606 BELL SYSTEM TECHNICAL JOURNAL 



vs. H\ the remarkable minimum occurs at a value of H departing by 

 less than 2 per cent from lirmcv/e. These data were obtained at 

 frequency 5.46-10^; others observed at 2.96-10^ and 3.45-10^ yielded 

 agreements almost as good, or better. 



Maxima of conductivity of ionized air have lately been found by 

 Jonescu and Mihul, at the predicted values of field strength H, for 

 various frequencies of the order 10^. Earlier Benner, in a note 

 deplorably brief, showed not only a curve of conductivity (or rather, 

 of something proportional to conductance) displaying a maximum, 

 but a curve of dielectric constant displaying a crinkle like that of a 

 dispersion-curve in the neighborhood of a region of anomalous disper- 

 sion. These are the curves of Fig. 9, already introduced into this 

 article to illustrate how these quantities vary in the vicinity of a 

 natural frequency. The "ionization-condenser" of Benner's experi- 

 ment consisted of the grid and plate of a triode, the space between 

 them populated with electrons emitted from the filament; it is to be 

 inferred that the tube contained some gas, but unfortunately nothing 

 definite is said about the kind or amount. The maximum of the one 

 curve and the crinkle in the other occurred at a value of H some 

 twelve per cent higher than the predicted value aforesaid. 



On applying a longitudinal magnetic field of about 21 gauss to the 

 tube of ionized hydrogen with which he had observed the peculiar 

 natural frequency mentioned above, H. Gutton found this one replaced 

 by two, well marked and well separated, one being shifted toward 

 higher frequencies from the original value and one toward lower. 

 The same phenomenon has been observed by Tonks, in his studies 

 of the natural frequencies which he identifies with the plasma-electron 

 oscillations predicted by the theory culminating in equation (24). 

 The doubling of the resonance is analogous to the Zeeman effect, 

 and is amenable to theory. 



The earth's magnetic field imprints a natural frequency upon the 

 electrons populating the air, in particular the upper strata thereof; 

 this afifects the transmission of radio waves in curious ways, which 

 were foretold in this journal seven years ago by Nichols and Schelleng, 

 and in England by Appleton and Barnett. 



In the second part of this article, I will treat of the conditions 

 under which a high-frequency field may initiate and maintain a 

 luminous discharge in a gas, and of the laws of these discharges. 



