286 niiLL SySTEM TECIIMCIL JOURX.IL 



Howe\er, if the bound electrons do not radiate energy while they 

 oscillate, but accumulate it and save it np and finally discharge it in a 

 single outburst when it attains some one of a certain series of values 

 hv, 2hv, 'Shv, etc. (// stands for a constant factor, v for the frequency 

 of vibration of the electrons and the emitted radiation) — then the 

 character of the radiation will agree with that which is observed, pro\'ided 

 a suitable value be chosen for the constant //. 



The \alue recjuired ' ff>r // in C.Ci.S. units (erg seconds) is (1. .■).'{. 10-'. 



Here, llii'ii. was a phenonirm in which ilie eleclroniagnctic Hu'cirx- 

 seemed to be fundamentall\ incapaljle of explaining, hor this 

 notion of a bound electron, which oscillates and does not meanwhile 

 radiate, is not merely foreign to the classical theory, but very dan- 

 gerous to it; one does not see how to introduce it, and displace the 

 opposed notion, without bringing down large portions of the structure 

 (including the numerical agreements which I cited in a foregoing foot- 

 note). Howe\er, Planck had arri\ed at this conclusion by an intricate 

 process of statistical and thermod>namical reasoning. Statistical 

 reasoning is notoriously the most laborious and perplexing in all 

 physics, and many will agree that thermodynamical reasoning is not 

 much less so. Planck's inference made an immense impression on 

 the most capable thinkers of the time; but in spite of the early ad- 

 herence of such men as Einstein and Poincare, I suspect that e\'en to 

 this day it might practically be confined to the pages of the more 

 profound treatises on the philosophical aspects of physics, if certain 

 experimenters had not been guided to seek and to discover phenomena 

 so simple that none could fail to apprehend them, so extraordinary 

 that none could fail to be amazed. 



Honour for this guidance belongs chielh- lo liinstein. Where 

 Planck in lilOO had said simply that bound electrons emit and absorb 

 energy in fixed finite quantities, and shortly afterwards had softened 

 his no\el idea as far as possible by making it apply only to the act 

 of emission, Einstein in IflOo rushed boldly in and presented the idea 

 that these fixed finite quantities of radiant energy retain their iden- 

 tity throughout their wanderings through space from the moment 

 of emission to the moment of absorption. This idea he offered as a 

 "heuristic" one — the word, if I grasp its connotation exacth-, is an 

 apologetic sort of a word, used to describe a theory which achieves 

 successes though its author feels at heart that it really is too absurd to 



* I take the numerical values of the constant /; scattered through this article from 

 Gcrlach. The weighted mean of the exix;rimental values, with due regard to the 

 relative rclialiility of the various methods, is taken as 6.55 or 6.56. 10~". None of 

 the individual values cited in these pages is definitely known to differ from this 

 average by more than the experimental error. 



