The singular property possessed by ammonium nitrate of decomposing in several distinct modes according to the rapidity of heating and the temperature to which the substance is raised, first indicated by Davy, has been minutely studied by M. Berthelot, who has shown that this comparatively simple salt may be decomposed in as many as six different ways. It may be (1) dissociated into gaseous nitric acid and ammonia; (2) decomposed into nitrous oxide and water; (3) resolved into nitrogen, oxygen, and water, (4) or into nitric oxide, nitrogen, and water, (5) or into nitrogen, nitrogen peroxide, and water; or lastly (6), under the influence of spongy platinum, it may be resolved into gaseous nitric acid, nitrogen, and aqueous vapour. These different modes of decomposition may be distinct or simultaneous; or, more exactly, the predominance of any one of them depends on relative rapidity and on the temperature at which decomposition is produced. This temperature is not fixed, but is itself subordinate to the rapidity of heating (_cf._ Berthelot’s “Explosives and Their Power,” translated by Hake and Macnab). The assertion of De la Metherie, that the gas produced by the solution of platinum in nitromuriatic acid was identical with the dephlogisticated nitrous air of Priestley (nitrous oxide), led Davy to examine the gaseous products of this reaction more particularly. He had no difficulty in disproving the statement of the French chemist; but his observations, although accurate, led him to no definite conclusion. “It remains doubtful,” he says, “whether the gas consists simply of highly oxigenated muriatic acid and nitrogen, produced by the decomposition of nitric acid from the coalescing affinities of platina and muriatic acid for oxygen; or whether it is composed of a _peculiar_ gas, analogous to oxigenated muriatic acid and nitrogen, generated from some unknown affinities.” The real nature of the gas, which has also been considered by Lavoisier as a particular species, not hitherto described, was first established by Gay Lussac, when Davy had himself proved that “oxigenated muriatic acid” was a simple substance. In the second section the combinations and composition of nitrous oxide are investigated, and an account is given of its decomposition by combustible bodies, and a series of experiments are described which are now among the stock illustrations of the chemical lecture-room. As to its composition, he says, “taking the mean estimations from the most accurate experiments, we may conclude that 100 grains of the known ponderable matter of nitrous oxide consist of about 36·7 oxygen and 63·3 nitrogen”--no very great disparity from modern numbers, viz. 36·4 oxygen and 63·6 nitrogen. He concludes this section with a short review of the characteristic properties of the combinations of oxygen and nitrogen, among which he is led to class atmospheric air.