MODIFIED ly COMPRESSION. 99 
a greater or a'lef§ degree, in all thefe experiments. The que- 
ftion naturally fuggefts itfelf, What has become. of the car- 
bonic acid, feparated in thefe partial calcinations from the 
earthy ‘bafis? Has it penetrated) the veffel, and efcaped en- 
tirely, or has it been retained within it in a gafeous, but 
highly compreffed ftate? It occurred to me, that this que- 
ftion might ‘be eafily refolved, by weighing the veffel before 
- and after the action of heat upon the carbonate. 
Wiru iron, a conftant and ‘inappreciable fource of irregu- 
larity exifted in the oxidation of the barrel. But with porce- 
lain the thing was eafy ; and I put it in practice in all my ex« 
periments with this material, which were made after the que- 
ftion had occurred to me. The'tube was weighed as foon as its 
muzzle was clofed, and again, after the breech had been expo- 
fed to the fire ; taking care, im both cafes, to allow'all to cool. 
In every cafe, I found fome lofs of weight, proving, that even 
in the beft experiments, the tubes were penetrated to a certain 
degree. Inext withed to try if any of the carbonic acid fepara- 
ted, remained within the tube in a gafeous form; and in that 
view, I wrapt the tube, which had juft been weighed, in a 
fheet of paper, and placed it, fo furrounded, on the fcale of 
the balance. As foon as its weight was afcertained, I broke 
the tube by a fmart blow, and then replaced upon the fcale 
the paper containing all the fragments. In thofe experiments, 
in which entire .calcination had taken place, the weight was 
found not to be changed, for all the carbonic acid had already 
efcaped during the action of heat. But in the good refults, I 
always found that a lofs of weight was the confequence of break- 
ing the tube. 
TueEse facts prove, that both caufes of calcination had ope- 
rated in the porcelain tubes ; that, in the cafes of {mall lof, 
part of the carbonic acid had efcaped through the veffel, and 
that part had been retained within it. I had in view methods’ 
N 2 by 
