Fig. 95.
The operation of the air-pump can be made clear by the plan in Fig. 95. But one pump-barrel, a, is represented, with a piston, c, working in it. In the piston there is a valve, i, opening upward, and also one at b, in the beginning of the passage leading to the centre of the plate where is the receiver, d. The working of the instrument is thus: If the piston be forced down, the air under it, being compressed, will close the valve at b, and will rush upward through the valve i in the piston. Let the piston now be raised; the resistance of the air above it will close the valve i, while the valve b will be opened by the air rushing from the receiver, d, through the passage, e, to fill the space between the piston and b. You see, then, that every time that the piston is drawn up air passes out of the receiver through the valve b into the space between this valve and the piston. None of this air which has passed out can go back again, for the moment that you press upon it by forcing downward the piston the valve b is shut down, and the air escapes from the pressure by passing out through the valve i. Each time, therefore, that you work the piston up and down you pump out some of the air from the receiver; and if you pump for some time there will be exceedingly little air left in it, and that will of course be diffused throughout the receiver. It will be thin, like that in the upper regions of the atmosphere.
Fig. 96. Fig. 97.
Fig. 98.
156. Experiments.—When the receiver is full of air it can be moved about on the plate easily, and can be lifted from it. But work the pumps a few strokes and you will find that the receiver is firmly fastened to the plate, for the air within, being made thin, presses with little force compared with the air outside. If the pumps be worked for some time no force could release the receiver from the pressure without breaking it. But loosen the screw, g, and thus let the air in, and the equality of the pressure on the outside and inside is at once restored. Take off now this large receiver, and place a small glass jar, open at both ends, on the plate, with the hand covering the upper opening, as represented in Fig. 96. On exhausting the air the hand is so firmly pressed into the glass that it requires considerable force to disengage it from the pressure. If we tie a piece of bladder or India rubber over this jar, as in Fig. 97, and then pump out the air, the bladder at first is pressed in as represented, and if we pump on it at length bursts with a loud report. It would make no difference in the result of the experiment if the jar were shaped as in Fig. 98, for the pressure is the same in all directions. The resemblance between air and liquids in this respect may be illustrated thus: Suppose that a flat fish covers with one of its sides the end of the tube of a pump. He feels no uncomfortable pressure, because the water in the pump and that below it press equally upon him. If, now, the pressure of the water in the pump could be suddenly taken off by the piston, the fish would be pressed upward into the tube, as the bladder is pressed upward in Fig. 98, or downward in Fig. 97, or as the hand is pressed downward in Fig. 96.
Fig. 99.
