8 Analysis of the residuum of fired gunpowder. [Jan. 



2d Expt. 



3d Expt. 



Sulphate of potassa, . . 

 Carbonate of potassa.. 

 Nitrate of potassa., . . . 

 Sulphuret of potassium 

 Sulphuretted hydrogen 

 Unconsunied charcoal. 



Earthy matter 



Experimental error.,. 



45.44. 

 29.75. 

 ,14,81 

 , .5.49 

 .1.38 

 , .0.56 

 ,.1.89 

 , .0.68 



46.83. 

 26.73, 

 15.34 

 0.65 

 1.52 

 0.60 

 2.00 

 3.33 



100.00 



100.00 



42. — Doubting however whether the result of saltpetre 

 was not too large, I tested a portion that had been collected, 

 and found that it contained sulphuric acid in considerable 

 quantity. After the precipitation by nitrate of silver and 

 while evaporating the solution of nitre, I fancied in every 

 experiment that there was a smell of sulphur, but I could not 

 detect it, for the precipitations by silver had been carefully 

 performed, so that the solutions afterwards exhibited no 

 traces of sulphur. Notwithstanding, I suspect that the solu- 

 tion of nitre did contain a minute portion of sulphur, and as it 

 also contained free nitric acid (15) the presence of these two 

 substances when the last portions of the saltpetre were eva- 

 porating, and when it was afterwards in fusion, will account 

 for the formation of the sulphuric acid, which would I appre- 

 hend combine with the potash. And as 15 grains of the 

 saltpetre gave sulphate of baryta equivalent to 1.71 of sulphate 

 of potassa, the quantity of saltpetre in the two foregoing ana- 

 lyses might be reduced in that proportion. But as this ulte- 

 rior result does not appear to legitimately affect them, I shall 

 not make the alteration. 



43. — Finally, to confirm or to remove the doubt just ex- 

 pressed, I made a synthetical mixture according to the analy- 

 sis, employing 15 per cent, of nitre as a mean between the 

 two experiments, and on submitting it to the action of the 

 blow-pipe on red hot charcoal, its action was as similar to the 

 action of the original substance as it could be. This verifi- 

 cation is satisfactory, and seems to point out that the analy- 

 sis has been correctly performed. 



10th .November, 1835. 



J. BRADDOCK. 



