The so-called "compound-interest principle," by which a very small charge of electricity might be employed to develop a very large one by the help of mechanical power, was first applied about a century ago in the revolving doubler. Long afterwards, Sir William Thomson availed himself of the same principle in the construction of the "mouse-mill," or replenisher. The Holtz machine, the Voss and Wimshurst machines, and the other induction-machines of the same class, all work on this principle. It may be illustrated as follows: Take two canisters, call them A and B, and place them on glass supports. Let a very small positive charge be given to A, B remaining uncharged. Now take a brass ball, supported by a silk string. Place it inside A, and let it touch its interior surface. The ball will, as shown by Franklin, Cavendish, and Faraday, remain uncharged. Now raise it near the top of the canister, and, while there, touch it. The ball will become negatively electrified, because the small positive charge in A will attract negative electricity from the earth into the ball. Take the ball, with its negative charge, still hanging by the silk thread, and lower it into B till it touches the bottom. It will give all its charge to B, which will thus acquire a slight negative charge. Raise the ball till it is near the top of B, and then touch it with the finger or a metal rod. It will receive a positive charge from the earth because of the attraction of the negative charge on B. Now remove the ball and let it again touch the interior of A. It will give up all its charge to A; and then, repeating the whole cycle of operations, the charge carried on the ball will be greater than before, and increase in each successive operation, the electrification increasing in geometrical progression like compound interest. A Leyden jar having one coating connected to A and the other to B, may thus be highly charged in course of time. A pair of carrier balls or plates, or a number of pairs, may be used instead of one. The carriers, just before leaving A and B, may be put in contact with one another instead of being put to earth; they may be mounted on a revolving shaft, and the forms of A and B modified to admit of the revolution of the carriers, and all the necessary contacts may be made automatically. We thus get various forms of the continuous electrophorus, and if the carriers are mounted on glass plates, and rows of points placed alongside the springs or brushes used for making the contacts, when the charges on the carriers become very strong, electricity will be radiated from the points on to the revolving glass plates, which will thus themselves take the place of the metal carriers. Such is the action in the Voss and other similar machines.