-44 M. Berzelius on some Compounds 
and the water to contain each an equal quantity of oxygen, and 
the magnesia two-thirds of that quantity. Although this last 
result might have been expressed by a formula possessing great 
probability, and although it proceeded from our greatest master 
in the art of correct analysis, I thought it necessary to repeat 
the experiment before founding ufKlti it, with confidence, a cal- 
culation respecting the real composition of magnesia alba. Sim- 
ple as this analysis may appear, it cost me much more time and 
precaution than many other researches at first sight more diffi- 
cult ; and I did not succeed in finding what I conceive to be 
the true result, till after sixteen different trials. 
I began with precipitating a solution of muriate of magnesia 
by carbonate of potash, the mixture being kept for some time 
in a state of ebullition. At first, I left the muriate of magnesia 
slightly in excess, because I had reason to thinks that when the 
liquid contains an excess of alkali, the precipitate carries with it 
a small quantity of this alkali, which water is unable to extract. 
I collected the precipitate upon a filter, and washed it, till the 
water which passed no longer acted upon the uitrate of silver. 
When dried aud examined in the manner indicated above, it 
gave. Magnesia, 41.60; carbonic acid, 86.58; water, 21.82. 
As a repetition of the analysis produced exactly the same re- 
sult, there could be no inaccuracy in the experiment. But 
when I dissolved the caustic magnesia, procured by the analysis^ 
in nitric acid, and mixed with the solution a little nitrate of sil- 
ver, the liquid became clouded, and deposited a small quantity 
of muriate of silver. The precipitate which I had analysed, 
must therefore have contained a portion of the muriate of mag- 
nesia. Another precipitate, formed by employing the carbo- 
nate of potash in excess, gave for its composition. Magnesia, 
42.37 ; carbonic acid, 37.17 ; water, 20.46. The magnesia in 
this case contained no muriatic acid. 
I repeated these experiments by precipitating the magnesia 
from its sulphate. When the sulphate was in excess, the pre- 
cipitate seemed to contain. Magnesia, 42.24 ; carbonic acid, 
37.00 ; water, 20,76. This magnesia being dissolved in muri- 
atic acid, formed a pretty abundant precipitate, by adding mu- 
riate of barytes. When, in the mixture of carbonate of potash 
and sulphate of magnesia, the former was in excess, the precis 
