Fig. 542.

position, since, as will be seen by what follows, the temperature of the liquid may, under circumstances which not unfrequently occur, rise somewhat above the true boiling point; but even when this is the case, a thermometer in the vapour will show the real boiling point of the liquid: under nearly all circumstances, the thermometer will stand lower in the vapour than it does in the liquid, if this is a mixture of two or more liquids of different boiling points instead of a pure, homogeneous substance. Through a second hole in the cork is inserted a glass tube, open at both ends, and bent at a right angle, as shown in fig. 542; by connecting this tube with a condenser, the loss of the liquid used for the experiment can be prevented. The liquid is heated either by applying a small flame to the outside of the vessel, or by means of a water-bath or sand-bath, care being taken that the sides of the vessel above the liquid do not get over-heated. The indications of the thermometer are observed during the whole time that the liquid is being slowly boiled away, until only a small quantity remains. The temperature thus observed is not, however, in most cases, the true boiling point of the liquid: usually, part of the mercury column in the thermometer rises above the cork, and is therefore exposed to a lower temperature than that of the boiling liquid; consequently, the upper extremity of the column stands at a lower point than it would do if the thermometer were completely immersed in the liquid. In order to find the correction which it thus becomes necessary to apply, a second thermometer is placed so that its bulb is in contact with the stem of the thermometer inserted into the cork of the boiling vessel, and is half way between the top of the mercury column of the latter thermometer and the middle of the cork. The temperature indicated by this second thermometer may be taken as the mean temperature of that portion of the mercury column of the principal thermometer which is not heated by the vapour of the boiling liquid. Let this temperature be to; let the uncorrected boiling point, directly indicated by the principal thermometer, be 7°; let N be the difference between T and the point of the scale situated at the middle of the cork, that is to say, the length, expressed in degrees of the scale, of that portion of the mercury column of the principal thermometer of which the mean temperature is ; lastly, let & be the coefficient of apparent expansion of mercury in the glass of which the thermometer is constructed. The correction to be applied to the directly observed temperature 7° is then

As already stated (p. 57), 8 may always be taken, in calculating the value of this expression, as = 0·0001545.

The table which follows on p. 86 gives the amounts of the correction in question for various values of N and of T-t. The amounts corresponding to other values of these factors can be easily deduced by interpolation from the numbers given in the table.

This table sufficiently shows that the correction in question can never be neglected in accurate experiments, and that in the case of liquids of high boiling points, its value may become very considerable.*

Since the boiling point of a liquid depends on the pressure to which the liquid is subjected, another correction becomes necessary in order to reduce determinations made under the varying pressure of the atmosphere, to the values which would be found if the atmosphere exerted always its normal pressure, equal to that of 760 millimetres of mercury at 0°. Strictly speaking, the correction to be applied to the boiling point of a liquid observed under an atmospheric pressure differing by a given amount from the above standard pressure, varies with the nature of the liquid; since equal alterations of pressure do not cause precisely equal changes in the boiling points of different liquids. Nevertheless, the greatest variations which ever occur in the pressure of the atmosphere are relatively so small, that they may, without any apprecable error, be regarded as affecting the boiling points of all liquids equally to the extent, namely, of 0-1° for a variation of pressure of 27 millimetres of mercury, this number being deduced from direct determinations of the boiling point of water under different pressures.

In what follows, whenever the boiling point of a liquid is spoken of without further explanation, it is to be understood to mean the boiling point under a pressure equal to that of 760 millimetres of mercury at 0°.

It is obvious that a precisely similar correction ought to be applied to all thermometric observations in which any portion of the mercury in the stem of the thermometer is at a different temperature from that in the bulb.

Circumstances which modify the Boiling Point.-Although, when a liquid is heated in such a manner that vapour can escape freely from some part of its surface, the vapour so formed has a tension equal to the pressure upon the free surface of the liquid as soon as the temperature of the latter reaches the boiling point, this temperature may nevertheless be attained, and even considerably exceeded, without the formation of a trace of vapour, if no portion of the surface of the liquid is freely exposed. These conditions can be realised by suspending the liquid to be examined in a second liquid of equal specific gravity, but higher boiling point.

The phenomena which take place under these circumstances have been particularly studied by Dufour (Ann. Ch. Phys. [3] lxviii. 378). In order to examine them in the case of water, he employed a mixture in the requisite proportions of oil of cloves (previously heated alone to about 200°) and linseed oil. The water, already heated to 80° or 90°, was dropped gently into the mixture of oils, so as not to disturb the film which coated the bottom of the vessel, and the temperature of the bath was gradually raised. Under these circumstances the ordinary boiling point of water, 100°, was passed without the occurrence of any perceptible change, and traces of ebullition scarcely began to show themselves below 110° or 115°. Even at these temperatures, ebullition seldom began except when the globules of water came in contact with the sides of the vessel or with the thermometer. A burst of vapour then occurred, and the globule, more or less diminished in size, was driven rapidly away, like a pith ball after touching an electrified conductor. These contacts were of course more difficult to avoid in the case of large than of small globules; hence the latter remained liquid, as a rule, to higher temperatures than the former.

