long. It also consumes considerable energy, the current at full voltage being about one-third of an ampere. Even at 110 volts this involves a loss which is by no means insignificant if it continues twenty-four hours a day throughout the year, particularly if several of these instruments are used in a plant. Sometimes they are provided with resistance coils in series, to multiply their range, and are employed to measure high potentials of 1,000 volts or

more, in which case the loss of energy is fully 300 watts, and becomes quite a large item.

Electrostatic Voltmeters. In these instruments the mutual electrostatic attraction of two oppositely charged bodies is utilized to measure the difference of potential existing between them. The principle is identical with that of the electrometers used in laboratory work. One of the simplest forms is Kelvin's (Thomson's) vertical electrostatic voltmeter, consisting of two fixed metal

lic plates, B, between which is pivoted a movable metallic plate, A, all three being parallel and vertical. The pair of fixed plates and the movable plate are respectively connected to the conductors between which the potential is to be measured. This causes

the plates to become oppositely charged; and the movable plate is deflected in its own plane against the force of a weight attached to its lower part, and the position of the pointer C is read on the scale E. The attraction is proportional to the square of the potential difference; hence electrostatic instruments have a disadvantage similar to that explained in the case of electrodynameters. In fact, it is very difficult to obtain a practical device which will measure pressures of less than about 50 volts. But in electric lighting the potentials used are very rarely less than 100 volts, and are often as high as 1,000 volts or more; consequently these instruments are well adapted to the purpose.

The advantages of electrostatic voltmeters are : —

1. They are applicable to either alternating- or direct-currents. 2. They are extremely simple in construction and action.

3. They are cheap.

4. Their accuracy does not depend upon any condition which is likely to change.

5. They consume no energy, since no current flows through them.

6. They are absolutely free from magnetic disturbance.

7. They can readily be made for measuring potentials up to 25,000 volts, or even more.

Up

The last advantage is so decided that this type is by far the best for very high potentials, and it possesses the six other advantages enumerated, even when employed for lower voltages. to the present time it has not been developed to the extent that its merits would seem to warrant, but it is now being perfected and applied more generally.

Electro-chemical Instruments are not commonly used for practical measurements, almost the only example being the Edison electrolytic meter, which will be described in Volume II., since it does not ordinarily form part of the generating-plant.

Special Alternating-Current Instruments. Forms of ampereand voltmeter have been devised in which effects peculiar to

the alternating current are utilized. For example, the action of repulsion, discovered by Professor Elihu Thomson, may be applied to purposes of measurement. These have no great advantage over the ordinary electromagnetic and electrodynamic effects, and are not only limited to alternating currents, but also in most cases depend upon the frequency.

Wattmeters. If the Siemens electro-dynamometer described on page 418 is modified by winding one of the coils with a great many turns of fine wire, and connecting it to the circuit so that the current received by it shall be proportional to the voltage, the other coil being as before in the main circuit, and having the total current passing through it, the result is that the mutual action of the two coils depends upon both the pressure (volts) and current (amperes) of the circuit. In other words, the force exerted between them will be directly proportional to the product of the voltage and current, which is the number of watts.

For direct-current purposes a wattmeter is not very important; since all that is necessary is to read the potential from the voltmeter and the current in the amperemeter, and multiply the two numbers together to obtain the power in watts. But in alternating-current measurements the case is different; since there is a lag of the current wave with respect to the E.M.F. wave, if the circuit contains self-induction, as it almost always does. On the other hand, the effect of capacity is to produce a lead of the current; but in either case the "apparent watts" obtained by multiplying the volts by the amperes, as indicated in different instruments, is not the real energy. This must be multiplied by the power factor, which is the cosine of the angle of lag, in order to find the "true watts." It is difficult to determine this power factor; hence it is far more convenient to employ a wattmeter, which eliminates the error due to lag or lead so that the actual power may be read directly.

Recording Volt-, Ampere-, and Wattmeters. These are employed in electrical generating-plants to make a continuous record of the output in volts, amperes, or watts. For example, a pen or other marking-device, is attached to the indicating part of a voltmeter, and a tape or circle of paper is slowly moved by clockwork so that a line is traced upon it showing the voltage at any time. These instruments require more positive action, and therefore

more energy, than the ordinary visual meters that merely move a pointer.

Ampere- and wattmeters are also made which integrate or sum up the total number of ampere-hours or watt-hours during a given period. These are chiefly used to measure the current supplied to each consumer, and they will be described under the head of Recording Meters in Volume II.

Regulating and fault-detecting apparatus, including rheostats, choke-coils, ground-detectors, and other similar devices and methods of using the same, which naturally belong to the subject of Electrical Distribution, will be treated under that head in Volume II.

The following are some of the most useful books on Electrical Measurements:

GRAY, A., Absolute Measurements in Electricity and Magnetism, 2 vols., London, 1888 and 1893.

KEMPE, H. R., Handbook of Electrical Testing, Fifth Edition, London, 1892. KEMPE, H. R., Electrical Engineer's Pocket Book, London, 1890.

MUNRO, J., and Jamieson, A., Pocket Book of Electrical Rules and Tables, London, 1896.

PRICE, W. A., The Measurement of Electrical Resistance, Oxford, 1894. WEBB, H. L., Testing of Insulated Wires and Cables, New York, 1891.

CHAPTER XXIV.

LIGHTNING-ARRESTERS.

THE various devices employed to protect electrical apparatus from lightning or atmospheric electricity are commonly termed lightning-arresters. This name, however, is not very appropriate; since they do not in any sense stop the discharge, but merely act to divert it, and convey it harmlessly to the ground.

Lightning-arresters are required wherever an electrical circuit, or any portion of it, extends out-doors; but they are obviously unnecessary if the conductors are wholly in-doors, underground, or submarime. The liability of an aerial wire to receive discharges of atmospheric electricity depends upon its length, its height above the ground, and its location, that is, whether it runs. over hills or mountains; and certain regions are far more subject to this trouble than others. In England, for instance, its occurrence is comparatively rare and insignificant, whereas in the Rocky Mountains it is a matter of serious and almost constant difficulty.

The troubles which are likely to be caused in an electrical plant by atmospheric electricity are:

1. Puncturing or charring of the insulation of the electrical machines, instruments, or conductors.

2. Melting off of wires, or fusing together of contact points. 3. Danger to persons.

4. Liability of starting a fire.

The first or second of these accidents will often almost ruin an electrical machine or instrument, and involve considerable time and expense in repairing. The last two, although of less frequent occurrence, may be even more serious in their results.

In many cases where a circuit is partly overhead and partly underground or under water, there is danger that the former portions will receive atmospheric discharges, which, running down