410 RECENT PROGRESS IN PHYSICS. 



and the latter, decomposed by the white heat, unites with particles 

 of air, the continuous mass is broken and separated, and only threads 

 of it remain, which are diffused more and more, and finally scattered 

 in the air.* Hence from flame conducting electricity threads issue, 

 which being separated from each other by the non-conducting gases 

 and hot air, necessarily flow away in ends and points. This being 

 granted, flame must be considered as a good conductor of electricity, 

 furnished with a number of points extending in every direction into 

 the air, and such, too^ as to exceed in perfection all the j)oints exist- 

 ing in nature. 



The quantity of electricity issuing from a conductor furnished with 

 points, is as much greater as the points are more perfect ; the least 

 traces of electricity are also removed by flame. The electrical density is 

 much greater on a point than on any other part of a conductor, at the 

 steam points of flame the density is therefore very great ; the elec- 

 tricity accumulated at the steam points acts then inductively upon 

 neighboring insulated conductors. If the accumulation of the elec- 

 tricity attracted to the parts of the insulated conductor, which lie 

 nearest the electrical steam points, is great enough, it will escape^ 

 and the conductor will remain charged only with the electricity which 

 is repelled by the electricity of the flame being of the same kind ; 

 the insulated conductor therefore remains charged with the same elec- 

 tricity which the flame has, without that of the flame having gone 

 over to it. 



On the other hand, if the flame be brought near to an electrified 

 body, the steam points will become electrical by induction, but their 

 electricity escapes, and the insulated conductor, on which the lamp is 

 placed, remains charged with the same electricity which the neighbor- 

 ing inducing electrified body possesses, without this electricity having 

 passed from the electrified body to the conductor provided with the 

 lamp. 



This view of the subject is justified by the following experiment: 

 A small metallic spirit-lamp, surrounded by a metallic cylinder 13 

 lines high, was placed on a properly insulated copper disk A, 3 inches 



11 lines in diameter. About 3f in- 

 '°" ■ dies from the lamp was placed a 



p-ill __^ second copper disk B, (fig. 45,) con- 



nected with the electroscope &, and 

 ^-tfno-1|^i;|j /ff^^^^^^ kept in a vertical plane by an insu- 



lating shellac handle. The point of 

 the wick and the middle of the disk 

 B were at the same height. 

 '^ When the lamp was lighted and 



A electrified by contact with one of 

 the poles of a dry pile, the electroscope immediately indicated a diver- 

 gence of 3 lines. Since the steam of the flame ascends vertically, a direct 

 transfer of electricity is improbable. If it does take place, A and B 

 must necessarily be in conducting connexion tlirough steam, and the 

 electroscope should collapse as soon as A is touched. But by touching 



[* This explanation is unsatisfactory, since super heated steam does not conduct better 

 than heated air ; all we know at present relative to this matter is, that a flame acts as an 

 assemblage of perfect points ] 



