ple device termed a commutator. This has various forms in different machines, but the principle involved is the same in all. The ends of the wire of the revolving coil are connected with the halves of a cylinder separated by an insulating substance. A metallic brush composed of bundles of wire, or thin strips, presses against each of these sections, and, so long as the cylinder remains stationary, the current taken off by the brushes will be an alternating one; but, when the cylinder revolves with the coil, the brushes will change from one half to the other at the moment of the reversal of the current, and its direction in the circuit will, therefore, always be the same. In most machines the armature has many coils, and the commutator cylinder a corresponding number of insulated sections.

The interest in the electric light during the past few years has resulted in greatly improving such machines, and in the devising of many new forms of varied excellence. A description of one of these, that has attracted wide attention and proved one of the most efficient, will suffice to indicate the general construction and mode of action of such devices. When a magnet is inserted in a closed coil of insulated wire, a momentary current is induced in the coil, and, when it is withdrawn, one of opposite direction occurs. If, instead of withdrawing the magnet, it is passed through the coil, currents will be induced in each of its spirals as the magnet passes them, which will be in one direction during the passage of the first half of the magnet, and in a reverse one during that of the latter half. If two magnets be placed end to end with their like poles in contact, and be bent into the form of a ring, currents can be continuously produced by revolving the ring within an inclosing coil. Mechanical difficulties prevent such an arrangement; but, if, instead of permanent magnets, a ring of soft ironFig. 15. be wound with insulated wire, and revolved between the poles of a magnet, the same results will be obtained and the difficulties avoided. The iron ring becomes a magnet by induction, and as it revolves its poles shift in the reverse direction, so as to always remain opposite those of the inducing magnet. The effect is the same as if the ring remained stationary and the coil revolved. This is the arrangement of the armature adopted in the Gramme machine, and the difference between it and others lies in this mode of inducing the current, instead of by insertion and withdrawal of a magnet from a coil. The manner in which the armature is actually constructed is shown in Fig, 15. The ring is composed of a bundle of soft-iron wires, about