206 



ROYAL SOCIETY OF CANADA 



curve, one in the violet and the ot^her in the green. It would seem 

 natural to expect the curve obtained 1 when the eye was acted upon by 

 both colours of wave leno;ths .577 ii and .474 j«, to have three elevia- 

 tions to con'espond' with, those in figures 8 and 9, viz: in the red, green 

 and violet. Instead of this, however, the persistency curve in figure 6 

 shows two eleva^tions — in the rod and violet — ^and Oi I depression in the 

 green: the anticipated doiihle elevation in the green is actjually replaced 

 by a depression. As I the duration of colour impressions on the retina is 

 an invei-se function of the luminosity, we have the pai-adoxical conclu- 

 sion that floubly fatiguing the green sensatiooi with blue and yellow light 



Fi(i. 9. Fig. 10. 



together causes green light to make a more luminous impression than it 

 does normally. Further,! since the parts of the curve disturbed by fatigu- 

 ing the retina with blue and yellow are those corresponding to red, green 

 and violet, it follows that noaie of ijthe colour sensations are independent 

 of each other; while the fact that the curve sjuffers disturbances in three 

 parts seems to favour the tjheory of three fundanienta.1 colour sensations. 

 Thq persistency curve in figure 7 was obtained, as explained before, 

 when the eye was fatigued with red and green of complementary hues. 

 Now, when the retina is acted on bfy red, or by green separately, a per- 

 sistency curve with but one elevation is obtained, which occurs in the 

 Ipart of the curve corresponding to the colour used in fatiguing the eye. 



