NINE-TENTHS of the distrust in which the gas meter is almost universally held is directly traceable to its obscure location in the dark, unpleasant surroundings of the cellar. We feel sure that the man who is loudest in his denunciation of this “instrument of extortion” makes no daily or even weekly trips down his cellar stairs for the purpose of verifying the totals footed up by the gas meter. If he did we are confident that much of his distrust would give way to ' respect. This same man may be most scrupulous about all his other accounts and would think it the height of absurdity to jot down his expenditures but once a month, with no memoranda to assist him in recalling the various items. Yet with ridiculous inconsistency he is ready to denounce any enlargement of the gas bill, forgetting that a whist party and a few friends to dinner or a succession of cloudy afternoons may have increased his gas-light hours—hours that are by no means overlooked by the mechanical accountant of the gas company. However, we are not writing a defense of the gas meter. It is by no means a perfect instrument; hence the necessity of checking its records periodically. Unfortunately not one man in a thousand knows how to read a meter, yet there is no great mystery about it. Directions are published either on the back of the gas bill or on cards issued by the gas company, which may be. had for the' asking. The meter is usually provided with three or four dials, each graduated in ten divisions. A small dial marked “Two feet” may be ignored for the present, as it is used only for testing purposes. The dial to the right is marked “1 thousand” and each of th(> divisions stands for 100 cubic feet, the entire circuit of the dial representing one thousand feet. On the next dial, marked “10 thousand,” the hand moves m the opposite direction and each division stands for a thousand feet. In the same way each division of the dial marked “100 thousand” stands for ten thousand feet. Fig. 1 shows a simple example which will illustrate the method of reading. From this reading, the reading of the previous month is subtracted to show the amount of gas used in the interim. It should be noted that when a hand lies between two figures, the lesser number is read. It is sometimes puzzling to know whether a hand lies on a figure or not, and this can be determined only by referring back to the dial on the right. For instance, if all three hands pointed to 9, we might carelessly put down the reading 99,900, which would be just 11,000 cubic feet out of the way. The true reading is 88,900, as a moment's consideration will show. The fact is made clearer by referring to the hands of a clock. When the hour hand of the clock points to X, and the minute hand to XI, we read not 10.55, but 9.55, because the minute hand shows us that the hour hand has not yet reached X. And so with the gas meter hands; the hand of the first dial shows that the hand of the second dial has not quite reached 9, although it may appear to be covering that number, while the hand of the second dial shows that the hand of the third one has not yet reached 9. It happens not infrequently that a meter inspector will make a mistake in such a reading, which may boost the gas bill considerably. Of course. the fault will right itself eventually, for at the end of the next month, there will be a larger subtrahend to deduct from tlie meter reading, and a correspondingly smaller bill. Sometimes an inspector, knowing about how much gas is used, will guess at the reading, if he finds it impossible or inconvenient to reach the meter when on his rounds, making nis estimate low so as to be on the safe side: Then when he does make a bona fide reading, the gas bill will show a sudden rise, making up for the accumulated deficiencies of the inspector's low estimates. In 1907, the Public Service Commission of the First District of New York took over the power and duties of State Inspector of gas meters in Greater New York, since which time the citizens of this metropolis have been protected by all manner of safeguards against the possibility of fraud. All meters are carefully tested by the Commission and must bear its .seal before being placed in a building. This seal indicates that the meter, when tested, was correct or not more than 2 per cent slow. A slow meter is one that does not register the full amount of gas passing through it, while a fast meter makes the consumer pay for more gas than he uses. No meter is allowed to be placed on a consumer's premises that runs at all fast, but if on a complaint a meter is not more than 2 per cent fast it is passed as correct. For the absurdly small fee of 50 cents (this applies to the smaller meters; for the larger meters from 20 to 300 lights and upward the charge is from 75 cents to $2.00), any one in New York can have his meter tested, and if the meter is found to be running more than 2 per cent fast, not only is the fee refunded and charged to the gas company, but the latter must make a settlement with the consumer for the unconsumed gas he has been paying for. From July, 1907, to July, 1911, 19,970 gas meters were tested on complaint; 45.8 per cent of these were found to be correct or within 2 per cent fast or slow; 10.3 per cent were found to be slow or running to the disadvantage of the gas company, while 43.9 per cent were fast. The variation of the meters averaged a very small figure, however, showing that the work of the meter itself was on the whole very creditable. When a complaint is made to the Commission, a complaint inspector visits the consumer and makes a thorough inspection of his premises, searching for leaks. Occasionally a leak will be discovered near an open window or close to a chimney flue. where the draught will carry off the gas into the open, letting it escape undetected. Sometimes a jet is left burning in an attic or in some unused room, such as an amateur photographer dark room, and the tenant comes back after a month or two of vacation to find a gas bill awaiting him. Not infrequently in apartment houses the meter is found to be connected with the wrong apartment and one man is then paying the other man's bills. Troubles of this sort account for the running of the gas meter when no gas is used. As a matter of fact, the meter can register nothing if no gas passes through it. When satisfied that there is no trouble with the pipes, the meter inspector seals and tags the meter and has it removed to the testing station. The manner of testing the meter is simple. A measured quantity of air is passed through the meter, and the two-foot testing dial of the meter is noted, to see whether it registers this amount correctly. The prover in which the air