586 BELL SYSTEM TECHNICAL JOURNAL 



foregoing laws. The waves are then absorbed — that is to say, their 

 amphtude tails off exponentially as they progress — and their phase- 

 velocity is altered. In the particular case in which g and therefore a 

 are negligibly small, there is no absorption, and the phase-velocity is 

 altered in the ratio 1 : Ve — a fact which is described by saying that 

 the index of refraction of the electron-populated gas is equal to Ve. 

 Since e is less than unity according to equation (8), the waves go 

 faster than they would if there were no free electrons roaming the gas. 

 Formulae for index of refraction and absorption-coefficient, in the 

 general case in which g and a do not vanish, are given further along 

 in this article (equations 15, 16). The transmission of radio waves 

 through the atmosphere of the earth, in which free electrons are 

 present in a concentration varying greatly with altitude, is much 

 affected by this refraction. 



Experiments on Oscillation of Electrons in Ionized Gases 

 There is a method which, in principle, is adequate for measuring 

 the dielectric constant and the conductivity of an ionized gas subjected 

 to high-frequency vibrations; adequate therefore, in principle, for 

 testing the expressions (7) and (8), for evaluating the quantities g 

 and N which figure in those expressions. I will describe the method, 

 first in the form in which it was applied to ionized air by Szekely, 

 with results which are of some interest in themselves but mainly 

 serve to illustrate how great a gulf intervenes between "adequate in 

 principle" and "adequate in practice." 



The gas — air at some pressure of the order of a few hundredths of 

 a millimeter of mercury — was contained in a long tube, having elec- 

 trodes near its ends whereby a direct current could be passed through 

 the tube, maintaining the air in a state of intense ionization. Inside 

 the tube and near its middle were a pair of parallel plane electrodes, 

 rectangular in shape (18 by 42 mm.) and 4.5 mm. apart; to these 

 the high-frequency voltage was applied ; I will speak of them hereafter 

 as "the ionization-condenser." When the direct current was flowing, 

 a self-sustaining glow-discharge existed in the tube, and the ionization- 

 condenser was submerged in the negative glow thereof. 



The ionization-condenser was shunted by an adjustable condenser 

 outside of the tube, and the two of them by an inductance and an 

 adjustable resistance; all this constituted a circuit — I will call it 

 hereafter the "high-frequency circuit" — coupled at one place to a 

 source of high-frequency E.M.F., at another place to a detector. 

 With such an arrangement, when one varies the capacity of the 

 adjustable condenser and leaves everything else unchanged, the 



