ON ACCIDENTAL OR SUBJECTIVE COLORS. 235 



width vary proportionally with that distance. If, now, we could demonstrate 

 that this apparent augmentation really corresponds to a visual angle constant for 

 the same eye and the same object, this demonstration would constitute one of 

 the most powerful arguments that could be brought in favor of the theory in 

 question; for this constancy of the visual angle is the principal character of the 

 phenomena of vision, which depend on a modification undergone by a portion 

 of the retina of a constant extent. This important fact M Plateau has suc- 

 ceeded in establishing by the help of an apparatus which we will not at present 

 stop to describe. It is sufficiently demonstrated, by a long series of observa- 

 tions made on many persons, that the angular value of the irradiation i.-i inde- 

 pendent of the distance from the object to the eye; and it thence resuUs, as a 

 necessary corollary, that the absolute width which we attribute to the irradiation 

 is, all things else being equal, proportional to the distance which exists, or which 

 appears to us to exist, between the object and ourselves. Irradiation had not 

 been observed hitherto by physicists and astronomers, except for remote objects. 

 It will be found, however, to exist for small as well as great distances — only, 

 the apparent effect diminishing proportionably with the distance, it strikes us 

 less in the case of near objects; but, from the experiments of M Plateau, it 

 may be accepted as an ascertained fact that irradiation manifests itself at all 

 distances, from the shortest susceptible of distinct division to any degree of re- 

 moteness. There was still an important question to resolve It was necessary 

 to determine approximately the irradiation in the case of the same person in 

 different circumstances. This M. Plateau has done, and has satisfied himself 

 that for the same individual, and with an object of the same brightness, irradia- 

 tion varies considerably from one day to another. 



In the third place, it is well established that irradiation increases with the 

 brightness of the object, and, again, the hypothesis of a propagation of the 

 impression on the retina renders the most satisfactory reason for this fact; for 

 it will not be disputed that the more energetic the excitation, the greater will be 

 the distance to which it is propagated. But observation further demonstrates 

 that the augmentation does not increase proportionally to the brightness ; that 

 its progress is much less rapid. For example, the angular diameters of the sun 

 and moon being little different from one another, the brightness of the solar 

 disk must be to that of the disk of the moon very nearly in the same ratio of 

 intensity as the light which reaches us from these two bodies. But the bright- 

 ness of the solar disk is nearly a thousand times that of the disk of the moon. 

 If irradiation, then, increased proportionably with the brightness of the object, 

 that which the sim developed would be enormous relatively to that of the moon, 

 and the first of these orbs would present to the naked eye the aspect of an im- 

 mense globe The aspect really presented to us by the disk of the sun compels 

 us, therefore, to admit that irradiation augments much less rapidly than the 

 brightness of the object which produces it. Hence it follows that if the law 

 which connects these two quantities be represented by a curve having for ab- 

 scissas the brightness, and for ordinates the corresponding irradiation, this curve 

 will turn its concavity towards the axis of the abscissa? ; and as the increase of 

 irradiation, at first very considerable when we commence with a feeble bright- 

 ness, becomes at last insensible when the brightness attains a certain limit, the 

 curve in question will have an asymptote parallel to the axis of the abscissaj. 



M. Plateau has sought to verify these conclusions by direct experiment, and 

 to obtain the outline of the curve in question. To do so there was but one 

 condition very difficult to fulfil — that, namely, of giving to the object successive 

 degrees of brightness, at (mce determinate and having a known ratio with one 

 another. This he accomplished in a very simple manner by taking advantage 

 of a principle of photometry which we shall examine when considering the 

 measurement of the intensities of light. 



It remained to appreciate the influence exerted on irradiation by the greater 



