2i COMPOUND ELECTRICAL INSTRUMENT. 



On the fpon- even though it has not been ufed for tivo or three months.— 

 Sfc^rof Ui'is ^^^ ^^ ^*- ^^^ ^^^^ ^^^^ within two or three hours, the e&'eck 

 inftrument. will take place with fewer motions of the lever, and if it has 

 been ufed within a few minutes, two or three motions will be 

 pore than fufficient to afcertain the quality of the ele6lricity. 

 It is to be obferved that the inftrument was always difcharged 

 (by the metal point as before obferved) between each trial. 



In the courfeof making the experiments on the fpontaneous, 

 electricity of this inftrument I found that it was always pofi- 

 tlve if the inftrument had not been ufed for two or three days, 

 U'hatever eledricity it was lall charged with. But the time it 

 nruft ftand unufed after it has been difcharged, to produce this 

 effeft, depends a great deal on the weather; if the air is very . 

 , humid, twenty-four hours is quite fufficient, but if it is very 



dry, it w'ill require four or five days. 



After I had obferved that after the inftrument was charged 

 with pofitive electricity its fpontaneous eledtricity was always 

 pofitive; and that after it was charged with negative eleftri- 

 city, its fpontaneous ele<Slricity was negative only within a 

 . , certain time after it had been difcharged, and then became 



pofitive; and alfo that it required a greater number of motions 

 of the lever to produce a certain eftedt on the ele6lrometer 

 with negative eledtricity, the longer it flood after it had been 

 charged with negative eleftricity, and that when it became 

 pofitive, the longer it flood the lefs number of motions oLthe 

 lever it required to efFe6l the ele<5lrometer with pofitive elec- 

 tricity to a certain degree, and this within fome certain limits. 

 I was at a lofs how to- account for this change. However, 

 after fome confideration, I began to fufpefl that the plates (al- 

 though all of the fame metal, copper) had each a property 

 peculiar to itfelf, of acquiring a certain fmaH charge of one 

 kind of electricity in preference to the other; and that if they 

 were left to themfelves they would naturally do fo. I therefore 

 began a fet of experiments to afcertain the probability of this 

 fuppofition. Firfl, direClIy after the inflrument had been ufed 

 for negative eleif^ricity, I difcharged it by touching each of the 

 plates with a metal point, which I held in contadl with each 

 plate for two or three feconds. An eleCtrometer was then 

 connected with the plate A, and while it was in this fituation, 

 it was made to diverge with pofitive eleClricity ;* it was then 



* The eleflrometer was made to diverge by bringing an excited 

 glafs or fealing wax near it. 



dif- 



/' 



