of heat which is again lost by the system as it passes back to its original configuration. We may picture to ourselves the action of heat as bringing about a separation of the molecules of the dissociating body into atoms, followed by a rearrangement of these atoms to form new molecules, the new system thus produced being dependent for its continued existence on supplies of energy from without itself; but when the supply of energy in the form of heat is stopped, we may suppose that the interatomic attractions bring back the system to its original molecular arrangement. When however a chemical decomposition occurs by the application of heat, the new configuration assumed by the atoms is stable, it does not require to be supplied with energy from without in order that it may continue to exist; hence there is no swinging back to the original state.

167. In the dissociation of calcium carbonate the system at any moment is composed of three distinct substances; a decrease in the amount of calcium carbonate is necessarily accompanied by an increase of lime and carbonic anhydride, and vice versa. If we study the processes of dissociation wherein more than one arrangement of the members of the system is possible, we shall find that the configuration assumed at any given temperature depends, as in the simpler case of calcium carbonate, solely on the pressure exerted by the gaseous products of the change.

Take, for instance, the two compounds, AgCl. 3NH, and 2AgCl. 3NH,, produced by the action of ammonia on solid silver chloride. If silver chloride is brought into an atmosphere of ammonia, at ordinary temperatures, the ammonia is absorbed with formation of 2AgCl. 3NH,, and the pressure falls; by increasing the quantity of ammonia absorption again proceeds, and so on. For every temperature there is a certain pressure whereat neither absorption or evolution of ammonia occurs; this equilibrium-pressure is independent of the relative

be made to return from the final state to the original state, under conditions which, at any stage of the reverse process, differ only infinitesimally from the conditions at the corresponding stage of the direct process." Clerk Maxwell; Article 'Diffusion' in Encycl. Brit. (9th ed.).

amounts of AgCl and 2AgCl. 3NH, present'. When the equilibrium-pressure is reached, if the pressure of the ammonia in the apparatus is largely increased, absorption occurs and the compound AgCl. 3NH, is produced. For this compound also there is a pressure, corresponding to each degree of temperature, whereat equilibrium is established. The following table gives some of the equilibrium-pressures as determined by Horstmann®.

The difference between the pressures corresponding to each of these compounds at any temperature is so great that it is comparatively easy to study the relations between pressure, temperature, and amount of chemical change for each compound: but when attempts are made to do this for hydrated salts, e.g. CuSO,5H2O, it becomes almost impossible to determine the equilibrium-pressures for various temperatures. When hydrated copper sulphate is heated to a given temperature in an enclosed space the pressure of the watergas evolved shews great fluctuations. At first sight it might be thought that the dissociation of this salt does not follow the ordinary rule, viz. that the equilibrium-pressure is independent of the relative amounts of decomposed and undecomposed solid substances present, and is dependent solely on the temperature. But on closer examination it is found that the solid undergoing dissociation is really a mixture of various hydrated salts, each with its own equilibrium-pressure, but that the differences between these pressures are small. In

1 Horstmann, Ber. 9. 749; Isambert, Compt. rend. 66. 1259: 70. 456.
2 loc. cit.

this case we have therefore a number of processes of dissociation proceeding simultaneously; hence we cannot expect to find a definite equilibrium-pressure for each temperature'.

The differences between the equilibrium-pressures of the various hydrates are so marked in the cases of some hydrated salts, that the phenomenon of dissociation can be shewn to follow the same course as with calcium carbonate, or the ammoniacal silver chlorides2.

168. In any dissociating system at the equilibriumpressure, the number of molecular decompositions and recompositions must be regarded as equal in number during any given interval of time. A general theory of chemical change, including gaseous dissociation, based on the molecular kinetic theory of gases, has been developed by Pfaundler. This theory will be considered when we come to deal in detail with the subject of chemical change. It is however difficult to reconcile Pfaundler's theory with the fact that the amount of dissociation of a solid, into solid and gaseous products, is independent of the relative quantities of the original substance and the solid product of dissociation. In the dissociation of calcium carbonate, e.g. we should expect, that the greater the amount of lime present, relatively to the amount of carbonate, the greater would be the chances that some of the CO, molecules should be caught and held fast, and that therefore the amount of dissociation at any temperature would depend upon the ratio between the lime and calcium carbonate actually present in the system at that temperature. Pfaundler has very ingeniously tried to get over this difficulty by referring the dissociation only to the molecules on the surfaces of the various members of the system. In gases all molecules may be regarded as being on the surface; but

1 (See Naumann, Ber. 7. 1573-) It should be noted that the equilibriumpressure for a given temperature is never attained immediately that temperature is reached; a little time must elapse before the entire system has settled down into equilibrium. See Horstmann, Ber. 9. 752, and Naumann, Annalen 160. 27. 2 For numbers see Debray, Compt. rend. 66. 194.

