0. N. Rood— Color System. 269 



Chrome yellow (paste) *846 



" " -933 



"Pale" " " -991 



Emerald green " 1*119 



Prussian blue (wash) __• *354 



" " ' " -392 



Cobalt blue (paste) *562 



Artificial Ult. blue (paste) .... *506 



Crimson lake (wash) 1*23 



The system which is here proposed can be built up by 

 experimenters and ought to be practically identical in all cases 

 if made by normal eyes, provided the same standard vermilion 

 is used, or the coefficient of it is known, or in general if the 

 starting point is made from any disc of bright color the coeffi- 

 cient of which is known. As before stated it may be possible 

 hereafter to refer the coefficient of the vermilion disc to white 

 card-board, which indeed has been done in the case of my own 

 disc with some approximation to correctness. The system 

 may then be considered a fairly reproducible one, but the 

 question still remains as to its nature. The thought that in it 

 we have one where complementary colors having a determined 

 coefficient of unity furnish the same gray, naturally suggests 

 itself and is rendered probable by the following considerations. 

 The standard red and its true complement when mixed furnish 

 a certain gray, and it is evident that other pairs situated at 

 small angular distances from them, such as 10° on either side, 

 will sensibly do the same. This furnishes two sectors of 20° 

 on either side of white ; but the coefficients of the colors 

 situated within them are reproducible with the aid of the 

 triangle, which would hardly be the case unless the colors 

 situated at its angles obeyed the same law, viz : furnished the 

 same white when mixed with their true complements ; this 

 again applies to the reversed triangle and the two provide for 

 about two opposite quadrants of the circle, and what holds 

 good of these ought, it would appear, to apply to the remain- 

 ing quadrants. 



The question whether in this system, all the colors located 

 on the circumference furnish by mixture the same white when 

 treated in pairs, is quite difficult to answer by actual experi- 

 ment, as in the case of most of the discs it is very hard to 

 ascertain the relative amounts of colored and white light 

 reflected, owing to the well known fact that in most cases por- 

 tions of the colored light mix, and produce what amounts to 

 an indeterminate amount of white light. Some attempts, with 

 the aid of the spectroscope, of colored glasses and of gray 

 discs were made to measure the actual amount of standard 



