552 THE BELL SYSTEM TECHNICAL JOURNAL, MAY 1954 



cleaned by filing then washing with methyl alcohol and distilled water. 

 The contacts were slowly brought together until glow discharge was 

 established. Before measurements were made the circuit current was 

 increased to allow the glow to cover the entire cathode flat surface as 

 well as a portion of its cylindrical surface. By allowing the contacts to 

 glow for about 20 minutes, the occasional arcing first observed was 

 eliminated and a steady glow was established. The cathode was observed 

 under a microscope and the current was adjusted to obtain a glow just 

 covering the flat cathode area. From the measured current and the 

 cathode area, the cathode current density was determined. The results 

 are given in Table II. 



POWER SUPPLY V " -«— MICROSCOPE 



Fig. 12 — Circuit for measurement of cathode current density in normal glow 

 discharge at static conditions. 



Observations on Glow Maintenance and Glow-arc Transitions 



The simphfied circuit in Fig. 13 was used. The contact cathode was 

 the flat end of a cyhnder. The cylindrical portion was tightly fitted into 

 a block of an insulating material allowing an exposure of the flat end and 

 a cylindrical area less than 10 per cent of the flat area. The anode was a 

 parallel plain surface of the same material. 



To avoid the unnecessary complications of a measuring circuit con- 

 nected to the contacts, the plates of a cathode ray oscilloscope were, 

 instead, connected across a capacitor, 10 times C, in series with the circuit 

 capacitor C. From the transients obtained, it was possible to identify 

 glow discharge, steady arcs and interrupted arcs. Four typical transients 

 are shown in Fig. 14. Transient A shows a case where glow discharge 

 was established and maintained for the entire half period of the circuit. 

 In transient B glow was not detected and, instead, interrupted arcs 

 occupied the entire half period. In transient C, glow discharge was 

 maintained for a short duration 1-2 followed by interrupted arcing, 2-3. 

 At point 3 the circuit current was high enough to maintain an arc and a 

 steady arc was obtained, 3-4. Transient D is similar to B where glow 

 discharge was undetectable. The multiple discharge in D, however, 

 lead to the steady arc 2-3. 



Before presenting our measurements and discussion, a review is given 



I 



