the Electricity of Flame. 177 



nexion with an experiment described by Edlund * — namely, 

 that when the incandescent carbon points of the electric lamp 

 are connected by a multiplier immediately after the extinc- 

 tion of the flame-arc, a strong thermoelectric current is indi- 

 cated. It follows also from the above-communicated experi- 

 ments that if the carbon points in the flame-arc are in different 

 degrees of ignition (it is well known that the positive is the 

 hotter; this condition is therefore fulfilled), an electromotive 

 counterforce must arise, the quantity of which essentially 

 depends on the nature of the conductors between which the 

 flame-arc passes — a deduction which has already been verified 

 by Edlund. 



The positive electrode, as the hotter, must behave like the 

 incandescent wire in our experiment — that is, be thermoelec- 

 trically negatively excited, — which indicates the rise of an 

 electromotive counterforce. 



§ 12. Dependence of the Electricity of Flames on the State of 

 Incandescence of the Electrodes* 



Having thus shown that platinum wires, as well as water 

 electrodes, in contact with hot air are electrically excited, we 

 return to the electricity of flame. 



The method discussed in the preceding section permits also 

 the determination of the electromotive force of the element 

 " incandescent platinum, hot air, flame-gases, incandescent 

 platinum," if the wire a b is introduced into the stratum of 

 hot air, and the wire c into the apex of the flame. 



As long as the wire a b (fig. 4) does not glow, it is positive 

 to the wire c in the flame ; but as soon as it becomes incan- 

 descent a negative value is added to the positive value of the 

 potential ; therefore the potential-difference between the tw T o 

 electrodes must undergo a diminution. This inference was 

 completely verified by experiment. 



In the following, E denotes the electromotive force of the 

 flame when c is incandescent and the wire a b not, and e the 

 electromotive force of the flame when both wires are incan- 

 descent. Of necessity e would be =E if the incandescence of 

 the wire a b had no influence; yet there resulted : — 



Series I. 



Series II. 



E. e. 



E. e. 



254 ' 148 



134 47 



246 147 



122 50 



216 116 



114 37 



* Pogg, Ann. cxxxi. p. 586 (1850), & cxxxiii k p. 353 (1851). 

 Phil. Mag. S. 5. Vol. 14. No. 87. Sept. 1882, N 



