71 
which depend on Weak Affinities. 
peated the experiment four times with every necessary precau- 
tion, I began to suspect that the sulphate of barytes had not ' 
been sufficiently washed. I had continued the purification, till 
the \vater which passed had entirely ceased to become turbid 
on the addition of sulphuric acid. Upon repeating the experi- 
ment once more, I took some drops of the water of lavage no 
longer containing barytes, and evaporated them in a spoon of 
platina ; — a method which I sometimes follow for discovering 
whether a precipitate is sufficiently washed. There remained a vi- 
sible spot. Having next tried the waters of lavage by nitrate of 
silver, I found muriatic acid in them. I continued to wash the 
precipitate with boiling water added each half hour during four 
days, and the action of the nitrate of silver was constantly the 
same as if the water were dissolving a muriate very sparingly 
soluble. I evaporated a great quantity of the water, and ob- 
tained a small residuum, which, when exposed to the continued 
action of heat, lost its solubility, and shewed itself to be magnesia. 
A small quantity of muriate of magnesia is, therefore, through 
chemical affinity, deposited along with the sulphate of barytes, 
from which it cannot be entirely separated by water. Having ascer- 
tained this fact, I took the precipitated sulphate from the filter, 
digested it in muriatic acid, then filtered and washed it anew. 
The excess of muriatic acid was soon removed, and the liquid 
which passed afterwards continued to act upon the nitrate of sil- 
ver, in the same weak degree as before the addition of the mu- 
riatic acid. Having finally dried it, and heated it to redness, I 
found its weight to be 19*44 grains, which is 0.236L grains more 
than I ought to have obtained. This proves, at all events, that, 
by the analytical method in question, no result worthy of confi- 
dence can be obtained, 
I have several times observed, that when one mixes the sul- 
phate of a weak base with muriate of barytes, a portion of this 
base is precipitated along with the barytes, combined in such a 
way as no longer to be separable even by an excess of acid. If, 
for example, we mix the sulphate of iron or of copper with mu- 
riate of barytes, the sulphate of barytes becomes, by calcination, 
reddish-yellow in the first case, and greenish-yellow in the last ; 
with the sulphate of cobalt it becomes reddish. In all these 
^cases, we must begin with acting upon the oxide, if we vish to 
