212 STUDIES IN LUMINESCENCE. 



be produced in the ionic concentration, and diffusion from the surface 

 layers inward will bring about a change in the decay curve similar in char- 

 acter to that caused by irregularities in the distribution of the active sub- 

 stance. Diffusion produced in this way would ultimately result in an appre- 

 ciable diminution in the surface concentration of the active material, and 

 we should therefore expect that the intensity of luminescence would be 

 diminished in such a substance by prolonged excitation. No effect of this 

 kind has been observed by us in the case of Sidot blende, but the obser- 

 vations of Werner 1 with a SrZn compound show evidence of fatigue resulting 

 from prolonged excitation. 



The change in the form of the decay curve due to changes in the duration 

 of excitation may be explained, at least in a general way, as a result of 

 diffusion. Diffusion of the ions will occur during excitation as well as 

 during decay. After prolonged excitation, therefore, the volume occupied 

 by ions will be greater than after short excitation, and the rate of decay in 

 other words the slant of the decay curve will be correspondingly reduced. 2 



Diffusion also offers an explanation of the phenomena of hysteresis in 

 phosphorescent substances. After prolonged excitation and subsequent 

 decay the neutral molecules that result from recombination will be dis- 

 tributed through a larger volume than before. Since the original distribution 

 of the active material was a stable one there will be a gradual return to the 

 normal distribution. But this will be a slow process and may well require 

 several days for its completion. In the mean time the material is in such 

 a condition that the decay following renewed excitation will be more 

 gradual than the normal, even for a short excitation. The spreading out 

 of the active substance by diffusion, which would normally require a long 

 excitation, has already been accomplished by the preceding excitation, 

 whose effects have not yet disappeared. Hysteresis effects such as those 

 discussed in Chapter IV are therefore to be expected. 



If this explanation of hysteresis is correct the effect of the infra-red rays 

 must be to facilitate the return of the substance to its normal condition; 

 in other words, to increase the rapidity of the diffusion by which the original 

 distribution of the active material is restored. It is natural to expect such 

 an effect in substances which are able to absorb the infra-red rays. But if 

 the restoration is accomplished by the diffusion of neutral ions the rapidity 

 of the action is surprising, for with strong infra-red rays we have found that 

 an exposure of only a few seconds is sufficient to restore Sidot blende to 

 its standard condition. 3 



INFLUENCE OF IONIC GROUPING. 



In discussing thermo-luminescence and the effect of infra-red rays upon 

 phosphorescence Wiedemann and Schmidt 4 have suggested that some of 

 the ions produced during excitation form semi-stable combinations or 



'A. Werner, Ann. cler Phys., 24, p. 164, 1907. 



! If the volume occupied by ions is vi for short excitation and j for long excitation, the ionic concentra- 

 tions will be n/vi and njvi respectively, and the two intensities of phosphorescence will be proportional, to 

 an 2 /vi and an/ 2 /vi. Prolonged excitation therefore produces an effect which is equivalent to a diminution 

 in the coefficient of recombination. 



a See Chapter V. 



"Ann. der Phys., 56, p. 247, 1895. 



