328 



ME. W. DUDDELL ON THE RESISTANCE AND 



opinion that traces of impurities, such as the vapours of the alkaline earths, are 

 essential to provide the carriers of the electric charges in the vapour column, so as 

 to render it conducting and the electric arc as we know it a physical possibility. 

 Unfortunately it has not, up to the present, been possible to obtain pure carbon 

 electrodes in order to test this theory. In favour of it is, however, the known fact 

 that, given an arc of fixed length and current between the best commercial solid 

 carbons, then any addition to it of such substances as potassium or sodium reduces 

 the P.D. required to maintain the arc and its resistance and increases its stability. 



The difference between the 11-millim. and 'J-millim. Electrographitic carbons 

 mentioned above is probably caused by the last traces of impurities having been 

 more completely expelled in the manufacture from the ll-millim. size than from 

 the 9-millim. 



TABLK IV. ^Various Carbon Electrodes. 



Arc length :l millims. Added alternating currents O'l ampere 



Direct current through arc if!) I amperes. Frequency 100,000 -- per second. 



Nature of Electrodes Varied. 



f This is an impedance, as the power-factor is not unity, the only one not found to be unity within the 

 limits of experimental error. 



