They have been investigated, from a theoretical point of view, with considerable success, by Mr. Rankine of Glasgow. The experimental inquiry I have undertaken in conjunction with my friend Mr. Tate, and a part of the results, comprising experiments up to a pressure of 60 lbs. per square inch, will appear in the Transactions of the Royal Society. We are now preparing to enter on the more arduous and dangerous task of ascertaining the density, volume, &c., at much higher pressures. The accumulation of facts on this subject, bearing directly upon the application of steam, cannot be otherwise than acceptable to the general reader, and I shall, therefore, without further preface, insert such an abstract as bears directly on the subject under consideration.

General Laws of Vaporisation.

When a liquid is heated in any vessel, its temperature progressively rises up to a certain point, at which it becomes perfectly stationary. At that point the heat continuously absorbed becomes latent, or is no longer registered by the thermometer; ebullition commences, and vapour, of a bulk enormously greater than that of the liquid from which it is formed, rises in bubbles and fills the vessel. In this condition the temperature of the liquid is perfectly constant; no urging of the fire will cause it to rise; the heat, absorbed continuously, expands itself in effecting that change in the state of aggregation of the liquid which we know as vaporisation.

This remarkable constancy in the temperature of liquids undergoing vaporisation in open vessels has long been known and applied to the graduation of thermometers. The point at which a liquid boils in an open vessel is called its boiling point. The following table gives the boiling points of some of the more important liquids :

Water

We have said that the boiling point of a liquid is constant when in an open vessel, that is, when subject to atmospheric

pressure. If we change the pressure the temperature of ebullition changes also. Thus, if we place a vessel of hot but not boiling water under the receiver of an air-pump, and rapidly exhaust the air, the liquid will after some time begin to boil, and we may notice that the lower its temperature the more perfect must we make the vacuum before ebullition commences. Or again, if water be subject to pressure greater than that of the atmosphere, its temperature must be raised higher than 212° before it will boil. Experiment, therefore, shows that the boiling point, constant at the same pressure, varies at different pressures, rising higher as the pressure increases, and vice versâ.

Strictly speaking, the pressure of the atmosphere is not always the same; it varies within narrow limits from day to day; it decreases as we ascend higher into it, and hence there will be a small but corresponding variation in the boiling point at different times and places. This last fact has afforded the means of measuring the altitude of mountains, by determining the difference of the boiling point at their base and their summit. Measuring the atmospheric pressure by the column it supports in the barometer, we may draw up the following table of the relation of the boiling point to the height of the barometer column and the altitude of the observer, assuming that the barometer stands at 29.922 inches, and water boils at 212° Fahr. at the level of the sea.

TABLE I.-EXHIBITING THE INFLUENCE OF CHANGES OF ATMOSPHERIC PRESSURE ON THE BOILING POINT OF WATER, AND THE BOILING POINT AT DIFFERENT ALTITUDES.

But, besides pressure, certain other circumstances exercise a slight but sensible influence on the boiling point. In a glass

vessel the boiling point of water is about 2° higher than in a metal one, owing apparently to some adhesion between the glass and the liquid. Dr. Miller states that, if the glass be varnished with shellac, the temperature of the water may be raised to 221° in the open air, when a sudden burst of steam will take place, during which the temperature falls to 212°. From a similar cause the presence of salts in solutions raises the boiling point in some cases considerably. A saturated solution of common salt boils at 227° Fahr., and a saturated solution of chloride of calcium, which has an enormous affinity for water, does not boil at a less temperature than 355° Fahr.

There is yet one other remarkable condition of evaporation which should be noticed here. If water be dropped upon a clean metallic surface heated sufficiently high, instead of entering into ebullition it assumes a globular form, and rolls about very slowly and quietly evaporates. This condition, known as the spheroidal state, has been investigated by M. Boutigny. He finds that the temperature of the liquid globule never rises so high as its boiling point, being indeed usually 5° to 10° below it; that the temperature of the plate necessary to cause the spheroidal state varies with different liquids, and depends in part on the conducting power of the plate; and he considers the temperature of the spheroid to be constant, being for water 205°-7, for alcohol 167°.9, and for ether 93°6.

