DECAY OF PHOSPHORESCENCE IN SIDOT BLENDE. 53 



In all of the experiments upon the decay of phosphorescence with which 

 we are familiar it is the total light that has been measured; so far as we 

 are aware no attempt has been made to determine the law of decay for 

 different portions of the phosphorescence spectrum. This fact complicates 

 the problem greatly, for in most cases of phosphorescence the spectrum 

 consists of two or more bands, which, in general, decay at different rates. 

 It can hardly be expected, therefore, that measurements of the total light 

 will be found to obey a simple law. The difficulties resulting from the 

 complexity of the luminescence spectrum were recognized by E- Becquerel 

 and were several times mentioned in the course of his classic researches. 

 In discussing his experiments on decay he suggested an expression of the 

 form 



I = Ae- at + Be- iii + ... (6) 



for the intensity of the total phosphorescence, with one term for each 

 band in the spectrum. But the difficulties of computation were such as to 

 lead him to abandon this expression and to employ instead the empirical 

 expression of eq. (2). Recognition of the fact that each band has its own 

 rate of decay is also implied in the law proposed by H. Becquerel (eq. 4). 



While it is possible to test the correctness of an expression of the form of 

 (4) or (6) by comparison with measurements of total intensity, such a test is 

 difficult, and can not be altogether satisfactory; for with an expression 

 containing several terms the number of constants is so great that the law 

 may be made to fit almost any data. It is clear that a much more severe 

 test of any given law of decay may be obtained from experiments with a 

 substance having only one band in its phosphorescence spectrum. The 

 Sidot blende screen used by us in earlier experiments was found to have 

 three bands in its luminescence spectrum. But since the two violet bands 

 do not appreciably overlap the green band the spectrophotometer enables 

 the behavior of the latter band alone to be conveniently studied. The 

 matter is still further simplified by using the violet rays of the arc in excita- 

 tion, since these rays, while they produce a brilliant green luminescence, are 

 incapable of exciting either one of the violet bands. 1 In view of the bril- 

 liancy of the green band, its long duration, and the ease with which it maybe 

 isolated, this band seems well adapted to the study of the decay of phos- 

 phorescence. 



EARLY STAGES IN THE DECAY OF PHOSPHORESCENCE IN SIDOT BLENDE. 



Our previous experiments on Sidot blende have shown that in the case 

 of the green band the phosphorescence spectrum shows no measurable 

 change in form during the first 10 seconds of decay. In other words, the 

 rate of decay is the same for all wave-lengths. To fix the behavior of the 

 whole band it would be sufficient, therefore, to determine the law for a 

 single wave-length. We have, however, made measurements at three 

 different wave-lengths, namely, at 0.483 ju, 0.512 n, and 0.547 M- The first 

 of these wave-lengths lies near the ultra edge of the band ; the second is not 

 far from the maximum; and the third is near the red edge. 



'See Chapter III of this memoir. 



