294 Photoelectric Properties of Potassium-Sodium Alloy. 



alloy through a galvanometer. These experiments and others 

 show that to obtain the effect well a very high vacuum is 

 necessary. In the process of preparing the tube the potassium 

 and sodium should be melted, and the alloy and tube well 

 heated for some time whilst the pump is going to drive off 

 all traces of water vapour, also of hydrocarbon derived from 

 the naphtha in which the metallic potassium is commonly 

 preserved, and especially to drive off the hydrogen which 

 seems to be occluded in considerable quantity by the alka- 

 line metals, no doubt derived from the decomposition of 

 water vapour. 



The effect of variation of pressure of the gas upon the 

 photoelectric effect was first investigated by Stoletow, and an 

 account of his researches is given in Prof. Sir J. J. Thomson's 

 treatise on the " Conduction of Electricity through Gases " 

 (p. 224). He found that the current increased rapidly as 

 the pressure diminished, which continued until the current 

 reached a maximum value, after which it began to decline, 

 but had a finite value at the lowest attainable pressures. 



Stoletow's experiments appear to have been made with 

 zinc plates and ultra-violet light. Owing to the greater 

 manipulative difficulty when dealing with the more oxidi- 

 sable and electropositive metals no attempt has been made to 

 push these investigations with potassium-sodium alloy very 

 far. The primary object in view in conducting them was to 

 ascertain if the photoelectric effects could be utilized as an 

 oscillation detector in radiotelegraphy; but as a type of glow- 

 lamp detector has now been found by the writer far more 

 efficient than that originally proposed by him, involving the 

 employment of a carbon filament, these photoelectric experi- 

 ments have not been pursued. They are put on record here 

 merely for the sake of aiding any who may wish to show 

 them as interesting lecture or class experiments, or pursue 

 the purely physical investigation of the effect itself still 

 further. 



The question of photoelectric effects is not without interest 

 in connexion with long distance radiotelegraphy. It has 

 been shown that perfectly dust-free air is not ionized by 

 ultra-violet light. If, therefore, the absorption of long radio- 

 telegraphic electric waves which is found to exist when they 

 pass through considerable distances of sunlit air is due to the 

 presence of free ions in the air, these may arise from the 

 photoelectric action of the light upon the dust particles. 

 This suggests the question whether these particles may not be 

 the same that create the blue colour of the sky. We know 

 that whenever photoelectric effects take place, light must be 



