116 BELL SYSTEM TECHNICAL JOURNAL 



high enough to make it probable that electrons would often acquire 

 the stated energy. But this is not to be taken as universally true 

 for all the values of fieldstrength which have been observed in high- 

 frequency glow-discharges.2* 



Certain data were obtained by Brasefield in experiments on air 

 over a frequency-range extending downward from Kirchner's, and 

 contained in Gutton's: that is to say, from 2-10^ down to 1.25-10^. 

 The electrodes — external belts of metal wrapped around a tube of 

 4.5 cm. diameter — were no less than 40 cm. apart; and instead of 

 measuring the least maintaining potential, Brasefield measured at 

 various pressures the amplitude V of the voltage existing between the 

 electrodes when a current of amplitude 100 mils was passing. The 

 resulting F-vs-^ curves for diverse frequencies had the customary 

 form, concave-upward with single minima. As the frequency was 

 raised from 1.25-10'' to 1.5-10^, the value of the minimum voltage 

 and that of the pressure at which it was attained both trended down- 

 ward, though with peculiar brief rises. As the frequency was further 

 raised from 1.5- 10^ to 2- 10^, there was a sudden tremendous upswing 

 of the minimum voltage, and a rise of the corresponding pressure, — 

 anomalies recalling the singularities of Gutton's curves. Under the 

 conditions prevailing at the minima of the curves for these two highest 

 frequencies, there was agreement (within the wide limits of uncer- 

 tainty) between Kmv and the ionizing-potential of hydrogen, and 

 between D from the first of equations (33) and the electronic mean 

 free path. 



In the direct-current glow-discharges in a cylinder of gas contained 

 in a tube, under certain conditions, there is a region (the so-called 

 "positive column") throughout which the fieldstrength is uniform 

 and low, and either decreases slowly as the current or the current- 

 density is increased, or else remains sensibly the same. This region is 

 apparently uniform in color and brightness. (I am not taking account 

 of cases where it is "striated," or cases in which it is visibly dimmer 

 near the wall than near the axis.) In the high-frequency glow- 

 discharge there is also, under certain conditions, a region of uniform 

 color and brightness occupying all of the tube except small portions 

 near the electrodes. Townsend and his school undertook to measure 

 the (alternating) fieldstrength in this region, and to compare it with 

 the values obtained in the direct-current glow. 



'''' If D computed by equations (33) should turn out to be very many times as 

 great as the electronic mean-free-path, the proper procedure would be to compute 

 the maximum energy of the oscillating electrons by the conventional method from 

 the general equation (equation 5 of Part I) for electrons moving in dense gases. 

 I fear, however, that in most cases the ratio of D to the electronic mean-free path 

 is not great enough to allow of passing to this limiting case. 



