664 PHILOSOPHICAL TRANSACTIONS. [ANNO 1800. 



opinion, a cast iron case was constructed, with a cylindrical chamber, of equal 

 length and diameter, calculated to hold 3± oz. Troy of the mercurial powder. 

 The case, being firmly screwed together, was charged through its vent-hole, and 

 introduced into a 12-pounder carronade, the bore of which it exactly fitted. The 

 powder was then inflamed, with proper precautions. The gun remained entire, 

 but the case divided ; the portion forming the upper surface of the chamber, was 

 expelled in one mass; that adjoining the breech, which constituted the rest of the 

 chamber, was cracked in every direction, and in part crumbled; yet it was so 

 wedged into some indentations which the explosion had made in the sides of the 

 piece, that the fragments were not removed without great labour. 



Exper. 2. Another cast iron case was prepared, of the same size as the former, 

 with a chamber also cylindrical, but wrought in a transverse direction, and of a 

 greater length than diameter; the thickness of metal at each extremity not being 

 more than ± of an inch. This case was filled with nearly 5 oz. Troy of the mer- 

 curial powder, and placed in the same carronade. Three 12-pound shot were next 

 introduced, and brought into close contact with the upper surface of the case, as 

 well as with each other. The gun a 2d time withstood the explosion: the case was 

 divided across the middle of the chamber, into 2 equal parts; that adjoining the 

 breech was, as in the former experiment, much flawed, and left immoveable; that 

 nearest to the muzzle was also much flawed, but driven out with the shot. All the 

 3 shot were broken; the 2 lower being divided into several pieces, and the upper 

 1 cracked through the centre. 



The report was so feeble, in both experiments, that an inattentive person, I am 

 confident, would not have heard it at the distance of 200 yards. 



Exper. 3. It was found so difficult to extract the fragments of the case remain- 

 ing in the carronade, after the last experiment, that a channel was drilled through 

 them, to the vent-hole of the piece. It was then charged with 6 oz. Troy of the 

 mercurial powder, made up as a cartridge, which did not occupy above 4- of the 

 diameter of the bore. A wad was placed over the powder, dry sand superadded, to 

 fill all vacuities, and the gun filled to the muzzle with 2 12-pound shot. A block 

 of wood was set at a small distance, to receive the impression of the shot, and the 

 powder was inflamed as usual. The carronade still resisted. One of the shot was 

 split into 2 pieces; and the block of wood was driven to a considerable distance, but 

 not penetrated by the shot above the depth of 1 inch. The report was somewhat 

 louder than the former ones. In all the 3 instances, a considerable recoil evidently 

 took place. I presume therefore, that in the first experiment related in § 5, there 

 must have been a recoil, though too trifling to be observed; and in the instances 

 where the gun and the proof were burst, it was not so much to be expected. 



Exper. 4. Finding that the carronade, from the great comparative size of its 

 bore to that of its length, required a larger quantity of mercurial powder to burst 

 it than we were provided with, we charged a -L-pounder swivel with J-fr oz. avoir- 

 dupois of the mercurial powder, (the service charge of gunpowder being 3 oz.) 



