MAGNETISM plest and most efficient, is as follows: Procure say ten flat bars of good steel bent into the usual form of a horse shoe; let these be well hardened and fitted with their flat sides together so as to form a compound magnet. Each of the members of this bundle may be magnetized separately to a small degree by supporting one of the legs on the lower end of a long rod of iron held nearly perpendicular in this latitude, and the other leg on the upper end of the same rod; or by rubbing one leg with the N. pole of a magnetized bar and the other with the S. pole. The several shoes, or bars, being in this way feebly magnetized, eight of them are joined together with their similar poles in contact, forming a compound magnet with which the remaining two bars are to be magnetized to a higher degree. For this purpose the latter are placed on a table on their flat sides, the N. pole of the one in contact with the S. pole of the other, so as to form a closed circuit; on any part of this circuit the compound horse shoe is placed perpendicular to the plane of the table, with its N. pole in the direction of the S. pole of the bar or shoe on which it rests, and then caused to slide in either direction entirely around the circuit, care being taken to retain its perpendicularity. After having gone over the surface of the two shoes in this way several times, they are turned over without separating their ends, and the process is repeated on the side which was previously under. By this method the two bars will receive a magnetic power nearly equal to the sum of the powers of the eight magnets in the bundle. Next these two bars are placed in the bundle, and two others are taken out and subjected to the same process. These in turn are put into the bundle, and two others are taken out and rubbed in the same way, until each pair of bars has been gone over two or three times in succession. By this method, with the most feeble beginning, the magnetism of the several shoes may be developed to their full capacity, and a magnetic battery produced of great power. A compound horse shoe of this kind is the most convenient instrument for magnetizing straight bars of hardened steel for practical uses. Suppose, for example, we wish to magnetize four bars, each 16 inches long, an inch wide, and an eighth of an inch thick; these are placed on their flat sides in the form of a rectangular parallelogram with their ends in contact; the compound horse shoe is then placed perpendicularly on the middle of one of the bars, and slid entirely around the parallelogram several times in succession; each bar is then turned over in its place so as to bring its lower side upward, and the process repeated, care being taken to keep the horse shoe perpendicular to the plane of the parallelogram, and its poles in the same relative positions to those of the bars. By this method, if the compound horse shoe is sufficiently powerful, the four bars can be magnetized to saturation in the course of a few minutes. If there are but two bars to be magnetized, the parallelogram is completed by joining the ends of these with two similar bars of soft iron, and the same process of rubbing performed as before.--We have seen, in the article ELECTROMAGNETISM, that the most powerful magnetic induction is produced in soft iron by transmitting around a bar of this metal a current of galvanism, and that temporary magnets of great power can be produced in this way. The same method affords the readiest means of strongly magnetizing steel bars. Whatever may be the nature of the change which takes place in iron at the moment of magnetization, we are certain that it pertains to the atoms or molecules of the body, and not to the assemblage of these as a whole. To be convinced of this, it is only necessary to magnetize a steel rod, for example a thick knitting needle, the polarity of which will be exhibited near its two ends, while no attraction will be manifested near the middle. If however we break this into two pieces, we shall find each half is a perfect magnet; the separated ends which were previously joined together in the middle of the whole length will now exhibit polarity. If each of these pieces be again broken in two, we shall have four perfect magnets; and however frequent the division or small the parts into which the needle is divided, each part will still exhibit a N. and S. pole. We may continue, at least in thought, this division, and we have no reason to doubt that however far it might be carried, the same result would be produced. We infer from this experiment that the reason why the middle of a bar exhibits no magnetism is not that none really exists there, but that it is neutralized by opposite polarities. We are also certain that magnetization is attended with at least a momentary motion of the atoms of the iron. This is proved by the fact that during the sudden magnetization of a bar of iron, by means of a current of electricity transmitted through a spiral conductor enclosing the bar, a sound is emitted; and if the bar be rapidly magnetized and demagnetized by an interruption of the current, a musical sound will be produced. This fact was first noted by Dr. Page of the United States, and subsequently experimented upon by De la Rive, Becquerel, and others in Europe. The fact that a change takes place in the molecules is also rendered evident by an experiment of Mr. Joule of Manchester, England, in which he found that, although the whole capacity of the iron bar did not change on being magnetized, yet its dimensions varied, its length being increased and its width correspondingly diminished. That the magnetic force resides on or very near the surface of a magnet has been shown by Jamin, who finds that for every magnet there is a certain relation between the quantity of magnetism and the solid and superficial contents, such as to establish a limit beyond which a given bar cannot exert magnetic
