Properties of Potassium-Sodium Alloy. 291 



battery of such cells which should give a high E.M.F. by the 

 mere impact o£ light upon its most electropositive elements. 



With the above described cells we can also verity easily a 

 number of interesting observations made by previous experi- 

 mentalists. 



As regards the nature of the radiation which is chiefly 

 effective, we find that for potassium alone, or for the 

 potassium-sodium alloy, the effective rays are the most 

 refrangible ones of the visible spectrum. The ultra-violet 

 light in the arc is considerably filtered out, if not altogether 

 stopped, by the thick glass condenser-lens-, and by the glass 

 walls of the tube. We can, however, effect a further 

 separation of visible rays by. screens of coloured glass, or 

 solutions or stained gelatine films. 



If we interpose in the path of the incident light a very 

 thin film of gelatine stained with a yellow dye or a sheet of 

 ordinary yellow glass, the deflexion of the galvanometer 

 drops almost to zero. The same reduction is effected by a 

 sheet of ruby glass or gelatine film stained reel. A green 

 glass cuts off a good deal of the deflexion, but a sheet of 

 cobalt glass reduced it only to about two-thirds, showing that 

 the blue glass is fairly transparent to those rays which can 

 produce this electromotive force. 



The galvanometer deflexions increase very rapidly with 

 the intensity of the incident light, and unless an extremely 

 sensitive galvanometer is used it is necessary to throw a very 

 intense beam of light upon the surface of the metal to obtain 

 any marked evidence of the production of electromotive force 

 by the mere action of light upon the surface of the alloy. 

 If the energy required to produce this photoelectric current 

 comes from the incident light, then the latter must in some 

 degree be absorbed, and the active rays must therefore be 

 those which are absorbed by the photoelectric metal or con- 

 versely emitted by it when it is heated. We know well that 

 the light due to the violet line in the name spectrum of 

 potassium salts passes easily through cobalt glass. Hence the 

 fact that light which has passed through cobalt glass is still 

 exceedingly active in producing the photoelectric effect with 

 potassium, may indicate that it is this ray which is absorbed 

 in its production. In the case of rubidium, Elster and Geitel 

 showed that the yellow and orange rays were relatively 

 more active in the production of the photoelectric effect, and 

 a glance at the flame spectrum of rubidium salts shows that 

 the lines are most numerous in the red, yellow, and green 

 region. 



Also, we can confirm easilv another observation of the 



