The effective values of the compensated coils are adjusted to give a range of resistances from 0.1 to 25.6 ohms on the binary scale in order to facilitate calibration, since their relative values are required with the greatest possible accuracy. (See CALIBRA TION.) The smallest coil, marked I, is equivalent to io cm. of the bridge wire, or to I° C with a thermometer having a fundamental interval of 10 ohms, and a zero resistance of approximately 25.6 ohms, which is that of the largest coil in the box.
Adjustment of Bridge-wire.—In using the bridge by the balance method, the resistance is balanced with the coils A, B, C, D, etc., to the nearest 0.1 ohm. The next three figures may be obtained by finding the balance point on the bridge wire to 0.1 mm. The bridge-wire is made of platinum-silver and is drawn down to a resistance slightly exceeding o•oo5 ohm per cm., giving an effective resistance (in terms of shift of the balance point) equal to double its actual resistance, or a little more than o•oi ohm per cm., so that a length of less than I o cm. is equivalent to the smallest coil I. Its whole length from L to R is then shunted with a suitable resistance until a length of I o cm. near the middle of the wire is precisely equivalent to the coil I. The whole length may then be tested for uniformity against coil I, but should not vary by more than i in 1,000 if the wire is of uniform quality and carefully drawn. It is seldom necessary to calibrate the bridge-wire more accurately than this, since it is employed only for measuring small differences, averaging 0.0025 ohm, for which an accuracy of about 0.0000025 ohm usually suffices. It is most important, how ever, to keep it stretched at a uniform and constant tension, and to ensure that it is not damaged by the sliding contact, and does not shift relatively to the scale SK. This cannot be secured by fix ing it with reference to the ebonite base at both ends, because the coefficient of expansion of ebonite is so much larger than that of the wire. Accordingly the bridge-wire LR is mounted on a frame, with the brass scale SK on one side and a steel bar OQ on the other, in a position corresponding approximately with its coeffi cient of expansion, which is intermediate between that of brass and that of steel. The right hand ends K,Q of these bars are fixed to the ebonite lid, but the left hand ends S,O, are free to slide so as to maintain the tension constant. The galvanometer wire MN is also of platinum-silver and similarly mounted. Contact is made between the two wires LR and MN at the sliding contact J by means of a short length of the same platinum-silver wire, with a sharp edge, rigidly fixed to the vernier carriage at right angles to the scale. Normally the bridge-wire is held just clear of the
contact piece by the flanges of the carriage, and the galvanometer circuit is open. Contact is made at any desired point by turning an ebonite screw, which locks the carriage, and presses the bridge wire down on the contact piece by means of an ebonite finger, with a renewable sleeve of soft rubber tubing, so that neither the wire, nor the sharp edge of the contact, is damaged. The contact is held by the screw while the battery is reversed to test the bal ance and eliminate thermal effects. If the balance is not exact, the deflection of the galvanometer is noted, and the contact is re leased and moved in the required direction. After two readings of deflection, the balance point may be found by interpolation.
Mercury Plug Mechanism.—Some of the plugs are removed from the box in figure 8 to show the mercury cups and connections. In practice the plugs are not removed, but merely raised vertically and kept suspended about half an inch above their respective cups, so that any mercury that may drip from the amalgamated rods falls back into the cup from which it came. To the centre of each 4-point plug is attached a brass rod sliding vertically in an overhead bar running the length of the box, and supported on pillars at either end. The plug is held in the raised position and prevented from turning by a ball and spring catch sliding in a groove in the brass rod. Four or five of the plugs can be raised or depressed simultaneously with one hand, so that the manipulation involved in changing the resistance is far quicker than with any other type of plug box. The mercury plugs and cups give more perfect contact and are much easier to clean than ordinary brass plugs and sockets. The overhead bar, together with all the plugs, can be removed in a few seconds by unscrewing two thumb-nuts, on the rare occasions when the cups require refilling or cleaning. These provisions are of special importance in industrial work, where the apparatus may be exposed to damp, or dust, or gas fumes, and the object is to save time. In such an environment it is often advisable to enclose the whole box in a dust-proof case, leaving only the operating rods exposed, and to abandon the use of the bridge-wire and balance method under such conditions in favour of the deflection method next described.