If, whilst the spheroid is rolling upon the metal plate, the temperature of the plate is allowed to fall below a certain tem.. perature (340° for water), the spheroid breaks, and is suddenly dispersed in vapour.

The temperature of the vapour rising from a liquid is necessarily identical with that of the liquid from which it rises, except in those cases in which the boiling point has been affected by adhesion, when the vapour at once adjusts itself to the normal temperature at that pressure.

So long as vapour is in contact with the liquid from which it has been formed, its temperature continues the same as that of the liquid, for if it be heated it takes up fresh liquid, and the temperature falls from the absorption of the heat rendered latent, until the normal temperature of the boiling point is regained.

The Vaporisation of Water and the Formation of Steam.

The temperature at which water boils is therefore constant at each pressure, and in consequence the temperature of the steam itself, when in contact with water, is constant at each pressure. The relation between the temperature and pressure of steam has been ascertained by experiment.

When in contact with the water producing it, steam is at the maximum density consistent with that temperature and pressure, and is then called saturated steam, or vaporous steam, and its temperature is called the maximum temperature of saturation at the given pressure. Usually when the pressure of steam is spoken of, the pressure of saturated steam is intended.

When isolated from the water producing it and heated, the steam expands, and decreases in density if the pressure be constant, or if the volume be constant it increases in pressure; it is then called variously anhydrous, gaseous, or superheated steam. The rate of expansion of superheated steam must be determined by experiment.

By the density of steam we mean the relative weight of a unit of volume. The specific volume of the steam is the reciprocal of the density, or the ratio of the volume of the steam to that of the volume of water which produced it. The density of saturated steam is constant at each temperature, and must be determined by experiment.

The latent heat of evaporation of steam is the quantity of heat which disappears in effecting the conversion of the water into vapour, or which reappears in the condensation of the steam. The latent heat of evaporation added to the sensible heat, or heat required to raise the temperature of the water up to the temperature of ebullition, is called the total heat of the steam.

The Relation between the Pressure and Temperature
of Saturated Steam.

Probably the earliest experiments on this subject were made by Watt, who tells us that when inventing the separate con

* Muirhead's Life of Watt, p. 76.

densation he made some trials (in 1774) from which he constructed a curve, of which the ordinates represented the pressures, and the abscissæ the temperatures of the steam, and thus enabled him to calculate the one from the other at sight, with sufficient accuracy for his purposes. Watt first surmised that the elastic force or pressure of the steam increased in a geometric progression for temperatures increasing in an arithmetical progression.

*

Robison made experiments upon the same subject at elevated temperatures, ascertaining the temperature at which the steam began to blow off from a safety valve loaded with weights, a proceeding susceptible of little accuracy. Dalton,† however, was more successful in devising an accurate method. He employed a barometer carefully purged of air, into which he introduced a small quantity of water. The barometer was surrounded by an outer water bath, by which the vapour in its chamber was heated to various temperatures. The mercury in the barometer tube adjusted itself so as to be in equilibrium at each temperature between the pressure of the atmosphere on the outside and the pressure of the vapour within, and the column fell as the temperature rose to an extent which is an exact measure of the pressure of the vapour within. The difference of height of an ordinary barometer and the barometer containing the water gives directly the pressure of the steam, so that by a series of careful measurements of a humid barometer and an ordinary dry barometer, the pressures corresponding to various temperatures may be observed.

This method, under various modifications, has been frequently employed, both for water and other liquids, at pressures which are less than that of the atmosphere. The chief difficulty is to maintain the liquid in the bath, by which the barometer is

* Mechanical Philosophy, vol. ii. p. 23.

† Memoirs of the Manchester Literary and Philosophical Society, vol. xv. p. 409. New Series, vol. v. p. 553.

PART I.