CONTEMPORARY ADVANCES IN PHYSICS 101 



Lassen's experiments, the fieldstrength E at breakdown is about 

 30 kv./cm., for every gap-width between 0.3 and 2 cm.; if for the 

 drift-speed of positive ions at this fieldstrength one puts 10^ and for 

 the amplitude of the oscillations puts the amount of the gap-width, 

 one gets 10^ for the critical frequency at gap-width 0.3 mm., and this 

 — as Fig. 13 displays — is the proper order of magnitude. A like 

 agreement is obtained with Reukema's data. But the values postu- 

 lated for the drift speeds are scarcely more than guesses (in Lassen's 

 case it is assumed that the mobility at 30 kv./cm. is two and a half 

 times what it is at one volt per cm.); and plausible as the theory 

 seems, the experiments help it but little. 



On the other hand, observations have been made on the number of 

 ions formed by an electron on its way across air at atmospheric 

 pressure, at fieldstrengths of the order of those existing in these experi- 

 ments. ^^ This is an exponential function of E, and small variations of 

 E thus make enormous differences in it. Lassen figures that just before 

 breakdown at frequency 2.45 • 10^ an electron crossing the gap (of any 

 width between 0.2 and 2 cm.) produces 36 ion-pairs, while just before 

 breakdown at constant voltage it would produce no fewer than ten 

 million. This is a striking result. 



Breakdown-Potentials in Gases at Low Pressures 

 Breakdown across a stratum of gas of low density — that is to say, 

 having a pressure of a few millimeters of mercury, or a few tenths or a 

 few hundredths of a millimeter — is normally followed by the establish- 

 ment of a durable self-sustaining discharge, oftenest of the type called 

 "glow." This rule, which for a gas at atmospheric pressure prevails 

 only if one at least of the electrodes is so much rounded that its 

 radius of curvature is decidedly smaller than the gap-width, is not thus 

 limited at those lower densities. For an obvious reason, the rarefied 

 gas is always confined within a tube, which in most of the experiments 

 with high frequencies (those on the ring-discharge excepted) is a 

 cylinder only a few centimeters wide; thus, to judge from experience 

 with direct-current discharges, the presence of the wall must have a 

 great influence on the phenomena. The electrodes are commonly 

 either discs inside the tube near its ends, or belts of tinfoil wrapped 

 around the outside of the tube; at high frequencies it often makes 

 surprisingly little difference which, and yet such differences as have 

 been reported are sometimes noteworthy. Breakdown-potentials are 

 generally determined by raising the amplitude of the high-frequency 



16 M. Paavola, Arch. f. Elektrolechnik, 22, 443-458 (1929); "Electrical Phenomena 

 in Gases," p. 278. 



