The table or stand on which the galvanometer is placed should be separate from that on which the rest of the apparatus is arranged, otherwise it is liable to be shaken when adjusting the other instruments. The table or stand must be arranged to be as free from vibration as possible. In some laboratories the room in which the sensitive galvanometers are kept is in the basement, the stands being built up specially from the foundation, in order to secure freedom from vibration. This cannot always be managed, and very often small brackets are attached to the walls, on which the galvanometers may be placed; provided the walls are thick, and there is no heavy running machinery near them, this makes a very satisfactory stand. All large masses of iron which would influence the galvanometers must be removed from the neighbourhood of the testing-room. Stationary masses of iron, such as iron pillars, are not so important, as the effect due to them will be constant, unless the magnetic field around them alters in value. Also movable magnets, or anything of a magnetic nature, such as knives, keys, watches, etc., must be placed where they will exert no influence on the instruments.

Great care must be taken to prevent the magnetic effect of the current in the other pieces of apparatus affecting the galvanometer. In order to ascertain whether or not there is any such action, it is advisable to send a current through the apparatus, having previously disconnected the galvanometer. Any deflection obtained under these circumstances points to the magnetic effect of some part of the apparatus on the galvanometer, and the apparatus must be rearranged until no such effect is observed. For this reason it is advisable in some cases to twist the wires leading to the galvanometer together for some distance from it.

4. In some experiments, as for instance the measurement of insulation resistance, the insulation of the galvanometer becomes very important, and the instrument must be tested for leakage. This can best be done by connecting up one terminal of the galvanometer to one pole of a battery of several cells, the other pole of the battery being earthed, whilst the other galvanometer terminal is insulated. A deflection of

the needle under these circumstances points to a leakage to earth from some part of the galvanometer coil. This test should be made to each terminal of the galvanometer separately, since the leak might be close to one, when of course no deflection would be obtained on the battery being connected to that terminal. Should a small leakage be found which would interfere with the accuracy of the measurements, the galvanometer should be placed on blocks of freshly scraped paraffin wax, one under each levelling screw; this will completely insulate the instrument from the earth.

5. The galvanometer must be set up so that the needle oscillates freely when disturbed from its zero position, the instrument being levelled so as to bring the needle into the centre of the field. In the case of a reflecting galvanometer, any friction between the needle and the coil can be detected by observing the motion of the spot of light on the scale when the needle is set oscillating, an irregular, jerky motion denoting friction between the two.

It is also advisable, whenever possible, to so arrange the galvanometer that the needle will point to the zero on the scale when under the influence of the earth's field alone. This is not essential, as the needle can always be brought to zero by means of a bar magnet, but it is often desirable.

6. Most galvanometers are supplied with a movable directing magnet, attached to the frame of the instrument, and by means of which the needle may be adjusted to zero by turning the magnet on its axis, or the controlling force may be altered by altering the distance between the needle and controlling magnet. Such an arrangement is shown in Fig. 1. Here the directing magnet is attached to the top of the case of the galvanometer, and the rotation is given to the magnet by means of a worm and worm-wheel, whilst the controlling force is altered by sliding the magnet up or down the vertical brass rod. The great objection to this arrangement is that it is almost impossible to alter the position of the magnet without seriously shaking the whole instrument, thus making it very difficult to rapidly vary the controlling field so as to give a long period of swing to the needle. This difficulty may be overcome by mounting

A simple form

the directing magnet on a separate stand. employed by the author is shown in Fig. 2. A represents a sliding board, moving in the frame F, between the guide. bars G, G; to it is fixed the vertical rod P, on which the magnet slides, so as to adjust it to the height of the gal

vanometer needle. The rod P is free to rotate in its socket, and is turned by means of a small band passing round the pulley S and the pulley T, to which a milled head is attached. The to-and-fro motion is given by means of a band, B, attached to A, and passing over two rollers, R, R, to one of which a milled head, M, is attached. The whole

apparatus may be made in the laboratory workshop. In use it is placed behind the galvanometer, and the magnet adjusted on the rod P until it is on a level with the needle. The milled head M is then turned so as to pull the magnet nearer the galvanometer until the required sensitiveness is obtained.

7. Suspension of the Magnetic System.The suspension of the needle of a galvanometer is a very important matter. The function of the suspending fibre may be twofold-either simply to suspend the needle, or, in addition, to supply the controlling force and of necessity the nature of the suspension varies with the above conditions. If a suspension only is required, the fibre must be as torsionless as possible; if

a controlling force is required, that must be supplied by the torsion of the fibre. In the first case it is necessary to get some substance which, when of sufficiently small diameter, is practically free from torsion, and is at the same time strong enough to support the weight of the suspended system. One such substance is unspun cocoon silk. This is produced by the silkworm as a double thread, each part having a diameter of about 0'0005 inch. The threads are separated from each other by washing them with warm water, painted on from a camel's-hair brush, in order to dissolve the gum by which they are fastened together. One such thread has been found to support a weight of about 5 grammes before breaking. A number of such fibres should be kept ready for use, suspended inside a glass tube, with small 1 See Gray's "Absolute Measurements in Electricity and Magnetism,"

vol. i. p. 241.

weights attached to them to keep them stretched. The torsion in silk fibres is very small, but the fibre is affected by both temperature and moisture, so that the zero of instruments in which such suspensions are employed is liable to alteration.

8. A much better material for a suspension is a thread from a garden spider's web; this, although possessing the properties of an almost ideal suspension, seems to be very little used. It is very strong, considering its size. Joule found that it could support a weight of 2 grammes without breaking.' The greatest point in its favour is that experiments go to show that it is almost absolutely torsionless.2 Experiments by Bottomley and Tanakadate go to show that a spider line capable of carrying a mirror and magnet weighing o'2 gramme has a torsional rigidity that of a single cocoon silk fibre. Its extension with temperature is also very small, a fibre 23 inches long, loaded with a weight of 1 gramme, not altering more than about 0'7% in length for a rise of over 50° C., a lengthening of about 2% taking place when dry air was changed for moist. This, however, need be no drawback, since, on account of its great freedom from torsion, very short fibres may be employed.

9. Another material for galvanometer needle suspension is quartz fibre, originally prepared by Professor C. V. Boys, F.R.S." Such fibres can be obtained of any desired degree of fineness. As regards strength, Professor Boys has shown that, for a fibre nearly 100 of an inch in diameter, the breaking stress was 517 tons per square inch. In addition to its other properties, quartz fibre is an almost perfect insulator even in a damp atmosphere, which makes it invaluable as a means of suspending charged bodies, such as the needle of an electrometer. The following hints regarding the method of suspending substances by quartz fibres have been taken from Professor Boys' Cantor Lectures on 66 Instruments for the Measuring of Radiant

Heat"

:

"Having chosen a fibre of the right diameter, and longer

Joule's "Scientific Papers," vol. i. p. 479.

2 Cantor Lecture on "Instruments for the Measuring of Radiant Heat," Boys, p. II; also Bennet, Phil. Trans., 1792.

3 See Pro. Roy. Soc., vol. 46, p. 253; also Phil. Mag., Aug., 1889.