EXPLOSIVE COMPOUND. 285 



raised, the compound may be drawn up in an uninterrupted 

 line to any mark on the stem that may be desired : the inch 

 measure of the stem of our syringe holds exactly 53 grains of 

 pure mercury 5 consequently it must hold 625 of a grain of 

 the explosive compound, the specific gravity of mercury being 

 13 568, and that of the explosive compound, according to our 

 experiments, being 1*6. 



The mode by which we obviated the errour arising from 

 having much water present, was to decompose the compound 

 over mercury in the following manner : 



A small stoppered phial was converted to an air receiver, by and the quan- 



having its bottom cut off — it was then sunk up to its neck in f"v of watei * 



. , . „ • , , , taken very 



the mercury contained in a small mercurial trough, the stopper small, over 



being first taken out. The capacity of the neck was then mercury, 

 filled with a few drops of water, into which was introduced the 

 '625 of a grain of the explosive compound — the glass stopper 

 was then put into its place, and the receiver, with its contents, 

 raised on to the shelf of the trough. Some potash was then 

 procured, which was free from carbonic acid, and had been 

 deprived of any combustible matter, by having undergone 

 igneous fusion j it was also free from any metallic oxide. Of Solution of 

 this potash a concentrated aqueous solution was made, and mis P otasn * d « cd > 

 solution passed up into the receiver to the compound, the de- 

 composition of which it occasioned. In performing this expe- 

 riment, it is of importance not to pass up the fused potash in 

 the solid state, as the heat which is occasioned by its solution 

 in the small quantity of water which it meets with, instantly 

 causes the compound to explode. 



The decomposition of the compound by liquid ammonia we or of ammo- 

 effected exactly in the same manner, passing up the solution ma * 

 of ammonia instead of the solution of potash j the solution of 

 pure ammonia must, however, be diluted with its own bulk 

 of water, otherwise it will immediately occasion an explosion 

 of the compound. 



By these arrangements, we believe that we have removed The analysis 

 every source of error j and, having repeated the analysis several e n "f t ™j t ^ 1S 

 times with the greatest care, and with scarcely any variation confidence, 

 in the quantities of gases obtained, we are enabled to give those 

 quantities with considerable confidence. This will be best 

 done by stating the particulars of two of our experiments. 



Ml 



