OF GUN-POWDER. 17

other grains, than it penetrates towards the centre of eachgrain.

49- The following observation will prove that a deter-mined time is requisite for the consumption of each grain.Make, of the common mixture, some grains of powder aslarge as pistol bullets; dry, and set fire to them : it will thenbe seen that the fire penetrates from the surface to the centre,in a longer or a shorter space of time according to the sizeof the grains ; so that if there be no difference betweenthese large grains and common ones than in size, it may beinferred, from analogy, that the smallest must require a cer-tain space of time, however short. It is likewise clear, that 1the flame, in spreading itself from the burning grains to thecontiguous ones, takes up a certain space of time; as maybe exemplified by setting fire to a train of powder.

50. An experiment will prove, that the action of fire isalso progressive when applied to powder confined in avessel; where the burning fluid being more dense, is alsomore active than when powder is burned in the open air.(Fig. .III.) To the funnel charged as before, (45.) letthe capital N O be fastened, with the aperture O of such.a size, that the receiver being exhausted, only i or 4 - of thepowder contained in the funnel may take fire. Introduce thered-hot rod, and leave it to cool ; then admit by degrees theair into the receiver, and loosen carefully the capital from thefunnel. If proper attention has been paid in this process, itwill be seen—

1. That nothing touches the point M of the rod; asthe powder being burnt all around, leaves a cavitypearly spherical.

2. That each of the grains which form the sides ofthe cavity are burnt a little towards the concave side; asmay be distinguished by the fixed nitre, and the smoothsurface of each grain.

3. That the grains which are between the interior sur-face of the funnel and those which form the sides of thecavity, are whitened by the flame of the sulphur, andthat the sides of the funnel are bronzed.

It is then proved, that the fire both spreads itself betweenthe intervals of the other grains, and penetrates from thesurface to the centre of each grain progressively.

51. But it is not sufficient so have proved that the burn-ing of each grain, and the inflammation of the contiguousones, are progressive; it is necessary further to strew, that the '

L activity