 light, we examine the parts of the organ which surround the space directly ex- 

 cited, we see that the normal state is not found except at a distance more or less 

 considerable from the contour of that space, and that this normal state is reached 

 in passing through the effects of irradiation and of the accidental colors of the 

 second class. 



Thus, on the one hand, the persistence of the primitive impression and the 

 accidental colors of the first class constitute the passage from the state of excita- 

 tion of a portion of the retina to the normal state, when we consider that pas- 

 sage in regard to time ; on the other hand, the irradiation and the accidental 

 colors of the second class constitute the passage from the state of excitation ot 

 this same portion of the retina to the normal state, when we consider this pas- 

 sage i7i regard to space. All the phenomena, therefore, with which we are oc- 

 cupied must be referable to the same principle; and M. Plateau has pointed out 

 that they are connected with one another in the most natural manner ; that 

 they are simple consequences of the law of" continuity. 



Let us examine, then, more closely the transition to the normal state in regard 

 to time. When the retina is suddenly withdrawn from the action of the rays 

 emitted from the colored object, the state of excitation of the organ at first con- 

 tinues for some time, growing feebler, yet without changing its nature, and thence 

 results the persistence of th3 piimitive impression; but presently this state of 

 the organ gives place to an opjwsite state, whence results the subjective image 

 of the object. This new sensation soon attains a maximum of intensity, and 

 grows feebler in its turn, usually manifesting, however, an oscillatory progress, 

 more or less regular, sometimes confined to a succession of disappearances and 

 reappearances, sometimes alternating with recurrences of the primitive impres- 

 sion. Now, it is impossible not to see a striking analogy between these phe- 

 nomena and the movement of a body removed from a position of stable equilibrium 

 and which returns to it by an oscillatory action. Are we not natui'ally led, by 

 the whole train of phenomena, to admit that the retina, diverted from its normal 

 state by the presence of a colored object and then suddenly left to itself, at first 

 .regains rapidly the point of repose, but that, impelled by its own movement, it 

 passes that point and attains an opposite state, whence it tends anew toward the 

 point of repose, at which it finally arrives in a permanent manner only after a 

 series of decreasing oscillations % 



If, now, we examine under the same point of view the transition to the normal 



