OF GUN-METAlI IZ§

that is rammed upon the powder, and by pressing earthclose round the tompion and the shell; the initial ve-locity will be thus increased.

3. By increasing the size of the chamber.

80. Stone mortars are used for throwing troops, work,men, &c. at sieges into confusion : the chamber is in formof a truncated cone, whose least base is towards the vent;they are sired with small quantities of powder, as with largecharges the stones would spread so much that few or nonewould have effect. Their diameter lhould not be less than15 inches, nor the chafe less than a diameter and a half mlength, that it mav contain a sufficient quantity of stones;by increasing the diameter and length of the chafe, theywould contain more stones, and consequently be more de-structive ; but then from their weight they would be very dis--ficult to transport.

CHAP, VI.

Of the Thickness of Metal in Fire-arms.

*-ts

81. i H E R E are two extremes to be equally avoidedin determining the thicknesses of metal in fire-arms ; the oneis, making them too thin and light; the other, too strongand heavy : the danger arising from the first is too obvious toneed insisting on; the second is comprized within certainlimits. The metal proportionally distributed throughout thewhole length of the piece, ought to be capable of relisting theforces which at each point tend to burst it; and of such aweight as to prevent any irregularity in the motion of theffiot, that would materially affect the justness of the siring(g) • we will endeavour to ascertain these limits.

82. No absolute and general rule can be laid dawn for thethickness of metal in guns, since several physical points areinvolved in the problem. In a gun of a given calibre, thethicknesses proper to resist the pressure of the fluid, may be{cnovvn by firing it very strongly reinforced several times withthe fame charge ; then thinning the metal and firing it againwith the fame charge ; and thus successively diminishing thethicknesses, till they he fa reduced that the piece burst at oncejn every point of its length. Such experiments cannot be :

i H coficlusivq