ELECTROMOTIVE FORCES OP THE ELECTRIC ARC. 335 



a reflecting moving-coil instrument of 20 ohms, having a resistance ot about 

 16,000 ohms in series with it. The resistance of both these instruments is so low 

 that no deflection due to the frictional charges could be observed on them. By 

 comparing the P.D.'s obtained when one or the other of the voltmeters was used an 

 estimate of the resistance of the vapour between the carbon tips could be made. 



If the flame was caused to impinge on the two carbon tips so as to heat them 

 equally, as judged by eye, then no P.D. was observed between them. If the flame 

 was now moved so as to heat one carbon more than the other, then a P.D. was 

 observed between the carbons, the hotter being positive to the cooler as indicated by 

 the voltmeter, that is, in the same direction as the back E.M.F. of the arc, assuming 

 that the positive crater is the hotter. 



The highest P.D. obtained was 1'5 volts when using cored carbons which had 

 been previously soaked in potassium carbonate solution. This P.D. was not in any 

 way due to differences in the quality of the two carbons, since by moving the jet so 

 as to heat either carbon more than the other the P.D. changed sign, the hotter 

 carbon always being positive. By setting one carbon in front of the other it was 

 found that the direction of the P.D. was unaffected by whether the stream of flume 

 gases flowed from the hotter to the cooler carbon or vice versa. 



With two pieces of the same solid carbon, about 2 millims. apart, no foreign bodies 

 being introduced in any way, the highest P.D. obtained was about -O'o volt on the 

 higher resistance voltmeter, and the resistance of the flame between the carbons was 

 deduced as about 4000 ohms, so that the E.M.F. between these solid carbons 

 was only about - G2 volt. After soaking this same pair of carbons in potassium 

 carbonate the P.D. obtained was I/O volt., the resistance of the vapour being only a 

 few hundred ohms, so that the effect of introducing this potassium salt was to 

 greatly increase the E.M.F. between the carbons. If these E.M.F.'s are due to 

 Peltier effects, then it would seem as if the introduction of the potassium greatly 

 increased the thermo-electric power of the junction. 



The above experiment is in agreement with the fact that the back E.M.F. of the 

 cored arc has been generally found larger than that of the solid arc, as the cored arc 

 undoubtedly contains more impurities. 



By varying the proportions in which the oxygen and coal gas were mixed before 

 burning, it was found that the highest P.D.'s appeared to correspond with the gases 

 being burnt in their combining proportions, so as to produce the hottest flame ; 

 if either gas was considerably in excess, a much lower P.D. was obtained. It is to be 

 noted, however, that by suppressing the coal gas altogether and allowing the carbon 

 to burn in the oxygen, P.D.'s up to 1'5 volt could be obtained, as before the hotter 

 electrode always being positive to the voltmeter. These latter experiments, though 

 not conclusive, are not in favour of the P.D. being due directly to a combination of 

 the carbon and either of the gases. 



The temperature of the positive crater of the arc is much higher than the highest 



