86 EFFECTS of HEAT 
the following general view: If, we take a, hollow tube or bar- 
rel (AD, fig. 1.) clofed at one end, and open at the other, of 
one foot or more in length; it is evident, that by introducing 
one end into a furnace, we can apply to it as great heat as 
art can produce, while the other end is kept cool, or, if necef- 
fary, expofed to extreme cold. If, then, the fubftance which 
we mean to fubjec& to the combined action of heat and pref- 
fure, be introduced into the breech or clofed end of the barrel 
(CD), and if the middle part be filled with fome refractory 
fubftance, leaving a {mall empty {pace at the muzzle (AB), we 
can apply heat to the muzzle, while the breech containing the 
fubjeét of experiment, is kept cool, and thus clofe the barrel by 
- any of the numerous modes which heat affords, from the weld- 
ing of iron to the melting of fealing-wax. Things being then 
reverfed, and the breech put into the furnace, a heat of any 
required intenfity may be applied to the fubject of experiment, 
now in a ftate of conftraint. 
My firft application of this fcheme was carried on with 
a common gun-barrel, cut off at the touch-hole, and welded 
very ftrongly at the breech by means of a plug of iron. 
Into it I introduced the carbonate, previoufly rammed into a 
cartridge of paper or pafteboard, in order to protect it from the 
iron, by which, in fome former trials, the fubjet of experiment . 
had been contaminated throughout during the action of heat. I 
then rammed the reft of the barrel full of pounded clay, previoul= 
ly baked in a ftrong heat, and I had the muzzle clofed like the 
breech, by a plug of iron welded upon it in a common forge ; 
the reft of the barrel being kept cold during this operation, 
by means of wet cloths. The breech of the barrel was 
then ‘introduced horizontally into a common muffle, heated to 
about 25° of WepGwoop. To the muzzle a rope was fixed, 
in fuch a manner, that the barrel could be withdrawn with- 
out 
