If we observe the height of the barometer at the time of making each experiment, we shall find a very remarkable correspondence between it and the boiling temperature. Invariably, whenever the barometer stands at the same height, the boiling temperature will be the same. Thus, if the barometer stands at 30 inches, the boiling temperature will be 212°. If the barometer fall to 29 inches, the thermometer stands at a small fraction above 211°. If the barometer rise to 30 inches, the boiling temperature rises to nearly 213°. The variation in the boiling temperature is, then, accompanied by a variation in the pressure of the atmosphere indicated by the barometer; and it is constantly found that the boiling point will remain unchanged, so long as the atmospheric pressure remains unchanged, and that every increase in the one causes a corresponding increase in the other.

(58.) From these facts it must be inferred, that the pressure excited on the surface of the water has a tendency to resist its ebullition, and to make it necessary, before it can boil, that it should receive a higher temperature; and, on the contrary, that every diminution of pressure on the surface of the water will give an increased facility to the process of ebullition, or will cause that process to take place at a lower temperature. As these facts are of the utmost importance in the theory of heat, it may be useful to verify them by direct experiment.

If the variable pressure excited on the surface of the water by the atmosphere be the cause of the change in the boiling temperature, it must happen, that any change of pressure produced by artificial means on the surface of the water must likewise change the boiling point, according to the same law. Thus, if a pressure considerably greater than the atmospheric pressure be excited on a liquid, the boiling point may be expected to rise considerably above 212°; and, on the other hand, if the surface of the water be relieved from the pressure of the atmosphere, and be submitted to a considerably diminished pressure, the water would boil below 212°.

Let B (fig. 17.) be a strong spherical vessel of brass, supported on a stand s, under which is placed a large spirit lamp L, or other means of heating it. In the top of this vessel are three apertures, in two of which are screwed a

thermometer T, the bulb of which enters the hollow brass sphere, and a stop-cock c, which may be closed or opened at pleasure, to confine the steam, or allow

T

A

Fig. 17.

it to escape. In the third aperture at the top, is screwed a long barometer tube, open at both ends. The lower end of this tube extends nearly to the bottom of the spherical vessel B. In the bottom of this vessel is placed a quantity of mercury, the surface of which rises to some height above the lower end of the tube A. Over the mercury is poured a quantity of water, so as to half fill the vessel B. Matters being thus arranged, the screws are made tight, so as to confine the water, and the lamp is allowed to act on the vessel; the temperature of the water is raised, and steam is produced, which, being confined within the vessel, exerts its pressure on the surface of the water, and resists its ebullition. The pressure of the steam acting on the surface of the water is communicated to the surface of the mercury, and it forces a portion of the mercury into the tube A, which presently rises above the point where the tube is screwed into the top of the vessel B. As the action of the lamp continues, the thermometer T exhibits a gradually increasing temperature; while the column of mercury in A shows the force with which the steam presses on the surface of the water in B,- this column being balanced by the pressure of the steam. Thus, the temperature and pressure of the steam at the same moment may always be observed by inspecting the thermometer IT and the tube A. When the column in the tube a has risen to the height of 30 inches above the level of the mercury in the vessel B, then the pressure of the steam will be equivalent to double the pressure of the atmosphere, because, the tube A being open at the top, the atmosphere presses on the

S

surface of the mercury in it. The thermometer T will be observed gradually to rise until it attains the temperature of 212°; but it will not stop there, as it would do if immersed in water boiled in an open vessel. It will, on the other hand, continue to rise; and when the column of mercury in a has attained the height of 30 inches, the thermometer T will have risen to 251°,-being 39° above the ordinary boiling point.

During the whole of this process, the surface of the water being submitted to a constantly increasing pressure, its ebullition is prevented, and it continues to receive heat without boiling. That it is the increased pressure which resists its ebullition, and causes it to receive a temperature above 212°, may be easily shown. Let the stop-cock c be opened; immediately the steam in B, having a pressure considerably greater than that of the atmosphere, will rush out, and will continue to issue from c, until its pressure is balanced by the atmosphere. At the same time the column of mercury in a will be observed rapidly to fall, and to sink below the orifice by which it is inserted in the vessel B. The thermometer T will also fall until it attains the temperature of 212°. At that point, however, it will remain stationary; and the water will now be distinctly heard to be in a state of rapid ebullition. If the stop-cock c be once more closed, the thermometer will begin to rise, and the column of mercury ascending in a will be again visible.

