According to Berthollet, the weight of the nitrogene in ammoniac is to that of the hydrogene as 121 to 29.[62] The difference between this estimation and mine is so small as to be almost unworthy of notice, and arises most probably from the slight difference between the accounts of Lavoisier and Monge, of the composition of water, and the different weights assigned to the gases employed. [62] Journal de Physique, 1786, t. 2, 177. We may then conclude, that 100 grains of ammoniac are composed of about 80 nitrogene, and 20 hydrogene. The decomposition of ammoniac by heat, as well as by the electric spark, was first discovered by Priestley. In an experiment[63] when aëriform ammoniac was sent through a heated tube from a caustic solution of ammoniac in water, this great discoverer observed that an inflammable gas was produced, though in no great quantity, and that a fluid blackened by matter, probably carbonaceous, likewise came over. [63] Phil. Trans. vol. 79, page 294. In my experiments the whole of the ammoniac appeared to be decomposed; the quantity of gas generated was immense, and not clouded, as is usually the case with gases generated at high temperatures. It is possible, that the larger quantity of water carried over in his experiment, by its strong attraction for ammoniac in the aëriform state, might have, in some measure, retarded the decomposition. It is however, more probable to suppose, that a fissure existed in the earthen tube he employed, through which a certain quantity of gas escaped, and coaly matter entered. Priestley found that the metallic oxides when strongly heated, decomposed ammoniac, the metal being revivified and water and nitrogene produced.[64] The estimations of the composition of ammoniac that may be deduced from his experiments on the oxide of lead, differ very little from those already detailed. [64] Vol. 2, page 398. II. _Specific gravity of Ammoniac._ From the great solubility of ammoniac in water, it is difficult to ascertain its specific gravity in the same manner as that of a gas combinable to no great extent with that fluid. It is impossible to prevent the existence of a small quantity of solution of ammoniac in the mercurial airholder,[65] or apparatus containing the gas; and during the diminution of the pressure of the atmosphere on this solution,[66] a certain quantity of gas is liberated from it, and hence a source of error. [65] Ammoniac generated at a temperature above that of the atmosphere, always deposits ammoniacal solution during its reduction to the common temperature. [66] By the introduction of aëriform ammoniac into the exhausted globe. To ascertain, then, the weight of ammoniac, I employed an apparatus similar to that used for the absorption of nitrous gas by nitric acid. 50 cubic inches of gas were collected in the mercurial airholder, from the decomposition of muriate of ammoniac by lime; thermometer being 58°, and barometer 29,6.