My experiments are, like Robins’, on a small scale; nor would I, like Hutton, try a brass gun of sixty calibres in length, carrying a one-pound ball; for one is strictly more limited than the other, and thus rendered the results laid down by him imperfect: for, as he says, “If you fill the tube with powder you get no greater velocity, as there is not a duration in the confinement to enable the powder to explode.” If he had assimilated the grain of his powder to the gun, he would have obtained a different result; and a knowledge of this fact, I apprehend, makes all the difference. The greatest velocity he obtained was with powder 1-1/2 times the weight of the ball in a gun of sixty calibres in length, and the velocity he then obtained was only 3,181 feet per second. The inferences that probably induced him to recommend others not to endeavour to obtain a greater velocity than 2,000 feet per second, were, like these experiments, drawn from imperfect data. With a ball of an ounce weight in a barrel of sixty calibres, and with 3-4ths the weight of ball in powder, or 12 drachms, a velocity can be given to the ball to equal it in force to 46,875 pounds. The velocity of this ball I leave to the calculations of the mathematical world. But, however, I will give the results of a round of experiments tried to ascertain this; and if the data laid down be correct, that the velocity of a ball must be multiplied by its weight to find the force, the result will be the establishment of a system of velocity never yet dreamt of. I cannot but imagine that there exists some error; though where it is I know not: every deduction I have drawn is consequent upon the results hereafter described. “The power required to force a punch 0·50 inch diameter through an iron plate 0·08 inch thick is 6,025 pounds, through copper 3,938 pounds. A simple rule for determining the force required for punching may thus be deduced:-- “Taking one inch diameter and one inch in thickness as the units of calculation it is shown that 150,000 is the constant number for wrought-iron plates, and 96,000 for copper plates. “Multiply the constant number by the given diameter in inches, the product is the pressure in pounds which will be required to punch a hole of a given diameter through a plate of a given thickness.” Now an idea struck me, that this would form a very good test of the comparative force of gunpowder, and I consequently commenced an extensive round of experiments.