of the Ionization Potential of Helium. 57l 



40 volts. This, together with the value of the primary 

 ionization potential taken from A, is a preliminary con- 

 firmation of the experiments of Franck and Hertz. 



After the cylinder and filament had been properly cleaned 

 up, results were obtained of which the curves in fig. 3 serve 

 as specimens. With the apparatus completely evacuated, 



Fig. 3. 



washed out repeatedly with air, and then reduced to a liquid- 

 air vacuum, the current-voltage curve has the form shown 

 in D. Saturation, which is nearly reached at —1 volt, is 

 complete at —10 volts, after which the current remains 

 practically constant. In B the apparatus contained pure 

 helium at a pressure of about '85 mm. Liquid air had been 

 on the U tube for six hours, but the green line of mercury 

 was still faintly visible in the spectrum. The electron 

 currents here were of such magnitude, due to the relatively 

 high temperature of the filament, that *3 microfarad was 

 necessary across the quadrants of the electrometer. Under 

 these conditions complete saturation was not attained. The 

 curve, nevertheless, shows clearly that impact ionization sets 

 in when the potential applied is 19*5 volts. Curves of which 

 A is a specimen were obtained after liquid air had been on 

 the U tube for 24 hours or more, and the spectrum contained 