In these experiments, it was a rare exception when ebullition occurred between 100° and 110; very commonly globules of 10 mm. in diameter reached 120° or 130°, and in one experiment the last temperature was attained by a globule of 18 mm. diameter, and therefore containing more than 3 c. c. of water. Spheres of 10 or 12 rum. diameter often reached 140°; those of 5 or 6 mm. reached 165°; and others of from 1 to 3 mm. attained 175° or even 178°, temperatures at which the elastic force of the vapour which forms at the freely exposed surface of water is between 8 and 9 atmospheres.

At these high temperatures, the contact of a solid body very generally occasioned the sudden partial or complete vaporisation of the globules, accompanied by a hissing sound like that produced on immersing red-hot iron in water. This invariably occurred when the globules were touched with pieces of wood or chalk, shreds of cotton, paper, &c., but not always on contact with a glass rod or metallic wire, the difference appearing to depend on the porous structure of the former substances. A platinum wire appeared to lose, to some extent, by frequent usage, the power of causing sudden vaporisation.

Sudden ebullition, amounting even to an explosion, if the temperature was above 120°, invariably occurred on passing the discharge of a Leyden jar or induction coil through a globule. A similar, but less violent, effect was produced by the passage of a weak galvanic current. These results are attributed by Dufour less to the contact of the globules with the conducting wires, than to the disengagement of gas at the extremities of the latter.

Saturated aqueous solutions of various salts--for example, chloride of sodium, sulphate of copper, nitrate of potassium, &c.-also remained liquid at temperatures much above their toiling points, when immersed in melted stearic acid resting on a layer of melted sulphur. In like manner, globules of chloroform (boiling point 61°), suspended in a solution of chloride of zinc, often remained liquid up to 97° or 98°; and globules of liquid sulphurous anhydride (Loiling point - 105°) could be heated in diluted sulphuric acid as high as + 8°. In all these cases, the same causes that operated in the case of Water, sufficed to❤ccasion the sudden complete or partial conversion of the overheated globules into vapour.

These results throw important light upon the nature of ebullition, and seem to indicate that it is to some extent an accidental phenomenon. In order to understand them, we must remember that, the globules being surrounded on all sides by liquid, evaporation cannot go on at their surface in the ordinary way. They are, however, in a state of tension, or unstable equilibrium, such that a very slight cause may occasion the sudden formation of vapour of more than the atmospheric tension. The most effectual of such causes would obviously be the contact of a minute globule of air or other gas: this globule, however small, would be a space into which vapour could be given off, and this vapour, having an elastic force greater than the pressure (that of the atmosphere and the upper layers of the liquid) whereby the globule was prevented from expanding, would force back the liquid walls of the bubble of gas, suddenly converting it into a large bubble of steam. Hence the unfailing efficacy, in causing the ebullition of the overheated globules of liquid, of the passage of an electric current or the contact

of porous substances, such as chalk, wood, paper, &c., which either allow air to escape from their pores when immersed in the heated liquid, or carry down into it small globules of air adhering to them. These globules afford space for the commencement of the formation of vapour, and this process being once begun, the space is increased by the force of the vapour already formed within it. In the absence of any such space, the liquid globule is in a condition somewhat analagous to that of a drop of melted glass which has been suddenly cooled in water (Rupert's drops), and which falls to powder on receiving the smallest scratch: there is no reason why the formation of vapour should begin at one point of the mass rather than at another, and thus the whole remains in a state of molecular tension until something occurs at some particular point to weaken the effect of the forces which oppose the formation of vapour, or until the tension increases (in consequence of rise of temperature) to such a degree that these forces are overcome simultaneously throughout the whole mass. It will be seen further on that these considerations exactly agree with the explanation which the mechanical theory of heat affords of the passage from the liquid to the gaseous state.

Even when liquids are heated in open vessels, the occurrence of ebullition, or the formation of vapour in the interior of the mass, appears to depend on like accidental causes. Thus the influence of the nature of the vessel wherein a liquid is contained has been long recognised. It has been observed, for instance, that water, which will boil steadily at 1000 in a metallic vessel, may often be heated to 105° or 106° without boiling, in a glass vessel previously washed with strong sulphuric acid. Under these circumstances, ebullition generally takes place very irregularly at a certain temperature a sudden burst of vapour occurs, and the temperature of the liquid falls at the same time to nearly the normal boiling point; the liquid then remains tranquil for a time, until the temperature having again risen considerably, another burst of vapour takes place, and so on. The "bumping" of heated liquids which results from this intermittent formation of vapour, is familiar to every chemist. It may be prevented to some extent by putting into the liquid a few scraps of platinum, or a globule of mercury; a small piece of charcoal is, however, much more effectual than either of these, though not always admissible. This irregular ebullition occurs much more frequently in some liquids than others, and to a greater extent in certain glass vessels than in others. For instance, methylic alcohol (boiling point 61°) may show a difference in its temperature of ebullition amounting, according to Kopp, to 5 or 6 degrees, depending on the vessel in which it is examined. When a liquid is boiling steadily in a glass flask or retort, it may almost always be noticed that the bubbles of vapour start from one or two particular points of the surface of the glass, indicating the existence of some irregularity of the glass at those points favourable to the formation of vapour. Another illustration of the necessity of some other cause than mere temperature in order to bring about the ebullition of liquids, is afforded by the remarkable observation of Professor Donny, of Ghent, that water thoroughly deprived of air and sealed up in a rather long glass tube quite free from air, may be heated to 138° at one end of the tube without boiling, and is then suddenly and violently thrown to the other end by a burst of vapour. An additional fact of the same kind may sometimes be observed during the distillation of liquids under the ordinary pressure. The writer has occasionally seen a liquid distil very rapidly, thus showing that vapour was being formed rapidly at the surface, although ebullition could not be maintained by the use of a more powerful flame than that which had sufficed, at an earlier stage of the experiment, to boil a larger quantity of the same liquid.

