CONTEMPORARY ADVANCES IN PHYSICS 117 



In the experiments (for instance those of Townsend and Nethercot) 

 the distance / between the electrodes was varied, the current main- 

 tained at some constant value, the voltage plotted as function of /. 

 The resulting V-vs-l curves were rising straight lines over large (but 

 not unlimited) ranges of conditions. In these experiments the gases 

 were nitrogen, helium and neon; the electrodes were external collars 

 surrounding the tube, one of which could be shifted. The same result 

 was later obtained by other pupils of Townsend (Hayman, P. Johnson, 

 F. L. Jones), who sometimes worked with internal-electrode tubes, 

 displacing one of the electrode-discs by a magnetic device. 



This result suggests that in the main part of the glowing gas there 

 is an alternating potential-gradient of constant amplitude, independent 

 of I. Denote its amplitude ^^ by b, those of current and voltage by i 

 and V: we have 



V =^ a + bl. 



Now a is to be interpreted as the sum of potential drops across regions 

 near the electrodes, where conditions differ from those of the middle 

 of the glow. 



Plotting V against / for various values of i, Townsend and Nethercot 

 found this important fact : the slope of the line, the potential-gradient 

 b, is independent of current over wide ranges (for instance, over the 

 range of i from 3 to 18 mils, in nitrogen contained in a tube of diameter 

 3.9 cm.). The difference (F — bl), however, increases with the 

 current; over a certain range of current-strengths, the increase is linear. 

 The value of the gradient b is of the order of a few volts per cm. 

 Townsend and Nethercot give for nitrogen in a tube of 3.1 cm. diameter 

 the values 13.2 (volts/cm.) at the pressure. 0.26 mm. and 19.3 at the 

 pressure 0.53 mm. For helium at 1 mm. they give 5.1; for neon at 

 1.06 mm. the value 3.5. These were obtained at the aforesaid fre- 

 quencies of 7.5 and 4 millions; and so we meet the question of the 

 dependence of b on frequency. 



The value of b was found to be nearly independent of frequency, 



so far as the rather scanty measurements go; in nitrogen, the same 



for the frequencies 4-10'' and 7.5-10'' (Townsend and Nethercot); in 



helium and neon, constant over the range from 4.7-10^ to 7.5-10^ 



cycles (Hayman); in neon, by further experiments, constant over the 



range from 2.5-10^ to 10^ cycles (Johnson). This brings us to the 



question: how does b, which is the amplitude of the alternating 



potential-gradient in the high-frequency glow, compare with the 



^^ Townsend's school give root-mean-square instead of amplitude-values for 

 sinusoidal quantities. 



