ACTIVATION OF ELECTKICAL CONTACTS BY ORGANIC VAPORS 809 



values leaves little doubt that we have correctly interpreted the in- 

 hibiting effect of glow discharges upon activation. 



6.4 ''Hysteresis'^ Effects 



Sometimes a pair of completely inactive electrical contacts of a noble 

 metal can be activated very quickly, and an apparently identical pair 

 of contacts cannot be activated at all under exactly the same experi- 

 mental conditions. In the first case a great amount of carbon may be 

 formed, and in the second case no detectable carbon at all. In order to 

 clear up this confusion, some controlled experiments were carried out 

 upon the activation and deactivation of silver and palladium electrodes 

 in air containing benzene vapor at various partial pressures. From these 

 experiments, it has been possible to relate the variability of earlier results 

 to previous history of the contacts, and the entire behavior is now quite 

 well understood. 



In these tests adjustable benzene vapor pressure was obtained by 

 first bubbling air at a controlled rate through benzene maintained at 

 constant temperature by a bath of acetone and dry ice, and then mixing 

 the saturated air with clean air in the proper proportions. In certain 

 tests silver contacts were operated in air flowing from this apparatus, 

 discharging a capacitor on each closure. The number of operations re- 

 quired to produce complete activation was measured for many different 

 values of the benzene vapor pressure. With contacts that had been 

 cleaned in a standard way before each test, it was found that the number 

 of closures required for activation rose extremely rapidly with decreasing 

 vapor pressure over a narrow range of pressures. There was always a 

 lower pressure limit below which it seemed impossible to activate the 

 contacts at all. All of the tests were made at the high operating rate of 

 60 closures per second. 



When the contacts had been cleaned by abrasion before each indi- 

 vidual test, the minimum pressure below which activation could not be 

 attained was of the order of 2 mm Hg. (This pressure was exceptionally 

 high because of the high operating rate, see Section 6.2.) A different re- 

 sult was found for contacts that had been previously activated and 

 then cleaned only by repeated arcing; for these contacts the minimum 

 pressure for activation was about 0.7 mm Hg. The factor of 3 between 

 these minimum pressures is doubtless related to the fact that, for those 

 electrodes which had been cleaned by arcing only, there existed neigh- 

 boring carbonized areas which were never cleaned. Each such area can 

 be expected to hold about three times as much adsorbed benzene on the 

 average as does the same area of clean metal, see Section 5.1, and espe- 

 cially lines 2 and 3 of Table IV. Thus for such surfaces more carbon can 



