30 Dr. H. E. Ive 



s on 



the nature of the objects, is hard to see, especially when in 

 motion. An increase of speed, which would increase the 

 separation, is no aid because of the spreading out and loss 

 of definition of both images. As a consequence the after- 

 image practically disappears, especially as a measurable 

 thing, until at very low illuminations the primary (blue) 

 image becomes, as noted, of the same order of brightness as 

 the after-image, when the latter again appears close in, but, 

 because of the low brightness, not easily susceptible of exact 

 measurement. 



A further obstacle to measurement is the fact that the 

 eye quickly loses its ability to see the after-image, and must 

 be rested. This does not appear to be at all a matter of 

 adaptation, in particular of dark adaptation, but of time 

 alone. It was found by experience that the eye was best 

 prepared by an interval between observations of fifteen or 

 twenty minutes at least of use in ordinary occupations of 

 reading, &c, in the normally day-lighted room. Probably 

 under these conditions the rod function is completely in- 

 hibited and permitted to recuperate, whereas at the illu- 

 minations where both primary and after-image are present 

 this same function is being more rapidly exhausted than its 

 normal rate of recuperation under conditions of pure rod 

 vision. 



The measurements obtained are shown in fig. 7. The dots 

 were observations made with the tungsten lamp, as described 

 above, the after-image being brought into coincidence with 

 the red image of the other aperture placed from ten to twenty 

 degrees behind. The crosses represent observations made 

 with an arc lamp with both apertures blue, a condition 

 presenting some advantage at those intensities where the 

 red aperture becomes decidedly brighter in appearance than 

 the blue, due to the Purkinje effect. All observations are, 

 however, reduced to give the interval between red and after- 

 image, by use of the blue-red interval previously determined 

 for the same conditions. It will be seen that these observa- 

 tions, limited in range and lacking in precision for the 

 reasons given, are in agreement, so far as they go, with 

 the hypotheses presented as to the nature of the after-image. 

 They are not sufficient in range or precision to prove it, and 

 so its chief support must be in the qualitative observations 

 described above. 



In an effort to find conditions of easier visibility for the 

 after-image a different form of apparatus was improvised, 

 consisting of two white strips carried on an axis like the 

 hands of a clock. One of these was turned through an angle 



