Fig. 4 is a photograph of the common dry meter, with the front, back, top and left side removed. It is called a ‘dry’ meter to distinguish it from those meters, having little vogue in this country, which employ a liquid in place of a valve motion. The apparatus shown consists of a case divided into three compartments by a horizontal partition one fourth of the way down from the top, and by a vertical partition centrally placed and extending upward from the bottom of the casing to the horizontal partition. The upper compartment contains the registering mechanism and a small valve chamber, the latter corresponding to the steam chest of an engine. In each of the lower compartments is a metal disk attached to the central partition by well-oiled flexible leathers, each disk, leather and the partition forming a bellows. As in a locomotive, the meter really consists of two separate mechanisms, set to operate out of phase and avoid dead centers. [Illustration: FIG. 5.] [Illustration: FIG. 6.] Considering one mechanism only, recourse may be had to a diagrammatic representation of the action (Fig. 5). Gas entering the inlet passes into the valve chamber. Here an ordinary D-slide-valve closes two of the openings, leaving a third through which the gas may flow into the bellows or inner compartment. The bellows expands, gradually filling the outer compartment, and forcing the gas out under the valve into the outlet pipe, as indicated by the arrows. When the bellows is fully distended the valve shifts into the position shown in Fig. 6, admitting the inflowing gas to the outer compartment and collapsing the bellows, whose contents are forced into the outlet pipe by the paths traced by the arrows. Thus, it will be observed, the meter is a volume measurer pure and simple, measuring cubic feet with as much deliberation as is required to deal water out of a cask by means of a pint dipper. Its percentage of error is the same at all pressures and under all loads within its capacity, and it measures cubic feet of gas regardless of whether that gas be expanded or compressed. And so we are obliged to realize, as another fallacy is exposed, that the meter does not spin around most energetically under the higher pressures, cheerfully and accommodatingly serving its masters by adding a mythical cubic foot or two to the count at each revolution.