of soft iron, changing its place in the bar with every change in the position of the bar, but is constant like that of a steel bar, retaining the same magnetism whatever be the position of the bar. By reversing the position of the bar and striking it a few blows with the hammer, its magnetism is reversed. The magnetism of the bar so struck resembles that of a steel magnet in all respects but this, that while, perhaps, no change can be remarked in hours or days, it infallably diminishes in a long time. To express this partially permanent character, the term Subpermanent Magnetism has been adopted.

196. A sphere of soft iron will be magnetised in the same way, however held. The diameter in the line of dip will be the axis of magnetism, and the lower and north half of the surface will be north, the upper and south half south.

In bodies of any other shape the effects will be similar.

197. In the northern hemisphere all vertical or upright bars, such as stanchions and angle irons composing the frames of ships, are magnetised by induction, their lower ends being north poles, the upper ends south poles, the upper ends attracting the north pole of the needle held near them. On the other hand, in the southern hemisphere, these conditions are reversed; the upper ends of vertical iron are north poles, repelling the north pole of a compass needle and attracting the south pole. On the magnetic equator, where there is no dip, vertical soft iron has no polarity, because its position is at right angles to the earth's line of force or dip. It is different with horizontal pieces of soft iron; they exert the same influence on a compass needle in both hemispheres, and in all latitudes.

198. The hull of an iron ship acts as a permament magnet on compasses placed outside the vessel as well as those placed inside; an iron ship must therefore be viewed in its effect on a properly placed magnet rather as one great magnet, than as an aggregation of smaller magnets.

Keeping in view that the inductive effect from the earth's magnetism is greatest in the line of the dip, and the existence of a neutral equatorial plane at right-angles to the line of dip in spherical bodies, we are prepared to see that each iron ship must have a distinct distribution of magnetism depending on the place of building, and the direction of the head and keel while building; the ship's polar axis and equatorial plane conforming more or less to the line of dip of the earth at the place where built, and a plane at right-angles to that line; abundant observation and experiment have proved this important general principle.

199. To illustrate this principle: let us suppose, as in the following figures 3, 4, 5, and 6, that four iron ships, or four composite built ships, with ribs, beams, stanchions, and deck girders of iron, are building on the cardinal points of the compass, in a port in England where the dip of the needle is 70°.

Fig. 3 shows the magnetic state of a ship built head North magnetic. The line marked Dip passes through the centre of the ship; it shows the direction of the line of the earth's magnetic force. The line marked Equatorial or Neutral line is the line of no deviation, and runs at right-angles to The after body of the ship, or the portion which is shaded, has

the Dip.

S. (blue) polarity, and the fore body, or white portion in the figure, N. (red) polarity; the upper part of the stern would have the S. (blue) polarity developed in a high degree; the lower part of the bows would have the N. (red) polarity equally developed. At the stern the north end of a compass needle would be strongly attracted; at the bow the south end of the needle would be strongly attracted; while a compass placed outside of the ship's topsides, above the line of no deviation, the north end of the needle will be attracted; if it be placed below that line the north end of the needle will be repelled and the south end attracted, in accordance with the law of magnetism. (No. 191).

Fig. 4 represents the magnetic condition of a ship built head South. It will be seen by comparing fig. 4 with fig. 3 that the conditions are reversed; in fig. 3 the magnetism of the after body of the ship is south (blue), while in fig. 4 the after part of the ship possesses north (red) polarity; now the fore body of the ship has S. (blue) polarity, while in fig. 3 it has N. (red) polarity; the upper part of the bow has S. (blue) polarity developed in a high degree, and the lower part of the stern N. (red) polarity equalled developed. At the

stern the N. end of a needle would be repelled, and also attracted to the strong S. (blue) pole at the bow. The dotted line crossing the equatorial line in figs. 3 and 4 shows the probable position of the neutral line after the ship has been some time afloat, with her head in an opposite direction to that in which she was built, or after she has made a voyage.

