The atmosphere is subject to changes of various kinds, and may be considered not only as a solvent, but a repository for different foreign bodies. Electricity, an agent so essential in the economy of nature, has its ends, its uses; and while, no doubt, it unites hydrogen with oxygen, in the most elevated regions of the air, and forms water, it may act under particular circumstances to produce a union of azote and oxygen so as to generate nitric acid. Dr. Priestley, (_Transactions of the American Philosophical Society_,) detected nitric acid in snow. But of all atmospheric phenomena, the formation of meteorolites, or meteoric stones, is the most wonderful. If they be really formed in the atmosphere, there can be no doubt, that the elementary principles which compose them must exist in it; and that the phenomenon denominated meteoric, in such cases, is no other than the operating cause, by which meteoric stones are generated.[15] Animal substances furnish azote, as it is one of their constituent parts; and in the act of its separation, by uniting with oxygen, principally furnished by the air, it forms nitric acid; which, attaching itself to the alkali of the vegetable matter, or the lime usually added to nitre beds, or to other salifiable bases, forms either nitrate of potassa, nitrate of lime, or a nitrate of the particular base. The lixiviation of the nitrous substances, and the use of wood ashes, or potash itself, will produce saltpetre.[16] Brongniart has given the following process for purifying or refining saltpetre: Pulverize the impure nitre, and wash it three times in cold water, in the proportion of 35 lbs. of water, to 100 lbs. of the salt, taking care to pour off the water before another portion is added. These washings separate the greater part of the muriate of soda, and the deliquescent salts, such as nitrate of lime. When thus washed, the nitre is to be dissolved in half its weight boiling water. On cooling, the salt begins to crystallize, and, by agitating the liquid during the process, minute crystals are obtained. These crystals when dried are to be washed in 5 lbs. of cold water for every 100 lbs. of the salt, and then dried in a temperature of forty-five degrees. In India, where nitrate of lime also occurs, but in situations different from those in the United States, the natives extract the saltpetre by a process similar to that we have described. They refine it by solution in water, evaporation, and crystallization. In France, the potash of commerce is used; and the nitrates which are decomposed, are those principally of lime and magnesia. According to the analysis of M. Pelletier, and the experiments of professor Vaizo, in 1781, they found the calcareous earth of the cave at Naples, to contain forty or forty-two to the hundred, of nitrate of potassa. (See _Annales de Chimie_, tome 23.)