ACTIVATION OF ELECTRICAL CONTACTS BY ORGANIC VAPORS 787 



7(c) which show clean palladium cathodes after constant current cathode 

 arcs of 4 amperes lasting, respectively, for 0.072 microsecond and for 

 0.14 microsecond. The striking potential in each of these arcs was 400 

 volts, the total arc energy being 40 ergs and 80 ergs. The three photo- 

 graphs of Figs. 7(a), 7(b) and 7(c) represent then the markings made on 

 the cathode by arcs of 40, 50 and 80 ergs respectively. The voltage of 

 400 was chosen for the two comparison photographs of Figs. 7(a) and 

 7(c) because this is above the minimum air breakdown potential, and 

 arcs on closure at striking potentials above this value are known to be 

 always cathode arcs (Reference 4, Fig. 4). 



Cathode markings such as those of Fig. 7(b) are occasionally pro- 

 duced by arcs at 50 volts on relatively clean palladium surfaces. In gen- 

 eral, however, an arc at this low striking potential between clean surfaces 

 is an anode arc, leaving a single well defined pit on the anode, and on the 

 cathode, a single roughened area with considerable metal spattered over 

 from the anode (Reference 9, Fig. 6). While a cathode arc, making on 

 the cathode the type of markings shown in Fig. 7, is rather rare between 

 clean palladium surfaces at a striking voltage as low as 50 (Reference 4, 

 Fig. 4) it is the usual kind of arc between surfaces upon which carbon 

 particles have been dusted, and by implication, it is the sort of arc that 

 occurs between active surfaces. That this arc should cause loss of metal 

 from the cathode is clear from the photographs of Fig. 7, and from the 

 fact that the damage done to the anode sometimes cannot be detected 

 and is always rather sUght.*^" Between clean surfaces, this sort of arc 

 occurs more frequently at higher striking voltages, and invariably on 

 closure when the potential is above the minimum breakdown potential 

 for air. It is the greater striking distance that favors the cathode type 

 of arc, and for active arcs also it is just this enhanced electrode separa- 

 tion, resulting from carbon particles, which can be thought of as the 

 reason for the arc being of this type. There is obviously a critical distance 

 above which arcs are of the cathode type, and for palladium electrodes 

 this critical distance is less than 1 X lO""* cm. Earlier experiments can 

 be used to define this critical distance better. From the data of Fig. 8 of 

 Reference 4 it appears that this distance for palladium is about 0.5 X 

 10-'^ cm. 



Markings made on the cathode by a single arc between active pal- 

 ladium surfaces are doubtless not easily distinguishable from those re- 

 sulting from a single arc that has been constrained to be of the cathode 

 tj^pe only by a high striking potential and the resulting great electrode 

 separation. Nevertheless, when many times repeated, the over-all results 



* See the footnote relating to Fig. 3, see page 776. 



