and Detonating Matches. 137 



The volume of these constituents is 55.5 ; by which if their 

 weight 100 be divided, the quotient is 1.80. 



The specific gravity of the first and second of the above 

 powders, including the interstices of their grains after being 

 well shaken down in a phial, is 1.02. This is a curious result, 

 as the size of the grains is extremely different. That of 

 Pigou and Wilks similarly tried is only 0.99; that of the 

 Battle powder is 1.03 ; and that of Curtis and Harvey is 

 nearly 1.05. Gunpowders thus appear to have nearly the 

 same weight as water, under an equal bulk ; so that an impe- 

 rial gallon will hold from ten pounds to ten pounds and a 

 half, as above shewn. 



The quantity of water that 100 grains of each part with on 

 a steam bath, and absorb when placed for 24 hours under a 

 moistened receiver standing in water, are as follows : 



100 grains of Waltham Abbey, lose 1 . 1 by steam heat, gain . 8 over water. . 

 ofHall 0.5 . 2.2 



Lawrence . . 1.0 . . .1.1 

 Pigou and Wilks 0.6 . . .2.2 



Curtis and Harvey 0.9 . . .1.7 



Thus we perceive that the large-grained government powder 

 resists the hygrometric influence better than the others ; among 

 which, however, Lawrence's ranks nearly as high. These two 

 are therefore relatively the best keeping gunpowders of the series. 

 The process most commonly practised in the analysis of 

 gunpowder seems to be tolerably exact. The nitre is first 

 separated by hot distilled water, evaporated and weighed. A 

 minute loss of salt may be counted on, from its known volati- 

 lity with boiling water. I have evaporated always on a steam 

 bath. It is probable that a small proportion of the lighter 

 and looser constituent of gunpowder, the carbon, flies off' in 

 the operations of corning and dusting. Hence analysis may 

 shew a small deficit of charcoal below the synthetic proportions 

 originally mixed. The residuum of charcoal and sulphur left 

 on the double filter-paper, being well dried by the heat of ordi- 

 nary steam, is estimated as usual by the difference of weight 

 of the inner and outer papers. This residuum is cleared off 

 into a platina capsule with a tooth-brush, and digested in a 

 dilute solution of potash at a boiling temperature. Three 



