In my first experiments on the production of nitrate of ammoniac, I endeavoured to ascertain its composition by decompounding carbonate of ammoniac by nitric acid; and in making for this purpose, the analysis of carbonate of ammoniac, I discovered that there existed many varieties of this salt, containing very different proportions of carbonic acid, alkali, and water; the carbonic acid and water being superabundant in it, in proportion as the temperature of its formation was low, and the alkali in proportion as it was high: and not only that a different salt was formed at every different temperature, but likewise that the difference in them was so great, that the carbonate of ammoniac formed at 300° contained more than 50 per cent alkali, whilst that produced at 60° contained only 20.[73] I found 210 grains of carbonate of ammoniac, which from comparison with other salts previously analised, I suspected to contain about 20 or 21 per cent alkali, saturated by 200 grains of nitric acid of 1,504. But though the carbonate was dissolved in much water, still, from the smell of the carbonic acid generated, I suspect that a small portion of the nitric acid was dissolved, and carried off by it. The solution, evaporated at about 200°, and afterwards exposed to a temperature below 300°, gave 232 grains of compact salt. But reasoning from the quantity of acid in 200 grains of nitric acid of 1,504, it ought to have given 245. Consequently 13 were lost by evaporation; and this loss agrees with that in the other experiments. [73] A particular account of the experiments from which these facts were deduced, was printed in September, and will appear in the first volume of the _Researches_. VI. _Decomposition of Sulphate of Ammoniac by Nitre._ As a cheap mode of obtaining nitrate of ammoniac, Dr. BEDDOES proposed to decompose nitre by sulphate of ammoniac, which is a well known article of commerce. From synthesis of sulphate of ammoniac, compared with analysis made in August 1799,[74] I concluded that 100 grains of prismatic salt were composed of about 18 grains ammoniac, 44 acid, and 38 water; and supposing 100 grains of nitre to contain 50 acid, 100 grains of sulphate of ammoniac will require for their decomposition 134 grains of nitre, and form 90,9 grains of compact nitrate of ammoniac. [74] And which will be published, with an account of its perfect decomposition at a high temperature, in the _Researches_.