376 On definite Proportions. 
a very correct basis for such computations; and-I have de- 
termined to repeat some of my former experiments with 
such accuracy as to entitle their results to the name of. 
Normal Analyses. Such however have been the difficulties 
which I have had to encounter, that [cannot yet venture to 
bring forward any of my analyses as perfectly entitled to 
this denomination. 
[The experiments on the sulphuret of lead, the oxides of 
lead, and the sulphuric acid, which were performed with 
this view, bave been already extracted from this communi- 
cation. Gillert.] 
Having inferred from my analyses that the acid of sul- 
phates and sulphites always contained either two or three 
times as much oxygen as the base of these salts, I have been 
induced to inquire if something of the same kind was nof 
also observable in other salts. The results of my investi- 
gation have confirmed this conjecture, and Ihave deduced 
from it a law which I shall partly demonstrate in this paper, 
and partly apply as already established. The law may be 
thus expressed : : 
In all neutral salts, the quantity of oxygen, which the 
acid contains, is an integer multiple of the quantity of oxy- 
gen in the base. Or,a little more generally, and, f believe, 
not less correctly : When two oxygenized substances saturate 
each other, the oxygen is always so proportioned, that its 
guantily in the substance which in the circle of the electrical 
column is attracted to the positive pole, is an integer multi- 
ple of its quantity in the other substance, which tends to- 
wards the negative pole. 
1. Correction of the analysis of the muriate of silver, and. 
of some others depending on it. 
I trust that 1 have demonstrated the superior accuracy of 
my analysis of the muriate of silver by the agreement of 
all my experiments with each other. But since it depends 
on a number of processes in which complete correctness was 
unattainable, | still entertained some doubt on the subject. 
Among many unsuccessful attempts to ascertain with greater 
accuracy the composition of the muriate of the protoxide 
of silver, I find only one which affords a tolerably satisfac- 
tory result. I prepared some pure sulphate of silver, and 
from this determined the constituent parts of the protoxide. 
When these were known, the quantity of muriatic acid in 
the muriate of the protoxide was casily deduced from them ; 
and this determination led to a number of corrections, which 
T shall here detail. 
SILVER 
