Experiment 1.--A piece about 5 inches long was cut off the breech-end of a common musket barrel. It was screwed at the part cut, and another plug fitted, so as to have two plugs, one at each end, leaving an internal space of about 3 inches. A percussion nipple was screwed into the end of one of these plugs. This being arranged, one of the plugs was turned out, and one drachm of gunpowder introduced. The plug was replaced, and the powder fired by putting a copper cap on the nipple, and striking it with a hammer. The whole force of the powder escaped at the hole in the nipple. Two, three, four, five, and six drachms were successively introduced, and fired in the same manner, without bursting or injuring the piece of barrel. At last, seven drachms forced out one end, in consequence of the screw having been carelessly fitted. This defect being repaired, Mr. Marsh, of Woolwich, repeatedly fired it with five drachms, merely holding it with a towel in his left hand, and firing it with a blow of a hammer. Six drachms of powder is the full service charge for a flint musket, and four drachms of a percussion musket; yet this immense pressure can be resisted by a cylinder of iron not more than one quarter of an inch thick, and not iron of the best quality. Experiment 2.--A good musket barrel had a cylinder of brass, three inches long, turned to fit the muzzle, and soldered in, so as to close it air-tight. The plug, or breech-screw, was removed, and a felt wad was pushed in with a short piece of wood, marked to the exact depth the charge would occupy, to prevent the ball rolling forward. A musket ball was then dropped in, and a cartridge, containing three drachms of powder, was introduced. The breech being screwed in, left the barrel loaded. It was fired by a percussion tube, but there was no report. On removing the breech-screw, the ball was found to be flattened. A repetition of this experiment, with four drachms, produced a similar result, but the ball was rather more flattened. With five drachms, the ball was perfectly round and uninjured. Six drachms burst the barrel close under the bayonet stud; the ball escaped through the opening, disfigured, but fell close to the barrel. In these experiments the barrel always advanced, instead of recoiling, as usual.