178 SCIENCE PROGRESS. 



they must be composed of quartz or other really trans- 

 parent substance. 



The statements here made about the need for extremely 

 ultra-violet rays are, however, subject to modification with 

 the nature of the o-as surrounding the metallic surface, for 

 Wiedemann and Ebert found that in an atmosphere of 

 carbonic acid even visible rays produced some effect ; and 

 a still more marked modification is dependent on the nature 

 of the metallic surface. I have spoken of zinc as a suitable 

 metal, and so it is ; but, with highly ultra-violet light, cop- 

 per, silver, and even platinum, do nearly as well. With 

 light of lower refrangibility, zinc is better than the less 

 oxidisable metals, but, as Elster and Geitel have found, 

 potassium or sodium is better still. With such metals in- 

 deed any kind of light serves, not only ordinary diffuse 

 daylight, but even the light of a candle. 



INFLUENCE OF THE KIND OF METAL. 



It appears that the order of metals in photo-electric 

 power, when tested by their discharging action for negative 

 electricity, is the same as their order in the voltaic series, 

 the alkali metals heading the list, and the noble metals 

 being the least effective. With light of ordinary wave 

 lengths, indeed, such as come easily through glass, only the 

 alkaline metals, or those above zinc in the voltaic series, 

 have any real power ; but with short enough wave lengths 

 all metals are competent to show the effect, though the noble 

 metals require light of an extremely high ultra-violet 

 character. 



This fact of the voltaic order of the metals clearly 

 suggests some chemical or semi-chemical explanation, but 

 it may be well to defer coming to a conclusion on that 

 point for the present. 



DISCHARGING ACTION OF METALS FOR POSITIVE 



ELECTRICITY. 



Hitherto we have spoken, as the earlier experimenters 

 in the subject did, as if negative electricity alone was dis- 

 charged while positive electricity was retained. 



