Second proof.--A sample of 1lb. from each lot, carefully weighed up, and a similar sample of the comparison powder, is exposed for three weeks in a box perforated with holes (called a damp chest), to the action of the atmosphere. This box is placed under cover, so that it is sheltered from the wet, but that the moisture can get to it. If, at the end of this time, there is a greater proportion of difference in range between them than one-twentieth, it is rejected. The pounds are also very carefully weighed up again, to ascertain the comparative absorption of moisture. This is called the hygrometric test. REMARKS ON THE PROOF OF POWDER BY THE EPROUVETTES. ~By eprouvettes or pendulum.~ By comparing the results of the proofs by the eprouvettes with those furnished by the cannon pendulum (vide plate 1, fig. 2 and 3), it will appear that the eprouvettes are entirely useless as instruments for testing the relative projectile force of different kinds of powder, when employed in large charges in a cannon. Powders of little density, or of fine grain, which burn most rapidly, give the highest proof with the eprouvettes, whilst the reverse is nearly true with the cannon. ~Real use of eprouvettes.~ The only real use of these eprouvettes is to check and verify the uniformity of a current manufacture of powder, where a certain course of operations is intended to be regularly pursued, and where the strength, tested by means of any instrument, should therefore be uniform. ~Best proof, by service charges.~ The only reliable mode of proving the strength of Gunpowder is, to test it with service charges in the arms for which it is designed; for which purpose the balistic pendulums (vide plate 3), are perfectly adapted. ~Best proof for small arms.~ For the proof of powder for small arms, the small balistic pendulum is a simple, convenient, and accurate instrument. ~Common eprouvette.~ The common eprouvettes are of no value as instruments for determining the relative force of different kinds of Gunpowder. OF THE SIZE OF GRAIN FOR GUNPOWDER. ~On size of grain.~ With regard to the particular size of grain for Gunpowder, I am confident great improvements might be made, both in obtaining greater regularity of effect and propelling force, by the adoption of a more uniform even grain. There are at present half-a-dozen different sizes in our cannon and musket powder; and I think it stands to reason, that the more equal the size, the more uniform will be the ignition of all the grains, and consequently the effect of the same charges will be much more regular. OBSERVATIONS ON THE MANUFACTURE OF GUNPOWDER ON THE CONTINENT AND AMERICA. It may not be uninteresting to have a slight knowledge of the method employed on the Continent, &c., for the production of Gunpowder. ~Proportion of the ingredients.~ The proportions of the three ingredients vary slightly all over the Continent and America, being as follows:--