dT

and the law that has the sign of

dt

dQ

dt

again holds good. Consider, now, the meaning of the condition (IV.). Evidently the meaning is that the potential of gravitation should be a continuous explicit function of volume alone. To every infinitesimal variation of volume corresponds an infinitesimal variation of the potential. This condition is realized when the body passes in a continuous manner from one configuration of mechanically stable equilibrium to another. It is evident that, when speaking of the contraction of a celestial body, we mean a "grosso modo" continuous passing from one configuration of stable equilibrium to another. Hence, resuming our reasoning, we may say that the temperature of a celestial body, uninfluenced by external forces and radiating heat, must be sinking. But a rain of meteorites (as Lord Kelvin has suggested) may bring quanta of energy sufficient to compensate the loss of heat by radiation, because the sun and a meteorite falling upon it constitute a mechanically unstable system, and the condition (IV.) is not fulfilled.

XXVII. Changes of Temperature caused by Contact of Liquids with Powdered Silica &c. By Dr. G. GORE, F.R.S.

IT

T is already known that the physical properties of the surface-layer of molecules of a solid or liquid substance are in some respects different from those of the molecules which constitute the inner portion of the mass, and that those of a very thin film of liquid are not exactly like those of a large bulk. It is further known that solid substances have the power of condensing extremely thin films of liquids, vapours, and atmospheric air upon their surfaces and in their pores, and that the physical properties of the liquids and vapours are thereby modified.

According to Rücker, the specific electric conductivity of a film of soap-solution increases as the thickness decreases (Nature,' June 1 and October 26, 1893, pp. 116, 625). The superficial film of a liquid is usually very much more viscous than the interior (Daniell, Text-book of Physics,' p. 258). According to Overbeck, distilled water offers more, and alcohol less, resistance to the vibrations of a pendulum at its surface than in the interior of its mass (Nature,' Feb. 24, *Communicated by the Author.

1881, pp. 397, 398). Pockels states that the surface of water has greater solvent power than the mass, and that a thin disk of camphor partly immersed in a vertical position in water having the cleanest possible surface is cut through in the course of a few hours (ibid. March 12, 1891, p. 439). We know that plates of metal partly immersed in a vertical position in corrosive liquids are frequently cut through at the air-contact line: this is often the case with anodes in electrolysis; and I have observed, whilst electrolytically depositing pure grey antimony from its solution, the deposit spread from the cathode over the surface of the liquid as if the surface conducted better than the interior of the liquid. Mensbrugghe, speaking of the cause of surface-tension and evaporation of liquids, concluded that the molecules of the surfacelayer of a liquid are farther apart than those of the mass (ibid. March 2, 1893, p. 428); he also observed that the boiling-point of carbon bisulphide contained in the pores of finely-divided carbon was higher than that of a mass of the free liquid (Phil. Mag. July 1877, p. 43). I have shown that when the molecules of a solid or liquid conducting substance are separated farther apart by solution or dilution, they acquire greater electromotive force due to increased molecular volume ("A Relation of Electromotive Force to Equivalent Volume and Molecular Velocity of Substances,' Proc. Birm. Phil. Soc. 1892, vol. viii. pp. 63-138). J. J. Thomson states that, with saline solutions, in some cases the surface-film of the liquid contains more and in other cases less salt per unit of volume than the interior; and that permanganate of potassium is removed from its aqueous solution by trickling through pure silica (Applications of Dynamics to Physics,' pp. 191, 192). It has long been known that vinegar is to a certain extent separated from water by pure quartz sand, and that when potato-brandy is filtered through that substance, water passes through first, then alcohol, and finally alcohol plus fusel-oil, as unaltered liquid (Gmelin's Handbook of Chemistry,' vol. i. p. 114). Further, in a recent research on the "Decomposition of Liquids by Contact with Powdered Silica," I have found that on agitating various aqueous solutions of acids, alkalies, and salts with finely precipitated pure silica and allowing the powder to subside, the latter substance in many cases abstracts a larger proportion of the dissolved compound than it does of the water, and in a number of cases less (see Proc. Birm. Phil. Soc. 1894, vol. ix. part 1, pp. 1-24), and that in some instances it abstracts more than 80 per cent. of the dissolved substance.

