166 MM. Elster and Geitel on 



Position of rest of the electrometer-needle: 511*0. 

 Daniell = 100. 



Position of 



Orientation of 



Electromotive 



electrodes. 



needle. 



force. 



a 



439-0 



+ 144 



b 



485-0 



+ 52 



c 



510-5 



+ 1 



d 



547-0 



- 72 



e 



572-0 



-122 



(The sign placed before E gives the direction of the electric 

 excitation of the electrode B.) 



Or, in words: — 



As long as B is outside of the flame it is positive, S negative 

 (fig. 2 a). 



When B is in the margin of the flame, the deflection dimi- 

 nishes, but B still remains positive, S negative (fig. 2 b). 



When B and S are equally immersed in the flame the needle 

 returns to its resting-position (511) ; deflection 510'5. There- 

 fore, with the position fig. 2 c, E = 0. 



If S is now drawn back into the margin of the flame 

 (fig. 2 d), it becomes positive, B negative; the polarity of 

 the flame is therefore reversed. And when, finally, S is quite 

 outside of the flame (fig. 2 e), it is strongly positive ; conse- 

 quently it behaves exactly as did the electrode B in the initial 

 position (fig. 2 a). 



With a suitable position of the electrodes, consequently, the 

 flame is shown to be not polarized lengthwise at all. This 

 proves that the longitudinal polarization of the flame is only 

 apparent, called forth by the unequal immersion of the two 

 electrodes. At the same time the second point, the constancy 

 of the value of the potential in the flame, is hereby explained. 



The reversal of the polarity can likewise be shown with a 

 gas-flame; only it does not bring back the electrometer-needle 

 quite to its position of rest, a small ± deflection of from 5 to 

 10 scale-divisions always continuing to subsist. 



A bisulpbide-of-carbon flame shows the reversal of the 

 polarity in like manner as a spirit-flame, which is interesting 

 inasmuch as in it the chemical process is fundamentally dif- 

 ferent. 



Leaving quite out of consideration provisionally a proper 

 electricity of flame, the electrical phenomena in question might 

 be essentially conditioned by contact of the metals with the 

 hot air and the gases of the flame. It might then be expected 

 that the electromotive force would depend on the nature of the 

 metals, as well as on that of the burning gases. 



