ether the Glafi fult of BleHfcmiters wtpalr their Effect, 431J 



but forms no ftar upon it. In this experiment I ufed a machine with a cylinder of fevcri 

 inches diameter. 



Hence it fecms to follow, that a hollow ball of paper, or a glafs globe covered with paper, 

 •might form an amufing part of the ele£lrical apparatus for experiments in the dark. 



The laws of a£lion with regard to weak eledlricities and thofe of confiderable intenfitr 

 appear to differ in various particulars, which require further invefllgatlon. Wlien a num- 

 ber of jars are charged in fucccflion, that is to fay, by infuliiting the whole feries, and 

 ' caufing the outer coating of the fird to communicate with the inner of the fecond, and the 

 outer coating of the fecond with the inner of the third, and fo forth, it is well known that 

 the whole may be charged by communicating eledtricity to the infide of the firft, provided 

 the laft hSve a communication with the earth ; but with this circumftance, that the charge 

 will reftore itfelf by explofion when the quantity of electricity is much lefs than the firft 

 jar would fingly have received and retained. From this refult it has with fome probabi- 

 lity been concluc'ed, that glafs refills the communication of dectric energy, and that the 

 jars are fucceffively lefs and lefs charged. Whether this lall conclufion be true, has not 

 ft believe been afcertained by experiment. 



Under all the uncertainties concerning the place occupied by the eleftric charge of 

 coated glafs, though it may feem unfair to make any inference refpefting glafs which is 

 uncoated, yet, upon the whole, there appears to be a probability that the interpofition of 

 naked glafs may impede the aftlon of eleclrlfied bodies. This queftlon more immediately 

 points at the tube In which the gold-leaf eIe£trometer of Bennet is inclofed. To determine 

 whether the tube of the eledrometcr does afFeft the ele£l:ric ftate of the included leaf, ei- 

 ther by compenfation or otherwife, I took a piece of window-glafs eighteen inches long, 

 two Inches wide, and one-twentieth of an inch thick, which I cleaned very well, and then, 

 paffed it feveral times through the hot air over the flame of a candle. In this flate one end 

 of the glafs was laid gently upon the eledlrlfied plate of Bennet's eleftrometer, and thea 

 fuddenly raifed by a turn of the wrill. It was fcarcely poffible to difcern that the leaves were 

 »t all affected ; but when the eledrometer was In the plus flate a very flight coUapfion was 

 produced by raifmg the glafs, and the contrary effedl was produced when the eledrometer 

 was negative. Some days afterwards the experiment was repeated, after the gold-leaf had 

 been changed for other pieces, which were very pointed and delicate In their movements. 

 The refult was, that the glafs was always fhewn by the eleflrometer to be in a weak pofi- 

 tive (late ; and, when the eleflrlcity of the dc£lroraeter wa^ made plus, the collapfion wa« 

 equal to the divergence when it was minus. 



In making thefe experiments I had previoufly fuppofed that the influence of the mt- 



tallic ftate of the ele£lrometer would produce fomewhat of the nature of a charge upoa 



the glafs ; and confequently that the intenfity of the leaves would have been diminlflied 



during the exiftence of that charge ; and alfo, that In fuch a cafe the aftlon of the metal 



through the glafs would be fubjedl to the fame diminution as in the feries of jars. But zi 



the glafs did not appear to aft in this manner, it feems proper to conclude that clean 



glafs does not afFe£t the eleclric ftate of bodies by Its vicinity, and that the divergence 



of the balls or the gold-leaf in the ele(ftrometers of Cavallo and Bennet Is not diminilhed 



by the tube which furrounds them. 



3 L 2 ^ Froirf 



