204 



ROYAL SOCIETY OF CANADA 



are depressed below 'tihe normal in the middle which corresponds largedy 

 to green. The intersections of the curves in each figure (figs. 6 and 7) 

 occur in the blue at ah^ut À = .455 /i. and in the orange ne; r the 

 yellow at about X = .62 /i. 



Though the effect on the retina of white light i of the comparatively 

 few wave lengths comprised in the complementai^ colours must of neces- 

 sity be different! from that of the complex mixture of waves forming the 

 white light from an electric arc, yet the eye is unable to see any differ- 



Fiu. 7. 



ence in the appearance of i the two lights when both are of the same inten- 

 sity. Measurements of [the persistenjce of vision made in the way die- 

 scribed are, however, capable of showing, in a most striking manner, the 

 difference in the effect of the two white lights. 



It is most remlarkable j that the curves in figures 6 and 7 are de- 

 pressed below the normal in the part corresponding to green. For, when 

 the eye is fatigued with yellow light {X = .577 fi) alone, for example, 

 the resulting persistency curve has two elevations, one in the red and the 

 other in the! green, as shown in figure 8, which is plotted from the data 

 in table 6. 



No persistency curve has been obtained when the eye was fatigued 

 with the complimentr.ry colour, blue, of wave length .474 /x. Figure 

 9, however, is a persistency curve obtained when the fatiguing stimulus 



