86 EFFECTS of HEAT 
the whole cradle ; when any fpace remained, it was filled up 
by a piece of chalk dreffed for the purpofe. (Fig. 4. reprefents 
the cradle filled, as juft defcribed). 
Tuincs being thus prepared, the gun-barrel, placed erect 
with its muzzle upwards, was half filled with the liquid fufible 
metal. The cradle was then introduced into the barrel, and 
plunged to the bottom of the liquid, fo that the carbonate was 
placed very near the breech, (as reprefented in fig. 5, the fu- 
fible metal ftanding at 0). The air-tube (B) being placed fo 
as to enter the liquid with its muzzle downwards, retained 
great part of the air it originally contained, though fome of it 
might be driven off by the heat, fo as to efcape through the 
liquid. The metal being now allowed to cool, and to fix round 
the cradle and ramrod, the air remaining in the air-tube was 
effectually confined, and all was held faft. The barrel being 
then filled to the brim with fufible metal, the apparatus was 
ready for the application of heat to the breech, (as fhewn in 
fig. 6.) 
In the experiments made at this time, I ufed a fquare brick 
furnace (figs. 7 and 8), having a muffle (7 s) traverfing it ho- 
rizontally and open at both ends. This muffle being fupported 
in the middle by a very flender prop, was expofed to fire from 
below, as well as all round. The barrel was placed in the 
muffle, with its breech in the hotteft part, and the end next 
the muzzle projecting beyond the furnace, and furrounded with 
cloths which were drenched with water from time to time. 
(This arrangement is fhewn in fig. 7). In this fituation, the 
fufible metal furrounding the cradle being melted, the air 
contained in the air-tube would of courfe feek the higheft po- 
fition, and its firft place in the air-tube would be occupied 
by fufible metal. (In fig. 6., the new pofition of the air is 
fhewn at pq). 
AT 
