1 wheel barometer. A small bell is connected with the interior of the holder, and its rise and fall is measured by a cord passing over a pulley which moves a pointer over a graduated circle. If the pressure increases as the holder rises, the weight G should be increased, and the contrary if it diminishes. The pressure to which the gas is subjected is varied by changing the weight F. Prove this, and determine the law. Do the same for different parts of the scale. To test the above work, fill the holder with air, and open H very slightly, or better, allow the air to escape through a minute aperture. The holder will now slowly descend, and by noting the time the index passes each tenth of a foot mark, a series of numbers is obtained whose first differences would be constant, if the apparatus was perfect. By varying the pressures, the orifices and the kind of gas in the holder, all the laws of the flow of gases may be verified. 57. GAS-METERS. Apparatus. Two gas-meters, one wet and the other dry, both graduated so as to read to thousandths of a foot. They are connected together so that the gas will pursue the following course. It leaves the pipe through a "valve, passes through a T to the wet meter, thence through a second T to the dry meter, and by a stopcock and third T to a fishtail burner. A short piece of pipe is screwed into the open end of each T, which may be closed by a cap, or connected with a gauge formed of a U-tube by a piece of rubber tubing. Experiment. Gas-meters are of two kinds, wet and dry. The former consists of a cylindrical vessel half full of water, in which is placed a rotary drum with four compartments. As these are filled in turn, the drum revolves, and the amount of gas consumed is measured by the number of revolutions. The dry meter resembles in principle a blacksmith's bellows reversed in such a way that air being forced into the nozzle, the handle moves up and down. The number of strokes is then recorded by clockwork and dials. The wet meter was first used, but is now superseded in houses by the dry meter, owing to the error introduced by any increase or diminution in the amount of water present. The former is, however, still in vogue for experiments, as by it small amounts of gas may be measured with much greater accuracy. To determine the amount of gas which has passed through the meter, subtract the reading at the beginning from that at the end of the experiment, or if the rate of flow is required, take readings at intervals of one minute, as in Experiment 5. Usually in the best meters, one revolution of the large hand equals one tenth of a cubic foot, and the dial being divided into a hundred parts gives thousandths. In meters intended to be used in houses, one revolution of the hands of the three lower dials equals 100,000, 10,000 and 1000 cubic feet, respectively, and a fourth dial is placed above, whose hand makes one revolution for every five cubic feet, and which is used in testing the metre. For this purpose, meters, turn it off The common method of testing a meter is to bring the upper hand to the zero, connect it with a gas-holder, and force air or gas through it until the reading is the same as before. The change of reading of the holder should now be just five feet, and the difference is the error of the meter. This experiment should be repeated two or three times. If the meter reads to thousandths of a foot an additional test is needed to see if the divisions of the large dial correspond to equal quantities of gas. allow the gas to flow very slowly through both and read them, dividing the thousandths into tenths by the eye. Allow a few thousandths to pass and read again, and so take a series of readings, until two complete revolutions have been made. Represent the results by a residual curve, in which abscissas represent the readings of the large hand of the wet meter, and ordinates the difference between the two enlarged, moving the origin down so as to bring the points on the paper. Two curves are thus obtained, one for each revolution, which should be coincident except for the accidental errors, and their deviation from a straight line shows the inequality of the thousandths, as given by the dry meter. The next thing to be determined is the loss of pressure due to each meter. Evidently a certain amount of power is necessary to overcome the friction, and this power is obtained at the expense of the pressure of the gas, which therefore leaves the meter with less pressure than it enters it. To measure this, connect the first and second T with the two arms of the gauge, and allow the gas to pass through the meter. The difference in level of the two tubes shows the loss due to the meter. See if this varies with different pressures and with different positions of the revolving drum. By using the second and third T, the dry meter may be tested in the same way. When a metre is placed between the outlet and the valve by which the gas is turned off, an error is introduced whenever the latter is opened or closed. This is due to the difference of pressure within and without the revolving drum, produced by gas flowing in or out without in some cases moving the hand. To show this, close both the valve and the stopcock near the third T. Open the valve, the gas will rush into both meters, moving the hands a small amount. Close the valve and open the stopcock, when the gas will rush out until the pressure within the meter equals that of the outer air. Take a number of readings, opening them thus alternately. Each meter is here affected by the error caused by the other, and by the intermediate pipe, to eliminate which the valve and stopcock should be placed close to the meter to be tested. The error may then be determined for different pressures and different positions of the hand. This error is not cumulative, and seldom exceeds one or two thousandths of a foot. To measure the amount of gas consumed by any burner under