NEW CLASSES OF GALVANIC CONDUCTORS. 327 



it surrounds. Indeed I have found, that, in an alcoholic ^ ct oil isset 



solution of soap diluted with water, a manifest separation 



of the oleaginous base of the soap will be effected after some 



hours, and it will be deposited on the wire of the positive 



pole. 



This fact is certain : yet it is easy to show, that the hy- Yet this cannot 

 . , , . , ., , . . , ./. ,1 • be the cause of 



pothesis to which it serves as a base is not less manifestly in the pheaome . 



contradiction with several particulars of the phenomena, na. 

 which it ought to explain. In reality, when several prisms 

 of soap, connected together by intermediate arcs, are ex- 

 posed to the action of the pile, there is no doubt a partial 

 insulation with respect to each point of insertion that cor- 

 responds with the positive effect : but we cannot thence 

 conclude, that this insulation is absolute, since the negative . 

 pole may be acted upon through all these prisms, and all 

 the points of insertion of their conducting arcs, so as 

 to take from it its charge. At the points of contact of the 

 positive wires therefore there is no absolute obstacle to the 

 passage of the electric fluid, and the hypothesis of an insu- 

 lating coat of oil falls to the ground. Besides, in perfectly 

 dry soap the positive insulating effect displays itself the in- 

 stant it is applied, when no preceding chemical decomposi- 

 tion can have taken place. 



On this hypothesis too how we shall explain the produc- Farther difficul- 

 tion of the same effect by the flame of phosphorus ? Must ties in &e W5Ly * 

 we recur to anew hypothesis to account for this single fact, 

 and say for instance, that here the oxiding action of the 

 acid in the state of vapour, being produced with more ener- 

 gy at the positive wire, renders it impermeable to the elec- 

 tric fluid ? But this hypothesis would be equally untenable, 

 since platina wires exhibit the phenomena of positive insu- 

 lation as well as any other metal ; and this effect manifests 

 itself the first moment of contact exactly in the same degree 

 as after the long-continued action of phosphorus in ignition. 

 Besides, on this supposition it would be difficult to explain, 

 why sulphur does not produce the same effect. 



I am persuaded therefore, that these hypotheses are The cause - 



completely erroneous : that the cause of the phenomena we * ere ore re ' 



1 -' ' r mains to be 



have discussed lies deeper, and is purely chemical \ and that shown, 



