A direct demonstration of this may be given by the following experiment: - On the mouth of a flask let a stopcock be fastened so as to be air-tight. The interior of the flask may then be put into free communication with the external air or that communication may be cut off at pleasure, by opening or closing the stop-cock. If a syringe be applied to the mouth of the flask, the stop-cock being open, a part of the air contained in it may be drawn out. After this, the stop-cock being closed, and the syringe detached, let the flask be placed in the dish of a good balance, and accurately counterpoised by weights in the other dish. This counterpoise will then represent the weight of the flask, and of the air which has remained in it. If the stop-cock be now opened, air will immediately rush in, and replace that which the syringe had withdrawn from the flask; and immediately the dish of the balance containing the flask will sink by the effect of the weight of the air thus admitted into the flask.

If the weight of quantity of air so small as to be capable of being withdrawn by a syringe from an ordinary flask be thus of sensible amount, it may be easily imagined that the vast mass of atmosphere extending from the surface of the earth upwards, to a height not ascertained with precision, but certainly not being less than thirty miles, must be very considerable. Such a force, pressing as it must constantly do, upon the surfaces of all bodies, whether solid or fluid, and resisting and modifying their movements, would play an important part in all mechanical phenomena; and it is, therefore, not sufficient merely to have recognised its existence, but it is most needful to measure its amount with that degree of certainty and precision, which will enable us to estimate its effects on those phenomena which we shall have to investigate.

(21.) The amount of the pressure of the atmosphere on each square inch of horizontal surface on which it rests, is obviously the weight of the column of air extending from that square inch of surface upwards to the top of the atmosphere. This force is measured by the following means:

Take a glass tube, ▲ B (fig. 8.), above 32 inches long, open at one end A, and closed at the other end в, and let it

Fig. 8.

Fig. 9

be filled with mercury (quicksilver). Let a glass vessel or cistern c, containing a quantity of mercury, be also provided. Applying the finger at A, so as to prevent the mercury in the tube from falling out, let the tube be inverted, and the end, stopped by the finger, plunged into the mercury in c. When the end of the tube is below the surface of the mercury in c (fig. 9.), let the finger be removed. It will be found that the mercury in the tube will not, as might be expected, fall to the level of the mercury in the cistern c, which it would do were the end в open, so as to admit the air into the upper part of the tube. On the other hand, the level D of the mercury in the tube will be nearly 30 inches above the level c of the mercury in the cistern.

B

The cause of this effect is, that the weight of the atmosphere rests on the surface c of the mercury in the cistern, and tends thereby to press it up, or rather to resist its fall in the tube; and as the fall is not assisted by the weight of the atmosphere on the surface D (since B is closed), it follows, that as much mercury remains suspended in the tube above the level c, as the weight of the atmosphere is able to support.

If the section of the tube were equal to the magnitude of a square inch, the weight of the column of mercury in the tube above the level c would be exactly equal to the weight of the atmosphere on each square inch of the surface c.

(22.) If such an apparatus be observed from time to time, it will be found that the column of mercury sustained in the tube will be subject to variation between certain limits, never falling below twenty-eight inches, and never rising above thirty-one inches. This variation of the mercurial column is produced by a corresponding variation in the weight of the atmosphere.

If the apparatus be transported to any height above its ordinary position, it will have a less quantity of atmosphere above it, and therefore the surface of the mercury in the cistern will be pressed by a less weight, and consequently the

column of mercury will fall proportionally. In virtue of this effect, such an instrument has been rendered a means of measuring heights, such as the heights of mountains, the ascents of balloons, &c. &c.

(23.) If a proper scale be attached to the tube containing the mercurial column, showing the absolute height of the column sustained at any time, and indicating its changes of height, the instrument becomes a BAROMETER.

Two cubic inches of mercury weigh very nearly one pound avoirdupois. Hence, when the barometric column measures thirty inches, the weight of the atmosphere resting on each square inch of surface is about fifteen pounds.

(24.) It is an established property of fluids, that they press equally in all directions; and air, like every other fluid, participates in this quality. Hence, it follows, that when the downward pressure or weight of the atmosphere is fifteen pounds on the square inch, the lateral, upward, and oblique pressures are of the same amount. But, independently of the general principle, it may be satisfactory to give experimental proof of this.

