CONTEMPORARY ADVANCES IN PHYSICS W 



axis of abscissae. The common frequency is 73 or 107 kilocycles 

 (Kampschulte seems to have found no difference between the behavior 

 of the two), except for the curve marked "50 cycles" which as before 

 may be regarded as the curve for steady voltage. The lowest of the 

 curves pertains to electrodes sharpened at their ends to cones, with an 

 angle of 30° at their points. 



Fig. 15 is still more interesting, although the data were obtained with 

 electrodes of a curious and inconvenient shape — collars or rings of 

 metal, sometimes with sharp edges and sometimes with rounded 

 edges, as the little sketches beside the curves suggest. Even for the 

 blunt-edged electrodes the breakdown occurs at notably lower voltages 

 for frequencies of the order of one hundred thousand than for 50- 

 cycle or steady voltage, unless the gap-width is small. For the sharp- 

 edged electrodes the difference is still more striking. 



Most interesting of all is the triad of curves at the bottom of Fig. 15. 

 If the gap-width between the sharp-edged electrodes is set (for instance) 

 at 10 cm., and voltage of frequency 10^ is applied and gradually in- 

 creased, three transitions follow one after another: first, the establish- 

 ment of a durable self-sustaining discharge of a certain aspect; then, 

 its sudden transformation into another durable self-sustaining dis- 

 charge of visibly different aspect ; lastly, the advent of the spark. The 

 first transition may be regarded as the breakdown of the initially un- 

 disturbed gas; the second, as the breakdown of the gas when ionized 

 in the peculiar way prevailing in the first of the self-sustaining dis- 

 charges; the third, correspondingly. With steady voltage likewise, 

 sparkover is anticipated by the onset of a durable self-sustaining dis- 

 charge if the ratio of gap-width to radius-of-curvature of one electrode 

 is sufficiently high.^* For the fifty-cycle A.C. applied to the sharp- 

 edged electrodes, Kampschulte displays in Fig. 15 only the curve for 

 sparkover; but he implies in the text that the other two discharges 

 were observed to precede the spark. 



Accurate explanation of these laws is lacking. The most that 

 has been achieved is a rough test of a certain rough inference from 

 theory. 



Say that we establish a certain gap-width : determine the sparking- 

 potential with steady or low-frequency A.C. voltage; and then apply 



""Electrical Phenomena in Gases," pp. 301-303, 443-445; note Fig. 64 on page 

 302 (taken from F. W. Peek, Jr., "Dielectric Phenomena in High- Voltage Engi- 

 neering") which shows the critical potentials for the durable discharge or "corona" 

 dropping below that for sparkover at a certain value of the ratio of gap-width to 

 radius. The terms "Glimm" and " Biirsten" used by Kampschulte would be 

 translated literally as "glow" and "brush," but usage in English is so uncertain 

 that I have done so with hesitation. 



