A. W. Swell — Air in an Intense Electric Field. 369 



should be definite, the air under investigation, between elec- 

 trodes and glass, was frequently renewed, the glass was cooled 

 whenever it became appreciably warmer than the electrodes 

 and the electrodes were maintained at a constant temperature 

 by ten liter water baths directly behind them. For the small 

 polar distances employed, such electrodes were able to keep air 

 and glass approximately at their own temperature. 



Electrode B was 30 c ' m X 30 cm . Electrode A consisted of an 

 inner portion 20 cm X 20 cm , soldered directly to the water tank, 



L10 Volts 



60C,cUs 



and an outer guard ring insulated from the tank, the separation 

 from the central portion being l mm and the outer edge 30 cm X 

 30 cm . The central portion and the guard ring were accurately 

 in the same plane. Without the guard ring there was consid- 

 erable spreading of the discharge at the edges, which became 

 particularly serious when the difference of potential was so 

 high that the electric intensity was sufficient to ionize the air 

 along paths bending far out from the direct line of the edges. 

 The current through the air opposite the central portion was 

 determined from the fall of potential across the non-inductive 

 resistance F. The insulation of the guard ring was assured 

 by testing with a telephone, and by measuring the current 

 through jT when F' was varied. If the insulation was perfect 

 the current through i^was practically independent of F. 



Am. Jour. Scl— Fourth Series, Vol. XXII, No. 13].— November, 1906. 

 26 



