710 Messrs. H. E. Ives and E. F. Kingsbury on the • 



was tested out experimentally, namely, that the two alter- 

 nated colours in the flicker photometer give no flicker when 

 the intensity of each is such as to call for the same individual 

 flicker. This means taking the two colours in the relative 

 intensities at which they have the same critical frequency. 



Fig. 2. 



Isochroinatics, showing variation in relative brightness of spectrum 

 colours due to variation of illumination, Flicker Photometer. (Kepro- 

 duced from previous paper, and to be compared with theoretical curves 

 fig. 8.) 



If the assumption were correct it would follow that if one 



plotted log - against log I s (where I s is the intensity of the 



■*-s 



standard illumination, and I c that of any colour) straight 

 lines should be obtained. An extremely careful experimental 

 test gave not straight lines but instead the curves of fig. 2. 

 Several factors were studied which might more or less 

 influence the results, but the real nature of the connexion 

 between the critical frequency phenomena, represented \bj 



