NEW CLASSES OF GALVANIC CONDUCFORS. 23Q 



two kinds of conductors act in opposite directions. If one 

 of the poles communicate with an electrometer by means of 

 a wire, an uninsulated flame, brought into contact with this 

 wire, will take from it, as has been seen, all the divergence 

 before imparted to the electrometer by the transient contact 

 of the opposite pole. But if a permanent metallic commu- 

 nication be established between this pole and the ground, 

 the electrometer will reach the maximum. of divergence, and 

 remain at it without any diminution, though the uninsulated 

 flame continue to touch the wire, by means of which the 

 electrometer is in communication with the pile. It is to be 

 observed, that this effect is precisely the same at both poles. 

 But how different would be the action of a metallic con- 

 ductor, if in this experiment it were substituted instead of 

 the flame! It is well known, that the application of unin- 

 sulated metal would prevent any intensity of electricity pro- 

 ducing divergence ; and that the application of a humid con- 

 ductor would at least diminish it extremely, if it did not 

 reduce it to nothing. Flame therefore, which has hitherto 

 been considered as a good conductor, does not here produce 

 the effect, that was to be expected from it. 



But this anomaly is of little importance, compared with It conducts po- 

 that which flame exhibits, when it is applied simultaneously S1 . nve e <ectri- 

 ^ i i i • i ii- i « C1> Y> anc * insu- 



to both poles, with a view to close by its means the galvanic ] a tes negative, 



circle. The following facts prove, that in this case it be- at tne same 

 longs entirely to the positive pole, and absolutely insulates 

 ull the negative effects, which has led me to place it in a 

 separate class. 



Let each pole of a well insulated pile, consisting of about Experiment in 

 a hundred and fifty pair of plates of silver and zine, be con- which it ap- 

 nected with a sensible electrometer. With each pole con- ductor. 

 nect a wire, supported by a completely insulating stand; and 

 let the extremities of the wires be brought so near together, 

 that one flame may be in contact with both. On an insulat- 

 ing stand place a spirit lamp, and commence the experiment 

 by bringing the flame into contact with the two metallic 

 wires. As long as the flame remains insulated, the electro- 

 meters of both poles will diverge nearly as if the two polar 

 wires were perfectly insulated. After some time, indeed, 

 the electrometer of the negative pole will exhibit a little 



stronger 



