ACTIVATION OF ELECTRICAL CONTACTS BY ORGANIC VAPORS 791 



separation, determining whether an arc is of the anode type or of the 

 cathode type, can be obtained from a simpHfied picture of the evapora- 

 tion of metal in an arc. One assumes a field emission arc just being estab- 

 lished between a cathode point and the anode surface, the initial ions 

 being supplied by oxygen and nitrogen of the air with as yet no metal 

 vaporization. The electrons from the point are assumed to travel in 

 straight lines to the anode and cover uniformly an area 7r(L tan dY where 

 L is the electrode separation. If i is the total electron current and v the 

 arc voltage, the power density on the anode is iv/iriL tan oy, decreasing 

 with increasing separation. A lower limit for the power put into the 

 cathode point is (p/d)i~ where d is the diameter of the point and p the 

 resistivity of the cathode metal. Whether the anode begins to vaporize 

 before the cathode, or vice versa, is determined in some way by the ratio 

 of these quantities BUp/d, where parameters unimportant for the present 

 discussion are grouped together in B. For Up/d greater than some critical 

 value, we shall have cathode evaporation and an ensuing cathode arc, 

 but for Up/d less than this value, the anode will begin to evaporate first 

 with a resulting anode arc. 



The resistivity that probably counts is the resistivity at the melting 

 point. At the temperature of melting, the resistivity of palladium is nine 

 times greater than that of silver. Thus one can expect from this simple 

 model that the critical distance which determines whether an arc is of 

 the cathode or anode type will be three times greater for silver than for 

 palladium. If a silver point is less sharp than a palladium point, d greater 

 for silver than for palladium, as it may be because of the well knowii 

 propert}^ of silver atoms to migrate at room temperature, the factor will 

 be greater than this value of three. Now we have the experimental esti- 

 mate of 0.5 X 10~^ cm for the critical distance for palladium. This simple 

 theory predicts that the critical distance for silver shall be greater than 

 this by a factor of three, or perhaps more. The experimental critical 

 distance for silver is between 2.5 and 5 X 10~* cm. 



Quantitative measures of the erosion of contacts of palladium and of 

 silver, which were given in Section 1 .4, are collected in Table II for ready 

 reference. 



From additional experiments, not reported in Section 1.4, it is kno^^^l 

 that these values of transfer apply approximately for potentials both 

 above and below the minimum breakdown potential for air. Not all 

 types of arcs occur, however, for both palladium and silver at potentials 

 above and below the minimum breakdown potential. At potentials that 

 give air breakdo^^^l, all arcs on closure are of the cathode type for both 

 metals whether active or inactive. At potentials that do not give air 



