250 Count Rumford's Experiments to determine 
those cases in which the weight was raised, and the stopper 
blown out of the bore, nothing was found remaining in the 
barrel. It was very remarkable that this hard substance was 
not found distributed about in all parts of the barrel indiffe- 
rently, but there was always found to be more of it near the 
middle of the length of the bore, than at either of its extremi- 
ties ; and the upper part of the vent tube in particular was 
always found quite filled with it. It should seem from hence, 
that it attached itself to those parts of the barrel which were 
soonest cooled; and hence the reason, most probably, why 
none of it was ever found in the lower part of the vent tube, 
where it was kept hot by the red-hot ball by which the powder 
was set on fire. 
I found by a particular experiment, that the gunpowder 
made use of, when it was well shaken together, occupied ra- 
ther less space in any given measure, than the same weight of 
water; consequently when gunpowder is fired in a confined 
space which it fills, the density of the generated elastic fluid 
must be at least equal to the density of water. The real spe- 
cific gravity of the solid grains of gunpowder, determined by 
weighing them in air and water, is to the specific gravity of 
water, as 1.868 to 1.000. But if a measure, whose capacity is 
one cubic foot, hold 1000 ounces of water, the same measure 
will hold just 1077 ounces of fine grained gunpowder, such as 
I made use of in my experiments ; that is to say, when it is 
well shaken together. When it was moderately shaken toge- 
ther, I found its weight to be exactly equal to that of an equal 
volume, or rather measure, of water. But it is evident that the 
weight of any given measure of gunpowder, must depend much 
upon the forms and sizes of its grains. I shall add only one 
