26 EOYAL SOCIETY OF CANADA 



tions slower than in any of the other cases. Curve (2) was obtained 

 after spending 16 minutes in darkness, in the middle of the day — it 

 does not show so great a periodicity and it is not so smooth as curve ( 1 ) . 

 The three curves for coloured light are not as smooth as either curve (1) 

 or curve (2), but they all denote more rapid oscillations. The curves 

 were obtained by " fixing]" respectively, the red, green and indigo 

 glasses in the daj^time and taking observations as already described. 

 The initial periodicity for indigo is the smallest, then comes green, and 

 then red. The curves for I B and G- overlap considerably during the 

 remainder of the runs, although the average periodicities of platted 

 points is in the same order as above (viz.: I. B. 566: G. 5S6 ; R. 652). 

 These cur\^es are not exactly comparable with one another as they were 

 made on different days. From these three lower curves no conclusions 

 regarding relative periodicity can well be drawn, and they are here 

 presented merely to show the nature of curves taken, in this way. 



58. These curves show that for all sensations (besides the gradual 

 slowing down of the oscillations) there are somewhat irregular periods 

 of increasing and decreasing periodicity. Undoubtedly, some of these 

 changes, especially those where a single point deviates considerably from 

 a smooth curve, arise from accidental errors in counting, or in press- 

 ing the key. Since these accidental errors cannot surely be detected 

 the curves were drawn through all points. 



59. In several sets, taken for the purpose, with assistant and stop- 

 watch, these changes in periodicity were found to be synchronous with 

 the apparent darkening and brightening up of the object or coloured 

 glass that was " fixed " : an increase in periodicity corresponding to a 

 darkening and a diminution of periodicity corresponding to a brighten- 

 ing up of the " background." The phenomenon of the alternate dark- 

 ening and brightening of an object, " fixed " for some time, is well 

 known.i 



60. The curves. Plates II, III and IV, were obtained in a sim- 

 ilar manner to those of Plate I; however, these curves represent the 

 average; of 10 sets each, more colours were used, and only three or four 

 readings taken in each set in order to obtain the initial periodicities 

 for different coloured light and. besides, in the first series (Plate II: 

 and Figs. 1 and 2, Plate III) changes were made in the intensity as 

 well as in the colour of the light. * 



61. Plate II shows the curves for red, green and blue sensations 

 with three different intensities, viz. : Fig. 1, the intensity of the light 



^ See Helmholtz " Physiologi.sche Optik."' 2nd Ed.. p. .511. With absolutely 

 steady " fixation " the intensity probably gradually (apparently) decreases; but 

 since it is impossible to absolutely " fix " the eyes for many seconds, the bright- 

 ening up may correspond to slight changes of fixation point. 



