530 M. H. Helmholtz on the Theory of Compound Colours. 



the phjenomenaj so far as I am able to see, completely; showing 

 that the mixture of the substances and the combination of their 

 colours are two processes altogether distinct, and hence that the 

 results obtained from the former furnish no conclusion regarding 

 the latter. Only when we have to deal with colours which stand 

 but slightly separated from each other in the spectrum does the 

 composition of the coloured light give nearly the same results as 

 the mixture of the pigments, for then the compound colour is 

 similar to the tones of the spectrum which lie between both the 

 simple ones. 



There are, however, two other methods of combining the light 

 proceeding from pigments, which yield results altogether in har- 

 mony with those obtained from the combination of similar 

 prismatic colours. The tirst of these methods is the imion of the 

 colours upon the rotating disc. It has been long noticed, that 

 results thus obtained are different from those derived from the 

 mixture of the pigments. I repeated the ex23eriment with yellow 

 and blue. For the former I cither made use of gamboge or 

 chrome-yellow, for the latter idtramarine or mountain-blue. 

 With quick rotation a pure gray is obtained. The difference of 

 the two methods is exhibited very strikingly when the middle of 

 the disc is coated with a mixture of both pigments, while the 

 rim is divided into sectors coloured by the pure pigments them- 

 selves. With quick rotation the middle of the disk appears 

 green, the rim gray. The former is much darker than the latter, 

 which according to the foregoing theory must be expected. 



Of the other method I have never yet found a description, but 

 can recommend it as very convenient. It is free from the defect 

 of the gray appearance of the mixed colours which is observed 

 upon the rotating disk, and admits, on the contrary, of the 

 generation of a perfect white from complementary-coloured 

 pigments. Let a glass plate, with plane and parallel surfaces, 

 be placed perpendicular to the leaf of a table, and let a coloured 

 wafer be placed before it. The image of the wafer is reflected 

 by the glass plate ; the appai'ent place of the image is at the 

 other side of the plate, and also on the surface of the table. Let 

 another wafer of a different colour be placed upon the exact spot 

 where the image is observed, this second wafer being seen through 

 the glass. The observer's eye will thus be affected by two de- 

 scriptions of rays, both of which appear to proceed from one and 

 the same body, one of which however belongs to the transmitted 

 and the other to the reflected light. Hence he observes a wafer 

 the colour of Avhich is compounded of those of the two wafers 

 actually before him. To make the experiment with greater con- 

 venience, it is only necessai'y to use a very small glass plate, as 

 thin as possible, and with plane parallel surfaces ; this is to be 



