282 Count Rumford’s Experiments to determine 
this is, in my opinion, a most decisive proof, not only that the 
combustion of gunpowder is by no means so rapid as it has ge- 
nerally been thought to be, but also (what will doubtless ap- 
pear quite incredible), that if a grain of gunpowder, actually 
on fire, and burning with the utmost violence over the whole 
extent of its surface, be projected with a very great velocity into 
a cold atmosphere, the fire will be extinguished, and the re- 
mains of the grain will fall to the ground unchanged, and as 
inflammable as before. 
This extraordinary fact was ascertained beyond all possibility 
of doubt by the following experiments. Having procured from 
a powder-mill in the neighbourhood of the city of Munich a 
quantity of gunpowder, all of the same mass, but formed into 
grains of very different sizes, some as small as the grains of 
the finest Battel powder, and the largest of them nearly as big 
as large pease, 1 placed a number of vertical screens of very 
thin paper, one behind another, at the distance of 12 inches 
from each other ; and loading a common musket repeatedly 
with this powder, sometimes without, and sometimes with a 
wad, I fired it against the foremost screen, and observed the 
quantity and effects of the unconsumed grains of powder which 
impinged against it. 
The screens were so contrived, by means of double frames 
united by hinges, that the paper could be changed with very 
little trouble, and it was actually changed after every experi- 
ment. 
The distance from the muzzle of the gun to the first screen 
was not always the same ; in some of the experiments it was 
only 8 feet, in others it was 10, and in some 12 feet. 
The charge of powder was varied in a great number of dif- 
