Platinic Chloride.-Gradually darkened in a very marked way, finally becoming blackish.

Ammonium Platinichloride gave same result.

Silver Tartrate.-When spread in a very thin skin over the mortar, each sharp stroke of the pestle left a black line behind it. This is a strong contrast with the complete resistance of this substance to simple pressure.

Silver Carbonate.-Action similar.
Silver Citrate.-Blackens very easily.
Silver Oxalate.-At least as easily.
Silver Arsenate.-Nearly as easily.

Silver Sulphite.-Effect visible in five minutes and gradually increasing. Very well marked.

Silver Salicylate.-No other silver salt appears to be so easily reducible as this. Every sharp stroke of the pestle leaves a brown mark behind it.

Silver Orthophosphate.-Affected easily. After the phosphate has been a good deal reduced, the unchanged part may be dissolved out with ammonia. The black residue after washing readily dissolves in dilute nitric acid, and the solution gives a white cloud with hydrochloric acid.

Potassium Ferridcyanide.-A crystal of the pure salt sharply ground in a mortar becomes in portions brown and in others blue. The quantity used must, as indeed in all of the above cases, be small, one or two decigrams. If a little distilled water be added an insoluble blue powder is left behind, and the solution formed strikes a blue colour when added to one of ferric alum. This indicates that the decomposition is twofold. The experiment is quite a striking one, and the result is easily obtained.

II.

This form of mechanical force, shearing-stress, may be applied to effect endothermic change in other ways. A very simple, and at the same time very efficient, method is that of pressure with a glass rod. Pure strong paper is to be imbued with a solution of the substance, if it is soluble, or if not, it is to be made into a paste with water and then applied with a brush. This paper is to be then very thoroughly dried, and is to be laid upon a piece of plate glass. Characters are to be marked on it with the end of a glass rod that has been rounded by heat, using as much pressure as is possible without tearing the paper.

More than twenty years ago I was able to show that marks made in this way on sensitive photograph-films could be developed, as an invisible image had been impressed. That,

however, is a somewhat different matter from actual and visible decomposition following each stroke of the rod, a result which may be obtained with various salts of gold, mercury, silver, and other metals.

Potassium Ferridcyanide.-Pure paper was imbued with a dilute solution of this salt. After thorough drying it was laid on a glass plate, and marks were made with a glass rod in the manner just described. These marks were immediately visible, and when the paper was plunged into dilute solution of ferric ammonia alum or of ferric chloride they became dark blue. It is probable that the decomposition here effected was twofold (see above).-The nitroprussides appear to be much. more stable than the ferrideyanides. When sodium nitroprusside paper was treated with pressure followed by appropriate reagents, no indications of decomposition were obtained. Auric Chloride.-Paper imbued with a solution of auric chloride and marked in the manner described was thoroughly soaked in water to remove, as far as possible, the rest of the gold salt. The marks were very distinct and gradually gained with time. Colour dark purplish grey.

Platinic Chloride.-After drying and marking, the paper was thoroughly soaked in water and dried. The marks were very distinct, of a yellow colour.

Ammonium Platinichloride.-Marks very visible. Continued to slowly deepen, and in a few weeks were almost black.

Silver Carbonate.-The traces of the rod were brown. When the paper was placed in ammonia the carbonate dissolved, but the marks resisted the action of the ammonia and remained.

Silver Phosphate.-Action very similar to the preceding. Silver Arsenate.-Similar action.

Silver Tartrate and Oxalate.-These salts gave analogous results to the preceding, but not so well marked. The carbonate, phosphate, and arsenate show this reaction best, and about equally well. What is rather curious is that silver chloride does not exhibit a visible action.

Mercuric Oxide.-Paper imbued with a saturated solution of mercuric nitrate and then treated with solution of potash and dried shows this reaction very distinctly. Mercuric oxide appears to be quite sensitive to light.

Turpeth Mineral.-Mercuric sulphate was dissolved in water with the aid of sulphuric acid. Paper was soaked in the solution, allowed to become nearly dry, and then washed. This paper showed the reaction very moderately, but the marks were brought out more strongly by immersion in ammonia.

Ferric Alum (Ammonia).-Paper imbued with solution of this salt, dried and marked, was immersed in solution of potassium ferrideyanide. The marks came out blue, showing that the ferric salt had undergone partial reduction.

It is easily conceivable that the action of shearing-stress should be enormously greater than that of simple pressure. For it seems probable that pressure can only cause decomposition when the resulting product is more dense, has a greater specific mass, than the original substance. With shearing-stress the case is altogether different. All matter is in a state of vibration, and it is easily conceivable that the forcible friction of a hard substance may increase vibration in somewhat the same way as does a bow drawn over a stretched cord. Both the elasticity and the tension of the atoms themselves are vastly greater than those of any stretched cord, so that the increased vibration may easily be sufficient to shatter the molecule.

