wliicli depend on Weak Affinities. 69 
magnesia iindissolved. This last remainder, when collected up- 
on a filter, w^ashed, dried and heated to redness, weighed 15.74 
parts. The alkaline liquid existing in the waters of lavage, 
was saturated with muriatic acid, in an apparatus calculated to 
avoid all loss from effervescence. Being afterwards evaporated, and 
exposed to the continued action of heat, it left S9.48 parts of mu- 
riate of potash. Water, when added to this mass, formed a turbid 
solution of it, from which a drop or two of caustic potash still 
precipitated a small quantity of magnesia, which, when separat- 
ed and heated to redness, was found to weigh 0.25 parts. Con- 
sequently, the whole of the magnesia amounted to 15.99 parts. 
.But as these 0.25 parts of magnesia were principally muriate of 
magnesia, a deduction of 0.58 must be made from the 29.48 
parts of muriate of potash ; the exact quantity of v/hich salt is, 
Bence, 28.9 parts, containing 18.28 parts of pure potash. 
Wherefore the analysis of this double salt gave. 
Potash, 18.28 
Magnesia, 15.99 
Carbonic Acid . 84.49 
Water, 31.24 
But in order to appreciate this result, we must know not only 
the composition of magnesia, but also that of the crystallised car- 
bonate of magnesia, and of the crystallised bicarbonate of potash, 
with as much certainty as we know tliat of carbonic acid or of 
potash. In some experiments, wdiich I published six years ago, 
upon the composition of msgnesia, I liad found that it contains 
from 38.8 to 39.8 per cent, of oxygen ; but, as a difference of 
1 per cent, is too great for being attributed to an ordinary error 
of observation, I determined to examine the point anew, and to 
try, if possible, to render it still more exact. The usual method 
is very simple. We dissolve a given weight of pure magnesia 
in sulphuric acid, likewise pure, but diluted ; w'e evaporate the 
.salt obtained, heat it to redness, and weigh it. 
As on several occasions, I had found, in analysing minerals^, 
that magnesia exerts a very strong affinity to silica, wliich tlie 
magnesia, recovered from the sulphuric acid by muriatic acid, 
and evaporated, always deposites before drying, in a gelatinou.s 
* Comprehending a small loss, which could not he avoided. 
