tions continually traceable. To a certain extent these are dependent on the varying position of the sun, and, to a much smaller extent, of the moon, with respect to the place of observation; but over and above all regular and periodic changes, there is a large amount of irregular fluctuation, which occasionally becomes so great as to constitute what is called a magnetic storm. Magnetic storms "are not connected with thunder-storms, or any other known disturbance of the atmosphere; but they are invariably connected with exhibitions of aurora borealis, and with spontaneous galvanic currents in the ordinary telegraph wires; and this connection is found to be so certain, that, upon remarking the display of one of the three classes of phenomena, we can at once assert that the other two are observable (the aurora borealis sometimes not visible here, but certainly visible in a more northern latitude)." They are sensibly the same at stations many miles apart, for example at Greenwich and Kew, and they affect the direction and amount

of horizontal much more than of vertical force.

688. Ship's Compass.-In a ship's compass, the box cc (Fig. 434) which contains the needle is weighted below, and hung on gimbals, which consist of two rings so arranged as to admit of motion about two independent horizontal axes tt, uu at right angles to each other, This arrangement prevents it from being tilted by the pitching and rolling of the ship. The needle a b is firmly attached to the compass card, which is a circular card marked with the 32 points of the compass, as in Fig. 435, and also usually divided at its circumference into 360 degrees. The card with its attached needle is accurately balanced on a point at its centre. The needle, which, in actual use, is concealed from view, lies along the line N S. The box contains a vertical mark in its interior on the side next

Airy on Magnetism, p. 204.

the ship's bow; and

COMPASS.

N.N.W.

N.W.

W.N.W.

N.

-N.N.E.

N.E.

677

E.N.E

E.

this mark serves as an index for reading off on the card the direction to which the ship's head is turned. Sometimes a reflector is employed, as m in the first part of Fig. 434, in such a position that an observer looking in from behind can read off the indicated direction by reflection, and can at the same time sight a distant object whose magnetic bearing is required. The origin of the compass is very obscure. The ancients were aware that the loadstone attracted iron, but were ignorant of its directive property. The instrument came into use in Europe some time in the course of the thirteenth century.

W.

W.S.W.

S.W.

S.S.W.

S.

S.S.E.

Fig. 435-Compass Card.

S.E.

E.S.E.

689. Methods of Magnetization.-The usual process of magnetizing a bar consists in rubbing it with or against a bar already magnetized. Different methods of doing this, called single touch, double touch, &c., have been devised, in which magnetized bars of steel were the magnetizing agents. Much greater power can, however, be obtained by means of electro-magnetism; and the two following methods are now almost exclusively employed by the makers of magnets.

1. A fixed electro-magnet (Fig. 436) is employed, and the bar to

be magnetized is drawn in opposite directions over its two poles. Each stroke tends to develop at the end of the bar at which the

motion ceases, the opposite magnetism to that of the pole which is in contact with it. Hence strokes in opposite directions over the two contrary poles tend to magnetize the bar the same way.

2. When very intense magnetization is to be produced, the electromagnet must be very powerful, and the bar then adheres to it so strongly that the operation above described becomes difficult of execution, besides scratching the bar. Hence it is more convenient to move along the bar, as in Fig. 437, a coil of wire through which a current is passing. This was the method employed by Arago and Ampère.

A bar of steel is said to be magnetized to saturation, when its magnetization is as intense as it is able to retain without sensible loss. It is possible, by means of a powerful magnet, to magnetize a bar considerably above saturation; but in this case it rapidly loses intensity.

Pieces of iron and steel frequently become magnetized temporarily or permanently by the influence of the earth's magnetism, and this action is the more powerful as the direction of their length more nearly coincides with that of the dipping-needle. If fire-irons which have usually stood in a nearly vertical position be examined by their influence on a needle, they will generally be found to have acquired some permanent magnetism, the lower end being that which seeks the north.

