^46 M. Berzelius on some Compounds 
ed strong alkaline properties. By boiling, it became turbid, 
and deposited a great quantity of magnesia. Now, it is evi- 
dent, that Avhen a salt, having magnesia for its base, is precipi- 
tated by an alkaline carbonate, the salt is decomposed into two 
portions. The first portion is entirely deposited ; but the se- 
cond, having gained from the former a superior quantity of car- 
bonic acid, is converted into a carbonate of magnesia, soluble in 
water. As the quantity which is dissolved of this latter portion 
must vary according to the quantity and temperature of the wa- 
ter employed, the composition of the precipitated substance will 
be different, on different occasions. What follows, will confirm 
this. Another portion of that precipitated in a cold state, al- 
ready analysed, being again mixed with water, at the tempera- 
ture of -f- 18° centigr. and agitated for some time, the water be- 
came alkaline, and gave an abundant precipitate by boiling. 
This last precipitate, after exposure to heat, left 38 per cent, of 
caustic magnesia, 1.6 per cent, more than was formerly obtained. 
A similar treatment of this last precipitate produced a similar 
result ; and, as of'ten as the operation of washing was repeated, 
the proportion of caustic magnesia continued to increase. 
From these facts I conclude, that when a neutral salt, having 
magnesia for its base, is decomposed by an alkaline carbonate, 
there results a corresponding carbonate of magnesia, which is 
partially decomposed by the presence of water, so as to form a 
carbonate, soluble in water by an excess of carbonic acid ; whilst 
one part of the carbonate is converted into magnesia alba, and 
another part is deposited without decomposition, — both in vari- 
able proportions, which depend upon the variation of the cir- 
cumstances in which the decomposition takes place. By each 
increase of the temperature and quantity of water, the propor- 
tion of magnesia alba is likewise increased, till nothing but itself 
is deposited, and consequently till the precipitate no longer 
changes its composition, by exposure to fresh quantities of boil- 
ing water. Granting that this reasoning is well founded, it on- 
ly remained for me to seek the maximum of this decomposition 
produced by water. Having, therefore, taken a quantity of 
magnesia alba, precipitated by long continued boiling with car- 
bonate of potash in excess, I boiled it with fresh quantities of 
wat^r; Its quantity was sensibly diminished;, the water be- 