Let four glass tubes, A, B, C, D (fig. 10.), be constructed of sufficient length, closed at one end, A, B, C, D, and open at the other. Let the open ends of three of them be bent, as re

* Exactly 15-68 oz. = 0.98 lb.

presented in the tubes B, C, D. Being previously filled with mercury, let them all be gently inverted, so as to have their closed ends up, as here represented. It will be found that the mercury will be sustained in all, and that the difference of the levels in all will be the same. Thus, the mercury is sustained in a by the upward pressure of the atmosphere; in B, by its horizontal or lateral pressure; in c, by its downward pressure;

+ This experiment with the tube a requires to be very carefully executed, and the tube should be one of small bore.

and in D, by its oblique pressure: and, as the difference of the levels is the same in all, these pressures are exactly equal.

(25.) The same arrangement by which the pressure of the atmosphere is measured by a mercurial column of equivalent weight, also supplies the means of measuring the pressure or elasticity of atmospheric air, or any other gas or vapour, whether in a more or less compressed or rarefied state; and as instruments constructed on this principlę are of considerable use in steam engines, we shall take this occasion to explain their principle and application.

In the experiments described in (21), the space D B in the top of the barometer-tube, from which the mercury descended, is a vacuum. If, however, it were occupied by a quantity of air in a rarefied state, or any other gas or vapour, such gas or vapour would press on the surface of the mercury at D, with a force determined by its elasticity. In that case, the atmospheric pressure acting on the surface of the mercury c in the cistern, would be balanced by the combined forces of the weight of the mercurial column sustained in the tube, and the elasticity of the gas or vapour in the upper part of it. Now if we know the actual amount of the atmospheric pressure, that is to say, the height of the column of mercury which it would be capable of sustaining,-we should then be able to determine the pressure of the rarefied air in the space C D.

For example, let us suppose that the barometric column, when BD (fig. 9.) is a vacuum, measures thirty inches: the atmospheric pressure, therefore, would be equal to the weight of a column of mercury of that height. Let us suppose that the elasticity of the gas or vapour occupying the upper part of the tube D B causes the column to fall to the height of twenty-six inches: it is evident, then, that the pressure of the air in the top of the tube would be equal to the weight of a column of mercury of four inches. In fine, to determine the pressure of the rarefied gas or vapour in the top of the tube, it is only necessary to observe the difference between the height of the column of mercury actually sustained in the tube, and the column sustained at the same time and

place in a common barometer: the difference of the two will be the column of mercury whose weight will represent the pressure of the vapour or gas in the top of the tube.

(26.) Whenever the air contained in any vessel or other enclosed space has by any means had its pressure reduced so as to be rendered less than that of the external air, the external air will have a tendency to rush into such vessel or enclosed space with a force proportionate to the excess of the pressure of such external air over that of the air within; and if any communication be opened between the interior of such vessel or enclosed space, and the external air, the latter will rush in until an equilibrium be established between the pressures within and without. It is evident that the force thus obtained by diminishing the pressure of air within a vessel may be applied to any mechanical purpose.

It is by such means that water is raised in an ordinary pump. A portion of the air contained between the piston of the pump and the surface of the water below, is withdrawn by the action of the piston, and the pressure of the air remaining under the piston is thereby diminished. The superior pressure of the atmosphere upon the external surface of the water in the well then forces up a column of water in the pump-barrel, and this is continued as the air is more and more rarefied by the action of the piston. By whatever means, therefore, the air can be wholly or partially withdrawn from any space, a mechanical power will be thereby developed, proportional in its amount and efficacy to the quantity of air so withdrawn. If, however, such air be withdrawn by any mechanical process, such as by a syringe, by a common pump, or by an air-pump, the quantity of force expended in withdrawing it is always equivalent to the amount of mechanical power obtained by the vacuum or partial vacuum so produced. Indeed the power expended is greater than the power so obtained, inasmuch as the friction, leakage, &c. of the exhausting apparatus must be allowed for.

(27.) There are, however, various other means by which air may be partially expelled from a vessel besides the direct application of mechanical force. Thus if heat be applied to