236 Count Rumford’s Experiments to determine 
ship guns on board the ships of war in which I made my pas- 
sage to and from America. 
It would take up too much time, and draw out this paper 
to too great a length, to give an account in detail of all these 
experiments, and of the various observations I have had oppor- 
tunities of making from time to time, relative to this subject. 
I shall, therefore, only observe at present, that the result of all 
my inquiries tended to confirm me more and more in the 
opinion, that the theory generally adopted relative to the ex- 
plosion of gunpowder was extremely erroneous, and that its 
force is in fact much greater than is generally imagined. That 
the position of Mr. Robins, which supposes the inflammation 
and combustion of gunpowder to be so instantaneous “ that 
“ the whole of the charge of a piece of ordnance is actually 
“ inflamed and converted into an elastic vapour before the 
“ bullet is sensibly moved from its place," is very far from 
being true ; and that the ratio of the elasticity of the generated 
fluid, to its density, or to the space it occupies as it expands, 
is very different from that assumed by Mr. Robins. 
The rules laid down by Mr. Robins for computing the ve- 
locities of bullets from their weight, the known dimensions of 
the gun, and the quantities of powder made use of for the 
charge, may, and certainly do, very often give the velocities 
very near the truth ; but this is no proof that the principles 
upon which these computations are made are just ; for it may 
easily happen, that a complication of erroneous suppositions 
may be so balanced, that the result of a calculation founded 
on them may, nevertheless, be very near the truth ; and this 
is never so likely to happen as when, from known effects, the 
