440 On EhSlrmeten and weo\ EkBriatteu 



From a variety of experiments it was clearly afcertained that the metallic coatingi^ 

 though by their vicinity tliey may diminifli the intenfity of the elcftric ftate in the leaves, 

 do neverthelefs increafc the angle of divergence by their attradion. 



When the goW-leaf eleftrometer is made with a very fmall tube, its fenfibility is fome« 

 what increafed by the nearnefs of the coatings ; but the chance of rendering it unfcr- 

 ticeable from cafual fri6lion, which excites the glafs, and caufes the gold leaf to ilick 

 to it, together with the lefs perfe£t view of the divergence through a tube of fmall cur- 

 Tature, afford reafons why a diameter of lefs than an inch fliould be rejected. Other 

 reafons of convenience indicate that the diameter of the glafs ihould not much exceed 

 tiiis quantity. 



I was once induced to think that the confiderable magnitude of the cap of Bennet's 

 ele£l:rometer might render it lefs capable of being adled upon by fmall quantities qf elec- 

 tricity. Experiment did not however give much countenance to this fuppofition. By 

 trials with heads of different fize, the fmalleft were found to be rather more fenfible to 

 extremely minute eledlricities, and lefs fo to fuch as were greater. The influence of very 

 weak eleflricity may produce the oppofite ftate in the whole of a fmall head, but only in 

 part of a larger ; the remaining part of this lalt afluming the oppofite flate, and robbing 

 the leaves of part of their intenfity. But in higher ele£lricities the whole of the 

 large head may be urged to give electricity to the leaves, in a quantity which the fmaller 

 head could not give without acquiring a higher degree of intenfity, and ccnfequently 

 more ftrongly refifting the dcfired procefs. It appears therefore that the maximum of 

 efFeft with a given ele£lricity, a£ling without communication, will not be obtained but 

 by an head of a definite figure and magnitude. 



From fome experiments of Hoadley and Wiifon with a number of prime condudors 

 feparately infulated, it was eftablifhed that an eleftrified body brought near one end of 

 fuch a fet of condudlors, formiing by contact a right line, will produce the contrary 

 flate in the nearer conduflors, and the fame ftate in thofe which are more remote. Or, 

 upon the hypothefis of a fingle fluid, the electricity of the body prcfented will repel that 

 of the compound conduftor; fo that by feparating them before the influence is removed, . 

 and afterwards examining them, the refpedtive ftates of the feveral portions of the whole 

 line may be afcertained. 



It was a conclufion obvioufly enough deducible from this experiment, that if two elec- 

 trometers of Bennet were' conneiSled by a metallic bar, there ought to be a difference 

 when an ele^rified body is prefented at either extremity, or at the middle of the bar. I 

 made this experiment with a brafs bar eighteen inches long. Both electrometers were 

 affecled at the fame ihftant. Their divergence was equal, and the effeft was the fame, 

 whether the glafs or fealing-wax was prefented at either end or at the middle of the bar. 

 This refult feems to indicate a difference in the mode of aftion between the very weak 

 «le£tricity I ufed and the much ftronger of Hoadley and Wiifon. 



The adtion of pointed bodies and of flame likewife exhibits a remarkable difference 

 between the ftrong and the weak eleiSlricities. The prime conduflor of an elefcrical 

 machine is very fuddenly deprived of the greatefl part of its eleCtricity by the operation 

 ef a metallic point j but does not feem to be much affected by the vicinity of burning 



candles. 



