208 



TRANSACTIONS OF THE 



[mar. 1, 



6. 



7. 



8. 



9. 



10. 



120 and green, 

 red and green, 

 39 and green, 

 blue and green, 

 red and yellow, 



red and green. 

 123 and green, 

 yellow and green. 

 128 and green. 

 123 and yellow. 



Each pair-wise combination of small disks was placed concen- 

 trically upon the corresponding pair-wise combination of large 

 disks. (See Fig. III.) The four disks thus combined were 



Fig. III. 



jDlaced upon the axis of the rotation machine and rotated at a 

 sufficiently high speed to cause all flickering to vanish. The 

 speed was then gradually diminished and in each case, as near as 

 could be judged, the flickering began inside and outside simulta- 

 neously. Now in each of the above combinations the difference 

 in luminosity between the disks opposed pair-wise is practically 

 the same for the small as for the large disks. On the one hand, 

 however, the difference is between two differently colored disks, 

 on the other hand, an equal difference in luminosity occurs be- 

 tween two gray disks or between a gra^^ and a colored disk. 



From these experiments we ma}^, therefore, conclude that, 

 under ordinary daylight illumination, the violence of a flicker, 

 due to a combination of two gray disks of diff'erent luminosities, 

 is not affected if for one or both of the gray disks a colored disk 

 of equal luminosity is substituted. It will be seen that the 

 above set of experiments constitute a rather severe test of the 

 assumption that under ordinary daylight illumination flicker de- 

 pends on luminosity and is independent of color. 



To further test the assumption the following experiments 

 were made : From a number of colored disks the luminosities 

 of which were unknown to me, I selected by the method of 

 Least Flicker a number of pairs which gave only a very feeble 



