DR. H. DEBUS ON THE CHEMICAL THEORY OF GUNPOWDER. 565 
calculated according to equation (VIII.). This mode of calculation is in Linck’s case 
not applicable, because large quantities of charcoal remained unburnt, the composition 
of which is not known. This charcoal having been exposed to the high temperature 
developed by the combustion of the powder can no longer have had the original 
composition, but was probably nearly pure carbon. 
The composition of the Austrian cannon powder can be represented by the symbols : 
16KN0 3 +19-8C+8-7S 
if hydrogen, oxygen, ash, and moisture of the charcoal are neglected. All the 
potassium and the oxygen of the decomposed saltpetre reappeared in the chief 
products of explosion, but more than 4 atoms of carbon and 3 atoms of sulphur 
remained free. The amount of the sulphide formed by this powder was very small 
in Karolyi’s experiment. 
The powder constituents which were transformed into the chief products are 
represented by the symbols 
16KN0 3 +14-32C+5-28S 
and 2'04 mols. of CO had been formed (p. 549). 
The values of these coefficients substituted for x, y and 2 in equation (VIII.) yield 
the following theoretical quantities of the chief products of combustion, to which are 
appended the quantities found by experiment:— 
16KN0 3 
14-32C 
5-24S 
k 2 co 3 . . . 
Theory. 
3‘06 
Experiments. 
3-05 
K 3 S0 4 . . . 
4-59 
4-61 
K 3 S 3 .... 
0-33 
0-33 
00, ... . 
9T2 
9‘23 
CO. 
2-04 
2'04 
n 2 . 
8-0 
7-55 
a striking confirmation of theory is thus presented. 
The analysis of the products of explosion of rifle powder carried out by Karolyi 
is contaminated by a considerable error. He finds, correctly, no oxygen in the by¬ 
products ; but no less than 4 atoms or yyth more oxygen appears in the chief 
products than in the decomposed saltpetre, and if from this the oxygen of the 
charcoal be subtracted there remain still 1*66 atoms of oxygen more than were 
contained in the entire powder. If these errors of experiment are corrected, by 
means of equation (VII.), a very good agreement of the numbers calculated 
according to equation (VIII.), and those derived from experiment, is obtained for 
rifle powder. 