It sometimes happens that, either from some peculiarity in the structure of a bar, or from some irregularity in the magnetizing pro

cess, a reversal of the direction of magnetization occurs in some part or parts of the length as compared with the rest. In this case the magnet will have not only a pole at each end, but also a pole at each point where the reversal occurs. These intermediate poles are called consequent points. Fig. 438 represents the arrangement of iron

filings about a bar-magnet which has two consequent points a', b'. The whole bar may be regarded as consisting of three magnets laid end to end, the ends which are in contact being similar poles. Thus the two poles at a' and the one pole at a are of one kind, while the two poles at b' and the one pole at b are of the opposite kind.

The lifting power (or portative force) of a magnet generally increases with its size, but not in simple proportion, small magnets being usually able to sustain a greater multiple of their own weight than large ones. Hence it has been found advantageous to construct compound magnets, consisting of a number of thin bars laid side by side, with their similar poles all pointing the same

Magnet.

way. Fig. 439 represents
such a compound magnet
composed of twelve ele-
mentary bars, arranged
4 x 3. Their ends are in-
serted in masses of soft
iron, the extremities of
which constitute the poles
of the system.

Fig. 440 represents a
compound horse-shoe mag-
net, whose poles N and S
support a keeper of soft
iron, from which is hung
a bucket for holding

weights. By continually

adding fresh weights day Fig. 440.-Compound Horse-shoe after day, the magnet may

Magnet.

be made to carry a much greater load than it could have supported originally; but if the keeper is torn away from the magnet, the additional power is instantly lost, and the magnet is only able to sustain its original load.

Much attention was at one time given to methods of obtaining steel magnets of great power. These researches have now been superseded by electro-magnetism, which affords the means of obtaining temporary magnets of almost any power we please.

690. Molecular Changes accompanying Magnetization.-Joule has

shown that, when a bar of iron is magnetized longitudinally, it acquires a slight increase of length, compensated, however, by transverse contraction, so that its volume undergoes no change.

If the magnetization is effected suddenly, by completing an electric circuit, an ear close to the bar hears a clink, and another clink is heard when the current is stopped.

These phenomena have been accounted for by the hypothesis that, when iron is magnetized, its molecules place their longest dimensions in the direction of magnetization.

The effect of heat in diminishing the strength of a magnet is another instance of the connection between magnetism and other molecular conditions. In ordinary cases, this diminution is merely transient; but if a steel magnet is raised to a white-heat, it is permanently demagnetized.

691. Action of Magnetism on all Bodies.-It has long been known that iron and steel are not the only substances which can be acted on by magnetism. Nickel and cobalt especially were known to be attracted by a magnet, though very much more feebly than iron, while bismuth and antimony were repelled. Faraday, by means of a powerful electro-magnet, showed that all or nearly all substances in nature, whether solid, liquid, or gaseous, were susceptible of magnetic influence, and that they could all be arranged in one or the other of two classes, characterized by opposite qualities. This opposition of quality is manifested in two ways.

1. As regards attraction and repulsion, iron and other paramagnetic bodies are attracted by either pole of a magnet, or more generally, they tend to move from places of weaker to places of stronger force. On the other hand, bismuth and other diamagnetic bodies are repelled by either pole of a magnet, and in general tend to move from places of stronger to places of weaker force.

2. As regards orientation, a paramagnetic1 body, when suspended between the poles of a magnet, tends to set axially; that is to say, tends to place its length along the line joining the poles; whereas a diamagnetic body tends to set equatorially, that is, to place its length at right angles to the line joining the poles.

1 The nomenclature here adopted was proposed by Faraday in 1850 (Researches, § 2790), and is eminently worthy of acceptance. Many writers, however, continue to employ magnetic in the exclusive sense of paramagnetic. To be consistent, they should call the other class antimagnetic, not diamagnetic. "The word magnetic ought to be general, and include all the phenomena and effects produced by the power."