23 G ON ACCIDENTAL OR SUBJECTIVE COLORS 



or less illumination of the field surrounding the luminous object. As the dis- 

 position of the retina to receive propagated impressions bears, in virtue of the 

 facts and observations above developed, an inverse ratio to the disposition of 

 the same organ to receive direct impressions, we are authorized to believe that 

 when the light acts directly on a portion of the retina, that portion becomes, 

 from this circumstance itself, less apt to receive a propagated impression. 

 Hence it is readily conceived that if the ground itself on which the luminous 

 object is projected conveys to the eye a certain quantity of light, the direct 

 impression which it produces -will counteract the irradiation of the border of the 

 object, and so much the more as the brightness of the ground is greater. This 

 being granted, let us imagine an object which is projected on a ground not 

 wholly deprived of light, and suppose that the illumination of that ground is 

 made to increase gradually. The irradiation produced along the outline of the 

 object will go on diminishing until the lustre of the field has become equal to 

 that of the object. This limit being passed, if the former continues to increase, 

 it is evident that the field will then produce, in its turn, an irradiation Avhicli 

 will encroach upon the object, and which will consequently be developed in an 

 inverse dii'ection to the former as regards the line which bounds the real 

 contour of that object. The irradiation will have passed, so to say, from posi- 

 tive to negative. This change, which is a direct consequence of the known 

 facts, justifies the assumption that at the instant when the lustre of the field 

 has beome equal to that of the object the irradiation of the latter is reduced to 

 zero; and as the effect must be reciprocal, if, instead of an object and surround- 

 ing field, two objects be supposed of equal brightness and in contact, the irra 

 diations of these two objects will be null at the point or line where they touch 

 This result has been verified. 



To evince directly the influence exerted by the greater or less illumination 

 of the field, recourse may be had to the following means : We begin by stain- 

 ing with black, for half its length, a rectangular piece of white paper, which is 

 thus divided into two rectangles, one completely dark, the other semi- 

 transparent; this latter half should be oiled for the purpose of increasing its 

 transparence. In this paper we cut with the point of a penknife a longitudinal 

 opening five millimetres in width, of which one half traverses the black, the 

 other the translucid space. We then stretch this paper on a frame, or, what is 

 better, paste it on a pane of glass. When this apparatus is placed against a 

 window, it is apparent that the longitudinal opening will form a luminous band, 

 one half of which is projected on a black ground, and the other on a ground 

 possessing a certain degree of brightness, although much less than that of the 

 band. If we now place ourselves at the distance of some metres, the two 

 halves will appear to be unequal widths, the first seeming considerably to 

 exceed the second. 



From the foregoing discussion it results that the hypothesis of the propaga- 

 tion of the impression to the parts adjacent on the retina explains in a satisfac- 

 tory manner all the laws of irradiation manifested to the naked eye. Unfor 

 tunately, adds M. Plateau, when, in order to observe the effects of' irradiation 

 we equip the eye with a lens, a wholly different order of facts is presented. 

 Thus, while a certain apparatus observed with a naked eye and at the distance 

 of distinct vision manifested a very apparent irradiation, the virtual image of 

 the same apparatus, produced by a strong magnifying glass situated at the 

 same distance, and possessing moreover perceptibly the same brightness 

 exhibits no appreciable irradiation. We are thus led to this singular conclu- 

 sion, that magnifying glasses appear to possess the faculty of considerably 

 diminishing ocular irradiation. Is not this because the lens, by restricting the 

 luminous pencil withoiit proportionably modifying its brightness, causes it to 

 occupy a smaller space on the retina? M. Plateau does not say so, and does 

 not attempt to explain this peculiarity. With a convergent lens of sufficiently 



