Effect of Heat on Yagnets.—When a magnet is heated to redness it loses permanently *very trace of magnetism; iron, also, at a red heat, ceases to be attracted by the magnet. At temperatures below -red heat the magnet parts with some of its power, the loss increasing with the temperature. The temperatures at which other substances affected `by the magnet lose their magnetism differ from that of iron. Cobalt remains magnetic .at the highest temperatures, and nickel loses this property at 662° F.
Ampere's Theory of Magnetism.—This theory forms the link between magnetism and galvanic electricity, and gives a simple explanation of the phenomena, of electromag netism and magneto-electricity. We shall therefore preface the short discussion of these two subjects by a reference to it. Ampere considers that every particle of a mag net has closed currents circulating about it in the same direction. A section of a rnag net according to this theory is shown in fig. 5. All the separate currents in the various particles may, however, be considered to be equivalent to one strong current circulating Tound the whole (fig. 6). We are to look upon a magnet, then, as a system, so to speak, of rings or rectangles, placed side by side, so as to form a cylinder or prism, in each of which a current in the same direction is circulating. Before magnetization the currents run in different directions, so that their effect as a system is lost, and the effect. of induction is to bring them to run in the same direction, The perfection of magnetiza tion is to render the various currents parallel to each other. Soft Iron, in consequence of its offering no resistance to such a disposition, becomes more powerfully magnetic under induction than steel, where such resistance exists. Experiment very strongly confirms the truth of this theory. Helices of copper wire, in which a current is made to circulate, manifest all the properties of a magnet. Such are shown, in skeleton, in figs. 7 and 8. Each convolution of the spiral may be taken as a substitute for one of the ring,s above spoken of. In helix fig. 7, the current, after entering, goes from right to left (contrary to the hands of a watch), and it is hence called left-handed; in fig. 8 it goes with the hands of a watch, and is right-handed. The extremities of both helices
act on the magnetic needle like the poles of a magnet while the current passes. The poles am shown by the letters N and.S. and this can be easily deduced from Ampere's ' rule (see GALVANISM), for, suppose the little figure of a naan to be placed in any part of the helix fig. 7, so that, while he looks towards the axis of the helix, the current enters by his feet, and leaves by his head, the north pole will be at his left hand, as shown in the figure. In the left-handed helix (fig. 8), the poles are reversed according to the same rule. If either of these helices be hung so as to be capable of horizontal motion, which, by a simple construction, can easily be done, as soon as the current is established, the north and south poles place themselves exactly as those of the magnetic needle would do; or, if they were hung so as to be able to move vertically in the magnetic meridian, they would take up the position of the dipping-needle (q.v.).
These movements can be still further explained by reference to the mutual action of electric currents on each other. It is found that when two currents are free to move, they endeavor to place themselves parallel to each other, and to move in the same direction, and that currents running in the same • direction, attract, and those running in oppo site directions repel. The apparatus fig. 9 is intended to prove this. The rectangle cdefis movable round the pins a and b, resting on two mercuty cups. The arrangernent is such that while the rectangle cd ef is movable about its axis, a current can continue steadily to flow in it. Further description is unnecessary, the diagram explaining itself. If a wire in which a cur rent passes downwards be placed vertically near c d, cd is attracted by it; but if the current pass upwards, it is repelled, and e f attracted. Place, now, the wire below and parallel to d e. If the current passes in the direction d to e, no change takes place, as the attrac tion cannot show itself ; but if the current moves from e to d, the whole turns round till it stands where e was, and both currents run the same way. If the wire be placed at right angles to d e, the rectangle turns round and comes to rest, when both currents are parallel, and in the same direction.