784 THE BELL SYSTEM TECHNICAL JOURNAL, MAY 1957 



tion. The experimentally observed field of O.G X 10^ volts/cm is the 

 field at which this motion becomes appreciable lor the very small sooty 

 particles. With the start of motion of this sort the field is increased, and 

 further motion is assured making the situation imstable. The gap is 

 greath^ decreased in length before electrical breakdown takes place, and 

 the field at electrical breakdown is probably as high as it is at any carbon 

 surface. 



2.2 Arc Voltage 



At the beginning of an active arc, at least one carbon particle is always 

 exploded by the arc current, but only when the surface is very heavily 

 carbonized is an enhanced arc voltage observed. Fig. 4. It must not be 

 thought that the higher arc voltage occasionally found at the beginning 

 of an arc is to be attributed directly to the presence of carbon vapor in 

 the arc during its ear^ stages, because carbon does not have an excep- 

 tionally high ionizatioii potential. P. Kisliuk has showii'^ that, in a field 

 emission short arc, the arc voltage should be just slightly larger than 

 the sum of the ionization potential of the metal of the electrodes and of 

 its thermionic work function. Now this result holds c^uite well for a num- 

 ber of different metal arcs, but does not hold at all for carbon. The short 

 carbon arc is apparently of a different type, and has no well defined arc 

 voltage. On the other hand, although carbon, unlike the noble metals, 

 gives out thermionic electrons copiously long before it is hot enough to 

 vaporize, a true thermionic arc cannot have the enormous current densi- 

 ties that occur in short arcs. (It does not seem impossible that thermi- 

 onic emission may help to maintain an arc when the current is lower near 

 its end.) The high arc voltage at the beginning and end of an active arc 

 between heavily carbonized surfaces may be due to a dearth of positive 

 ions, requiring a higher applied field to maintain the field emission. In 

 any case, it is like the higher arc voltage of carbon which w^e do not under- 

 stand. When the higher arc voltage is not detected, the vaporization of 

 metal must be profuse, and only when vaporization is reduced, as it is 

 when the current is very small near the beginning and end of an arc at 

 the discharge of a capacitor into an inductive circuit, is the higher arc 

 voltage observed. On rare occasions heavily carbonized surfaces show 

 a suddenly enhanced arc voltage for a short interval near the middle of 

 an arc. That this should occur very much less often than at the beginning 

 or end of an arc is understandable. 



The observation that the arc \'oltage sometimes becomes high near 

 the end of an arc suggests strongly that an active arc is moving con- 

 tinually during its life. Only when the current is insufficient to vaporize 

 carbon and underlying metal freely, and thus to maintain the large ion 



I 



