Glass becomes conducting as temperature rises 1 8 3 



DBbd was fastened on the tube, so that as the outside coating was made 

 to extend as far as Dd, that is three or four inches above the mercury in 

 FGMN, where the tube was very little heated, and as the inside coating 

 reached still higher, that is to the bottom of the upper ball, no sensible 

 error could proceed from thence. 



The use of the upper ball was to prevent the mercury within the tube 

 from overflowing when hot. 



368] By a mean between the experiments made while the ball was 

 heating and while cooling, its charge answering to the different degrees 

 of heat was as follows. 



369] At 295 the electricity passed through the glass pretty freely, 

 but at 305 much faster. It appears, therefore, that the charge of glass 

 is considerably greater when heated to such a degree as to suffer the 

 electricity to pass through than when cold, but that its charge does not 

 begin to be sensibly increased till it is heated to a considerable degree*. 



370] On the charges of plates of several different sorts of glass, and also 

 of plates of some other substances which do not conduct electricity, charged 

 in the manner of Leyden vials. 



The result of the experiments I made on this subject is contained in 

 the two following tables: 



TABLE OF GLASS PLATES f. 



[Note 26.] 



t [See Art. 673.] 



