which depend 07i weak Affinities. ^47 
came alkaline ; and, after evaporation, left a residuum nf mag- 
nesia. I then found that the time, during which the boil- 
ing continues, has less influence than the quantity of water, in 
determining the decomposition of the carbonate which is mixed 
with the magnesia alba. After the water has produced a cer- 
tain effect, the remaining precipitate is but very slowly changed 
by continued boiling, A portion of magnesia alba, which had 
at first yielded a residuum of 43 per cent., after exposure to 
boiling water, gave 43.4 per cent. After a second exposure, it 
gave 44.4 ; after a third, 44.5 ; a fourth, 44.58 : the fifth and 
sixth trials produced no alteration, though the water, even 
when cold, continued to dissolve a portion of the precipitate 
suflicient to produce a white spot, when a few drops of it were 
evaporated in a platina spoon ; — a circumstance well worth the 
attention of those who engage in mineraiogical analyses. 
After having thus found a fixed point of composition, I ana^ 
iysed the substance, according to the method ^ready detailed, 
and obtained 
Magnesia, - 44.58 
Carbonic acid, - 35.70 
Water, - - 19.72 
We can now understand how the analyses of magnesia alba 
Iiave exhibited results so variable, and why the skilful Klaproth 
could obtain no more than 40 per cent, of magnesia, whilst 
Bergman and Kirwan obtained even 45 per cent., which pro- 
bably was in fact no more than 44.6 ; for in those times one 
third of a grain was regarded as a quantity, small enough to be 
rejected for the convenience of having round numbers. 
I remarked another circumstance relative to the analysis of 
magnesia alba. At the first application of fire, the powder 
produces a kind of ebullition, in which the vapours of the water 
carry with them a quantity of the earth, in the form of dust. 
When the experiment is made in a crucible, as the loss occa- 
sioned by this circumstance cannot be observed, it deceives the 
operator. In the apparatus which I used, this ebullition is 
instantly discovered, and checked by diminishing the temperature. 
By calculating the result of this analysis, it is found that 
the ojsygen of the magnesia is 17.26 ; that of the water, 17.47 ; 
^nd that of the carbonic acid, 25.73 : or one and a half times 
