SOMF. CO.V77:,U/'()A'./A'r . //)r. /.V(T.V IX rilVSlCS- III JQl 



interprt'tatioii, IniC it si-rms iinprDtitaMi* to i-nltT into ilio oilicrs. It 

 is tluTt-fori' /i,„.,x whii'li apiH-ars to nu'rit llu- most .iiti-ntinn. 



Now the miTo fact that then- is a ina\iiiuin) M-liuitN- of thi- i-sraprd 

 I'kvlrniis, that there is an /i,„„,. is not in itself of a nature to snyjuest 

 that the classical theory is ina(le(|iiate. It is tiie peculiar depeiiden^-e 

 of this quantity on the two most important controllable (lualities 

 of the lijiht — on its intensity and on its frecpienc}*- wliidi awakens 

 the hrst faint suspicions that something has at last been disco\ered. 

 which the classical theor\' is ill adapted to explain. One would 

 pretlict with a k'xxI tli'jd <'f confidence that tiie greater the intensity 

 of the light, the i^reater the energy acfjuired by the electron in each 

 cycle of its forced oscillation would be. the greater the energy with 

 which it woukl finally break away, the greater the residuum of energy 

 which at the end would be left to it. F3ut /i|„;,x '^ found to be inde- 

 j)endent of the intensit>' of the light. This is strange; it is as though 

 the waxes beating upon a beach were doubled in their height and the 

 powerful new waxes disturbed four times as many r>ebbles as before, 

 but did not displace a single one of them any farther nor agitate it 

 any more violently than the original gentle waves did to the pebbles 

 that they washed about. As for the dependence t)f £niax "" tht> fre- 

 quency of the light, it would be necessary to make additional assump- 

 tions to calculate it from the classical theory; in any case it would 

 probably not be ver\' simple. But the actual relation between £ma.x 

 and V is the simplest of all relations, shf>rt of an absolute proportion- 

 ality; this is it : 



E,„..^ = bv-P (1) 



Fig. 3 shows the relation for sodium, obserxetl l>y Miliikan. 



The maximum energy of the photoelectrons increases linearly with 

 the frecjuency of the light. P is a constant which varies from one 

 metal to another. In the terms of the simple foregoing interpreta- 

 tion, P is the energy which an electron must spend (more precisely, 

 the energy which it must invest or convert into potential energy-) 

 when it passes through the frontier of the metal on its way outward. 

 Comparing the xalues of P for several metals with the contact poten- 

 tials which they display relatix'ely to one another, one finds powerful 

 evidence confirming this theory. Having discussed this particular 

 aspect of the question in the fifth article of this series, I will not 

 enter further into it at this point. 



The constant // is the same for all the metals which ha\e been 

 use<l in such exfieriments. The best determinations have been made 

 upon two or three of the alkali metals, for these are the only metals 



