which depend on Weak Affinities. S49 
adheres to the salt by a slight affinity ; and, in most cases, al- 
lows itself to be disengaged by a heat very little exceeding that 
of -f- 100°. In every case it is disengaged long before the acid. 
Now, in the analysis of magnesia alba, it is to be remarked, 
that in the first place a certain quantity of water disengages it- 
self, as if it were water of crystallisation; but that afterwards 
the carbonic acid and the water are disengaged together, and 
that the last portion of magnesia alba, though developed at a 
red heat, still deposites a drop of water, whilst passing through 
the slender tube which conducts it to the cold recipient. 
This circumstance, which I have fully demonstrated by experi- 
ments made on purpose to verify k, proves that the water is re- 
tained by the magnesia alba, with a force greater than that 
which retains the water of crystallisation, even in those salts 
which are distinguished by their strong affinity for water. But 
this affinity can be no other than that by which a part of the 
magnesia, as a base or substance positively electrified, is com- 
bined with water, as an acid, or at least a body negatively elec- 
trified, in the form of a hydrate. But in this case, the mag- 
nesia alba must be composed of hydrate of magnesia and of 
carbonate of magnesia ; nor is it difficult to conceive that the 
last portions of the weaker negatively electrified substance may 
not be expelled by heat, before those of the stronger, since each 
is combined to its base separately. 
In magnesia alba, fths of the magnesia combines with the car- 
bonic acid to form a carbonate of magnesia ; the remaining Jth 
forms the hydrate. But as the oxygen of the water is equal to 
that contained by the whole of the magnesia, it may be asked 
whether the water is all combined with the hydrate, or partly 
supplies the carbonate with water of crystallisation. No posi- 
tive answer can be returned to this question, but it is very pro- 
bable that the carbonate contsiins no water ; Jirsty Because this 
carbonate is often found, in a natural state, perfectly dry ; and 
because the water, in magnesia alba, is not sufficient to furnish 
the water of crystallisation necessary for this salt, in an isolated 
state : Secondly.^ As the water, in this case, performs the func- 
tion of an acid, (a function more active than that of water of 
crystallisation), it seems highly probable that, in order to coun- 
terbalance the stronger agent, the carbonic acid, the atoms of 
