ACTIVATION OF ELECTRICAL CONTACTS BY ORGANIC VAPORS 779 



each of the individual arcs was 40 ergs, and for the longest arc time 600 

 ergs. The results indicate no significant variation of the erosion per unit 

 of energy over this range. 



There is some evidence that arcs at the break of active palladium sur- 

 faces give significantly lower cathode erosion per unit of energy (1 or 2 X 

 10""'* cc/erg) than do arcs at closure. The reason for the difference is 

 not clearly understood, but widely different currents and electrode sepa- 

 rations may be significant factors. 



By examining contacts of palladium or platinum after many active 

 arcs (on either break or closure), it is found that the erosion tends to be 

 uniform over the surface, wearing each electrode down smoothly, with 

 much less loss from the anode than from the cathode. This type of wear 

 is quite different from that produced by arcs at clean surfaces. Erosion 

 by arcs at clean surfaces always gives a mound of metal on one electrode, 

 with a corresponding pit in the other ; the loss of metal from one electrode 

 is not appreciably greater than the gain by the other, the entire erosion 

 consisting simply of transfer of metal between the contacts. 



Now the inactive arcs at clean surfaces are known to be of two types 

 which have been called "anode arcs" and "cathode arcs."^- "^ In anode 

 arcs, the transfer of metal is about 4 X 10~^* cc/erg and is from anode 

 to cathode, with a resulting pit in the anode and a matching mound on 

 the cathode (Reference 9, page 1085-1086). In inactive cathode arcs, 

 measurements made in the same way and not yet published have shown 

 that the transfer is smaller — about 1 X lO"'* cc/erg — and is in the 

 opposite direction, from cathode to anode, with a resulting pit in the 

 cathode and a matching mound on the anode.'** It will be shown later. 

 Section 2.4(a), that arcs at active palladium surfaces are of the cathode 

 type, each individual arc being not readily chstinguishable from an inac- 

 tive cathode arc in the effect it produces on the cathode surface. The 

 reason for the net cathode loss being greater in an active cathode arc 

 than in a cathode arc at clean surfaces is due, at least in part, to some 

 reverse transfer in an arc at clean surfaces. 



1 .4 ih) Silver and Gold. The erosion of silver surfaces is quite complex, 

 and an adequate description of all of the phenomena encountered is re- 

 served for later publication.'" A simplified description of the main fea- 

 tures of the erosion of silver contacts is given here. Tests upon acti\'e 

 gold contacts have been less extensive than upon active silver contacts, 

 but as far as the observations go, gold has been found to behave just Hke 

 silver. 



At active silver surfaces the erosion is, in most cases, from the anode, 

 as it is at inactive surfaces. The arcs are active anode arcs, see Section 



