204 Prof. Millikan and Mr. Winchester on Influence of 



the rate of discharge of each metal is a perfectly definite 

 and invariable quantity, while Ladenburg could obtain no 

 constancy at all except in the case of aluminium {cf. § L0). 

 It should be said, however, that since our metals were in 

 contact with air for some hours before the tube was closed 

 and exhausted, any changes in the photo-electric sensitiveness 

 which may be due to such oxidation as might occur in that 

 time have not been eliminated from our observations. 



A second reason for the differences in the relative photo- 

 electric sensitiveness shown in Ladenberg's results and our 

 own, is to be found in the fact that if, as Lenard's experi- 

 ments and those herein recorded seem to demonstrate, photo- 

 electric discharge is due to atomic resonance, the order in 

 which the different metals exhibit the phenomenon should 

 depend upon the nature of the source, and indeed Lenard 

 obtained an indication of such a dependence*. The subject 

 is, however, one upon which further experiments are desirable 

 and to which attention is being at present directed. It is 

 sufficient for the present purpose to point out that our expe- 

 riments add conclusive evidence to that already brought 

 forward by Ladenburg, that photo-electric effects in a 

 vacuum bear no relationship whatever to the Volta contact 

 series of the elements. 



The apparent existence of such a relationship in air, as 

 shown by Elster and Geitel's and Ramsay and Spencer's 

 experiments, is doubtless to be accounted for as follows f : — 

 In air there is formed at the surface of metals in contact 

 with oxygen a double-layer which is more strongly positive 

 on the side of the metal in the case of copper, gold, silver, 

 and other metals which stand at the negative end of the Volta 

 contact series, than in the case of aluminium, zinc, lead, and 

 other metals which stand near the positive end of this series. 

 This double-layer is doubtless due to the incipient action of 

 oxygen on the metal, just as, in the case of the contact 

 potential between a liquid and a metal, it is due to the action 



* Lenard, Ann. d. Phys. viii. p. 163 (1902). 



f In these considerations it is assumed that the nature of the Volta 

 contact effect is as follows : — When a condenser is made of two different 

 metals, for example zinc and copper, which are immersed in air and 

 separated by shellac, that metal which is most highly positive with 

 respect to oxygen, namely copper, will be the one which, when the plates 

 are connected, receives from the other an excess of negative electricity. 

 When, then, the connexion is broken and the plates separated, free 

 negative electricity will be found upon the plate which is most strongly 

 positive with respect to oxygen, i. e. upon the copper. This is the plate 

 which, according to convention, is called electro-negative with respect to 

 the other. 



