upon Gun- powder , &ic. 293 
{till confiderably deficient. But the manifeft irregularity of the 
velocities in thofe inftances affords abundant reafon to conclude, 
that it muif have arifen from fome extraordinary accidental 
caufe, and therefore, that little dependance is to be put upon 
the refult of thofe experiments. I cannot take upon me to 
determine pofitively what the caufe was which produced this 
irregularity; but I ftrongly fufpedt, that it arofe from the 
breaking of the bullets in the barrel by the force of the explo- 
fion : for thefe bullets, as has already been mentioned, were 
formed of lead, inclofing leffer bullets of plafter of Paris ; 
and I well remember to have obferved at the time feveral final 1 
fragments of the plafter which had fallen down by the fide of 
the pendulum. Iconfefs, I did not then pay much attention to 
this circumftance, as I naturally concluded, that it arofe from 
the breaking of the bullet in penetrating the target of the pen- 
dulum, and that the fmall pieces of plafter I faw upon the 
ground had fallen out of the hole by which the bullet entered. 
But if the bullets were not abfolutely broken in pieces in firing, 
yet, if they were confiderably bruifed, and the plafter or a part 
of it were feparated from the lead, fuch a change in their form 
might produce a great increafe of the refiftance, and even their 
initial velocities might be affedted by it, for their form being 
changed from that of a globe to fome other figure, they might 
not fit the bore, and a part of the force of the charge might 
be loft by the windage. 
That this adtually happened in the 87th experiment feems 
very probable, as the velocity with which the bullet was pro- 
jected, as it was determined by the recoil, is confiderably lefs 
iii proportion in that experiment than in either of thofe that 
precede it in that fet, or in thofe which follow it, as will ap- 
Vol. LXXL R r pear 
