The products, instead of being simply carbonic acid, and nitrogene, are carbonic acid, nitrogene, nitrous acid, probably ammonia, and sometimes nitrous gas. The nitrous acid is disengaged from the base by the intense heat. Concerning the formation of the ammonia, it is useless to reason till we have obtained unequivocal testimonies of its existence; it may be produced either by the decomposition of the water contained in the nitre, by the combination of its oxygene with the charcoal, and of its nascent hydrogene with the nitrogene of the nitric acid; or from some unknown decomposition of the potash. As neither Lavoisier nor Berthollet found nitrous gas produced in the decomposition of nitre by charcoal, when a water apparatus was employed; and as it was not uniformly evolved in my experiments, the most probable supposition is, that it arises from the decomposition of a portion of the free nitrous acid intensely heated, by the mercury. In none of my experiments was the whole of the nitre and charcoal decomposed, some of it was uniformly thrown with the gases into the mercurial apparatus. The nitrogene evolved, as far as I could ascertain by the common tests, was mingled with no inflammable gas. If we consider experiment _f_ as accurate, with regard to the relative quantities of carbonic acid and nitrogene produced, they are to each other nearly as 20 to 8; that is, allowing 2 for the nitrous gas, and consequently, reasoning in the same manner as Lavoisier, concerning the composition of nitric acid, it should be composed of 1 nitrogene to 3,38 oxygene. But though the quantity of oxygene in this estimation is far short of that given in his, yet still it is too much. From whatever source the errors arise, whether from the evolution of phlogisticated nitrous acid, or the decomposition of water, or the production of nitrous gas, they all tend to increase the proportion of the carbonic acid to the nitrogene. I am unacquainted with any experiment from which accurate opinions concerning the different relative proportions of oxygene and nitrogene in the nitric and nitrous acids could be deduced. Lavoisier’s calculation is founded on his fundamental experiment, and on the combination of nitrous gas and oxygene. Dr. Priestley’s experiment mentioned in section 12, on the absorption of nitrous gas by nitrous acid, from which Kirwan[51] deduces the composition of the differently colored nitrous acids, was made over water, by which, as is evident from a minute examination of the facts[52], the greater portion of the nitrous gas employed was absorbed. [51] Essay on phlogiston.