In like manner Edison developed a "pressure" or carbon relay, adapted to the transference of signals of variable strength from one circuit to another. An ordinary relay consists of an electromagnet inserted in the main line for telegraphing, which brings a local battery and sounder circuit into play, reproducing in the local circuit the signals sent over the main line. The relay is adjusted to the weaker currents likely to be received, but the signals reproduced on the sounder by the agency of the relay are, of course, all of equal strength, as they depend upon the local battery, which has only this steady work to perform. In cases where it is desirable to reproduce the signals in the local circuit with the same variations in strength as they are received by the relay, the Edison carbon pressure relay does the work. The poles of the electromagnet in the local circuit are hollowed out and filled up with carbon disks or powdered plumbago. The armature and the carbon-tipped poles of the electromagnet form part of the local circuit; and if the relay is actuated by a weak current the armature will be attracted but feebly. The carbon being only slightly compressed will offer considerable resistance to the flow of current from the local battery, and therefore the signal on the local sounder will be weak. If, on the contrary, the incoming current on the main line be strong, the armature will be strongly attracted, the carbon will be sharply compressed, the resistance in the local circuit will be proportionately lowered, and the signal heard on the local sounder will be a loud one. Thus it will be seen, by another clever juggle with the willing agent, carbon, for which he has found so many duties, Edison is able to transfer or transmit exactly, to the local circuit, the main-line current in all its minutest variations. In his researches to determine the nature of the motograph phenomena, and to open up other sources of electrical current generation, Edison has worked out a very ingenious and somewhat perplexing piece of apparatus known as the "chalk battery." It consists of a series of chalk cylinders mounted on a shaft revolved by hand. Resting against each of these cylinders is a palladium-faced spring, and similar springs make contact with the shaft between each cylinder. By connecting all these springs in circuit with a galvanometer and revolving the shaft rapidly, a notable deflection is obtained of the galvanometer needle, indicating the production of electrical energy. The reason for this does not appear to have been determined.