the Electricity of Flame. U9 



Electrode S (fig. 2 c) -white-hot. Base-electrode B 



In the flame (fig. 2 c). In air (fig. 2 a). 



Dark . . . E'=+4 

 Red-hot . . E'=+3 E=+150 



White-hot. . E'=-6 



The sign prefixed refers to electrode B. 



Also when the electrodes are in one cross section of the 

 flame and at the same time dip completely into it, in spite of 

 great differences of temperature the electromotive force is very- 

 small. If the wires ab and c (fig. 4) were brought into the 

 coolest (lowest) part of an alcohol-flame. E was ascertained to 

 be 23. By a suitable shifting of the wire c this could be 

 reduced to 6, notwithstanding that c was not red-hot while 

 a b was put into a dazzling white incandescence by an electric 

 current. The reason for this surprising behaviour appears to 

 be that the flame-gases, being relatively good conductors in 

 comparison with the hot air, prevent the electrical difference 

 from being completed. This would also be confirmed by the 

 fact that, in the experiments with wires in air, the maximum 

 is found when the wire a b is bright red-hot. "\Vhite-hot wires, 

 when the electrodes are in the same position, constantly give 

 lower values for the electromotive force, as we have already 

 mentioned above — which, we think, can only be accounted for 

 by the conductivity of the surrounding air being so augmented 

 by the strong heating that it forms as it were a secondary 

 closing of the circuit. 



§ 14. Cases in ichich the Thermoelectric Excitation predomi- 

 nates over the Electrolytic, and vice versa. 



The view last discussed explains also very naturally the 

 experiment mentioned in § 4, that the electromotive force 

 rises from 12 up to 190 as soon as the electrode B (fig. 2 a) 

 is drawn quite out of the flame. The moment this takes place, 

 the secondary closing formed by the flame is removed, and 

 the thermoelectric force corresponding to the temperature- 

 differences of the electrodes comes fully into action. 



If this explanation be not admitted as the correct one, it 

 may appear doubtful if with a flame the thermoelectric exci- 

 tation does not entirely fall away, and the electrolytic exclu- 

 sively condition the electric behaviour of the flame. But if 

 the experiments related in § 12 tell against this, others also 

 can be instituted which it would be difficult to explain without 

 admitting a thermoelectric excitation. They are the follow- 

 ing : — 



If the apex of the flame be conducted-from to earth bv a 

 N2 



