nected, which total must always be maintained in the the test, which is conducted as follows:-The galvanometer G has its free terminal connected in turn to the points a, b, c, and d, and is balanced to zero in each case by withdrawing plugs from the box r, say, and inserting them in r1 such that r + rl still equals 10,000 ohms. Let the values withdrawn in this manner from r be respectively a, b, c, and d, corresponding to the balancing required when G is connected to each of these Ad -C

points in turn, then B=

For the practical measurement of resistances over a range extending from zero to five megohms, Evershed's Ohm-meter is especially applicable, and has withstood the test of time. It is shown in general view in the accompanying illustration, and consists of two essential parts, viz., the ohm-meter proper and the generator. The former consists of an astatic system of magnetic needles delicately suspended at the centre or point of intersection of two coils placed at an angle of 45 degs. with one another. One coil (outer) is connected in series, and the other (inner) in shunt with the circuit containing the resistance to be measured. In the latter pattern of instrument the astatic needles are magnetised by the actual current from the generator, so that, by obtaining a mean of two readings consequent on turning the generator handle first in one direction and then in the other, the instrument may be employed with a fair degree of accuracy in the immediate neighbourhood of strong magnetic fields due to dynamos, etc., in addition to being, through its astaticism, independent of the earth's magnetic field.

The generator, which is contained in a separate case, consists of a special magneto machine capable of producing a range of voltages varying from 10 to 500 at a moderate speed imparted to it by means of a convenient handle, in either direction.

The instrument is direct reading, and only requires connecting up according to the directions supplied with it, and a subsequent rotation of the generator handle at a moderate speed.

To check the accuracy of an ohm-meter, it is best to first measure a resistance of convenient current-carrying

capacity by the ordinary bridge method, and then measure its resistance by ohm-meter, rather than compare with the bridge coils direct, in order to eliminate possible temperature errors due to heating of the coils.

Before departing altogether from the resistance section of this series, we will briefly consider its near neighbour or reciprocal, in the capacity of:

Evershed Ohm-Meter Testing Set.

7. Conductivity Measurement.-It frequently happens, in dealing with conductors of various kinds, that we require to discover what is known as the "percentage conductivity," i.e., the conductivity of the sample conductor under test, as compared with an exactly similar sample of absolutely pure copper, regarded as possessing a conductivity of 100. If my readers have met with many current specifications for electrical work they will in all probability have experienced the term "all copper used to have a conductivity of %." This means that a margin of so much per cent. is allowed for impurities and other causes affecting the conducting power of otherwise pure copper.

A method of determining the percentage conductivity of any given sample consists in cutting off a suitable length, such as 100 feet, for example, and carefully determining its resistance in ohms, by means of the Wheatstone bridge method previously described. It is then weighed in a delicate scale pan, the resultant weight being reduced to grains. The temperature also is carefully observed at the time of making the test; then the 12 × 22.61 percentage conductivity =

w kr

resistance in ohms.
temperature co-efficient.

The numerical values of k at various temperatures are given in tabular form in Kempe's "Handbook of Electrical Testing," and are reproduced herewith.

Co-efficients for correcting the observed resistance of pure copper wire at any temperature to 75° F., or at 75° F. to any temperature:

Table of multiplying co-efficients for reducing the observed resistance of ordinary copper wire at any

temperature to 60° Fahrenheit :

:

In some cases the diameter of the conductor in mils. is more readily obtainable than the weight of the sample tested. In such cases, the percentage conductivity 7 x 1065.6 where 1, k, and r represent the same quan

d2 kr tities as in the previous equation whilst d is the diameter of the sample in mils, or thousandths of an inch.

When dealing with conductors of fine gauge in which the correct diameter is somewhat difficult to determine, it is far better to resort to the weight method, by means of which, given a fairly sensitive balance, great accuracy can be attained.

For rapidly conducting a large series of conductivity tests on conductors of various sizes, Messrs. Nalder Brothers have designed the composite apparatus shown in the accompanying illustration. Figure 48 represents a working diagram of the apparatus, which consists in the main, of a series of ten carefully calibrated standards a, b, c, &c., of lw, w, w, respectively, down to 1/512w.

F

These are connected at one end to a common 'bus bar A, and at the other to individual studs on a circular double contact switch B, which connects one end of them, respectively with a galvanometer terminal 1, and variable resistance switch C, the movable arm of which is connected to one of the main terminals X. S is an adjustable slider, working on the common 'bus bar D, which is connected to No. 2 galvanometer terminal. and H are hinged steel knives enclosing the space of one metre between their respective edges. L is a metre scale, over which the slider S indicates. Galvanometer terminals 3 and 4 are connected to the 'bus bar A, and the knife F respectively. G is the high resistance galvanometer which, by means of the double switch J, can be connected across terminals 1 and 2, or 3 and 4, at will.