506 PHILOSOPHICAL TRANSACTIONS. [aNNO 1788. 



nature, cannot be completely destroyed in any definite time, may be applied to 

 explain the action of the present instrument. When the plates a and b are op- 

 posite each other, the 2 fixed plates a and c may be considered as one mass ; and 

 ' the revolving plate b, together with the ball d, will constitute another mass. 

 All the experiments yet made concur to prove, that these 2 masses will not pos- 

 sess the same electric state; but that, with respect to each other, their electri- 

 cities will be plus and minus. These states would be simple and without any 

 compensation, if the masses were remote from each other; but as that is not the 

 case, a part of the redundant electricity will take the form of a charge in the 

 opposed plates a and b. From other experiments it appears that the effect of the 

 compensation on plates opposed to each other, at the distance of -^ part of an 

 inch, is such that they require, to produce a given intensity, at least 100 times 

 the quantity of electricity that would have produced it in either, singly and apart. 

 The redundant electricities in the masses under consideration will therefore be 

 unequally distributed : the plate a will have about gg parts, and the plate c 1 ; 

 and, for the same reason, the revolving plate b will have 99 parts of the oppo- 

 site electricity, and the ball d 1 . The rotation, by destroying the contacts, pre- 

 serves this unequal distribution, and carries b from a to c, at the same time that 

 the tail k connects the ball with the plate c. In this situation, the electricity in 

 B acts on that in c, and produces the contrary state, by virtue of the communi- 

 cation between c and the ball ; which last must therefore acquire an electricity of 

 the same kind with that of the revolving plate. But the rotation again destroys 

 the contact, and restores b to its first situation opposite a. Here, if we attend 

 to the effect of the whole revolution, we shall find that the electric states of the 

 respective masses have been greatly increased: for the 99 parts in a and in b 

 remain, and the 1 part of electricity in c has been increased so as nearly to com- 

 pensate 99 parts of the opposite electricity in the revolving plate b, while the 

 communication produced an equal mutation in the electricity of the ball. A 2d 

 rotation will of course produce a proportional augmentation of these increased 

 quantities: and a continuance of turning will soon bring the intensities to their 

 maximum, which is limited by an explosion between the plates. 



If one of the parts be connected with an electrometer, more especially that of 

 Bennet, these effects will be very clearly seen. The spark is usually produced by 

 a number of turns between 11 and 20; and the electrometer is sensibly acted on 

 by still fewer. When one of the parts is occasionally connected with the earth, 

 or when the adjustment of the plates is altered, there are some variations in the 

 effects, not difficult to be reduced to the general principles, but sufficiently curious 

 to excite the meditations of persons the most experienced in this branch of na- 

 tural philosophy. If the ball be connected with the lower part of Bennet's elec- 

 trometer, and the plate a with the upper part, and any weak electricity be com-^ 



