But on introducing a particle of sulphuric acid into the solution, sufficient to dissolve some of the red precipitate, and then pouring a little of it into a solution of prussiate of potash, it gave a fine blue prussiate of iron. Hence the red precipitate was evidently red yellow oxide of iron. I now examined the gas, suspecting that it was nitrous oxide. On mingling a little of it with atmospheric air, it gave red vapor, and diminished. Solution of sulphate of iron introduced to the remainder, almost wholly absorbed it: the small residual globule of nitrogene could not equal one thirtieth of a cubic inch. Consequently it was nitrous gas, nearly pure. Caustic potash was now introduced into the solution, till all the oxide of iron was precipitated. The solution, when heated, gave a strong smell of ammoniac, and dense white fumes when held over muriatic acid. It was kept at the heat of ebullition till the evaporation had been nearly compleated. Sulphuric acid poured upon the residuum gave no yellow fumes, or nitric acid vapor in any way perceptible; even when heated and made to boil, there was no indication of the production of any vapor, except that of the sulphuric acid. _h._ This experiment, compared with the others, seemed almost to prove, that nitrous gas combined with solution of pale green sulphate of iron, at the common temperature, without decomposition; and that when the impregnated solution was heated, the greater portion of gas was disengaged, whilst the remainder was decompounded by the green oxide of iron; which attracted at the same time oxygene from the water and the nitrous gas; whilst their other constituent principles, hydrogene and nitrogene, entered into union as ammoniac. Whilst, however, I was reasoning upon this singular chemical change, as affording presumptive proofs in favor of the exertion of simple affinities by the constituent parts of compound substances, a doubt concerning the decomposition of the nitrous gas occurred to me. As near as I could guess at the quantity of nitrous gas contained by the impregnated solution, at least ¾ of it must have been expelled undecompounded. More than a quarter of a cubic inch of common air had been present in the mattrass: the oxygene of this common air must have combined with the nitrous gas, to form nitric acid. Might not this nitric acid have been decomposed, and furnished oxygene to the red oxide of iron, and nitrogene to the small quantity of ammoniac found in the solution, as in _d_? _i._ I now introduced to a solution of green sulphate confined by mercury, nitrous gas, perfectly free from nitric acid. When the solution was saturated, a portion of it was introduced into a small mattrass filled with dry mercury, in the mercurial trough. The curved tube was closed by a small cork at the top, and filled with nitrous gas; it was then adapted to the mattrass, which was raised from the trough, and the solution thus effectually preserved from the contact of the atmosphere.