The whole machine is so simple, easy, and expeditious, that, as Dr. Hutton remarks, the weighing of the powder is the chief part of the trouble; and so accurate and uniform, that the successive repetition, or firings, with the same quantity of the same sort of powder, hardly ever make a difference in the recoil of the one-hundredth part of itself. Gregory (_Treatise of Mechanics_, vol. ii, p. 178) has given a more particular description of the eprouvette of Dr. Hutton; namely, that it is a small brass gun, 2-1/2 feet long, suspended by a metallic stem, or rod, turning, by an axis, on a firm and strong frame, by means of which, the piece oscillates in a circular arch. A little below the axis, the stem divides into two branches, reaching down to the gun, to which the lower ends of the branches are fixed, the one near the muzzle, the other near the breech of the piece. The upper end of the stem is firmly attached to the axis, which turns very freely by its extremities in the sockets of the supporting frame; by which means, the gun and stem vibrate together in a vertical plane, with a very small degree of friction. The charge is the same we have mentioned, usually about two ounces, without any ball, and then fired; by the force of the explosion, the piece is made to recoil or vibrate, describing an arch or angle, which will be greater or less, according to the quantity or strength of the powder. To measure the quantity of recoil, and consequently the strength of the powder, a circular brazen or silver arch of a convenient extent, and of a radius equal to its distance below the axis, is fixed against the descending two branches of the stem, and graduated into divisions, according to the purpose required by the machine: _viz._ 1st. Into equal parts, or _degrees_, for the purpose of determining the angle actually described in the vibration. 2nd. Into equal parts, according to the _chords_, being, in fact, 100 times the double sines of the half angles, and running up to 100, as equivalent to 90 degrees. 3d. Into unequal parts, according to the versed sines; they are, in truth, 100 times the versed sines of our common tables, 141-1/2 corresponding with 90 degrees. These serve to compare the forces. The divisions in these scales are pointed out by an index, which is carried on the arch during the oscillation, and then, stopping there, shows the actual extent of the vibration. Two ounces of powder, give, on an average, according to the experiments of professor Gregory, about 36 on the chords, or about 21° on the arch. A more detailed account, with diagrams, may be seen, by consulting Hutton's Tracts, vol. iii, p. 153.