240 NEW CLASSES OF GALVANIC CONDUCTORS. 



stronger divergence than that of the positive, though every 

 thing else will appear to indicate an absolute insulation : for 

 if a communication be established between either of the 

 poles and the ground, its electrometer will lose all its diver- 

 gence, and that of the opposite pole will attain its maximum ; 

 and on touching both poles at the same time, as strong a 

 shock will be received, as if the two poles were insulated by 

 a stratum of air. It appears, that hitherto philosophers 

 have contented themselves with this single experiment, to 

 affirm that flame insulates all galvanic effects ; but the fol- 

 lowing facts prove, that this insulation is partial, and that 

 flame continues to be an excellent conductor for the positive 

 pole. 

 Proof that it Every thing remaining as in the preceding experiment, let 

 conducts posi- a communication be made between the flame and its sup- 

 tiveeec nu y. p 0r ^. . or? wmcn j s m0 re simple, touch the flame itself with 

 an uninsulated metallic rod. Immediately all the diver- 

 gence passes to the negative pole, and the positive is abso- 

 lutely discharged. If the strongest divergence possible have 

 been previously given to the negative electrometer, by touch- 

 ing the opposite pole, no application of a good conductor to 

 the flame will take off the least part of this negative diver- 

 gence ; while the same application will instantly destroy 

 every vestige of divergence before imparted to the positive 

 pole, and transfer it to the negative side in the strongest de- 

 gree possible. 



Whatever extent be given to the flame, and however near 

 to the negative wire it be touched, it still remains impossi- 

 ble to act through its medium on the negative side, so as to 

 take away the divergence. Flame belongs wholly therefore 

 to the positive pole, since by touching it this pole is imme- 

 diately discharged, and the negative pole is mediately 

 brought to a maximum of divergence. 

 This farther This paradoxical property is confirmed by the following 



confirmed by experiments. The two polar wires being united in the same 

 insulated flame, immerse in this flame the hook of a sensi- 

 ble electrometer, and it will acquire a weak positive diver- 

 gence, if the two poles had not previously arrived at an 

 equilibrium of intensity. But this positive divergence at- 

 tains its maximum, the moment the negative pole is made 



to 



:penments. 



