862

GLADSTONE'S EXPERIMENTS ON THE

portion of iron dissolved in the same bulk of water (each of the solutions employed contained a quantity of iron corresponding very nearly to 1 grain of sesquioxide of iron in 1000 grain-measures of water). A solution of pure sulphocyanide of potassium was then prepared of such a strength, that when I measure of this solution and 4 of that of the iron salts were mingled, the proportion of sulphocyanogen should be exactly sufficient to convert the whole of the iron into sulphocyanide, if complete mutual decomposition occurred: thus the proportions of the two salts employed were such that it would be possible for exact mutual interchange to occur as represented in the following equation: Fe, 6 No ̧ + 6 KScy=Fe,Scy +6 KN→ ̧. On making the experiment in this manner, it was found that the iron was never wholly converted into the red salt, for the tint was deepened by the addition of more either of the iron salt or of the sulphocyanide. In order to obtain a quantitative estimate of the amount of these effects, definite measures of the solutions of ferric nitrate and of sulphocyanide of potassium were mixed together, and the liquid so obtained was diluted with water until it occupied a known, but arbitrary volume. This diluted mixture furnished a liquid of a certain depth of colour which was employed as a standard of comparison. Another measure of the solution of ferric nitrate, equal to that used in the standard solution, was mixed with regulated additions of the sulphocyanide of potassium, and the liquid thus obtained was diluted with measured quantities of water after each addition of sulphocyanide, until, as far as the eye could distinguish, this solution had the same depth of tint as that employed as the standard; it was then assumed that the quantity of sulphocyanide of iron formed was proportionate to the bulk of the two solutions.* Suppose that the standard solution occupied a volume of 880 measures: it was found that if twice the quantity of the sulphocyanide of potassium employed in the standard liquid were made use of in the new solution, this mixture would require dilution till it occupied 1270 measures. The proportion of sulphocyanide of iron formed in these two cases was assumed to be as 880 to 1270, or

as I to 1'44. The excess of sulphocyanide thus employed had therefore withdrawn an additional quantity of iron from its combination with the nitric acid.

*Gladstone found that simple dilution of the sulphocyanide of iron reduced the tint in a proportion greater than could be accounted for by mere dilution; but this source of error was eliminated, and was not found to present itself in other cases which he employed to test the accuracy of the general

conclusion.

INFLUENCE OF MASS ON CHEMICAL ATTRACTION.

863

In this manner experiments were made with quantities of the sulphocyanide of potassium, progressively increasing from one-fifth of an equivalent of the sulphocyanide to each equivalent of nitrate of iron, up to 375 equivalents of sulphocyanide to 1 equivalent of nitrate of iron, and it was found that the quantity of sulphocyanide of iron which was formed, continued to increase with every addition of sulphocyanide of potassium, though the effect of each consecutive addition became less and less marked.

It was ascertained, as indeed it was to be expected, that the proportions of sulphocyanide of iron, which are formed by the mixture of equivalent quantities of other salts of iron with given amounts of the sulphocyanide of potassium, vary with the nature of the acid radicle contained in the ferric salt. For example, it was found that when nitrate of iron was mixed with sulphocyanide of potassium, in the proportion of equivalent quantities of each, that o194 of an equivalent of the red salt was formed. When an equivalent of ferric chloride was used, o'173 of an equivalent was formed; when the sulphate was employed, o'126 of an equivalent was produced; with ferric acetate o'04 only was formed; and when citrate of iron was employed, the quantity of sulphocyanide of iron which it yielded was too small to admit of being estimated. The iron therefore retained the radicles of these different acids with degrees of force which vary inversely with the quantity of sulphocyanide which is formed, whilst the potassium in the sulphocyanide attracted them with a power in direct proportion to these quantities. Various attempts have been made to obtain relative numerical expressions for the force of chemical attraction by which different compounds are united, but they have all hitherto failed. Experiments conducted upon the principle of those of Gladstone appear to offer the fairest prospect of solving this interesting and important problem.

Besides the sulphocyanide of iron, Gladstone examined a variety of other coloured compounds; one of these was the scarlet bromide of gold, which becomes yellow when mixed with the chlorides of potassium and sodium, to an extent varying with the proportion in which these salts are added: the sulphate of quinia, when mixed with a soluble chloride, bromide, or iodide, also afforded similar indications, as it loses its fluorescent character (110) in proportion to the quantity of chloride or bromide with which it is mixed. From these and from a variety of other experiments, it appears that when two or more compounds in solution are made to act upon each other, provided that the products which they form by their mutual action are also soluble, the following conclusions may

864

BUNSEN'S EXPERIMENTS ON THE

be drawn:-1. That mutual interchange between the bodies which are mixed takes place in determinate proportions. 2. That these proportions are independent of the manner in which the compounds were originally combined: thus, if sulphate of potassium and ferric nitrate be mixed in equivalent quantities, the result is the same as if nitrate of potassium and ferric sulphate had been employed in equivalent quantities. This is a fundamental point in these inquiries; but if Struve's observations (quoted by Graham, Elem. Chem. 2nd Ed. p. 232) be correct-viz., that in the preparation of mineral waters, the taste of the liquid varies, not only according to the nature of the salts, but also according to the order in which they are added-it cannot be a general law. 3. That these proportions are dependent partly upon the strength of the mutual attractions of the components for each other, and partly also upon the mass, or relative proportion of each compound which is present in the mixture. 4. That the alteration of the mass of any one of these compounds alters the amount of all the other compounds which co-exist in the mixture, in a regularly progressive ratio; and these quantities admit of being represented by regular curves. In most cases this adjustment of the relative quantities of the different bodies takes place immediately that the mixture is made.

