110 



BELL SYSTEM TECHNICAL JOURNAL 



such a curve as those of Fig. 10, and call Vmm the "minimum of the 

 least maintaining potential" or simply the "minimum maintaining 

 potential" for the frequency in question (we must choose between 

 lengthiness and lack of precision in our terms!). The value of pmm 

 and the value of Vmm both decrease with increase of v, after the 

 variation begins; consequently, a curve of Vmm vs v, such as we will 

 now consider, corresponds not to a single pressure but to as many 

 different pressures as there are points. ^^ Disregarding this com- 

 plication, notice the curve of Fig. 22. 



This is the curve of minimum maintaining-potential versus frequency 

 for air in a tube of 24 mm. internal diameter, with electrodes 19 mm. 

 apart. It is taken from Rohde, who says that the curves for oxygen. 



50 100 



V X 10-5 



500 1000 2000 



Fig. 22 — Minimum maintaining-potential vs. frequency, in air, under conditions 

 described in the context. (Rohde.) 



nitrogen, hydrogen, helium, neon and argon are similar. The com- 

 parative constancy of Vmm at frequencies below about half a million, 

 the rapid decline thenceforward as far almost as 10^, are evident. 

 Beyond, there is a definite hint that the voltage again becomes inde- 

 pendent of V, Sit a value far lower than its low-frequency or direct- 

 current amount; for each of the aforesaid gases, hydrogen alone 

 excepted, Vmm was sensibly the same at 1.39-10^ (Rohde's highest 

 frequency) as at 6.95-10^. 



One is struck by the resemblance to what Reukema and Lassen 

 observed of the sparking-potential across atmospheric air : approximate 

 constancy up to a certain critical frequency, ensuing decline, eventual 



'' The only extensive set of published curves of Fm-vs-j* is that of H. Gutton, which 

 is more complex than one would expect (see below). 



