ATMOSPHERIC ELECTRICITY. 299 



If the influence of the electricity of the ground be fuf- Thefe effefts 



^, ,ii- i r cc a. it i, cannot fake place 



cient to prevent the divergence, the lame effect ought to - n a chanibe r r 



take place in every clofe chamber, becaufe a roof or ceiling &c. 

 may be confidered as a prolongation of the ground. Hence 

 it is that this polarity does not manifeft itfelf in a chamber ; 

 it is alfo erroneous to pretend that the atmofphere produces 

 electricity; and it is equally erroneous to alfert, that electric 

 repulfion does not take place in a vacuum; Mr. Errnan pro- 

 poses to eftablifh in an inconteflible manner in another me- 

 moir, that the repulfive force of electricity, as well as of 

 magnetifm, are alfo manifefied in a vacuum. 



If the electricity of the rod of the electrometer could be The electric 

 afcribed to that of the atmoinhere, it would follow, that it J gns f" no [, 



» ' come from the 



(hould be equally electric through its whole length; but if it air; .'or the-rod, 

 depend on the earth, the rod will not be equally eletfrified, ^g^rou?" 

 and its different parts will be varioufly modified accordingly as out. 

 they may be more or lei's remote from the earth ; and this in 

 fact is the cafe. Two electrometers, each having a pointed 

 rod of three feet, one of which was twice bended in a right 

 angle, fo as to have an horizontal portion of three or four 

 inches before it proceeded upwards, were held in fuch a . 

 manner, that the point of the Straight rod of the one touched 

 the horizontal part or elbow of the other, which was lower. 

 When they were touched, to difcharge them, no divergence 

 was feen, but when they were afterwards Separated horizon- 

 tally, the leaves of the lower electrometer diverged, while 

 thofe of the upper were not at all moved. 



It is very remarkable that the nearer the contact is made to Its ftate differs 



the upper point of the compound conductor to take its elec- as f , he . con ^ ft ,s 

 . . ,, . . . . . . ,. . mode in different 



tricity from it, the greater is the pohtive divergence of the parts of its 



lower electrometer; and if the difcharge be made very near ien S^« 

 the upper point, the electrometer will indicate plus but very 

 feebly in comparifon with that below. 



If thefe two electrometers, the conducting parts of which Electric polarity, 

 form together a length of fix feet, be raifed perpendicularly 

 fome feet above the earth, the plus electricity will be oblerved 

 in both ; but if they be feparated horizontally, the upper one 

 will retain its divergence, and the lower will augment itfud- 

 denly, even to double. From this, it is evident that the con- 

 ductor has not the fame charge of electricity through ail its 

 length, and this is precifely what is called polarity ; and as the 



pre fence 



