[ 99 ] 



magncfia. It is difficult to account for this phenomqnon, as ma- 

 rine fait both impedes the ignition and leffens the cxplofion of 

 gun-powder; and I believe it may be dcmonftrated that magnefian 

 or calcareous nitre produces at leaft the laft of thofe eiFedls, if we 

 coniider the faintnefs of its own detonation, when it has any ; 

 and that dcliquefcent quality, which muft communicate a degree 

 of humidity to the compofition, inimical to a forcible explofion ; 

 and (what is in my opinion of much greater confcquence) which 

 muft be noxious in the extreme to the durability of gun-powder : 

 I have reafon to believe (as far as my experience can eftablifh the 

 fad) that powder made with falt-petre, oftener than four times 

 refined, is of inferior ftrength, though probably more durable, 

 than that which has been only thrice depurated : If the elaflic 

 and expanfive fluid contained in nitre partakes at all of a fpi- 

 rituous nature, may not repeated evaporation liberate a portion of 

 it? Stahl afferts, that the nitrous acid is a combination of the 

 vitriolic acid with the principle of inflammabih'ty, effcdted by the 

 agency of putrefadion ; and Pietch of the Berlin Academy feems 

 to prove this theory by his experiment of moiftening a calcareous 

 fione with vitriolic acid and urine, which being expofed for fomc 

 time to the adion of the atmofphere, was found ftrongly impreg- 

 nated with nitre. If the aforefaid experiment be accurate, we 

 muft admit that falt-petre is a compound fubftance ; and it may 

 not be a very improbable dedudion to fuppofe that repeated 

 chxation in part deprives this fait of that elaftic fluid which con- 

 ftitutes the ftrength of gun-powder : And this opinion is ftrongly 

 corroborated by two well-known fads ; firft, in purifying a large 

 quantity of nitre there is a diftind deficiency of weight after the 

 proccfs, which cannot be accounted for by the weight of the 



O 2 refiduum; 



