62 Experiments 2ipon Giinpozvder. 



in a second, is extremely near '^^'^k^ feet in a second, — the 

 diminution of the velocity by the resistance as here de- 

 termined. 



If the diminution of the velocities of the bullets in 

 the two subsequent experiments be computed in like 

 manner, it will turn out in the 86th experiments 65 

 feet in a second, and in the 87th experiment = 33 feet 

 in a second; and making these corrections, the com- 

 parison of the two methods of ascertaining the veloci- 

 ties will stand thus : — 



Velocities shown by the pendulum. 

 Add the diminution of the velocity 



by the resistance of the air. 

 Making together 

 Velocity by the recoil. 

 The difference = 



So that it appears, notwithstanding these corrections, 

 that the velocities in the 86th and 87th experiments, and 

 particularly in the last, as they were determined by the 

 pendulum, are still considerably deficient. But the 

 manifest irregularity of the velocities in those instances 

 affords abundant reason to conclude that it must have 

 arisen from some extraordinary accidental cause, and, 

 therefore, that little dependence is to be put upon the 

 result of those experiments. I cannot take upon me 

 to determine positively what the cause was which pro- 

 duced this irregularity ; but 1 strongly suspect that it 

 arose from the breaking of the bullets in the barrel by 

 the force of the explosion ; for these bullets, as has 

 already been mentioned, were formed of lead, inclosing 

 lesser bullets of plaster of Paris ; and I well remember 

 to have observed at the time several small fragments of 

 the plaster, which had fallen down by the side of the 



