NEW CLASSES OP GALVANIC CONDUCTORS. 317 



«f metal, a wet finger, soap wetted at the point of contact, 



and soap perfectly dry; and I found them all equal in 



degree. It will soon appear why, notwithstanding this, 



no shock is obtained on employing a prism of soap, unless 



it be wetted at the point of contact. 



If now two wires, issuing from the two poles of the If a wire from 



pile, have their extremities fixed in a perfectly insulated each H e con }- 

 r > 7 . municate wuh 



prism of soap, into which they should penetrate a few lines, an insulated 



no remarkable effect is perceived : that is to say, after having p * ece . of s . oap » 

 r J ' ° the circuit will 



brought the two poles to the same intensity, by applying to not be com- 

 them an insulated metallic rod, the electrometers of the ple 

 two poles will act as they did before the intervention of 

 the soap, and when a stratum of air insulated them per- 

 fectly with respect to each other. But the instant a free If this soap be 

 communication is established between the soap and the " nmsu ' at f d > 



, , ... , ., . . „ tne negative 



ground, the positive electrometer exhibits a maximum of electricity will 



divergence, and that of the negative side loses all signs of it, be conducted 



. , ., . , , , , ,. , T , ' off, but not the 



precisely as if a communication had been established bet positive. 



tween the ground and the negative pole itself. Consequently 

 the soap, which insulates the positive effect, is a perfect 

 conductor for the negative; to which it belongs throughout 

 its whole extent, for if you touch the soap with a fine point 

 ever so near the place into which the positive wire is in- 

 serted, it is impossible to take from it any portion of elec- 

 tricity, so perfect is the insulation of this pole. 



A very striking proof of this paradoxical property is, Touching the 



if one fineer be applied to the wire of the positive pole, P 0Sltlve wire 



9 rr * r J Wltri a dry 



and another finger wetted to the soap, no shock is felt, finger and the 



and the electrometers do not show the least change in their s , oa P wltn wet 



,. • r> •/. , . , does not form a 



respective divergencies, cut if the experiment be repeated communica- 



by establishing a communication between the positive pole tion : 



. , .*i_i_ A i_-i! > i « . ifbothfingers 



and the soap with both fingers wetted, a very perceptible be wet a shock 



shock will be felt, and the two electrometers will arrive at is ^ eIt - 



an equal and a very weak degree of intensity. 



These facts are sufficient to establish the existence of this 

 fifth class of substances; but on pursuing our researches 

 farther we meet with many interesting phenomena. 



To discern these the better, the continuity of one of the An apparatus 



wires should be interrupted, and an apparatus for extricating f? rdeconi P°- 



, , . ° sing water 



gasses be interposed between its parts. In this case no forming part of 



hemical the P osltlYe 



