HERSCHEL' LAW FOR LIGHT. gg 



The reason that a yellow substance is bleached by blue rays, is because it absorbs 

 tJiose rays, for the very same reason, therefore, that it looks yellow. A purple vegeta- 

 ble body is bleached by the yellow and green rays, and because it absorbs those rays it 

 looks purple. 



356. As respects light, the phenomena of coloration are obvious to our organs of vision ; 

 as respects the dark beams ot" heat, of chemical action, and of phosphorescence, they 

 must be hypothetical or ideal ; but, hi the same manner tbat MELLOM has found the 

 admission of them for the calorific rays of such admirable advantage, so, in this depart- 

 ment of science, similar and palpable advantages arise. Who could for a i^oinrnt 

 doubt that light and the tithonic rays were agents totally distinct, so soon as he came 

 to understand that to one of them iodide of silver is yellow, to the other orange ; to 

 one chloride of silver is white, to the other red ; to one bromide of silver is white, to 

 the other it is black ! 



357. These principles undergo a severe test when we examine the phenomena that 

 arise when sensitive surfaces of different degrees of thinness are used. As is well 

 known, if a silver plate be exposed to the vapours of iodine, it passes through severa" 

 orders of colour, red, yellow, blue red, yellow, blue, &c., &c. In each one of these 

 series, as I formerly showed, the yellows are chemically the most sensitive (Ap., 622, &c.). 



358. If, therefore, we successively expose to the solar spectrum thin plates of iodide 

 of silver of the above-named tints, we might infer that the resulting iuipression should 

 change its position with the colour of the plate ; that, if that colour was yellow, the 

 rays corresponding to the blue should be absorbed, and the spectrum impression be 

 found among the more refrangible rays ; that, if the colour was blue, a tint arising from 

 absorption of the yellow, a spectrum stain should be found corresponding to the posi- 

 tion of the yellow rays, and not to the blue, these last undergoing reflexion ; and, 

 finally, as the colour of the plate changed, so should the photographic spectrum shift 

 its position. 



359. But, on making the experiment, I found that this result does not arise; it is 

 immaterial what the colour of a Daguerreotype plate may be, the spectrum leaves upon 

 it an impression in an invariable position. In the absence of diffused light, this impres- 

 sion is entirely among the more refrangible rays. 



360. But as the colour of the plate changes, although the photographic impression 

 remains invariable in position, it undergoes variations in intensity. It exhibits the 

 deepest stain when the plate is yellow, is more faint when the plate is red, and be- 

 comes hardly perceptible when the plate is of a grayish metallic aspect. Time, of 

 course, enters as an element into these results ; a gray colour will receive as deep an 

 impression as a yellow, if the period of exposure be inordinately prolonged. 



361. From this we gather, that on exposing films of iodide of silver of different 

 thicknesses, and therefore of different colours, to the spectrum, the resulting impression 

 does not shift its place, but, remaining fixed in position, undergoes variations in sensi- 

 tiveness variations which are exhibited by differences in the depth of the stains. And 

 this result is a striking instance of the doctrine of ideal coloration. Upon the princi- 

 ples of that doctrine, it meets with a beautiful explanation. 



M 



