292 Considerations on Colours, 



containing sensibly opnke parts disseminated throughout a 

 pellucid substance. By the same reasons he explains the 

 colours under which the grayish dust collected bv age on 

 old paper or coloured stufls shows itself, and deduces the 

 same consequences in regard to the blucish appearance of 

 the veins of the human body. 



Ffc proposes, likewise, a new method of rendering very 

 sensible the colours of contrasts more lively even than by 

 the known process of accidental colours, and yet without 

 occasioning extraordinaiy fatigue to the eye. The latter 

 condition is of importance, for it is known that it is dan- 

 gerous to expose to forced exercise an organ so delicate as 

 the eve. 



This method consists simply (when one is in an apart- 

 ment and in the open light) in placing before the window 

 the painted pieces of paper on which you inteiul to observe 

 the contrasts, as in the example before mentioned. As the 

 coloured paper, then, wliich serves as a field has a semi- 

 transparcncv, and bv these means is more illuminated, 

 while the small band of another colour placed over it is, on 

 account of the double thickness, more opake and in the 

 shade ; the colour arising from the contrast becomes thus 

 more striking. 



It is this disposition \\hich produces the singularly strik- 

 insj cflcct of the contrast of a small piece of white card ap- 

 plied successively to paper, glass, or stuff of any colour 

 whatever. When the transparent body is red, the opake 

 white appears of a blueish green ; it is then seen decidedly 

 blue if the ground be orance ; then of a sort of violet on a 

 yellow ground, or grcn on crimson, 8cc. ; alwavs accord- 

 ing to the exact correspondence of the coniplcvicniary co- 

 lours. 



It is here to be observed, that according to the rule in- 

 dicated, if from the white wiiich is formed bv the unioxi 

 of all the coloured rays we suppress, ior cxauiplc, the red 

 ravs, the remaining bundle ought to be seen under the co- 

 lour of a vcrv pale blucish green ; but as the small white 

 piece in the prcccdinci; experiment is in ihe shade, the block 

 which results irom it may be of the degree proper for de- 

 stroving the cilect of the while, and then the llucish green 

 appears of a briglit shade. The same reasoning is appli- 

 cable to the cases of all the other colours. 



To produce well the eflects here announced, in repeating 

 thtse experiments, it is necessary, when the f)ppor!uniiy of 

 clear weather has been obtained, to guard ag.::r..>i. liic reUcc- 

 tioas of nci-hbouring bodies, and against duuble contrasts. 



Thus. 



