196 TRANSACTIONS OF THE [mAR. 1, 



dent of color. The method depends upon the shock which the 

 retina experiences when one surface is withdrawn from view and 

 replaced by another of different luminosity. If, for example, a 

 disk, one-half of which is of different luminosity from the other, 

 is rotated before the eye a flickering is observed with appropriate 

 rates of rotation. Professor Rood states that the flickering dis- 

 appears if the two halves are of equal luminosity, no matter what 

 their differences in color may be. For the determination of the 

 luminosity of a colored disk a series of one hundred gray disks 

 were used. The gray disks of the series varied quite uniformly 

 in luminosity from the luminosity of white bristol board to the 

 luminosity of lamp black. The luminosity of a given colored 

 disk was then determined by successively combining it " pair- 

 wise " with various gray disks until one would be found with 

 which it would give no flicker on rotation. The luminosity of 

 the gray disk could then be determined in the ordinary way and 

 this would represent the luminosity of the colored disk. The 

 number thus obtained to represent the luminosities of the differ- 

 ent colored disks were controlled in the following way : Two 

 disks of complementary colors were selected. Their luminosi- 

 ties were measured by the flicker method. These disks were then 

 combined so as to form a gray on rotation. The luminosity of 

 this gray was then measured in the ordinary way and also cal- 

 culated from its color equation. In all cases it was found that 

 the measured luminosity of the gray agreed with the calculated 

 luminosity, the latter being dependent upon the flicker, within 

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



The flicker method has been adapted to use on the optical 

 bench by Frank P. Whitman.* In the summary of the results 

 obtained by his investigations he remarks : 



"The flicker photometer used to compare lights of any color 

 approximates in convenience and accurac}'^ any of the ordinary 

 photometric appliances used with lights of the same color. 

 Different observers whose vision is normal obtain like results. 



" The instrument gives a true measurement of the luminosity 

 comparable with that obtained in other trustworthy ways." 



The results obtained by Professor Rood and Mr. Whitman 

 promise a solution of one of the most difficult problems of pho- 

 tometry, viz., the comparison of the luminosities of differently 

 colored surfaces. While the ordinary photometric methods have, 

 in general, been most thoroughly worked over, the application of 

 the phenomena of flickering to photometric measurements being 

 entirely new opens a very promising field for investigation. 

 The present research was undertaken at the suggestion of Pro- 



* Physical Review. Vol. III. No. 4. Page 241. 1896. 