If, instead of a stop-cock being at c, the aperture were made to communicate with a valve, like the safety-valve of a steam engine, loaded with a certain weight, — say at the rate of 15lbs. on the square inch,—then the thermometer T, and the mercury in the tube A, would not rise indefinitely as before. The thermometer would continue to rise till it attained the temperature of 251°; and the mercury in the tube a would rise to the height of 30 inches. At this limit the resistance of the valve would be balanced by the pressure of the steam; and as fast as the water would have a tendency to produce steam of a higher pressure, the valve would be raised and the steam suffered to escape; the thermometer T and the column of mercury in a remaining stationary during this process. the valve were loaded more heavily, the phenomena would be

A

If

the same, only that the mercury in T and A would become stationary at certain heights. But, on the other hand, if the valve were loaded at a less pressure than 15 lbs. on the square inch, then the mercury in the two tubes would become stationary at lower points.

(59.) These experiments show that every increase of pressure above the ordinary pressure of the atmosphere causes an increase in the temperature at which water boils. We shall now inquire whether a diminution of pressure will produce a corresponding effect on the boiling point.

This may be easily accomplished by the aid of an air pump. Let water at the temperature of 200° be placed in a glass vessel under the receiver of an air pump, and let the air be gradually withdrawn. After a few strokes of the pump, the water will boil; and if the mercurial gauge of the pump be observed, it will be found that its altitude will be about 23 inches. Thus the pressure to which the water is submitted has been reduced from the ordinary pressure of the atmosphere expressed by the column of 30 inches of mercury, to a diminished pressure expressed by 23 inches; and we find that the temperature at which the water boils has been lowered from 212° to 200°. Let the same experiment be repeated with water at the temperature of 180°, and it will be found that a further rarefaction of the air is necessary, but the water will at length boil. If the gauge of the pump be now observed, it will be found to stand at about fifteen inches, showing, that at the temperature of 180° water will boil under half the ordinary pressure of the atmosphere. These experiments may be varied and repeated; and it will be always found, that, as the pressure is diminished or increased, the temperature at which the water will boil will be also diminished or increased.

(60.) The same effects may be exhibited in a striking manner without an air pump, by producing a vacuum by the condensation of steam. Let a small quantity of water be placed in a thin glass flask, and let it be boiled by holding it over a spirit lamp. When the steam is observed to issue abundantly from the mouth of the flask, let it be quickly corked and removed from the lamp. The process of boiling will then cease, and the water will become quiescent; but if the flask be plunged

in a vessel of cold water, the water it contains will again pass into a state of violent ebullition, thus exhibiting the singular fact of water being boiled by cooling it. This effect is produced by the cold medium in which the flask is immersed, causing the steam above the surface of the water in it to be condensed, and therefore relieving the water from its pressure. The water, under these circumstances, boils at a lower temperature than when submitted to the pressure of the uncondensed vapour.

(61.) There is no limit to the temperature to which water may be raised, if it be submitted to a sufficient pressure to resist its tendency to take the vaporous form. If a strong metallic vessel be nearly filled with water, so as to prevent the liquid from escaping by any force which it can exert, the water thus inclosed may be heated to any temperature whatever without boiling; in fact, it may be made red-hot; and the temperature to which it may be raised will have no limit, except the strength of the vessel containing it, or the point at which the metal of which it is formed may begin to soften or to be fused.

(62.) The following table will show the temperature at which water will boil under different pressures of the atmosphere corresponding to the altitudes of the barometer between 26 and 31 inches.

From this table it appears, that, for every tenth of an inch which the barometric column varies between these limits, the boiling temperature changes by the fraction of a degree expressed by the decimal 176, or nearly by the vulgar fraction .

(63.) In the experiment already described, by which the la

I