EFFECT OF LIGHT ON ELECTRICAL CONDUCTIVITY OF SELENIUM. 379 



light ; at a depth x the intensity is I e~ ax and the energy 

 absorbed in a thickness dx is a I e~ ax clx. 



If e x is the effect produced at a depth x we may assume 

 e x = f(al e- ax ) 



?X 



and the total effect produced is E = \ f (a I e~ ax ) dx 



Jo 



X being the thickness of the selenium. 



The effect is not proportional to the absorption of energy, 

 for if it were the increase in conductivity would be pro- 

 portional to the intensity of the light incident on the cell, 

 whereas it rises more slowly than the intensity as is seen 

 in fig. 2. 



Let us assume tentatively that the effect produced is 

 proportional to the nth power of the energy absorbed where 

 n < 1. 



rX 



Then E =[ pa n l n e~ anx dx, (p being a constant) 



_ p G n ~ I" /y e —anX\ 



n 



If X is very small E = pa n I^X, so that the greater the 

 value of a, the coefficient of absorption, the greater is the 

 effect produced, that is the cell has maximum sensitiveness 

 for the light which is most strongly absorbed. 



^ a n— 1 jn 



If X is infinitely great E = t 1 and as n < 1, the 



n 



smaller a is the greater the effect ; that is the light which 



is least absorbed produces the maximum effect if n is the 



same for all intensities. 



For intermediate thicknesses the effect will be a maxi- 

 mum for some value of a which is neither the maximum or 

 the minimum, but which would seem to depend on the 

 thickness, the value of a for which the effect is a maximum 

 being greater the thinner the layer of selenium in the cell. 

 With many of the cells tested the increase in conductivity 



