To ascertain this point with precision, I distilled a small quantity of the same kind of water, which had been near an hour in ebullition, into a graduated cylinder containing mercury. To this I introduced about one third of its bulk, i. e. 12 measures of nitrous oxide, which had been carefully generated in the mercurial apparatus. After the absorption, a small globule of gas only remained, which could hardly have equalled one fourth of a measure. On admitting to this globule a minute quantity of nitrous gas, an evident diminution took place. Though this experiment proved that in proportion as the water was free from air, the residuum was less, and though there was no reason to suppose that the ebullition and distillation had freed the water from the whole of the air it had held in solution, still I considered a decisive experiment wanting to determine whether nitrous oxide was the only gas produced in the slow decomposition of nitrate of ammoniac, or whether a minute quantity of oxygene was not likewise evolved. I received the middle part of the product of a decomposition of nitrate of ammoniac, under a cylinder filled with dry mercury, and introduced to it some strong solution of ammoniac. After the white cloud produced by the combination of the ammoniacal vapor with the nitric acid suspended in the nitrous oxide, had been completely precipitated, I introduced a small quantity of nitrous gas. No white vapor was produced. Now if any gas combinable with nitrous gas had existed in the cylinder, the quantity of nitrous acid produced, however small, would have been rendered perceptible by the ammoniacal fumes; for when a minute globule of common air was admitted into the cylinder, white clouds were instantly perceptible. It seems therefore reasonable to conclude, 1. That the residual gas of nitrous oxide, is air previously contained in the water, (which in no case can be perfectly freed from it by ebullition), and liberated by the stronger attraction of that fluid for nitrous oxide. 2. That nitrate of ammoniac, at temperatures below 440°, is decompounded into pure nitrous oxide, and fluid. 3. That in ascertaining the purity of nitrous oxide from its absorption by water, corrections ought to be made for the quantity of gas dispelled from the water. This quantity in common water distilled under mercury being about ¹/₅₀; in water simply boiled, and used when hot, about ¹/₃₆; and in common spring water, ¹/₁₂. IV. _Specific gravity of Nitrous Oxide._ To understand accurately the changes taking place during the decomposition of nitrate of ammoniac, we must be acquainted with the specific gravity and composition of nitrous oxide. 90 cubic inches of it, containing about ¹/₃₅ common air, introduced from the mercurial airholder into an exhausted globe, increased it in weight 44,75 grains; thermometer being 51°, and atmospheric pressure 30,7.