IV. Apparent Exceptions to the Law of Avogadro The above method of the determination of molecular weights is founded upon the principle that molecules occupy two volumes in the state of gas or vapour, an atom of hydrogen occupying one volume. Now, the densities of a number of vapours are at variance with this proposition. Thus, judging from their vapour densities taken at a sufficiently high temperature, the molecules of the following compounds would occupy four volumes instead of two:-Ammonium chloride and similar compounds, phosphonium iodide, sulphuric acid, phosphorus pentachloride, iodine trichloride, calomel, amylene hydrochloride and hydrobromide, chloral hydrate, &c. But we must endeavour to discover whether the vapours of the compounds in question are not decomposed at the temperatures to which they are raised in order to take their densities, a point to which H. Kopp, Kekulé, and Cannizzaro long ago drew attention. If this is the case, it is obvious that the densities determined at these temperatures do not refer to these compounds themselves, but to the mixture of the products of their decomposition. Thus, for example, we should not be authorised in saying that the molecule of ammonium chloride occupies four volumes if it could be shown that at 360°-the temperature at which the density was taken-this molecule is entirely decomposed into two new molecules-hydrochloric acid gas and ammonia gas-which exist side by side in a state of mixture, each occupying two volumes. It has been proved that this decomposition does take place in the case of some of the compounds mentioned above, and we propose to give, in some detail, the facts and arguments upon which this proof rests. I. It is unmistakable in the case of amylene hydrobromide, CH10.HBr=C,H,,Br. At a temperature which is not more than 40° or 50° above its boiling point, the vapour of this body presents a density (5.2) which agrees with the normal condensation into two volumes, and this density is constant between 150° and 180°. But from 180° upwards it decreases by degrees till at 360° it has sunk to one-half. The vapour is, therefore, completely dissociated into amylene and hydrobromic acid gas, which recombine upon cooling. The same phenomena are observed in amylene hydrochloride. Nevertheless the recomposition of the dissociated elements is not complete, especially in the case of amylene hydrobromide, for, when the flasks are opened under mercury, there is always a residue of a certain quantity of acid gas, testifying to the dissociation which has taken place at a high temperature. There can be but one interpretation of this fact. Amylene hydrobromide cannot possess several vapour densities. The true vapour density of its molecule is that which indicates a condensation into two volumes. The other or halved density indicates a halving of its molecule, and is not a true vapour density. It represents a mixture of decomposition products, and is, as we say, an apparent or anomalous vapour density. 5° II. The case is the same with phosphorus pentachloride, PC. It was generally thought a few years ago, upon the authority of some very accurate but wrongly interpreted experiments of Cahours, that the molecule of phosphorus pentachloride answered to a condensation of elements into three volumes (H,O=2 vol.) It is more correct to say that we have here a phenomenon of partial decomposition or dissociation, according to the beautiful conception of Sainte-Claire Deville, and that, at the temperature at which this vapour is partly dissociated, of the two molecules which occupy four volumes, one is still intact and occupies two volumes, while the other is entirely decomposed into phosphorus trichloride, PC, and into chlorine, Cl2, these products of decomposition occupying four volumes; hence the apparent condensation of two molecules into six volumes, or of one molecule into three volumes. Recent experiments have greatly strengthened this interpretation. The dissociation of phosphorus pentachloride has, in fact, been prevented by diffusing its vapour either into an atmosphere of phosphorus trichloride or into an atmosphere of chlorine. Thus the vapour of the pentachloride being formed in a saturated medium of one or other of its products of dissociation, the latter is retarded, the product having become more stable. We may conceive, in fact, that the pentachloride being dissociated into trichloride and chlorine by heat, the trichloride will have less tendency to separate from the chlorine in an atmosphere saturated with trichloride, |