Fig. 5 is intended to show the magnetic state of a ship whose head has been east on the building slip. The whole of the upper part of the ship would have S. (blue) polarity; the whole of the lower part would have N. (red) polarity; but the magnetism of the starboard side of the upper works would be developed in higher degree than the port side, and the N. end of a compass needle, if carried at the usual height of a compass along the amidship line of the upper deck from end to end, would be attracted to the starboard side.

In fig. 6, ship built head west, the magnetic conditions of fig. 5, head east, are reversed; the whole of the upper part of the ship has still S. polarity, and the lower N. polarity; but the magnetism of the port side of the upper works is developed in a higher degree than the starboard side, and the N. end of a compass needle, if carried along the upper deck from end to end, would be attracted to the port side.*

Fig. 7 represents an iron ship built head North in Australia, with a dip of about 68° South. In this ship the shaded part showing S. polarity lies below the equatorial line. It will be useful to compare this figure with figure 3, and mark the difference in the magnetic state of the two ships.

200. A little attention to the above diagrams will give the seaman a rough idea of the distribution of magnetism in iron ships; but it must be borne in mind that all large detached pieces of iron in a ship, such as iron masts, funnels, cylinders, and other masses of vertical iron are independent magnets; in north magnetic latitude, their lower ends being north poles, their upper ends south poles.

201.

The compasses of composite ships with iron frames and iron deck beams, are affected in the same way as those of ships built wholly of iron.

From the special magnetic properties developed in a ship according to her position when building, it follows that a compass aft, in the usual place of the steering binnacle, the character of the deviation-though not the amount-may be approximately represented in a tabular form, as follows:

202.

DEVIATION OF THE COMPASS.

The deviation of the compass is the angle through which the magnetic needle is deflected from its natural position by the disturbing force of iron near it, that is, the angle included between the magnetic meridian and a plane passing through the poles of a compass needle.

The deviation is named East or West according as the north point of the compass so disturbed is to the east or west of its natural position.

Deviation consists of two principal parts, the Semicircular and the Quadrantal, following different laws, and requiring two different kinds of compensation; there is sometimes a third part of small amount called the Constant.

203. In the case of iron ships, as in that of iron bars (195), percussion and vibration by hammering in rivetting render the iron of which the vessel is constructed more susceptible to the inductive force of the earth, and causes the magnetism which the iron of the ship thus acquires to partake more of the character of permanent magnetism. Still this sub-permanent magnetism undergoes a considerable diminution by being submitted to percussion, with the ship's head in a different position to that in which it was when she was being built, and especially if in a contrary direction. But the iron of which a ship is constructed always retains a large amount of this sub-permanent magnetism as long as it remains in the form of a ship. The deviation arising from sub-permanent magnetism is greater than that which is the result of transient induced magnetism. The polarity of the ship's magnetism, while she remains on the stocks, takes the direction of the earth's line of force or dip, and its effects on compasses will evidently depend on the direction of the ship's head was whilst being built. Taking the case of a ship built head north (fig. 3, page 119), the fore part of the ship has acquired north magnetism, and its action will be precisely the same as that of the north pole of a magnet; hence, on northerly courses, the north end of the compass needle will be repelled, and the directive power of the needle will be diminished. On southerly courses the north end of the needle points towards the stern, which has acquired sub-permanent south magnetism, then the directive power of the needle is increased. On easterly and westerly courses the effects on the compass are greatest, since the force acts at right-angles to the needle; and on all intermediate positions of the ship's head the disturbances due to such positions are intermediate. As the ship's head is brought east of north, repulsion of the north end of the needle takes place, and westerly deviation is the result, and it reaches its maximum value when the fore-and-aft line of the ship is at right-angles to the needle; beyond that position the fore part of the ship attracts the south end of the needle, and westerly deviation is still the result. This attraction continues until the ship's head reaches south, when the line of action of the ship lies in the same direction as the needle, and no disturbance occurs, but the directive power of the needle is greater. On bringing the ship's head round west of south, the south pole of the needle still continues to be attracted, which causes easterly deviation, and it again attains its maximum when the fore