264 VASIL OBRESHKOVE 



For, as the intensity factor increases and the reaction-time 

 becomes short, a secondary period, for secondary processes after 

 sensitization, of appreciable vakie may cause an increase in the 

 intensity-reaction-time products. From another standpoint it 

 may be claimed that in the lower range of intensities the Bunsen- 

 Roscoe law holds good, but with higher illumination the velocity 

 of changes taking place in the receptors is no longer proportional 

 to the intensity factor. In such a case a proportionally longer 

 period is required to produce the amount of stimulating substance 

 necessary in order to activate the nerve endings. 



In searching for the duration of a period occupied by secondary 

 processes after sensitization, during which the animal need not 

 be illuminated in order to respond to the stimulus, the following 

 method was employed. By increasing the intensity of light, 

 the reaction-time was reduced to approximately three seconds. 

 It was found possible to reduce further the reaction-time by 

 employing still higher degrees of stimulation. Thus the average 

 value obtained with 200 candle-meters was found to be 1.7 

 seconds and with 500 candle-meters 0.76 second. 'With higher 

 intensities than 500 candle-meters, it was found difficult to 

 measure accurately with a stop-watch the reaction-time, because 

 of its extreme shortness. The actual values obtained with 60, 

 200, and 500 candle-meters are given in table 5. 



If the reaction-time is composed of two periods, a sensitiza- 

 tion period and a period for other processes following sensitization, 

 then from the reaction-time with 500 candle-meters it is seen 

 that 0.76 second must represent both these processes. From 

 this it can be concluded that the duration of this secondary 

 period must be very short and that, whatever its value is, it 

 must lie at or below 0.76 second. If the deviation noted with the 

 higher intensities in the intensity-reaction-time products is due 

 to such a period, then a value lying between and 0.76 second 

 ought to be found which when subtracted from the average 

 reaction-times for the various intensities would yield results 

 capable of right-line plotting. That this is not the case is seen 

 from figure 7, in which the curves were plotted by the same 

 method as was employed in figure 6, except that from average 



