’ 
of the Constituents of Water, &e. 207 
OXYZen ceceeeeeseseceseseers 1.10359 
Hydrogen ...seeeeeeseeeeeeess 0.07321 
Carbonic acid ..esseeeeseeeees L019 
Azote .cutslicssiededeseseses = 0.969 
If we take the aboye proportions in weight of the elements of 
_ water, and take the density of oxygen as obtained by our experi- 
ments at 1.1626, the specific gravity of the hydrogen will turn 
out to be 0.0688, but by direct experiment it gave us 0.0687. 
It appears, then, that the greatest difference between our results 
and those of Messrs. Biot and Arrago relates to the density of hy- 
drogen ; which confirms what we have said above on the cause 
of this difference. The increase in the number which we have 
given for carbonic acid, though small, is sufficient, however, to 
jnfluence in a sensible manner the number expressing the den- 
sity of the vapour of carbon on account of its levity; and it ap- 
pears to us to accord better with the results of the analysis of ve- 
getable substances. Lastly, the density of azote, calculated from 
our observations, approaches more nearly to that which is de- 
duced from the composition of the nitrates. 
For the convenience of those who engage in analytical re- 
searches, we have colleeted in the following table the densities 
and proportions by weight of several compounds calculated from 
the bases above given. These numbers should be preferred to 
those that are obtained by direct analysis, which hardly ever bear 
the same degree of approximation that may be obtained by in- 
ference from the above-mentioned data.—Before we conclude, 
we may observe that our new results differ but little from those 
which are given in an anonymous memoir, inserted in the Annals 
of Philosophy for November 1815 and February 1816; but 
the English author has added no observations, and the hypo- 
theses which have served him to correct the established numbers 
being absolutely gratuitous or false, no confidence can be placed 
in his results, 
Table 
