ELECTRICITY AND RAIN.

651

falls may have an opposite charge to the portion which is left suspended.

Lord Rayleigh has found that drops of water which are slightly electrified have a tendency to coalesce, whereas unelectrified drops usually rebound after collision.1 It is probable that electricity may in this way give rise to showers.

The coalescence of small drops to form large ones, though it increases the electrical density on the surfaces of the drops, does not increase the total quantity of electricity, and therefore (§ 611) cannot directly influence the observed potential.

Thunder-storms and other powerful manifestations of atmospheric electricity seem to be accompaniments of very sudden and complete condensation which gives unusu

ally free scope to the causes of irregular distribution just indicated.

He

659. Hail.- Hail has sometimes been ascribed to an electrical origin, and a singular theory was devised by Volta to account for the supposed fact that hailstones are sustained in the air. imagined that two layers of cloud, one above the other, charged with opposite electricities, kept the hailstones continually moving up and down by alternate attraction and repulsion. An experiment called electric hail is sometimes employed to illustrate this idea. Two metallic plates are employed (Fig. 413), the lower one connected with the earth, and the upper one with the conductor of the electrical machine; and pith-balls are placed between them. As the machine is turned, the balls fly rapidly backwards and forwards from one plate to the other.

1 The observation was made upon jets discharged from a nozzle directed upwards, and it was found that connecting the discharging vessel with one pole of a single Grove's cell (the other pole being to earth) was sufficient to produce coalescence of the drops. Proc. Royal Society, Feb. 27, 1879.

660. Waterspouts.-Waterspouts, being often accompanied by strong manifestations of electricity, have been ascribed by Peltier

and others to an electrical origin; but the account of them given in the subjoined note appears more probable.1

1 "On account of the centrifugal force arising from the rapid gyrations near the centre of a tornado, it must frequently be nearly a vacuum. Hence when a tornado passes over a building, the external pressure, in a great measure, is suddenly removed, when the atmosphere within, not being able to escape at once, exerts a pressure upon the interior, of perhaps nearly fifteen pounds to the square inch, which causes the parts to be thrown in every direction to a great distance. For the same reason, also, the corks fly from empty bottles, and everything with air confined within explodes. When a tornado happens at sea, it generally produces a waterspout. This is generally first formed above, in the form of a cloud shaped like a funnel or inverted cone. As there is less resistance to the motions in the upper strata than near the earth's surface, the rapid gyratory motion commences there first. . . This draws down the strata of cold air above, which, coming in contact with the warm and moist atmosphere ascending in the middle of the tornado, condenses the vapour and forms the funnel-shaped cloud. As the gyratory motion becomes more violent, it gradually overcomes the resistances nearer the surface of the sea, and the vertex of the funnel-shaped cloud gradually descends lower, and the imperfect vacuum of the centre of the tornado reaches the sea, up which the water has a tendency to ascend to a certain height, and thence the rapidly ascending spiral motion of the atmosphere carries the spray upward, until it joins the cloud above, when the waterspout is complete. The upper part of a waterspout is frequently formed in tornadoes on land. When tornadoes happen on sandy plains, instead of waterspouts they produce the moving pillars of sand which are often seen on sandy deserts."-W. Ferrel, in Mathematical Monthly.

...

MAGNETISM.

CHAPTER LI.

GENERAL STATEMENT OF FACTS AND LAWS.

661. Magnets, Natural and Artificial.-Natural magnets, or lodestones, are exceedingly rare, although a closely allied ore of iron, capable of being strongly acted on by magnetic forces, and hence called magnetic iron-ore, is found in large quantity in Sweden and elsewhere. Artificial magnets are usually pieces of steel, which have been permanently endowed with magnetism by methods which we shall hereafter describe. Magnets are chiefly characterized by the property of attracting iron, and by the tendency to assume a particular orientation when freely suspended.

is

662. Force Greatest at the Ends.-The property of attracting iron very unequally manifested at different points of the surface of a magnet. If, for example, an ordinary bar-magnet be plunged in iron-filings, these cling in large quantity to the terminal portions, and leave the middle bare, as in the lower diagram of Fig. 415. If the magnet is very thick in proportion to its length, we may have filings adhering to all parts of it, but the quantity diminishes rapidly towards the middle. The name poles is used, in a somewhat loose sense, to denote the two terminal portions of a magnet, or to denote two points, not very accurately defined, situated in these portions. The middle portion, to which the filings refuse to adhere, is called neutral.

Fig. 415.-Magnets dipped in Filings.

663. Lines Formed by Filings.-If a sheet of card is laid horizontally upon a magnet, and wrought-iron filings are sifted over it, these will, with the assistance of a few taps given to the card, arrange

themselves in a system of curved lines, as shown in Fig. 416. These lines give very important indications both of the direction and intensity of the force produced by the magnet at different points of the

space around it. They cluster very closely about the two poles pp, and thus indicate the places where the force is most intense.

664. Curve of Intensities.-Some idea may be obtained of the relative intensities of magnetic force at different points in the length of a magnet, by measuring the weights of iron which can be supported at them. Much better determinations can be obtained either by the use of the torsion-balance, or by counting the number of vibrations made by a small magnetized needle when suspended opposite different parts of the bar, the bar being in a vertical position, and the vibrations of the needle being horizontal. The intensity of the force is nearly as the square of the number of vibrations; on the same principle that the force of gravity at different places is proportional to the square of the number of vibrations of a pendulum (§ 120). Both these methods of determination were employed by Coulomb, who was the first to make magnetism an accurate science; and the results which he obtained are represented by the curve of intensities A M B (Fig. 417). M is the middle of the bar, O one end of it, and the ordinates

1 The lines formed by the filings may be called the lines of effective force for particles only free to move in the plane of the card. The lines of total force cut the card at various angles, and are at some places perpendicular to it, as shown by the filings standing on end. For the definition of lines of magnetic force, see § 672.

of the curve (that is, the distances of its points from the line OX) represent the intensities of force at the different points in its length. The curve was constructed from observations of the force at several points in the length; but in dealing with the observation made opposite the very end, the force actually observed was multiplied by 2. Perfect symmetry was found between the intensities over the two halves of the length. In the figure we have inverted the curve for one

M

Fig. 417.-Curve of Intensities.

half, in order to indicate an opposition of properties, which we shall shortly have to describe. The curves of intensities for two magnets of different sizes but of the same form are usually similar.

665. Magnetic Needle.-Any magnet freely suspended near its centre is usually called a magnetic needle, or more properly a magnetized needle. One of its most usual forms is that of a very elongated rhombus of thin steel, having, very near its centre, a concavity or cup by means of which it can be balanced on a point. When it is thus balanced horizontally, it does not, like a piece of ordinary matter, remain in equilibrium in all azimuths,1 but assumes one particular direction, to which it always comes back after displacement. In this position of stable equilibrium, one of its ends points to magnetic north, and the other to magnetic south, which differ in general by several degrees from geographical (or true) north and south. This is the principle on which compasses are constructed.

Fig. 418.-Magnetized Needle.

1 All lines in the same vertical plane are said to have the same azimuth. Azimuthal angles are angles between vertical planes, or between horizontal lines. The azimuth of a line when stated numerically, is the angle which the vertical plane containing it makes with a vertical plane of reference, and this latter is usually the plane of the meridian. Some