3 See post, chap. 11. par. 187.

See this difficulty stated by Horstmann, Ber. 9. 757

5 Ber. 9. 1152.

when a solid is present, only those molecules of the solid which can be directly bombarded by the gaseous molecules, are, in Pfaundler's language, on the surface. The number of surface-molecules is very small compared to the total number, hence a change in the relative amounts of the solid compounds present (call these AB and B) will be accompanied by a change in the ratio of surface-molecules of AB to those of B, small in comparison with the number of molecules of the gaseous body present (call this A); hence only a small change of pressure will occur, and this will quickly be rectified by the absorption (or evolution) of a little more of A. But all molecules of A must gradually come to the surface of AB, and take part in the exchange which is going on between AB and A; but the molecules of A which pass into the interior of the solids AB and B will remain there a comparatively long time, and hence will exert but a small influence on the pressure of the gas A.

Horstmann' has rather endeavoured to develop a general theory of dissociation from thermodynamical principles; he has deduced a formula from the second law of thermodynamics applicable to cases of dissociation, and generally to chemical changes brought about by the action of heat, and he has sought to shew that when the positive and negative changes are equal in a process of dissociation, i.e. when the equilibrium-pressure is reached, the entropy of the system has attained its maximum value. But the application of thermodynamical methods to questions of chemical equilibrium, including dissociation, will be considered in another chapter.

169. The special characteristics of dissociation which, taken together, mark it off from decomposition, are then briefly these: (1) heat is absorbed, and the temperature of the dissociating system increases throughout the entire process; (2) the original configuration of the system is returned to, if the products of dissociation are allowed to cool in contact with each other; (3) the change is gradual, and

1 Annalen 170. 192; see also do. Supplbd. 8. 112.
2 See Neues Handwörterbuch der Chemie, 2. 999.

therefore at any given temperature the dissociation is partial, although the whole mass of the dissociating substance is submitted to the same thermal conditions as regards supply of heat from without; (4) the amount of dissociation is dependent on the temperature and pressure of the gaseous product or products; and (5) is independent of the ratio between the quantities of the solid products of the change present at any temperature1.

170. It is evident that the possible occurrence of dissociation must have an important bearing on determinations of the densities of gaseous compounds from which the molecular weights of these compounds are deduced?.

We have already learned (par. 101) that the density of the vapour of acetic acid decreases, as temperature increases, from the boiling point to about 100° above this point, after which it becomes constant. If the temperature is kept constant and the pressure is increased, the density of the vapour increases. The rate of change in the value of the density of this vapour, which accompanies change of temperature and pressure, is much more rapid than the rate of change in the value of the density of air under similar conditions; there are however no abrupt changes in the value under discussion. The smallest molecular formula assignable to acetic acid,

1 It is probable that relatively less energy is transformed into heat in the formation of a dissociable compound from its constituents, than in the formation of an analogous compound which decomposes, but does not dissociate, when heated. If this is so, then perhaps the relatively small evolution of heat which attends the formation of the former compound may be connected with the existence of atomic groups in the molecule (or in the reacting unit) of this compound. When the compound is heated, it separates, on this supposition, into those groups the parts of which hold together; whereas when the other compound is heated much energy is absorbed, and the result is a separation of the molecules into their constituent atoms, which at once pair off into new molecules, so that the conditions required for the re-formation of the original compound no longer exist. This view may, I think, be shewn to be in keeping with Pfaundler's theory of chemical equilibrium (see post, chap. II. par. 187). The ratio of the energy of rotation of the parts of the molecules to the energy of agitation of the molecules as wholes, would, on this view, be partly dependent on whether these molecules were built up of individual atoms, or of groups of atoms each of which was more thermally stable than the molecules themselves.

2 See ante, book 1. chap. I. par. 16.