234 ON ACCIDENTAL OR SUBJECTIVE COLORS. 



in relation to time. On one hand, wlien the retina, after having been excited 

 for a certain time by the light emanating from an object, is suddenly withdrawn 

 from that action, the impression still persists during some moments. On the 

 other hand, while the retina is subjected to the action of the light emanating 

 from an object, the impression will extend to a small distance around the image 

 of that object. Both one and the other phenomenon would be results of a sim- 

 ple law of continuity, in virtue of which, when a portion of the organ is divert- 

 ed from its normal condition, the dynamic state which results can neither be in- 

 stantaneously annihilated, nor subsist in immediate contiguity with a state of 

 perfect repose. 



To these purely rational considerations, Mr. Plateau adds a great number ot 

 facts, a few of which only will be here cited : 1st. We know that the small 

 space which, in the retina, corresponds to the insertion of the optic nerve, and 

 which is called the punctum ccectim, appears insensible to the action of the direct 

 light ; hence, if a small white or colored object be placed on a black surface, 

 and, closing one eye, we direct the other in such a manner that the image of 

 this object falls on the point in question, the object disappears. Now, if instead 

 of a colored object on a black ground, we place a black object on a colored 

 ground, the object equally disappears, and the color of the surface extends itself 

 over the place which it occupies. The most insensible part of the retina, there 

 fore, is impressible by communication ; in other words, the surrounding impres- 

 sion is propagated laterally over this small space ; the same then, with stronger 

 reason, must be the case in the sensitive portions. 2d. It is a constant fact 

 that irradiation augments with the duration of the contemplation of the object ; 

 it is admitted, moreover, by physicists that by continued contemplation the 

 retina becomes less and less susceptible of impression by the direct light; if 

 these two facts be considered, the conclusion must be that in proportion as the 

 direct light seems to lose its power over the retina, the propagated impression 

 will be more strongly developed. 3d. On a colored ground place a narrow strip 

 of white paper, or else trace a black line on a white surface, and let the eye be 

 then fixed on another point distant from this small object some seven or eight 

 centimetres, so that the latter shall be only seen indirectly. At the end of some 

 instants, if the eye be kept motionless, the object completely disappears, and 

 the color of the surrounding surface seems to extend over the place which it 

 occupies. The lateral propagation of the impressions to the adjacent parts of 

 the retina is therefore as nearly as possible demonstrated. 



On the whole, then, the theory which attributes irradiation to a propagation 

 of the impression is supported by considerations of reason, analogy, and expet- 

 iment. This same theory, moreover, explains the different laws to which the 

 phenomenon is subject. Hence we infer, in the first place, that irradiation, a 

 phenomenon of sensation, cannot, as experience confirms, be the same in different 

 persons ; and that in the same eye it will be sometimes more, sometimes less, de- 

 veloped. It results, in the second place, that irradiation must appear so much 

 the greater as the luminous object is more remote; for, first, the absolute width 

 of the small band of propagated impression which surrounds on the retina the 

 image of the object, cannot depend on that object, provided the latter preserves 

 the same brightness and continues to be projected on a ground equally dark; 

 hence it follows that the visual angle, subtended by this same width, must re- 

 main equally constant whatever be the distance. Now, as the observer neces- 

 sarily refers the resulting appearance to the object itself, he must attribute to 

 the small luminous band, which to him seems added to the contour of the object, 

 an absolute width proportional to the distance which exists, or which he sup- 

 poses to exist, between the object and his eye. In effect, if we imagine for an 

 instant the appearance to be a reality, and that a similar small band is really 

 added to the outline of the object, it can evidently not be seen under a visual 

 angle constant and independent of the distance, except by making its absolute 