or gas is contained is built just like a gasometer, consisting of a cylinder, with open bottom which telescopes within another cylinder filled with water to form a seal between the two. The inner cylinder is provided with a port which may be opened to admit air as it is raised. This port is than closed, sealing the air within as in a diving bell. The meter is now connected with the cylinder, and after care has been taken to make all the joints tight, the air is allowed to pass out through the meter. When two cubic feet of air have passed through, as indicated by the small dial, an indicator on the prover should show that two cubic feet of air have actually issued from the cylinder. If over two per cent plus or minus is indicated, the meter is rejected and must go to the repair shop. Usually, three readings are taken, and the average of the three determines the condition of the meter, this precaution being observed so as to do away with the personal equation, which might be a factor of serious importance if but a single reading were used. If we follow the meter to the repair shop, we shall have a chance to investigate its interior construction and learn something of the ailments to which it is subject. The drawing in Fig. 3 shows us how a meter would look if the front wall of the tin casing were torn off. Probably th'” first object to catcli ine eye would be the large circular bellows that occupies the lower part of the meter at A. There is another bellows of the same size and shape on the opposite side of the central vertical partition B. These be.llows or “diaphragms,” to use the technical name, consist of a ring of metal solder I'd to the partition B and connected bv a strip of flexible leather to a metal disk: The djaphragma are the lungs of tnometer which inhale and exhale measured quantities of gas, while a counting mechanism is actuated by the heaving lungs, .and indicates the cOunt on the dials. The counting mechanism is situated in the upper chamber or “gallery” formed by the horizontal partition C, and here also an, the slide valves D that admit the gas to the diaphragms. A somewhat diagrammatic section through one of these valves is shown in Fig. 4. Gas enters the meter through the tube E and flows into the valve chamber through the port l:'. The valve port G leads to one of the diaphragm compartments, while the port H leads to the interior of the diaphragm in that compartment. The outlet port is shown at 7, and it leads through pipe J to 'the burners in the house. The valve D is shaped to provide communication between the outlet port 7 and one or other of «the ports G and H at a II times. so that there is no interruption in the flow of gas out of the meter. With the valve in the position shown by full lines in FF:_ 4, the gas is slowly filling the diaphragm chamber and flattening the diaphragm, which in turn is exhaling the gas it contains by way of the ports H and I. When the valve moves to the position shown by dotted lines, gas enters the diaphragms through port H and, as the dianhragm expands, under the pressure of the mains, it causes a proportionate amount of gas to flow out of the chamber through ports G and 7 to the outlet J. As there are two sets of diaphragms and valves operating in alternation, there is no appreciable variation in the pressure of the outflowing gas. Each diaphragm is connected to a vertical rock shaft provided with a crank arm, and connected by a link L to a sleeve 11 adjustable on an arm N projecting from the valve shaft O. The combined motions of the rock shafts as they are oscillated by the inhaling and exhaling diaphragms, cause the va 1 ve shaft to rotate, and by means of connecting rods 7', to operate the valves. By means of a worm and gear R, the counting mechanism is operated. The dog Fi serves to prevent the mechanism from running backward, and it is the action of the arm N striking this dog that produces the clicking noise which may plainly 'w heard when gas is flowing through the meter. Adjustment of the meter is provided by sliding the sleeve 11 in or out upon the crank arm N. It is not very reassuring to the consumer to know that the.majority of troubles that affect the gas meter tend to make it run fast rather than slow. As the parts become loose with wear there is a tendency for them to run faster, and as the leather bellows are saturated with moisture condensed from the gas, they shrink and deliver less gas at each aspiration. Sometimes so much water collects in the meter case that it seriously hinders the opening and closing of the bellows, and they operate only when partially filled. Here again the consumer loses. There is another difficulty which should be considered, but which, however, can be overcome by a very simple device. Normally the pressure used in New York is “15 to 25 tenths.” to use the trade expression, which is equivalent to about one-half to one pound water pressure, U or siphon gage measurement. If, for instance, the pressure should be increased to, say, 40 tenths, 50 per cent more gas would be consumed at the burner and. although there would be increase in light, it would, not be sufficient to compensate for the increase in the gas bill. On the other hand, the meter would not register the full amount of gas passing. There would be a loss of about 2 per cent, which the company would have to stand for. To obviate this trouble, a governor is provided, which may be placed upon the service pipe before it enters the meter. This will hold the pressure down to the normal limit. There was a period, from 1907 to 1909, when the pressure in New York was confined by law between the limits of 15 and 25 tenths. This limitation was finally withdrawn on the representation of the gas companies. who claimed that owing to friction when conducting gas over a long range, it is necessary to use a high initial pressure in order to maintain sufficient pressure at the end of the line. Still another factor to be considered is >the elevation of i- burner. The pressure at the crest of a hill is more than that in the valley below; on the ground floor of a skyscraper it would be much less than at the top. There is a difference of one-tenth pressure for every 14 feet of elevation, so that at the top of a 700-foot structure like the Metropolitan tower, twice as much gas would flow through a five-foot burner as at the street level. However, it is only in naked gas flames that there is a loss due to high pressure. Incandescent man tie lamps are self-governing and will consume only a fixed quantity of gas, while if the gas is used for cooking purposes there is no loss; for a rise, in Testing a complaint meter. The Meter pressure will result in a corresponding rise in temperature,