ELECTRICITY AND MECHANICAL FORCE. 401 



inserted a small unit-jar of 9 square inches surface, the knobs 

 of which are 0*2 inch apart. 



The balance of the electrometer is now fixed by a stiff wire 

 inserted between the attracting knobs, and the Leyden jar 

 charged by discharges from the unit-jar. After a certain 

 number of these (twenty-two in the experiment performed in 

 the theatre of the Institution), the discharge of the large jar 

 takes place across the ^-inch interval ; this may be viewed as 

 the expression of electrical power received from the unit-jar. 

 The experiment is now repeated, the wire between the balls 

 having been removed, and therefore the ' tip ' or the raising of 

 the weight is performed by the electrical repulsion and attrac- 

 tion of the two pairs of balls ; at twenty-two discharges of 

 the unit-jar the balance is subverted, and one knob drops 

 upon the other, but no discharge takes place, showing that some 

 electricity has been lost, or converted into the mechanical 

 power which raises the balance. By another mode of expres- 

 sion the electricity may be supposed to be masked or ana- 

 logous to latent heat, and would be restored if the ball were 

 brought back, without discharge, by extraneous force. 



The experiment I believe to be new, and to be suggestive 

 of others of a similar character, which may be indefinitely 

 varied. Thus, two balls made to diverge by electricity should 

 not give to an electrometer the same amount of electricity as if 

 they were, whilst electrified, kept forcibly together. An experi- 

 ment of this sort I have made since my lecture in the following 

 manner: To a thick brass wire, two feet long, insulated and 

 terminated by knobs, are suspended, by fine platina wires, 

 two pairs of discs of paper coated with tinfoil, and four inches 

 in diameter. The apparatus is electrised in a dry atmosphere 

 by sparks from a machine, and the discs of each pair respect- 

 ively diverge. To one of the pairs a silk thread is attached, 

 by which the discs can be forcibly approximated, and as rfften 

 as this is done the divergence of the other pair increases. 



Another mode of showing the same effect is the following : 

 On the top of an ordinary gold-leaf electroscope place two 

 brass plates, such as those commonly used for a condenser ; 

 connect them by a long fine wire, and electrify them by a 

 rubbed rod of glass or sealing-wax, so that the gold-leaves 



D D 



