the hydrogen. When potassium is heated in the vapour of the substance, there is generally a small expansion of volume, and inflammable gas produced; but it is very difficult to cleanse potassium both from naphtha and an adhering crust of moist potash; and either of these, though in extremely minute quantities, would give fallacious results.
A more unexceptionable experiment made with iron has been already described; and the inferences from it are against the presence of hydrogen in the compound.
Some of the substance in vapour was electrized over mercury by having many hundred sparks passed through it. Calomel was formed and carbon deposited. A very minute bubble of gas was produced, but it was much too small to interfere with the conclusions drawn respecting the binary nature of the compound; and was probably caused by air that had adhered to the sides of the tube when the mercury was poured in.
The most perfect demonstration that the body contains no hydrogen, and indeed of its nature altogether, is obtained from the circumstances which attend its formation. When the fluid compound of chlorine and olefiant gas is acted on by chlorine and solar light in close vessels, although the whole of the chlorine disappears, yet there is no change of volume, its place being occupied by muriatic acid gas. Hence, as muriatic acid gas is known to consist of equal volumes of chlorine and hydrogen, combined without condensation, it is evident that half the chlorine introduced into the vessel has- combined with the elements of the fluid, and liberated an equal volume of hydrogen; and as, when the chloride is perfectly formed, it condenses no muriatic acid gas, a method, apparently free from all fallacy, is thus afforded of ascertaining its nature.
I have made many experiments on given volumes of chlorine and olefiant gases. A clean dry retort was fitted with a cap and stopcock. Its capacity was 25–25 cubic inches. Being exhausted by the air-pump, it was filled with nitrogen (24–25) cubic inches being required), and being again exhausted, 5 cubic inches of olefiant gas, and 10 cubic inches of chlorine, were introduced. It was then set aside for half an hour, that the fluid compound might form, and afterwards being placed again over a jar of chlorine, 19–25 cubic inches entered; so that the condensation had been as nearly as possible 10 cubic