A theory due to Weber and later improved by Maxwell which is useful in the correlation of the various phenomena observed in the magnetic behavior of iron is at once suggested by a sim ple experiment. Take a magnetic steel needle and which shows distinct polarity. Upon being broken into two parts it will be found that in stead of securing two isolated poles, that each piece possesses a plus and a minus pole prac tically identical with the poles of the original needle. Carry this process to any length and each little piece, however small, will be found to possess two poles, one positive and the other negative. If we assume that this process could be carried on to the smallest conceivable particle of the iron we should say that each molecule of the iron is by itself a magnet. We may further suppose that the molecular magnets in a neutral piece of iron are entirely devoid of regular ar rangement as regards position of the poles; such a chaotic condition may be indicated roughly by Fig. 7. It may be supposed that these molecular magnets are partly held in.position by the action of forces analogous to friction, which also tend to hold them in any new position to a greater or less extent in case the original arrangement is disturbed. Under the action of a weak mag netic field these friction forces would prevent the turning of the molecular magnets into paral lelism with the field lines. As soon as the field is strong enough to overcome this sort of fric tion we might expect the same tendency to ar rangement of these minute magnets that is ob served in the case of iron filings in the mapping of magnetic fields. As long as a considerable number of the axes of these molecular magnets make fairly large angles with the field lines the leverage by which turning is produced would be considerable; if however they approach parallel ism with each other and the field lines, the ef fective twisting would be very materially re duced. This would correspond to the approxi mate saturation of the iron and no considerable change in position could be produced by increas ing the field strength. The general arrangement may be indicated approximately in. Fig. 8, where it will be observed that there is a tendency for free positive poles to appear at one end, namely, where the field lines leave the iron, and for un compensated negative poles to appear at the other end.
The facts in favor of this sort of explana tion may perhaps be briefly summarized as follows: (1) The general shape of the mag netic curve is explained. (2) As friction or other molecular forces tend to prevent a return to the original chaotic condition after magneti zation this arrangement would in part persist after removal from the field, or permanent mag netism would be explained. (3) Soft iron i) should be easier to magnetize and less erma nent than hardened iron or steel. (4) Jarring as by blows tends to reduce friction to as sist in the process of magnetization and also to reduce permanent magnetism. (5) High molec ular activity consequent on rise of temperature decreases magnetic action, in fact, at a dull red heat iron is non-magnetic. (6) Rapid reversals of magnetism involves work against molecular forces and the production of heat, this heat is proportional to the area of the loop. (7) A tube of iron filings or a set of pivoted magnets shows the same behavior in a rising or falling magnetic field as a solid bar. The precise agree ment between experimental facts and the indi cationsfrom theory shown above makes this conception extremely useful. Just why mole
cules of iron should be permanently endowed withetic properties is a subject for specu has been indulged in by numerous prominent scientists. It has been supposed for example that electrical currents flow around these molecules, that they consist of vortex rings or that small electrically charged parts are in vibration in such a way as to produce the phenomena of permanent molecular magnetism. The general usefulness of the hypothesis is is no way connected with the truth or falsity of such speculations any more than the facts re garding free fall are dependent on our view of gravitation.
The general statements noted above regard ing magnetism of iron are of importance in the manufacture and utilization of magnets for various purposes. Where a considerable amount of permanent magnetism is undesirable, soft iron or steel is always used. For the manufac ture of permanent magnets special steel is se lected and hardened and is then magnetized by its insertion into a solenoid carrying a power ful electric current. Severe shocks or blows are frequently given in order to assist in the molecular rearrangement. The interaction of all the elementary magnets together with tem perature changes and mechanical shocks will tend to weaken a magnet. This loss is very considerable at first but finally an almost permanent state is reached. When used in electrical measuring instruments magnets are artificially aged by subjecting them to con siderable changes of temperature and a series of mechanical shocks. It should be noted in this connection that the permanence of the mag net will be somewhat increased by joining its poles, when not in use, by a piece of soft iron. The induced magnetism at the ends of the iron helps to hold the poles of the small molecular magnets in position, and counteracts the tend ency of these poles to demagnetize the bar.
Aside from the extended use of permanent magnets in electrical instruments their practi cable application is comparatively limited. The electro-magnet is widely used where it is de sired to cause a temporary force action at a i distance from the operator, as for example in the telegraph, etc. Powerful electromagnets are now frequently used to lift large masses of iron during manufacturing•processes. It may be noted in this connection that the lifting force of a magnetic piece of iron depends on the square of the number of lines per unit area at the contact face. Only so large an area should be used in contact as can be very highly mag netized by the current available. In the con struction of dynamos, motors and transform ers, the magnetic quality of the iron used is of great importance. The total number of lines set up (flux) must be sufficient for the opera tions involved and saturation should not he approached in any part. Where the cross sec tion may be made large, cast iron can be used, hut where the flux must he concentrated, special soft iron or steel i.. required. All air gaps are made as small as mechanical and electrical con siderations of construction will permit, in order that the required flux may be more easily set up by the electromagnets. Or as electrical engi neers say, the magnetic °reluctance is made small by use of properly proportioned iron parts and small air gaps in order that the mag neto-motive force required may not be ex cessive.