different pressures, connect it with one of the meters, and attach the gauge to the third T. Turn on the gas and take a five minute observation; that is, take six consecutive readings of the meter at intervals of one minute, also the pressure as given by the gauge. Do the same with several other pressures, and see if the flow is proportional to the square root of the pressure, or if the curve formed by the readings of the gauge and volumes of gas burnt is a parabola. A similar experiment may be performed with an aperture in a plate of platinum, and the height of the flame measured corresponding to different pressures and rates of consumption. A regulator of some form, such as will be described under the photometer, should be introduced to prevent accidental variations in the pressure, if great accuracy is expected in the last experiment. The laws of the efflux of gases may then be tested, or the uniform division of the meter, by allowing the gas to escape very slowly, and seeing if the volume is proportional to the time. 8 58. BAROMETER. Apparatus. Two barometer tubes, one already filled and placed in its cistern, some pure mercury, and a stand by which the height of the mercury column may be measured. This stand may be made in a variety of ways. Thus a half metre steel bar divided into millimetres is fastened to an upright, and a slider attached to it, so that it may be set at any desired height. This slider carries, first a steel point about 40 centimetres below, to determine the height of the mercury in the cistern; secondly, a vernier or a brass plate with a single line cut on it and resting against the steel scale, and finally two index plates of brass between which the tube is placed. The slider is raised or lowered until a thin line of light is just visible between the top of the mercury and the bottom of the index plates, and the reading then taken by the vernier. The tubes are held in position either by rings of brass, or by strips fastened by hinges. A steel rod three or four decimetres long and a tall jar of water, are also needed. Experiment. First, to find the distance from the steel point to the index plates. This may be done by the cathetometer, or by the second steel rod. Place the jar of water close to the upright, and bring the points of both rod and slider just in contact with the surface of the liquid. Read the vernier, lower the slider until the index plates are just on a level with the top of the steel rod, and read again. The difference added to the length of the rod equals the distance from the index plates to the lower point. The length of the steel rod is found by bringing the index plate first to its upper and then to its lower end. Now place the filled barometer in its proper position between the index plates, move the slider down until the point just touches Raise the its reflection in the mercury, and read the vernier. slider, until placing the eye on a level with the index plates the light is just cut off between them and the mercury. The difference between these readings is the height through which the slider has been raised, and this added to the distance from the plates to the point gives the height of the mercury column. Now measure the height of the standard barometer placed with the other meteorological instruments. Reduce it to millimetres (1 metre 39.37 inches), and the difference of the two is the error of the barometer first measured. It is probably due to a little air in the top of the tube. If the Now fill the empty tube in the following manner. mercury is not perfectly pure, it must be cleaned as described in Experiment 9. Hold the tube with the left hand in an inclined position, the closed end resting on the table. Pour in mercury slowly to within a few inches of the top. To prevent spilling, the stream should be guided by the forefinger and thumb of the left hand held at the opening of the tube. Next close this opening with the finger and raise the closed end so that the bubble of air shall move slowly along the tube. Make it pass from end to end, until all the small adhering air-bubbles are removed. Then fill it full of mercury, and closing the end again invert it, and immerse in the cistern, removing the finger under the surface of the mercury. The latter will now descend in the tube until its height is about 30 inches, leaving a vacuum at the top. Its pressure at the bottom of the tube is then just equal to that of the atmosphere, which by pressing on the outside mercury, supports the inner column. The vacuum at the top is known from its discoverer as the Torricellian vacuum, and is one of the most perfect that can be obtained artificially. To see if any air has entered, incline the tube, and notice if the mercury rises to the top, remaining at the same level throughout, and if when made to oscillate gently it strikes the top with a sharp click; if not, air has entered, and the experiment must be repeated. Next, put this tube in the place of that previously filled, measure its height and determine the error. Allow a bubble of air to enter one of the barometer tubes (the one in which the error is the greatest), and notice that it increases in volume as it rises until it reaches the top, when it causes a considerable depression of the mercury column. Repeat, until this has fallen eight or ten inches. Then measure the height with care, and suppose it to be an observation taken at the top of a mountain. Compute the height on this supposition by the method given below. The temperature of the air and mercury at the upper station may be assumed equal to that of the thermometer outside the window, and at the lower station to those obtained by direct observation. This work may well be supplemented by a determination of the altitude of a real mountain. For this purpose the pressure of the air must be measured at the top and bottom, either by an aneroid, or more accurately by a mercurial mountain barome |