The transformation of light, heat, and electricity into mechanical energy, as well as the converse transformations, are extremely familiar. That mechanical energy may be transformed into chemical energy is proved by the results described in these papers. The converse transformation, that of chemism into work, is in an industrial point of view by far the greatest chemical problem now awaiting solution. But it is by no means certain that such a transformation is practically possible. At least it seems probable that the improvement in our method of obtaining work from the chemism of carbon may be in the direction of substituting electricity for heat as the intermediary.

IV. On an Approximate Law of the Variation in the Pressure of Saturated Vapours. By the late K. D. KRAEVITCH*.

THE

HE dependence between the pressure of a vapour in a state of saturation and the temperature is probably expressed by an exceedingly complex function, which, notwithstanding the endeavours of many renowned physicists, has not yet been determined. Even the interpolation (empirical) formulæ, with the exception of Biot's formula, cannot be regarded as satisfactory. The cause of failure is to be found in the endeavour to attain the dependence sought for in all its exactitude, which is probably excessively difficult (if not impossible) to do, because the pressure of a vapour in a state *Communicated by Prof. Mendeléeff. Translated by George Kamensky.

of saturation depends, not only on the temperature, but also on a number of other circumstances: on the properties of the liquid (coefficient of expansion, latent heat of vaporization, specific heat, &c.), and of the vapour (coefficient of expansion, specific heat, molecular weight, &c.). In the present stage of science it would be well to search for even an approximate dependence between vapour-pressure and other thermal quantities, and there is no necessity to limit this dependence to a function of temperature; and it might be deemed a considerable scientific success were an expression found for the pressure of a vapour in a state of saturation, in dependence upon any thermal quantity whatever. The finding of such a dependence is naturally easier than that of a precise function, and one, especially, of temperature only*. Such approximate inexact laws and formule would serve as guiding clues for further researches and for the discovery of more exact laws.

In this paper it is my object to show that the pressure of a vapour is in a somewhat simple, although only approximate, dependence upon other observable quantities; or, in other words, if certain quantities are known, corresponding to a definite temperature, having a special significance, then it is possible to calculate with sufficient accuracy the vapourpressure for temperatures near to it, and approximately also for temperatures far removed.

1. Let us imagine a kilogramme of a liquid at a temperature of vaporization t under a pressure p, and let us convert it by two methods into vapour saturating a space, at a temperature t+dt and under a pressure p+dp, and calculate each time the increase of internal energy in the material.

(a) Let us increase the pressure p on the liquid by dp† and heat it to the new boiling-point t+dt, corresponding to the new pressure. The internal energy of the substance will then increase by cdt, where c is an amount of heat which is slightly less than the specific heat of the liquid. In the case of a small external pressure we can neglect this difference, because the coefficient of expansion and specific volume of the liquid, upon which the amount of heat cor sumed in external work depends, are exceedingly small. This difference only becomes significant

I have written more fully on the importance of researches of this kind in my papers, "Remarks on Van der Waals' Formula" (Journal of the Russian Physico-Chemical Society, vol. xix.), “On the Dependence of the Latent Heat of Vaporization upon other Factors" (Idem, xxi.), and in the Repertorium der Physik (vol. xxvi. p. 589).

+ The internal energy of a liquid then decreases, but so slightly that we do not take it into consideration.

for abnormally high pressures, or when the coefficient of expansion of the liquid becomes great. On the liquid attaining the temperature t+dt under the pressure p+ dp, let us convert it into saturated vapour at the same temperature and pressure. If p denote the internal (potential) latent heat of vaporization for the temperature t, we find that the internal energy of the liquid increases by p+dp on its conversion into vapour at a temperature t+dt. Thus, if a kilogramme of a substance in a liquid state, having a temperature t and subjected to a pressure P, be converted into saturated vapour at a temperature t+dt and under a pressure p+dp, then the energy of the substance increases by

cdt+p+dp.

Let us heat the

(6) Let us bring the given liquid, which is under a pressure p and at a temperature t, wholly into the state of saturated vapour at the same pressure and temperature. The internal energy of the substance is increased by p. vapour by dt without altering the volume v. By this means the internal energy is further increased by kdt, where k is a quantity equal to the specific heat of the vapour at constant volume. The vapour will now be in a superheated state. Let us increase the pressure p, under which it occurs, by dp without changing its temperature t+dt; the vapour then passes into a state of saturation. Moreover, if it follows the laws of Boyle and Gay-Lussac, the internal energy of the vapour will not change on compression. If, on the contrary, it does not follow these laws, then the internal energy will decrease by some quantity S, because as a rule the internal energy of substances is decreased by contraction. Hence, in order to convert a kilogramme of a liquid at t and p into vapour at t+dt and p+dp, it is necessary (not counting external action) to augment the internal energy of the liquid by p+kdt-S.

And as both expressions represent one and the same thing, therefore

whence

=

If it be allowed that a saturated vapour is, at a certain pressure, subject to the laws of Boyle and Gay-Lussac, then S 0, because the energy of a perfect gas is not dependent upon the pressure. This is the fundamental proposition of the present paper; and it is not an arbitrary one, * Gorny Journal, 1869, vol. ii. p. 389.