774 THE BELL SYSTEM TECHXIf'AL JOl'KXAL, M\Y 1957 



neath the electrodes. It is clear that the increased arcing of activation 

 is caused by solid carbonaceous material made b}' decomposition of 

 organic vapor and not by the vapor itself. Clean metal contacts can, in 

 fact, be made to show all of the symptoms of activation by allowing soot 

 from a flame to settle on their surfaces.*! Activation produced in this 

 way is, of course, temporary, lasting only until the deposited soot is 

 burned away. 



When one looks for characteristics of arcs between active surfaces to 

 which numerical values can be attached, four features come at once to 

 mind — the electric field at which an arc strikes, the voltage across the 

 arc after it is established, the minimum arc current (which is just the 

 current at which the arc goes out), and finally, after the arc is over, the 

 amount of metal that was gained or lost by each of the electrodes during 

 the arc. All of these ciuantities have been measured for active arcs as 

 well as for arcs at clean surfaces, and a brief summary of the results of 

 the measurements is given here. 



1.1 Striking Field 



To measure the electrode separation at which an arc strikes between 

 closing electrodes, relay contacts were operated repeatedly, discharging 

 on each closure a capacitor charged to a measured voltage. An arc at 

 each closure was assured by using short leads between the capacitor and 

 the contacts to keep the circuit inductance very low. The time from the 

 initiation of the arc to the touching of the contacts was measured on an 

 oscilloscope. J 



Fig. 2 illustrates the results of measurements made b}^ F. E. Ha- 

 worth^ upon palladium electrodes closing at 30 cm sec to discharge a 

 veiy small capacitor charged to 50 volts. Before the start of the experi- 

 ment the electrodes had been cleaned by repeated arcing in air, and the 

 first experimental point represents the closure of these clean electrodes. 

 All of the other measurements were made in air containing a fairly high 

 partial pressure of limonene vapor. Each point plotted on the curve rep- 



* Unpublished work of P. P. Kisliuk. 



t It is interesting to point out that a surface is not made active by rubbing 

 petrolatum upon it, although activation will occur very quickly if an electric 

 current is made or broken at such a greasy surface, so that some of the grease is 

 decomposed. 



t For inactive arcs, it is necessary to make a correction for the height of the 

 mound of metal thrown up by the arc (Reference 6, page 1136). After the contacts 

 become active, there is no appreciable mound thrown up (at palladium surfaces), 

 and the electrode separation at the initiation of the arc is calculated at once from 

 the closure time and the previously measured electrode velocity. The height of 

 the mound produced before the contacts are active was minimized b.y using a 

 capacitance of only 40 X 10^'^ farad. 