(1003) Experiments of Bunsen and of Debus.-The results are different if the products of the chemical combination be at once removed from the sphere of action,-as by the formation of gaseous compounds, or of an insoluble precipitate when two liquids are mixed. Bunsen has investigated the results obtained in some cases of gaseous combination. He found that when a mixture of hydrogen and carbonic oxide was detonated with oxygen in quantity insufficient for its complete combustion, the oxygen divided itself between the two gases in such a manner that the quantities of water and of carbonic anhydride produced were in very simple atomic relations to each other (Liebig's Annal. lxxxv. 137). He exploded together mixtures of oxygen, hydrogen, and carbonic oxide, in varying proportions, the hydrogen and carbonic oxide being each in considerable excess over the oxygen: under such circumstances water and carbonic anhydride were formed; but the quantity of carbonic anhydride was greater, in proportion as the carbonic oxide preponderated, according to a certain law. Similar results were obtained by detonating cyanogen with a quantity of oxygen insufficient for its complete combustion; in such case nitrogen and a mixture of carbonic anhydride and carbonic oxide in simple proportions were obtained: and when a mix

INFLUENCE OF MASS ON COMBINATION.

865

ture of carbonic anhydride and hydrogen was detonated with a quantity of oxygen insufficient for the consumption of the hydrogen, a certain proportion of the carbonic anhydride was reduced to carbonic oxide, according to the terms of the same law.

The following is the law deduced by Bunsen from his experiments:-. When two gaseous bodies, A, B, are mixed with a third body, c, and fired by means of the electric spark, the body c takes from A and B quantities which always stand to one another in a simple atomic relation: so that for I atom of a c, 1, 2, 3 or 4 atoms of в c are produced; for 2 atoms of a c, 3, or 5, or 7 atoms of B c are formed. If atom of the compound a c, and one of в c be formed in this manner, the mass of A may be increased in the presence of в, up to a certain point, without any change in that atomic proportion; but if a certain limit be passed, the relation of atoms, instead of being as 1 : 1, suddenly becomes as I: 2, or as 2:3; and so on.

B

2. When a body, a, acting upon an excess of any compound, B C, reduces it, so that a c is formed, and в is set at liberty; thenif в in its turn can reduce the newly-formed compound, a c-the final result is, that the reduced part of a c is in simple atomic proportion to the unreduced part. In the case of these reductions also, the mass of one of the ingredients of the mixture may be increased up to a certain point without altering the relative proportions of the compounds obtained; but if increased beyond this limit, a sudden alteration in the relative proportions of the products occurs; but these proportions still admit of being represented by simple ratios. This second portion of the law needs confirmation by more extended experiments.

The following experiments illustrate the first part of the foregoing law-On exploding mixtures of carbonic oxide and hydrogen with oxygen, in the following proportions, Bunsen found that the quantities of carbonic oxide and hydrogen which were oxidized were in the proportions stated below:

These experiments show that, as the proportion of carbonic oxide to the hydrogen in the mixture decreased, the proportion oxidized on detonation decreased also, but it decreased per saltum,

II.

3 к

866

INFLUENCE OF MASS ON COMBINATION.

not gradually, and these proportions were found to be uniformly the same on repeating the detonation with the same mixture, although the degree of compression to which the mixture was subjected during the detonation was considerably varied in different experiments.

The following are Bunsen's principal experiments in support of the second part of the foregoing law :-When carbonic anhydride is driven over ignited charcoal, it is wholly converted into carbonic oxide; but when steam is transmitted over ignited charcoal, a mixture of hydrogen, carbonic oxide, and carbonic anhydride is produced, in the proportion of 4 volumes of hydrogen, 2 of carbonic oxide, and 1 volume of carbonic anhydride. Again, when a mixture of cyanogen with atmospheric air and oxygen was detonated in the eudiometer in the proportion of 6.2 of cyanogen to 10 of oxygen,* the cyanogen yielded 3 volumes of nitrogen, 2 of carbonic oxide, and 4 of carbonic anhydride: and when a mixture of 407 of carbonic anhydride, 33.25 of hydrogen, and 10 of oxygen was detonated, a portion of the carbonic anhydride yielded oxygen to the hydrogen, and was reduced to the state of carbonic oxide; 3 volumes of carbonic oxide being formed, whilst exactly 2 volumes of carbonic anhydride remained unacted upon, although a large excess of hydrogen was present.

Debus arrived at substantially the same results with precipitates as those indicated by Bunsen for gaseous mixtures:-he precipitated a mixture of lime and baryta water, by small proportions of a solution of carbonic acid; and experiments upon a large excess of a dilute solution of the mixed chlorides of calcium and of barium to which a dilute solution of carbonate of sodium was added, led to a similar result.

In the experiments of Bunsen, it must be recollected that the first products of the chemical combination are immediately removed from the sphere of action: carbonic anhydride, and carbonic oxide, and water will not mutually react upon each other; and in the experiments of Debus, the carbonates of the metals of the earths are insoluble-they are therefore at once withdrawn from further action upon the mixture.

(1004) Adhesion.-The influence of adhesion in aiding chemical action is often exerted by overcoming the opposite force of elasticity: this is exemplified by the manner in which water

Cyanogen requires for the complete combustion of its carbon twice its volume of oxygen, so that 6'2 of cyanogen would have required 124 instead of 10 of oxygen; there is therefore more oxygen than would suffice for the conversion of the carbon into carbonic oxide.