From these experiments it follows clearly, that nitric acid is a compound of nitrous gas and oxygen. The nature of nitrous gas itself Lavoisier did not succeed in ascertaining. It passed with him for a simple substance; but what he did ascertain enabled him to explain the action of nitric acid on metals. When nitric acid is poured upon a metal which it is capable of dissolving, copper for example, or mercury, the oxygen of the acid unites to the metal, and converts into an oxide, while the nitrous gas, the other constituent of the acid, makes its escape in the gaseous form. The oxide combines with and is dissolved by another portion of the acid which escapes decomposition. It was discovered by Dr. Priestley, that when nitrous gas and oxygen gas are mixed together in certain proportions, they instantly unite, and are converted into nitrous acid. If this mixture be made over water, the volume of the gases is instantly diminished, because the nitrous acid formed loses its elasticity, and is absorbed by the water. When nitrous gas is mixed with air containing oxygen gas, the diminution of volume after mixture is greater the more oxygen gas is present in the air. This induced Dr. Priestley to employ nitrous gas as a test of the purity of common air. He mixed together equal volumes of the nitrous gas and air to be examined, and he judged of the purity of the air by the degree of condensation: the greater the diminution of bulk, the greater did he consider the proportion of oxygen in the air under examination to be. This method of proceeding was immediately adopted by chemists and physicians; but there was a want of uniformity in the mode of proceeding, and a considerable diversity in the results. M. Lavoisier endeavoured to improve the process, in a paper inserted in the Memoirs of the Academy, for 1782; but his method did not answer the purpose intended: it was Mr. Cavendish that first pointed out an accurate mode of testing air by means of nitrous gas, and who showed that the proportions of oxygen and azotic gas in common air are invariable.