5i< On Vegetable and Animal Analysis . 
that we fill successively with gas, two and sometimes three 
flasks of the same capacity; that these gases are identi 
cal, and always proceed from one and the same weight of 
materials. 
We might add, that the exactness of any analysis con 
sists rather in the accuracy of the instruments, and of 
the methods which we employ, than in the quantity of 
matter upon which we operate. The an ilysis of the air 
is more exact than any analysis of the salts, and yet it is 
performed upon two or three hundred times less matter 
than the latter. This is because in the former, where we 
judge of weights by volumes which are very considera- 
ble, the errors which we may commit are perhaps one 
thousand or twelve hundred times less perceptible than 
in the latter, where we are deprived of this resource. 
Now as we transform into gas the substances which we 
analyse, we bring our analyses not only to the certainty 
of the common mineral analyses, but to that of the most 
precise mineral analyses : more particularly as we collect 
at least a litre of gas, and as we find even in our way of 
proceeding the proof of an extreme exactitude and of the 
most trifling errors. 
We have already methodically analysed, with all the 
precautions just mentioned, sixteen vegetable substances; 
viz. the oxalic, tartarous, mucous, citric, and acetic acids ; 
turpentine in resin ; copal, wax, olive oil ; sugar, gum, 
starch, sugar of milk, oak and ash wood, and the crystal- 
lizable principle of manna. The results whieh we ob- 
tained seem to us to be of the first rate importance, for 
they led to three very remarkable laws to which the com- 
position of vegetables is subjected, and which may be 
thus expressed : 
First Law. 
A vegetable substance is always acid when the oxy 
