206 New Determination of the Proportions 
pour; but that, in passing the humid gases from the jar to the 
balloon glass, it is difficult entirely to avoid the condensation of 
a minute portion of aqueous vapour, when the sudden expan- 
sion of the transferred gas causes a reduction in its tempera- 
ture. 
M. Biot, to avoid Jong calculations and corrections, often uu- 
certain, has proposed to weigh the exhausted balloon both before 
and after the weighing of the gas, and to take the mean of these 
two determinations as the true weight of the balloon at the mo- 
ment in which it is weighed full of the elastic fluid. For this pro- 
ceeding to be accurate, the atmospheric changes must go on uni- 
formly, and the first and third weighings should be made at a 
distance of time nearly equal to the intermediate weighing. For. 
short intervals, this method is not exposed to the risk of any im- 
portant error; but when elastic fluids obtained by long and diffi- 
cult processes are operated upon, a considerable time may elapse 
between the first and the second weighing, and during the dou- 
ble of this interval, the uniformity of the variations may no longer 
have taken place. We therefore preferred taking the weight of 
the exhausted balloon immediately after each weighing of it when 
full of the required gas. A few minutes are sufficient to 
make the vacuum, and during this short interval, it is very rare _ 
that any change in the circumstances of the atmosphere can 
occur. 
The following observations relate to oxygen, hydrogen, azote, 
and carbonic acid: The oxygen was extracted from chlorate of 
potash, and was passed through a strong solution of caustic pot- 
ash, to extract any portion of carbonic acid with which it might 
be contaminated. ‘The method of obtaining hydrogen has been 
already described. The carbonic acid disengaged from white 
marble by means of nitric acid, was made to traverse a long co- 
lumn of powdered crystals of subcarbonate cf soda before it 
reached the vessel-that was to receive it. Lastly, the azote was 
obtained by decomposing ammonia by chlorine, and passing the 
gas through an acid and an alkaline solution alternately. 
The following are the results of the specific gravities of the 
gases according to our experiments, the gases being perfectly 
dry, and atmospheric air being = 1,000. 
CN EEA andthe cwdidlnimes mca tad dy tisia Wel One 
Ar BPD: peers sbi wasters 65919 «akon s EN 
CRAIG BING. riety tyarsesis fied Rh Cale Loe 
PERT Rabati avelvics wie a Sfeieivic a x.e eaten 
The gravities of the same gases, as determined by Messrs. 
Biot and Arrago, are as follows: 
Oxygen 
