respecting the Phcenomena of Electricity. lO/* 



cation with conductors, can part with none. How then can its 

 outer surface receive the great quantity we find in it V The 

 answer is however obvious, for this is the common effect oi fric- 

 tion. The cylinder receives constantly -e/ec/^v'c ^?«W from the 

 rubber, especially when this is covered with a metallic amatgamd ; 

 for it is in communication with the ground. In that case it is 

 not necessary that the opposite surface of the glass should part 

 with any electric fluid ; the whole process takes place on the 

 outward stirface. The rubber gives electric fluid to the ^lass 

 cj/linder, which parts constantly with it, in meeting the prime 

 conductor; but the rubber communicating with the ground, fur- 

 nishes also constantly a new cjuantity of the fluid. 



41. It remains only to state a very essential point in electric 

 phaenomena, namely, How does the electric fluid communicate 

 itself through space ? This was the object of an experiment 

 which I made in 1/74 in Mr. Walsh's laboratory, which experi- 

 ment in the first volume of my work, Idees sur La Metcorologie, 

 p. 521, I left under Mr. Walsh's name, because he had published 

 it, without my knowledge, in the Philosophical Transactions for 

 1785. But these experiments were concerted between Dr. 

 Franklin and myself, and only in the house and in presence of 

 Mr. Walsh, with an apparatus of which the doscrijition will 

 show the purpose and its origin. 



42. In my experiments for producing a truly comparable la,- 

 rometer, repeating those of a French academician, Du Fay, for 

 producing light in the vacuum of barometeis, an exjjeri'ment 

 related in the Mhn. de I'u^cadcr/iie des-Sciences de Paris for 

 1/23, I found, as has been explained in my work, Recherches 

 sur les Modifications de l' Alrnoslihere, torn. i. p. 43, that when 

 the Torricellian vacuum was procured by making the mercmv 

 hoil in the tube, no light appeared at the top of the baroinetor ; 

 \vhence I concluded, that a perfect vacuum vvas not a conduclor 

 of the electric fluid. 



43. Having expressed this idea to Dr. Franklin, ihe proposed 

 to me an experiment, very difficult to execute, but whii'Ii he en- 

 couraged me to undertake. The apparatus consisted of a glass 

 syphon, the legs of which were about three feet distant from each 

 otlier; the curve began about three inches above the point at 

 which stood the common barometer at that moment. When 

 the syphon was filled with mercury it was inverted, with its two 

 legs ])hmged into separate cups, each resting on an insulating 

 stand : tliese cups received the mercury descending from the 

 upper curve, and the column thus separating, there \iere two 

 barometers with a common vaccum. 



44. Ill that situation, when a spark was given with a Leyden 



vial 



