when we consider that our dynamo may send a destructive arc to follow up the lightning, it is apparent that arresters should always be separated and as far as possible removed from combustible material, and should be placed where they can be readily inspected and repaired when injured by a discharge, as often happens. It is the judgment of experts who have most thoroughly investigated lightning discharges, as well as of engineers who have observed the effects of lightning, that it is better to steer the lightning to earth outside of the station than to invite it inside and then attempt to direct its movements. Whether we have lightning arresters on our switch board or not, we should have them on our overhead line, and if we can get arresters that will work all right on a pole, it would seem that a pole just outside the station was a better place than a switch board for the arresters to protect our machines and station.

Rule "c" requires lightning arresters to be connected with ground by two separate wires, and this is desirable as it is often very difficult to get a good earth contact (when you want one). A good connection to a grounded. pipe in a building might invite unknown trouble, and a poor connection would be the means of leading the lightning into the building without giving it an adequate means of escape. In case the pipe should become disconnected from the earth, its use as an earth connection would simply lead the lightning into the building and turn it loose upon a network of pipes, leaving it to make its escape at its own convenience. This kind of thing has been done very often, and if the results have ·

not always been disastrous, it has only been due to the element of chance, which seems to play such a large part in the movements of lightning. Rule "d" is called forth by the fact, as above stated, that a dynamo will maintain an arc which lightning has started. Such an arc would, of course, destroy the lightning arrester even if it did no other damage. There are many devices for either breaking the arc thus formed or for preventing an arc being formed, and these devices constitute the only essential points of difference between lightning arresters. The type of lightning arrester which is best adapted to any given plant depends upon the pressure and current of the dynamos and circuits and the kind of work to which the current is to be applied. We need only state here that there are upon the market satisfactory devices adapted to all conditions of practice.

CHAPTER IV.

CENTRAL STATIONS FOR LIGHT AND POWER.

PART III.

TEXT OF CODE COVERED BY THIS CHAPTER. 7. TESTING:a. All series and alternating circuits must be tested every two hours while in operation, to discover any leakage to earth, abnormal in view of the potential and method of operation. b. All multiple arc low potential systems (300 volts or less) must be provided with an indicating or detecting device, readily attachable, to afford easy means of testing where the station operates continuously. c. Data obtained from all tests must be preserved for examination by insurance inspectors.

These rules on testing to be applied at such places as may be designated by the association having jurisdiction.

8. MOTORS:-a. Must be wired under the same precautions as with a current of the same volume and potential for lighting. The motor and resistance-box must be protected by a double pole cut-out and controlled by a double pole switch, except in cases where one-quarter horse-power or less is used on low tension circuit a single pole switch will be accepted. b. Must be thoroughly insulated, mounted on filled dry wood, be raised at least eight inches above the surrounding floor, be provided with pans to prevent oil from soaking into the floor, and must be kept clean. c. Must be covered with a waterproof cover when not in use, and, if deemed necessary by the inspector, be inclosed in an approved

case.

9. RESISTANCE BOXES:-a. Must be equipped with metal or other non-combustible frames. b. Must be placed on the switchboard, or at a distance of a foot from combustible material, or separated therefrom by a non-inflammable, non-absorptive, insulating material.

Testing. We have thus far treated of methods of securing good insulation. Proper maintenance is just as important as proper installation. It is always difficult to maintain the insulation upon the circuits of a Central Station, especially where there are many and long circuits. We must always remember that a ground upon a wire of one polarity puts a strain upon the insulation of all the wires of opposite polarity. If a second ground comes upon a wire in a building, it may cause serious trouble, even if the first ground is out in the street. Rule "a" is not very specific, as it simply states that the circuit shall be "tested." This might mean a rough test, which would only indicate the presence of trouble. The insulation of the circuits should be measured frequently, and these measurements recorded. These measurements can be made while the circuits are in operation by using a suitable volt meter, or an instrument specially designed for that purpose.

Series Circuit.-If we attach one end of a wire to one brush or pole of a dynamo and the other end to the other pole, it will form what we call a circuit. If the dynamo is in motion, a current will flow through the wire. The strength of current will depend upon the electrical pressure or E. M. F. between the two poles of the dynamo, and upon the resistance of the wire (this resistance being determined by the size and length of the wire and the material of which it is made). If we cut this wire at any point, and bridge over the gap thus formed by inserting therein a lamp, our current will now flow through the lamp. If we cut the wire in

a second place and insert there a second lamp, the current will flow through both lamps. We describe such an arrangement by saying that the lamps are connected in "series."

By making more cuts and inserting a lamp in each gap, we can connect up any number of lamps in series, and as the current always flows in a closed circuit, the same current will flow through each and all the lamps. This method of connection is commonly used for arc lamps. As each lamp offers some resistance to the flow of the circuit, we will increase the resistance of our total circuit or path every time that we add a lamp. The increased resistance will decrease our current unless at the same time we increase the pressure of our dynamo. For example, the current required to properly operate a two thousand candle power arc lamp, such as is commonly used in street lighting, is 10 amperes. It requires a pressure of 50 volts to send this current through one lamp.

If now

If

we insert a second lamp in series with the first, it will require an additional 50 volts, or a total pressure of 100 volts to send 10 amperes through both lamps. we insert 10 lamps in the series, it will require 50 volts for each lamp, or a total of 500 volts to push 10 amperes through the entire series. It naturally follows, therefore, that if we wish to run a great many lamps in one series, we must have a very high pressure. The code regards anything over 300 volts as a high pressure, or "high potential." As 300 volts will only operate a series of 6 arc lamps, nearly all arc lighting circuits come under the head of "high potential." It is common practice to install arc circuits of 50 lamps requir