of British and Foreign Salt. 1 2 3 
ensued ; and in these, to the watery solution was added a 
mixture of equal parts of saturated solutions of carbonate of 
ammonia, and phosphate of soda. A precipitate more or less 
copious was produced, which was collected, dried at po° Fah- 
renheit, and weighed. 
(D. c. ) By direct experiments I had determined, that qo 
grains of this precipitate result from the decomposition of 10c 
grains of sulphate of magnesia, of such a degree of dryness, 
as to lose 44 grains out of 100, by exposure to a low red heat. 
Hence 100 grains of ammoniaco-magnesian phosphate indi- 
cate 111 grains of crystallized, on 62.2 of desiccated, sulphate 
of magnesia.* From the v/eight of the ammoniaco-magnesian 
phosphate, it is easy, therefore, to infer the proportion of sul- 
phate of magnesia in any mixture of the two earthy sulphates, 
(D. d.) It was possible, however, that in addition to the 
sulphates of lime and of magnesia, the quantity of which had 
been determined by the foregoing process, the specimen of 
salt under examination might contain also an alkaline sulphate. 
To decide this point, it was necessary to compare the amount 
of the acid, deducible from the weight of the sulphate of ba- 
rytes (C. b. ), with that which ought to exist in the sulphate 
of lime, and sulphate of magnesia actually found by experi- 
ment. But to make this comparison, some collateral experi- 
ments were previously necessary. 
(D. e.) By these experiments, I found that sulphate of 
lime prepared by double decomposition, then calcined in a low 
* The assumption that crystallized sulphate of magnesia contains only 44 per cent, 
of water, though it was correctly true with the specimen on which I operated, is be- 
low the average, which, I find from several experiments, is about one half the weight 
of the salt. Mr. Kirwan states the water of crystallization to be 53.6 in too grains; 
but this, I believe, a little exceeds the truth. 
MDCCCX. 
