22 



Dr. H. E. Ives on 



blue image merges into the rod image the red and blue will 

 again approach each other. 



Fig. 3. 



*7S 



>so 



•25 



■ 

 • — T B -T R 



/ 2 3 4 5 



Loo b 



Time-interval between blue and red, and between red and after-images, 

 calculated from data of fig. 1. 



3rd Case. 



A coloured (photopic) image, and a colourless (scotopic) 

 image. 



It has been a debated question whether the colourless 

 rod image is actually present at high intensities. On the 

 assumption that it is 5 we have the condition 



a\o i 



£ A = — , 

 9 



a 



g<a log I + &, 

 1 



. (9) 

 .(10) 

 • (11) 



a log I + o ' 



I£ we call the rod image the after-image, it should lag 

 behind the photopic image by an interval which increases 

 with the intensity in the manner shown in fig. 3. If the 

 interval is measured after a blue image it becomes zero 



