Visual Diffusivity. 31 



of ten to twenty degrees, and the two then rotated together 

 until the after-image of one appeared in line with the other. 

 The advantage of this apparatus was that the eye could be 

 much more easily kept fixated at the centre of rotation, half 

 of the phenomenon of interest being on each side of the 

 centre. In the apparatus constructed illumination was 

 necessarily from the front, which made it difficult to have 

 a perfectly black background ; perhaps substantially better 

 results could be obtained with a similar device arranged for 

 transmitted light, for the improvement in ease of fixation is 

 considerable. However, there was no indication that the 

 gap between high and low illumination conditions would be 

 appreciably better filled by this scheme, so that the con- 

 struction of such an apparatus was not undertaken. 



Some observations made with this clock-hand device, 

 using white light *, indicated a bending back or return of 

 the after-image just before its disappearance at the highest 

 illumination. This might be due to a failure of the recti- 

 linear logarithmic relation between diffusivity and illu- 

 mination, such as some earlier work on critical frequency 

 suggested, or again it might be due to the rod function 

 showing a tendency to increase its diffusivity just before 

 failure. The clock-hand observations differed in magnitude 

 from the ones plotted, but in the opposite direction to what 

 better fixation would have accounted for, probably due to 

 the different part and area of retina used. 



5. Discussion. 



The lagging of a blue image after a red is such a striking 

 sight that it seems unlikely to have escaped observation, 

 although I can find no mention of it. As an effect predicted 

 from a physical theory of the transmission of impressions 

 from the retina to the seat of consciousness it may, I think, 

 claim novelty. 



The positive after-image has, of course, been known for a 

 very long time, and it is, moreover, a part of von Kries's 

 Duplicitats-theorie that this image is the rod image, delayed 

 in perception by the greater reaction time of the rods. The 

 existence of a relationship between the illumination and the 

 time-interval separating the after-image from the primary, 

 such as is here suggested and in part borne out, appears to 

 be new. 



* When the white hands are illuminated by the yellow-white light 

 of the tungsten lamp, the after-image appears blue ; if a blue glass is 

 placed over the eye the after-image appears brown, pointing clearly to 

 the colour being a subjective contrast effect. 