6

It is generally recognized by physicists and chemists that

the surfaces of mutual contact of two different substances (or of two masses of the same substance in different physical or molecular states) are usually localities of change and transference of energy. Two such substances can rarely be brought into mutual contact without such change occurring, and the changes which occur are more or less different in every different case. In multitudes of cases the changes are of a chemical character, in others they are physical. Volta showed that the mere contact of two different metals rendered them electropolar. Many years ago Pouillet (Annales de Chemie, 1822, vol. xx. pp. 141-162), also Regnault (Gmelin's 'Handbook of Chemistry,' vol. i. p. 300), and recently Meissner ('Nature,' Nov. 4, 1886, p. 24), found that when finely divided dry solid substances, such as silica, metals, emery, carbon, &c. were wetted by liquids, such as water, benzol, or alcohol, which have no dissolving or corrosive action upon them, heat is evolved. Pouillet employed many oxidizable substances. Daniell states that when a liquid begins to flow through a dry tube, heat is evolved ('Text-book of Physics,' p. 284).

In the present research I have endeavoured to extend our knowledge of the thermal phenomena of this subject. The following is a sectional sketch and a description of the apparatus and method of experimenting.

A is a glass vessel usually containing 50 centim. of the liquid and the bulb of a sensitive thermometer, and is supported in a wire ring having three slender legs. B is a wide glass tube containing 100 grains of finely precipitated pure and dry silica or other fine powder. C is a narrow glass tube open at both ends, and loosely enclosing the stem of the thermometer; it is supported firmly at its upper end by the clip D, and has fixed securely upon its lower end a conical and very smooth piece of cork which accurately fits the outer tube.

Both the glass vessels A and B are thickly coated with cotton wool. The thermometer is capable of distinctly indicating a one-hundredth of a Centigrade degree change of temperature, and its bulb is supported a quarter of an inch above the bottom by a small piece of cork.

The apparatus is placed in a situation in which the temperature remains sufficiently uniform, and is entirely surrounded by a movable screen of cardboard.

In making an experiment, the powder and liquid were placed near each other in a covered glass vessel in a sheltered situation during a period of sixteen hours, in order to acquire the same temperature. The liquid was then poured into A, and the powder into B, the cup A placed in position beneath with the thermometer immersed in the liquid, and the entire arrangement allowed to stand a sufficient period of time, usually about half an hour, for the solid and liquid to regain the same temperature. The outer tube B was then raised so as to allow the whole of the powder to fall slowly in a thin annular stream into the liquid. The powder entirely surrounded and covered the bulb to a height of one fourth of an inch above its upper end. If the bulb was not sufficiently covered by it, the changes of temperature were less owing to the liquid above being less heated than the powder. The liquid was not stirred in any instance. The act of pouring the liquid into A caused a rise of temperature of about 03° C., but that of pouring the silica into B had no apparent effect. The maximum rise by contact of the powder and liquid usually occurred in about three minutes. The powders employed were almost wholly different from those used by M. Pouillet, and were such as would not oxidize, and had no recognized chemical action upon the liquid: it is important also to use powders which do not agglomerate or change by

use.

EXPERIMENTS.

Silica with various Liquids.

In all the following experiments with silica, except those otherwise indicated, the weight of silica employed was 100 grains, the volume of solution was 50 centim., and the proportion of dissolved anhydrous substance was 10 per cent. by weight of the water.

The silica was thoroughly purified and ignited previous to each experiment, and was kept in a covered glass vessel during the cooling until it was used. The proportion of cyanogen in the cyanide agreed with that of 100 per cent. KCy; but by evaporating a solution of it to dryness with an excess of sulphuric acid, the weight of sulphate obtained indicated a mixture of 14.45 per cent. NaCy and 85.55 KCy. The following Table shows the results of the experiments with silica.

Phil. Mag. S. 5. Vol. 37. No. 226. March 1894. Y

Note "A," exp. 42.—In this case the silica was exposed in an uncovered vessel to the air during sixteen hours, but this had only a minute influence upon the amount of rise of temperature (compare with No. 39).

Remarks.-In every one of the instances of Table I. a rise of temperature occurred; as, however, water alone gave a rise and was used as the solvent in all cases, this circumstance may explain that uniformity of result. Whilst many of the solutions gave larger rises than water, quite as many gave smaller ones; and we may regard all those which gave smaller ones as containing substances which absorbed heat and diminished the rise, whilst those which gave larger ones evolved heat and added to the amount of rise produced by water alone. Probably by further research some substances may be found which will more than neutralize the temperatureeffect of the water and give negative numbers.

The series of results obtained with potassium cyanide in exps. 52, 53, 54, and 57 show that the magnitude of rise of