70 
CAPTAIN NOBLE AND MR. F. A. ABEL ON FIRED GUNPOWDER. 
tion of carbonic anhydride from it, by addition of sulphuric acid, and for the transmis- 
sion of the gas through small weighed absorption-tubes containing respectively sulphuric 
acid, calcium chloride, and solution of potassium hydrate. The increase in weight of 
the latter corresponded to the proportion of carbonic anhydride in the solid residue. 
j. Potassium sulphide, potassium carbonate , and potassium hydrate. — Pure man- 
ganous chloride or sulphate was added in excess to the aqueous solution of the residue, 
and the amount of manganese, in the thoroughly washed precipitate, determined as 
red oxide. If the amount obtained exceeded those which would be furnished by the 
potassium sulphide and carbonate (deduced from the previous determinations), the excess 
was taken to correspond to potassium hydrate existing in the residue. If it was less, 
the sulphur existing as monosulphide of potassium was calculated from the weight of 
the manganous oxide, and the difference between it and the sulphur found in the cupric 
oxide (in determination e) was taken to represent excess of sulphur, or free sulphur, 
and was added to the result of determination d, the necessary correction being made in 
the number furnished by determination e. 
k. Total amount of potassium. — The solution of the residue, after treatment with 
cupric oxide, was evaporated with excess of sulphuric acid, and the residue repeatedly 
treated with ammonium carbonate and ignited, until the weight of potassium sulphate 
was constant. Or water and sulphuric acid were added to about 4 grms. of the residue, 
and after boiling to expel sulphurous acid, two or three drops of nitric acid were added 
to peroxidize the little iron in solution and excess of ammonia to precipitate the latter. 
The precipitate and insoluble matters (glass &c.) were then filtered off, and the solu- 
tion evaporated, the weight of potassium sulphate being ascertained by treatment of 
the residue as already described. In this way the amount of potassium arrived at 
indirectly, by the determinations of the several substancesjvith which it existed in com- 
bination, was controlled by direct estimation. 
l. Ammonium sesqidcarbonate. — The solution of about 12 grms. of the residue was 
diluted to one litre the liquid was then carefully distilled until about 250 cub. centims. 
remained in the retort, the distillate being allowed to pass into dilute chlorhydric acid. 
As some minute quantities of potassium salt might have passed over, the distillate was 
returned to a retort, mixed with excess of sodium carbonate and again distilled, the 
product passing into dilute chlorhydric acid. This second distillate was evaporated, 
and the ammonium determined as platinum salt with the usual precautions, the weight 
of the latter being controlled by ignition and determination of the weight of the 
platinum. 
m. Potassium nitrate. — The portion of solution remaining in the retort, after the 
first distillation above described, was acidified with sulphuric acid; a piece of thin 
sheet zinc was then placed in the liquid and allowed to remain for a week, a small 
quantity of sulphuric acid being occasionally added. After the lapse of that time the 
zinc was removed, and the ammonia produced from any nitrate existing in the liquid 
was determined exactly as at 1. 
