Rifle shells may be extemporised by having little tin tubes the length of the conical bullet and the size of the gun nipple made. These, when their ends have been plugged with wood, are placed one by one in the mould, held upright by a bit of very fine brass wire, and the lead cast round them, so that the lead at the base of the bullet may extend beyond the end of the tube and cover it. The thin end of the tube should project just a trifle beyond the apex of the cone, as a rest for the cap. The wooden plug at this end is now taken out; the tube filled with the best sporting powder, and a strong, well made percussion-cap put on the tube, and secured there with strong varnish, sealing-wax, or pitch. The loading of a breech-loader with these is accomplished with no danger, but with a muzzle-loader the very greatest caution is required. The end of the ramrod or loading stick should be very deeply countersunk, in order to take all pressure off the cap; and even with this precaution it is well to make use of an overhanging branch of a tree to place the rod against, whilst the rifle is thrust muzzle upwards until the ball is home. Shells somewhat on this principle were first brought into notice by the late General Jacob, of the Scinde Irregular Horse. They are tremendously destructive when skilfully used: destroying large animals by exploding in them, and blowing up magazines of gunpowder at all but incredible distances; but we have known them explode _outside_ the elephant and other large pachyderms, thus failing completely in the object they were used for. Mr. Metford's improvement on the Jacob shell is worthy of remark. Finding that it did not always explode, he mixed equal parts of chlorate of potass and sulphate of antimony; the two can be mixed on a plate with a _bone_ paper-cutter or a quill pen. The more they are mixed, the more sensitive is their detonating power. The bullets are moulded with a hole from point nearly to base, as for Jacob's shells, but no copper tube is used. The powder is filled in with a quill to the top, and settled down by a few taps of the base of the bullet on a table, and the end is then stopped with wax. But it is very questionable whether, in close encounter with large animals, it is not better, after all, to rely on the more certain effects of heavy balls of ordinary construction, with strong charges of powder behind them. With the numerous improvements in rifle shells we cannot deal, as many of them are too complex in their component parts for a wandering hunter or explorer to be able to imitate successfully.