We pass, therefore, to electrical communication as effected by induction--the influence which one conductor exerts on another through an intervening insulator. At the outset we shall do well to bear in mind that magnetic phenomena, which are so closely akin to electrical, are always inductive. To observe a common example of magnetic induction, we have only to move a horseshoe magnet in the vicinity of a compass needle, which will instantly sway about as if blown hither and thither by a sharp draught of air. This action takes place if a slate, a pane of glass, or a shingle is interposed between the needle and its perturber. There is no known insulator for magnetism, and an induction of this kind exerts itself perceptibly for many yards when large masses of iron are polarised, so that the derangement of compasses at sea from moving iron objects aboard ship, or from ferric ores underlying a sea-coast, is a constant peril to the mariner. Electrical conductors behave much like magnetic masses. A current conveyed by a conductor induces a counter-current in all surrounding bodies, and in a degree proportioned to their conductive power. This effect is, of course, greatest upon the bodies nearest at hand, and we have already remarked its serious retarding effect in ocean telegraphy. When the original current is of high intensity, it can induce a perceptible current in another wire at a distance of several miles. In 1842 Henry remarked that electric waves had this quality, but in that early day of electrical interpretation the full significance of the fact eluded him. In the top room of his house he produced a spark an inch long, which induced currents in wires stretched in his cellar, through two thick floors and two rooms which came between. Induction of this sort causes the annoyance, familiar in single telephonic circuits, of being obliged to overhear other subscribers, whose wires are often far away from our own.