strument may be synonymous with that of the instrument itself.

Resistance of a Partial Earth Fault.-A rough method for the determir ation of fault resistance by the aid of a high resistance voltmeter is due to Swinburne, and consists in first measuring the testing voltage by means of the said voltmeter connected across its terminals, and then obtaining a second, lesser reading from the voltmeter in series with the fault, then

Resistance of Leak + Resistance of Voltmeter
Resistance of Voltmeter

Ammeter and Voltmeter Calibration.-This series would not be complete without a more or less detailed reference to the subject of calibration, or, in other words, the comparison of a commercial instrument with a standard, for the purpose of preparing a scale for, or correct system of, reading the indications of the former. The usual plan adopted in factories where commercial instruments are constructed consists in first standardising one individual instrument by one of the methods to be described, and then adopting it as a standard of comparison with the remainder, checking it from time to time by the original method, as circumstances and previous experience dictate. I will now proceed to deal with one or two methods for calibration of the commercial standards, involving the use of simple apparatus, and, as an opening, we will consider the all-important section included under the heading of Voltmeters.

For voltmeter calibration we require a source of E.M.F. capable of supplying a voltage at least equal to the maximum reading on the scale of the instrument under calibration, and this is best supplied by a battery of accumulators. If an ordinary current-yielding set of accumulators be not obtainable, the requirement may be met by one of the many patterns of testing battery on the market, most of which consist of a pair of lead strips, pasted or "formed" in the usual manner, and immersed in an electrolyte contained in a small test tube. If such a set be adopted for the purpose in hand, it must be borne in mind that no appreciable current can

be taken from them without materially lowering the ter-
minal voltage, and, to this end, it is usual to employ
them in opposition to the standard testing voltage which
is commonly provided by a standard cell or cells.

A simple method of voltmeter calibration, described
by Swinburne, and adopted, I believe, by Messrs. Cromp-

ton and Co. in the earlier days of instrument manufacture, is depicted in Fig. 90, where a, b, and c represent, as usual, the proportional and adjustable arms respectively of a Post Office Wheatstone bridge. V is the voltmeter to be calibrated, E the source of E.M.F. previously alluded to, Es a standard cell (usually Clark's). K2 is an auxiliary circuit key, R an adjustable resistance arranged in convenient form, so as to give a wide range of adjustment, r being an auxiliary rheostat or slide resistance for final balancing in reading to single degrees of the voltmeter scale. rl and r2 are special resistances of 54.6 and 145.4 B.A. ohms respectively, introduced in order to render the arrangement direct reading in B.A. volts, their total resistance being 200 B.A. ohms.

In the event of standards other than B.A. being adopted, the same sum of resistances must be adhered to. G is a galvanometer.

As to the actual manner of conducting the operation, I cannot do better than quote Mr. Swinburne's own remarks on the subject:—

"R, r, in Fig. 90, are adjustable resistances and rheostat coupled up as shown. This resistance box is supposed to have coils for the thousands, so that when all the plugs are out 11,110 ohms are in circuit; rl and r2 are two specially made up resistances of 54.6 and 145.4 ohms respectively. The Clark standard cell is in series with a galvanometer G. The standard cell circuit is from between rl and r2 to the keys, so as to shunt the resistance of 145.4 ohms. If the plugs be withdrawn between the extremities of arm a, there is a resistance of 1,110 ohms in the standard cell circuit. The object of the resistance rl is to make up 200 ohms with r2, so that the apparatus may be direct reading.

166

Suppose the voltmeter is to be calibrated up to 100 volts. To get 50 volts on it, plugs are drawn in the resistance box to give 4,800 ohms, which, with rl and r2, make up 5,000 ohms. The resistances R are then adjusted till approximately 50 volts are on the voltmeter. The key K1 is then pressed for a moment. The galvanometer will be deflected, showing that there is too much or too little electromotive force. The switches R can regulate within 1 per cent., so that a pair of positions will be found, one on each side of the desired resistance. The rheostat r is then adjusted till there is no deflection on pressing Kl. The key K is then pressed, and a final adjustment made. The key K1 and its resistances are put to prevent a large current passing through the cell in either direction, and the key K must only be used for final adjustment. The 50-volt reading having been taken, 5,100 ohms are drawn, and the 51-volt reading is taken, and so on.”

An alternative method, which, however, possesses the attendant disadvantages that it is not direct reading, and that it cannot be adopted except in the case of high resistance voltmeters, the maximum current through the coils of which will not be sufficient to appreciably heat the bridge coils, is represented in Fig. 91, and is based

upon Ohm's familiar law, E equals C.R. In this method, the resistance of the voltmeter is first measured by one of the ordinary methods at its normal temperature, i.e., at that temperature which its winding will have when connected across an electromotive force equal to that which it is ultimately intended to measure. This resistance having been duly obtained and noted, attention must be given to Fig. 91, in which the same lettering is used as in the previous figure. It will be seen on reference to this figure that the voltmeter V, testing E.M.F., E., and resistances R, r, are connected in series with the bridge, whilst the standard cell Es and galvanometer G are in shunt.

The method of conducting the test in this case consists in inserting resistances in the bridge, and adjusting R and r until there is no deflection on the galvanometer G, then, by Ohm's law, the current will be equal to the E.M.F. of the standard cell Es, divided by the resistance as read on the bridge coils. The value for C thus obtained, multiplied by the previously measured voltmeter resistance, gives the true reading of the voltmeter V at the time.

A somewhat rougher method of voltmeter calibration by the aid of a metre bridge is represented in Fig. 92, where A B is the homogeneous slide wire of an ordinary metre bridge, the resistance of which, per unit of length, must be known. R is an adjustable resistance of such a type that it will not appreciably heat with the current passing through it under the maximum E.M.F. to be recorded on the voltmeter V. E is the testing battery, as