206 TRANSACTIONS OF THE [MAR. 1, 



98.1 per cent. D. G. + 1.9 per cent. W. = 15.3 per cent. G. + 1.1 

 per cent. B. + 77.0 per cent. Bk. 



The luminosity of the dark green disk calculated from this 

 equation is 9.6 per cent. 



The luminosity of each of the above four colored disks was 

 then independently determined by the method of Equal Flickers, 

 the average of six determinations for each disk being taken. The 

 agreement between the results obtained by the two methods is 

 shown in the following table. 



Description of the Luminosity determined by the Luminosity calculated from 



colored disks. method of Equal Flicker. the color equations. 



Straw Yellow 34.4% 34.3% 



Light Orange 54.6% 56.6% 



Dark Brown 23.2% 23.0% 



Dark Green 10.1% " 9.6% 



The greatest difference between the luminosities of any one of 

 the disks determined by the two methods is 1.1 per cent. 



The determination of the luminosity of a colored disk b}^ 

 matching it, using a few standard disks, is based upon the as- 

 sumption that the luminosities obtained by the flicker method 

 satisfy Grassman's law for color mixtures. The experiments of 

 Professor Rood* and Mr. Whitman "j" show that the assumption 

 is correct within the experimental error which was about 2 per 

 cent, of the luminosity of white bristol board. Similar experi- 

 ments made by myself gave like results. The luminosities of 

 several gra3's, formed by combinations of three colors, were 

 measured in the ordinary way and also calculated from the color 

 equations. The diflerence between the numbers obtained by the 

 two methods for the luminosity of a given gra^^ was in no case 

 over two per cent, of the luminosity of white bristol board. The 

 agreement between the luminosity of a given colored disk de- 

 termined by the method of matching and the luminosity of the 

 same disk determined directly bj' the flicker method is another 

 proof of the correctness of the assumption. The experiments 

 referred to in this paragraph constitute, so far as I know, the 

 only experimental proofs so far given of the correctness of 

 Grassman's law for color mixtures. 



Flickering is Found to be Independent of Color. 



In the preceding section it was stated that for the determina- 

 tion of luminosities by means of the phenomenon of flicker- 

 ing, the assumption must be made that the violence of the 



* American Journal of Science, September, 1893. 

 t Physical Review, Vol. III., No. 4. 1896. 



