Di Daubeny on separating Lime from Magnesia. 121 
same quantity of magnesia as it did on the first; but a little 
consideration will explain to us the cause of this occurrence. 
When the ammoniacal salt is added to a mixed solution of 
the two earths, it would seem that the magnesia is divided be- 
tween the ammonia and the lime, somewhat in proportion to the 
degree of its affinity to the two, the larger portion being held in 
solution with the former, in the state of a triple salt, the smaller 
carried down with the latter as a simple carbonate. 
When, however, the precipitate is redissolved in an acid 
and submitted a second time to the action of the same re-agent, 
it is evident that only a small proportion of the magnesia exist- 
ing in the compound will be thrown down, as the carbonate of 
ammonia will exert the same affinity as before, and consequent- 
ly form, with the larger portion, a triple salt soluble in water. 
From the facts above stated, it seems to follow, that carbo- 
nate of ammonia is not to be depended on as a complete preci- 
pitate of lime. Let us now go on to consider, whether phos'- 
phate of soda can be employed to determine the proportion of 
magnesia. 
1 dissolved 1 gr. of crystallised sulphate of magnesia in 8000 
gr. of water, and found that the addition of carbonate of am- 
monia and phosphate of soda produced a very sensible precipi- 
tate, in a portion of the liquor set aside in a watch-glass. In- 
deed, this salt, even when diluted with 12,000 gr. of water, was 
detected in this manner by a slight cloudiness. When a weak 
solution of sulphate of lime was substituted for pure vvater, the 
test proved less delicate ; but even then it detected 2 gr. of sul- 
phate of magnesia in a pint of the fluid. 
Now, crystallised sulphate of magnesia, according to the latest 
experiments -f*, (which I am happy to find confirm those I have 
myself had occasion to make), consist of about 
* It would seem to follow in theory, that, by adding a larger quantity of the 
carbonate of ammonia, the whole of the magnesia might be retained in solution in 
the first instance. But it must be recollected, that, by increasing the quantity of 
the re-agent, more of the lime would probably be held dissolved. 
-f- M. Longchamp, in his paper, states the proportions of water lower, and the 
base higher, than this ; but Gay Lussac has since contradicted him, and given an 
estimate that agrees very nearly with that above. 
