636 Prof. E. C. C. Baly on 



These examples are sufficient to show not only that the 

 constant frequency differences exist, but also that these 

 differences are of ten 1/X = 400 or some multiple of 400. It 

 is not easy to measure the centre of an absorption band from 

 an absorption curve as obtained with a spectrograph ; and 

 therefore great accuracy has not been possible in the above. 

 Somewhat the same is true of fluorescent observations, for 

 very little accurate work has been carried out on the position 

 of the maximum. Investigations are now in progress in this 

 laboratory both on absorption and fluorescence in order to 

 determine whether this constant frequency difference is a 

 general law. It is hoped that these results will shortly be 

 ready for publication. 



Now in any case of the selective absorption of light by a 

 substance, it may be presumed that the light energy is 

 absorbed in definite quanta, defined by the well-known 

 expression hv, where h is the Planck constant and v the 

 oscillation frequency of the absorption band. If now the 

 absorption of the light energy is not accompanied by any 

 photochemical change whereby new substances are produced, 

 it is reasonable to make the simple assumption that the 

 energy thus absorbed is transformed into heat and evolvf d 

 again as such. Furthermore, if the substance possesses a 

 characteristic absorption band in the infra-red, it is to be 

 expected that the light energy absorbed will be evolved in 

 radiation having the same wave-length as that of the infra-red 

 absorption band. If this be the case, then this heat energy 

 will also be evolved in quanta. The whole of the energy 

 absorbed in the form of light should be evolved in the form 

 of heat, and therefore for every quantum of light energy 

 a certain number of complete quanta of heat energy must be 

 evolved. If v be the oscillation frequency of the absorption 

 band in the light region, and v x be that of the band in the 

 infra-red, then clearly 



hv = nhv x and v — nv x , 



where n is some whole number. 



In words, the oscillation frequency of the ultra-violet 

 absorption band should be a multiple of the oscillation 

 frequency of the infra-red absorption band. 



If the above assumption be justified, namely that a quantum 

 of light energy absorbed is completely evolved as a whole 

 number of quanta of light energy, and that the relation holds 

 good that 



nv x = z> , 



