Prof. Buff on the Electrical Properties of Flame. 147 



observed to pass from the warmer to the less warm wire through 

 the intervening space of gas. 



The author concludes from these experiments, that air and 

 other gases, when heated, and thus rendered conductible, excite 

 electrically bodies plunged in them. Gases thus range them- 

 selves in the same list as other conductors of electricity. When 

 two metallic wires, or other conductors which are connected at 

 one end, are brought into contact with a sufficiently heated gas, 

 we have, properly speaking, a closed circuit. If one of the 

 places of contact with the gas be more strongly heated than the 

 other, a thermo-electric current is the necessary consequence. 



There is, however, another source of electrical excitation in the 

 flame, as is proved by the following experiment : — One platinum 

 wire was introduced into the obscure centre of the flame, the 

 other was brought near its outer surface ; a current immediately 

 exhibited itself, which passed through the flame from the interior 

 to the exterior wire. It continued to pass in the direction even 

 after the outer wire had attained a bright red heat, while the 

 inner one glowed but feebly. It is evident that the thermo- 

 current which would have passed from the hotter to the cooler 

 wire, was in this case overcome by a current, the source of which 

 was the place of contact of the flame and the air. The electricity 

 here developed is so feeble, that the condensing electrometer 

 is better suited to its examination than the multiplying galva- 

 nometer. It is easy to see, observes the author, how experi- 

 menters who have neglected to separate these two sources of 

 excitation may have arrived at contradictory results. 



By properly connecting a platinum wire, which was dipped 

 into the centre of the flame, with a condensing plate, the latter 

 became charged with negative electricity, and hence the author 

 concludes that positive electricity is given off by the outer surface 

 of the flame. The charging here is exceedingly slow, and can be 

 greatly accelerated when a second wire, which is connected with 

 the other plate of the condenser, is held over the flame. 



One end of the galvanometer wire was connected with the 

 platinum wire which dipped into the centre of the flame, the 

 other end of the same was connected with the earth. The cur- 

 rent thus obtained was too feeble to cause the slightest motion 

 of the galvanometer needle. But when a spacious platinum dish 

 containing water was brought over tin* flame and connected with 

 the other end of the galvanometer wire, it required no very sen- 

 sitive instrument to demonstrate the existence of a current. 



" Hence/' observes the author, "as the strength of the flame- 

 current by an equal chemical activity and equal conduction of 

 the inner portion of the flame is essentially dependent on the 

 nature of the conduction from its upper portion, it must be coll- 

 ie 2 



