178 Spreading is dependent on strength of charge 



For this purpose I first found what proportion the charge of a tin 

 cylinder 15 feet long and 17 inches in circumference bore to that of the 

 two plates D and E together when the electricity was very weak. This 

 I did in the manner represented in Fig. 27, where AB is the tin cylinder 



Fig. 27. 



supported horizontally by non-conductors. DC is a brass wire 37 inches 

 long and about inch in diameter supported also horizontally by non- 

 conductors, the end C being in contact with the cylinder, and a pair of 

 fine pith balls being suspended from the other end D. FE is a piece of 

 wire communicating with the prime conductor, and between it and DC 

 is suspended by a silk string the wire FT in a vertical situation. 



359] The cylinder AB, and consequently the wire DC, were first 

 electrified negatively to such a degree as to make the pith balls separate 

 to the distance of one diameter of the balls. The prime conductor and 

 wire FE being then charged to the usual degree, as shewn by the usual 

 electrometer hung down from it, one end of the wire W was brought in 

 contact with E so as to be electrified by it, and was then immediately 

 removed and brought in contact with DC so as to communicate its elec- 

 tricity to the cylinder*. 



Now I found that if the wire W was 29 inches long, and $ in diameter, 

 and its electricity was twice communicated in this manner to the cylinder, 

 the pith balls would separate as much positively as they before did nega- 

 tively, consequently the cylinder AB and the wires DC and W together, 

 when electrified to such a degree as to make the pith balls separate one 

 diameter, contain as much electricity as the wire W alone does when 

 electrified in the usual degree. 



The cylinder AB was then removed, and the two glass plates D and E 

 placed under the wire DC in its room, their upper coatings communicating 



* [See plan at Art. 539.] 



