“MODIFIED by COMPRESSION. 113 
~ ebullition: But as that did not often occur, I feldom took 
the trouble of this laft precaution. 
‘Ir was now of confequence to protect the tube, thus prepa- 
red, from being touched during the experiment, by any fub- 
ftance, above all, by the carbonate of lime, which might adhere 
to it, and thus confound the appreciation by weight. This was 
provided for as follows : The {mall tube (Fig. 23. 7%), with its 
pounded carbonate (4), and its cylinder of lump-chalk (7,), 
- was dropt into a large tube of porcelain (/ #, Fig. 24.). Upon 
this a fragment of porcelain (/), of fuch a fize as not to fall in 
between the tubes, was laid. Then a cylinder of chalk (7) was 
dreffed,fo as nearly to fit and fill up the infide of the large tube, 
one end of it being rudely cut into the form of acone. This mafs 
' bemg then introduced, with its cylindrical end downwards, 
was made to prefs upon the fragment of porcelain (/). I then 
dropped into the fpace (7), between the conical part of 
this mafs and the tube, a fet of fragments of chalk, of a fize 
beyond what could poffibly fall between the cylindrical part 
and the tube, and prefied them down with a blunt tool, by 
which the chalk being at the fame time crufhed and rammed 
into the angle, was forced into a mafs of fome folidity, which 
effectually prevented any thing from pafling between the large 
ma{s of chalk and the tube. In praétice, I have found this 
method always to anfwer, when done with care. _I covered 
the chalk, thus rammed, with a ftratum of pounded flint (0), 
and that again with pounded chalk (f) firmly rammed. In 
this manner, I filled the whole of the large tube with alternate 
layers of filex and chalk; the muzzle being always occu- 
' pied with chalk, which was eafily prefled into a mafs of to- 
lerable firmnefs, and, fuffering no change in very low heats, 
excluded the fufible metal in the firft ftages of the experiment. 
Tue large tube, thus filled, was placed in the cradle, fome- 
times with the muzzle upwards, and fometimes the reverfe. I 
Vou. VI.—P. I. ip have 
