780 THE BELL SYSTEM TECHNICAL JOURXAL, MAY 1957 



2.4(b), which have never been observed at palladium contacts. The 

 metal lost from the anode after a great many active anode arcs tends 

 to be eroded smoothly over the entire surface, like the cathode loss in 

 arcs of the cathode type at palladium contacts. At a moderate pressure 

 of activating vapor, almost all of the metal eroded from a silver anode 

 is transferred to the cathode, but at a high pressure much of it is lost. 

 Whether the metal from the anode is transferred or lost is correlated 

 with the amount of carbon formed by the active arcs; if the production 

 of carbon is small, metal is transferred, but in the presence of much 

 carbon, the metal does not stick to the cathode and is lost. The amount 

 of carbon formed (in air) by active anode type arcs at silver surfaces is 

 very much less than the amount formed by active cathode type arcs at ' 

 palladium surfaces, and this difference accounts for the fact that a great 

 deal of the eroded metal is transferred at active sih'er surfaces, although 

 there is always very little transfer at active palladium surfaces.* The 

 erosion of a silver anode by active anode arcs may be as great as 10~^^ 

 cc/erg, but is lower than this whenever the carbon formation is suffi- 

 ciently slight to permit much transfer of metal. 



Long, sustained break arcs at active silver surfaces become cathode 

 arcs when the electrode separation becomes sufficiently great. Such arcs 

 give cathode erosion resembling that at active palladium surfaces. For a 

 long sustained break arc, the cathode erosion suffered when the electrode 

 separation becomes very large may be greater than the anode erosion 

 occurring when the electrodes are closer together, so that the net loss 

 from the cathode may be the greater. There may even be a small net 

 anode gain. ^ 



Measurements of transfer at electrical contacts have sometimes been 

 very confusing in the past, both because of their complexity and because 

 of their apparently erratic character. Now, with well developed insight 

 into the mechanism of short arcs, this complexity of transfer and its 

 varied character have been most useful in improving our understanding 

 of short arcs and of the transfer of metal to which they give rise. 



The over-all picture of activation will be given in the following pages. 



2. INTERPRETATION OF ACTIVATION ■ 



After one has concluded that activation is due to solid carbonaceous 

 material, it is natural that tests should be made upon contacts of solid 



* At extremely low pressures of activating vapor, active anode arcs at silver 

 surfaces may not only transfer to the cathode practically all of the metal lost from 

 the anode, but the type of erosion may even he changed to the mound anfl pit 

 type characteristic of inactive arcs.' 



I 



