The unit of measurement of gas pressure is that pressure which will cause a difference of water level of one tenth of an inch in the two legs of a V-shaped tube when one end is connected with the gas main and the other end is left open to the outer air. Ten tenths, or one inch, is the standard, or normal pressure. [Illustration: EXHIBIT 1.] Any appliance--even a gas-burner--operates to best advantage under certain well-defined conditions. Depart from these conditions, and the efficiency of the device is impaired to an extent depending largely upon the nature of the appliance under consideration. For example, burn an incandescent lamp at fifty per cent. above normal voltage and it breaks down; burn a gas jet at two hundred per cent. above normal pressure, and it still operates--how satisfactorily ‘deponent sayeth not.’ Now, the gas-burner is supposed to operate to best advantage at ten tenths of an inch. At this pressure the flame is neither so wavering as to be affected by every chance draught, nor so rigid as to permit the gas to blow through without being properly consumed. Below the normal the flame decreases; above, the light is increased somewhat, but not by any means in proportion to the increase in the gas flow. Thus we see that the satisfactory employment of gas as an illuminant depends upon the maintenance of a pressure high enough to deliver the required amount of gas, but not so high as to cause wasteful consumption. [Illustration: EXHIBIT 2.] Turning back now to the gas main, let us consider the pressures actually existing. Exhibit 1 is a photograph of a twenty-four-hour record of pressure at a point not far from the works. The radial lines represent time, and there is a line for each quarter of an hour. The circles represent pressure, there being one circle for each tenth of an inch. Starting at _E_, the point at which the record begins, and following the irregular line clockwise, one may readily determine the fluctuations of pressure and the time of their occurrence. Interpreting the diagram, we find that the pressure was slightly above the normal until 4.30 P. M. (_A_), when the works began to raise the pressure little by little, in order to compensate for the increased loss due to increased flow through the mains. At 6.15 P. M. (_B_), the works ceased increasing the pressure. While this increase lasted--from 6.15 P. M. (_B_) to 10.15 P. M. (_C_)--our friend near the works suffered under twenty-one tenths pressure, the gas blowing merrily through the tips and the meter conscientiously registering gas wasted as well as gas utilized. From 10.15 P. M. (_C_) the pressure falls by steps during the ensuing two hours, finally reaching eleven tenths just after midnight (_D_), which latter pressure is quite steadily maintained until the following forenoon. The service from bedtime to dinner time should have proved quite satisfactory. One would naturally expect to find this consumer complaining of high bills, however.