tems which are not undergoing what is usually called chemical change) are really in some one of those phases of relative instability which are easily overthrown by contact with small quantities of matter in other phases. The more complex the possible actions and reactions between the components of any chemically heterogeneous system, the more probable will be the occurrence of relatively unstable phases, and the more easily will what may be called the normal course of the chemical change be turned aside by small changes in the magnitudes or intensities of the system. 'Chemical induction' will be a marked feature of such processes.

Molecular compounds may be regarded as systems in phases of indifferent equilibrium2.

194. The considerations regarding chemical equilibrium which have been sketched in the preceding paragraphs shew the great importance of accurate determinations of the course and rate of chemical changes. A considerable amount of work has been done in this direction, but much more regular and systematised research is needed before many generalisations can be made.

195. In 1855, Gladstone studied various reactions in which ferric salts reacted with potassium sulphocyanide, &c. in aqueous solutions, with the production of reddish coloured compounds. The amount of change was determined by measurements of the depth of colour produced3.

Gladstone concluded from his experiments that in chemical operations wherein all the reacting bodies and all the possible products are in solution, the rate of change depends on the rate of mutual diffusion of the various substances, and the 'coefficients of affinity' of the reacting bodies.

1 When we have more data regarding the differences between the quantities of energy associated with different isomerides, it is possible that the whole theory of isomerism, regarded from the thermodynamical point of view, may be developed from the fundamental principle of equilibrium as laid down by Gibbs.

2 See especially Graham's work on the colloidal and crystalloidal states of matter, more particularly his paper in Phil. Trans. for 1861. 183. See also many interesting observations in a small book by Dr Ord On the influence of colloids upon crystalline form and cohesion.

3 Phil. Trans. for 1855. 179; and C. S. Journal, 9. 54.

196. Harcourt and Esson1 examined two chemical processes, viz.

(1) H2O2+2KI+H2SO1 = 2H2O+K2SO+I1⁄2 ; and

(2) K,MnO + 5H2C2O4 + 3H2SO1=K2SO1+2MnSO4+8H2O+10CO2 Their experiments shewed that the amount of the first change varied directly with the quantity of iodide present, other conditions remaining constant. If the quantity of H2SO, was made relatively large, then the amount of change varied with (1) the quantity of iodide, (2) the quantity of dioxide, (3) the time, (4) the total volume of the reacting substances, and (5) with 'some function of each of 'the other conditions under which the change occurs.' Among these other conditions, the influence of varying the quantities of acid, and of varying the temperature, were examined. It was found that the change was accelerated in proportion to the increase of the quantity of H2SO. Other acids were tried, and the conclusion was arrived at that each acid has a definite acceleration-coefficient.'

The reactions between permanganic acid and oxalic acid in presence of sulphuric acid were found to be very complex; it appeared to be possible to analyse the change into four principal parts occurring simultaneously, and the results obtained pointed to the conclusion that for each part of the total operation the statement held good that 'when any 'substance is undergoing a chemical change of which no con'dition varies excepting the diminution of the changing sub'stance, the amount of change occurring at any moment is 'directly proportional to the quantity of the substance.'

197. Menschutkin has studied the velocities and limits of the typical change which occurs when an alcohol and a carbon acid react to produce an ethereal salt and water2.

Certain fairly definite connections between the molecular weights and the structure of the acids and alcohols on the one hand, and the reaction-values of the change studied, on

1 Proc. R. S. 14. 470: 15. 262; and C. S. Journal, (2) 5. 460.

2 For references, details of the methods, and of the plan adopted for stating the results, see book 1. chap. IV. par. 157.

the other hand, have been established by Menschutkin. Thus, by determining the velocity and limit of etherification of the various acids of the acetic series in the reactions between these acids and isobutylic alcohol, it is shewn that replacement of H in the CH2n+1 group of these acids is accompanied by an increase of the limit with a decrease of the velocity. These changes in the etherification-values are more marked when secondary acids are employed, and they reach their greatest values when tertiary acetic acids react on isobutylic alcohol.

When the acid is unchanged but the alcohol is varied, the velocity of etherification up to the point whereat equilibrium is established, increases by a constant amount for each increase of CH, in the molecular weight of the alcohol used, provided the latter always belongs to the class of normal alcohols.

198. Kajander's' experiments on the rate of evolution of hydrogen by the action of various acids on thin plates of magnesium shew that the rate of this change varies with the temperature, the concentration and the nature of the acid employed.

199. But in all these cases the reaction chosen for examination was so complicated that no generally applicable conclusions could be deduced. If conclusions as to the equilibrium of chemical systems are to be deduced from observations of the velocities of changes undergone by these systems, then the simplest changes must be chosen, and, if possible, such as are unmixed with subsidiary physical operations.

One general conclusion appears fairly deducible from the experiments of those chemists whose work has been referred to in this section, namely, that each chemical substance which forms a member of any changing system exerts a specific action on the course of the changes which that system undergoes'.

1 Abstracts in Ber. 14, 2053 and 2676 (original papers are in Russian).

2 See post, chap. 111. pars. 223 and 227. The more important memoirs, besides those already referred to, on the subject of this chapter are as follows: BER

CHAPTER III.

AFFINITY.

Introductory.

200. FROM the beginning of the eighteenth, until the early years of the present century, chemists busied themselves with constructing tables of affinity. The conception which found expression in these tables was of the same kind as underlies such terms as relationship, kinship, &c. As there are degrees of relatedness, so, it was said, there are degrees of affinity. The same substance exhibits different degrees of affinity according to the nature of the other substances with which it reacts. When potash is heated with salammoniac, ammonia is produced, but when sand or silica is heated with the same salt there is no change; this, said Glauber, is because the potash 'loves and is loved by' the acid in the salammoniac. Geoffrey, in 1718, drew up tables of affinity of which the following is a specimen.

THOLLET, Statique Chimique 1. 409 et seq. WILLIAMSON, Proc. R. S. 16. 72. HURTER, Chem. News 22. 193. VAN'T HOFF, Ber. 10. 669; see also his Études de Dynamique Chimique (1884). POTILITZIN, Ber. 12. 2370. LEmoine, Ann. Chim. Phys. (5) 12. 145. HOOD, Phil. Mag. (5) 6. 371: 8. 121: 13. 419. BERTHELOT, Essai de Méc. Chimique 2. 13, 58, 92, &c. HELL and URECH, Ber. 13. 531. The researches of Guldberg and Waage and of Ostwald are considered in detail in chap. III.

M. C.

26

Each substance was said to have a greater affinity than those which came after it for the compound at the head of the column. Thus a compound of sulphuric acid and copper would be decomposed by the action of iron, or of any other substance placed above copper in the column headed sulphuric acid.

But it was gradually found that more than a single table was required for each substance, because the affinity of any substance for any other was not the same at all temperatures, and it also varied according as the reacting substances were solids or in solution. In 1775 Bergmann constructed tables of affinity for 59 substances, two for each, one representing the affinities at low temperatures when the substances reacted in solutions, and the other the affinities between the solid substances at high temperatures. Bergmann's table for potash, for instance, was constructed thus;

Bergmann also fully recognised that each constituent of any reacting substances exhibits affinity for each constituent of the other substances. This point was more fully insisted on in the tables of Guyton de Morveau (1786), of which the following is an example.

SULPHURIC ACID. NITRIC ACID. HYDROCHLORIC ACID. ACETIC ACID.