DR. H. DEBUS ON THE CHEMICAL THEORY OE GUNPOWDER. 
533 
E. L. G. 
)!> 
F. G. 
P.. . 
First analysis by Noble and Abel . 
Second ,, 
16KNO s +19-5lC+6-92S 
16KNO 3 +22-40C+6-83S 
16KNO 3 + 20-8C+6-8S 
16KNO s +21-86C+6-79S 
and for the mean 16KN0 3 + 21'14C+6 , 83S. 
If also the analyses of the late Mr. Wills are taken into consideration, then the 
mean composition of the powders of Waltham Abbey would be represented by the 
symbols 
16KNO s +21-18C + 6-63S 
The differences of composition found by analysis for the same description of powder 
are also of great importance in the calculations of the analytical results of the products 
of explosion. Noble and Abel determine the potassium in x parts of the solid 
powder residue, and by means of the number so obtained calculate the weight of the 
total residue. The amount of potassium in the quantity of powder exploded is known 
from the analysis of the powder, and as the whole of this potassium must reappear in 
the solid residue, it is easy to find the total residue if the weight of potassium in x 
parts of it are known. 
The total weight of the powder residue subtracted from the weight of the powder 
used in the experiment gives as difference the total weight of the gases produced by 
the explosion. 
This mode of calculation requires that the exact composition of the powder used in 
each experiment should be known. Messrs. Noble and Abel assume the composition 
of the powders of Waltham Abbey, P., E. F. G., and F. G., to be constant; they also 
consider in their first memoir that of E. L. G. to be so, but in their second memoir 
they base the calculations of the later experiments on the second analysis of this 
description of powder. This assumption is, as a matter of fact, not correct; on the 
contrary, we may take it as highly probable, that in any two experiments made by 
Noble and Abel with the same class of powder, the powder used in the one experi¬ 
ment was not exactly of the same composition as the powder used in the other. 
Accordingly the weight of the solid residue as calculated by Noble and Abel will 
have been found in some experiments too high, in others too low, and as a consequence 
of this the total weight of the gaseous products cannot have been exact. The correct¬ 
ness of this is proved by the differences between the composition of the powders 
calculated from the products of combustion, and the composition deduced from direct 
analysis (see pp. 530, 531). These errors will, however, compensate each other if we 
take the mean of the analytical results of all the experiments. 
Before we proceed to do so it will be desirable to consider another circumstance 
which is not without influence on the final results of Messrs. Noble and Abel. 
They burnt their powders in steel cylinders in quantities from 100 to 750 grins., so 
that the solid products of combustion after explosion remained from 60 to 120 seconds 
