e+a: 



e — x . 



e. 



+ 9-5 



-9-0 



+ 0-5' 



-1-7 



-0-2 



-1-9 



+ 7-3 



-5-5 



+ 1-8 



182 MM. Elster and Geitel on 



the electricity does not take place through the introduction of 

 the transverse wires. 



Further, in contradiction to the existence of a proper electri- 

 city of flame is the fact that when platinum wires as homoge- 

 neous as possible are used as electrodes, and are also approxi- 

 mately in the same state of incandescence, the electromotive 

 force of the flame sinks to a minimum. The electromotive 

 force of the combination 



White-hot platinum | Hot air + Hot air | Flame-gases + 

 Flame-gases | White-hot platinum 

 was ascertained to be 0*0013 daniell — a value which lies within 

 the limits of errors of observation, and consequently may be 

 put = 0. In this determination, of course, it was necessary 

 to employ the method discussed in § 11. 



For e the following values were obtained: — 



Daniell = 100; 



consequently e=0'0013 daniell. According to this, the elec- 

 tromotive force of the member 



Hot air | Flame-gases 

 may be put =0, and consequently a proper electricity of the 

 flame be left out of consideration. 



The most important circumstance contradicting the existence 

 of a proper electricity of flame may be that the reversal of the 

 polarity of the flame is not connected with the reversal of the 

 combustion-process. 



A flame of air burning in an atmosphere of illuminating- 

 gas exhibits the same polarity as illuminating-gas burning 

 in air. 



An incandescent platinum wire in burning air was strongly 

 negative to the metal (copper) out of which the flame issued ; 

 and a second wire, introduced into its sensitive stratum, 

 received a strong positive charge, just like the base-electrode 

 of an ordinary flame. The details of this experiment were as 

 follows : — 



First the electromotive force of an ordinary gas-flame 

 issuing from a glass tube provided with a copper jet, with the 

 electrodes in a determined position (fig. 2 a), was measured. 

 A current of air was then passed through the tube, which was 

 placed in a space filled with illuminating-gas, and the air- 

 current ignited by the spark of an induction-apparatus. No 

 source of error was given rise to by this (as we convinced 



