upon Gun-powder, &c. 259 
where it is neceflary to attend to the recoil of the guns, in 
order to prerent very dangerous accidents that might be occa- 
fioned by it, the conftant practice has been in our navy, and, I 
believe, on board the fhips of all other nations, to lefien the 
quantity of powder after the firft four or five rounds i our 32 
pounders, for inftance, are commonly fired with 14 lbs-, of 
powder at the beginning of an action, but the charge is very 
loon reduced to 1 1 lbs. and afterwards to 9 lbs., and the filled 
cartridges are prepared accordingly. 
By the recoil it fhould feem, that the powder exerted a greater 
force alfo in the fourth experiment, being the fecond upon the 
fecond day, than it did upon the third,, or the firft upon that 
day ; but the pendulum giving way, it was not poffible to com- 
pare the velocities of the bullets in the manner we did in the 
two experiments mentioned above. 
This augmentation of the force of powder,, when it is fired 
in a piece that is warm, may be accounted for in the following 
manner. There is no fubftance we are acquainted with that 
does not require to be heated before it will burn y even gun- 
powder is not inflammable when it is cold.. Great numbers of 
fparks or red-hot particles from the, flint and fteel are frequently 
feen to light upon the priming of a mufket, without fetting 
fire to the powder,, and grains of powder may be made to pafs 
through the flame of a candle without taking the fire ; and 
what is ftill more extraordinary, if large grains of powder are 
let fall from the height of two or three feet upon a red-hot 
plate of iron,, laid at an angle of about 45° with the plane of 
the horizon, they will rebound intire without being burnt, or 
in the leaft altered, by the experiment.. In all thefe cafes the 
fire is too feeble, or the duration of its aSIon is not fufEciently 