The whole of these phenomena cease to be unintelligible if, with Dufour, we distinguish between the boiling point of a liquid, and the temperature at which the elastic force of its vapour becomes equal to the pressure of the atmosphere, and define the boiling point as the lowest temperature at which ebullition can occur, instead of as the temperature at which, under normal conditions, it must occur.

Spheroidal State.— When a drop of water is allowed to fall upon a piece of iron, the temperature of which considerably exceeds 100°, it retains its globular form, moves about rapidly on the surface of the iron without wetting it, and evaporates with comparative slowness. As the iron cools, a point is reached at which the globule of water wets it, spreads over its surface, boils and quickly disappears. The condition of the globule first described has been distinguished as the spheroïdal state. This condition can be assumed by all volatile liquids when they come in contact with the surface of either a solid or a liquid body heated considerably above their boiling points. The temperature of a liquid in the spheroïdal state is always below its ordinary boiling point, notwithstanding the higher temperature of the surface on which it rests. The absence of ebullition is therefore due in this case to some other cause than that which produces the phenomena of deferred ebullition which were considered in the last paragraph. It is the result of a want of perfect contact between the liquid and the heated surface. Many experiments prove that the liquid globule rests upon a sort of cushion of its own

vapour, produced by the heat radiated from the hot surface against its under side. As fast as this vapour escapes from under the globule, its place is supplied by a fresh quantity produced in the same way, so that the globule is constantly buoyed up by it, and never comes into actual contact with the heated surface. If, however, the temperature of the latter is allowed to fall, the formation of vapour at the under surface of the globule becomes less and less rapid, until at last it is not supplied fast enough to prevent the globule touching the hot metal or liquid on which it rests: as soon as contact occurs, heat is rapidly imparted to the globule, it enters into ebullition, and quickly boils away.

According to Boutigny's experiments, the lowest temperature at which a metallic vessel will cause the spheroïdal state to be assumed by water is 142°; for alcohol the lowest temperature is 134°, and for ether 61°.

Solid bodies which evaporate without becoming liquid also assume a condition analogous to the spheroïdal state of liquids, when they are placed upon a surface whose temperature is sufficiently high to vaporise them very rapidly. This is very distinctly seen on throwing a piece of solid carbonate of ammonium into a red-hot platinum crucible, and when a fragment of solid carbonic anhydride is placed upon any good conducting surface at the ordinary temperature.

Effect of Substances in Solution on the Boiling Point of Liquids-Liquids holding solid bodies in solution boil generally at higher temperatures than they do in the pure state. Thus sea-water, containing on an average about 3.5 per cent. of saline matter, boils at about 103°. On the other hand, the boiling point of water is lowered by mixing it with alcohol, and that of alcohol by mixing it with ether; the boiling point of a mixture being always intermediate between the boiling points of its components. The effect of salts in raising the boiling point of water may be considered as also coming under this rule. The following table gives the boiling points of saturated solutions of several salts, according to Legrand.

(For Legrand's determinations of the boiling points of weaker solutions of these and some other salts, and for the similar experiments of Griffiths and Faraday, see Gmelin's Handbook, i. 269, 270.)

Very contradictory opinions have been maintained by different observers with regard to the temperature of the vapour which issues from boiling saline solutions. It has been said, on the one hand, to be the same as that of pure water boiling under the same pressure; and, on the other hand, to be equal to that of the highest stratum of the solution. According to the recent experiments of Magnus (Pogg. Ann. cxii. 408), the latter statement appears to be nearest the truth. These experiments prove decisively that the vapour of boiling solutions is hotter than that of pure water, and that its temperature rises as the solutions become more concentrated, and therefore boil at higher temperatures; nevertheless, the vapour was always found a little colder than the mass of the boiling solution, and the difference was greater at high temperatures than it was at low ones.

Relations between the Boiling Point of substances and their Chemical Composition.—Many attempts have been made to trace some connection between the boiling points of different liquids and their chemical composition. The most extensive and important series of observations that have been made upon this point are due to Kopp (Ann. Ch. Pharm. xcvi. 2, 330; xcviii. 267, 367; Phil. Trans. 1860, 257). The principal conclusions deducible from these investigations are as follows:

1. Analogous compounds, presenting the same difference of composition, very frequently differ by the same amount in their boiling points, or the interval between their boiling points is proportional to their difference of composition. A compound con