180 MM. Bister and Geitel on 



platinum pan, and a fine platinum wire, dazzlingly white-hot, 

 be brought below it into the extreme margin of the flame, 

 the platinum wire shows free positive electricity. According 

 to the previous experiments, however, a white-hot wire ought 

 to appear negatively charged with respect to the red-hot pan. 

 In this case, therefore, the electrolytic excitation outweighs 

 the thermoelectric. But if the difference of temperature 

 between the two electrodes be made still greater (which can 

 easily be done by cooling the platinum pan with water), the 

 polarity of the flame is reversed, the white-hot platinum wire 

 is consequently now negative, as the thermoelectric theory 

 requires it to be. 



In an experiment of this sort the following values were 

 obtained : — 



(1) Pan red-hot . ._ . . . (~)].j]=+20 

 Platinum wire white-hot . ( + ) J 



(2) Pan cooled by H 2 . . ( + ) I jj = -32 

 Platinum wire as above . ( — ) J 



The values of E are considerably lower, because in this 

 experiment both electrodes are in the flame. At the same time 

 the margin appears negative to the interior of the flame — a 

 behaviour which cannot be observed under ordinary con- 

 ditions. 



It can, further, be shown that an intensely white-hot plati- 

 num wire in air is negative to one not red-hot in the flame. 

 If the wire a b (fig. 4) was stretched at about 3 millim. distance 

 from the margin of the flame, and c introduced into the foot 

 of the flame so as to be completely enveloped by the flame- 

 gases but at the same time not to become red-hot, then like- 

 wise the polarity of the flame was reversed as soon as a b was 

 rendered brilliantly incandescent by the current. With the 

 wire a b dark the electromotive force amounted to about 1 

 daniell ; with it white-hot, to about 0*1 7-0*2 daniell, but with 

 the deflection in the opposite direction to the former : there- 

 fore in this case the thermoelectric again outweighs the elec- 

 trolytic excitation. 



From these and all the preceding experiments it follows 

 that we cannot explain the electric behaviour of flame by 

 assuming either an exclusively thermoelectric or an exclusively 

 electrolytic excitation, but that we must perforce regard both 

 as cooperating in producing the total electrical state of the 

 flame. 



15. On the proper Electricity of the Flame, 



If we imagine two platinum wires introduced into a flame 

 and the stratum of air which envelopes it, there arises, accord- 





