54 Experiments upon Gunpowder. 



second. In the iO4th and io5th experiments, the recoil 

 was 12.92 and 13.28, and the velocity computed from 

 the mean of those chords is 1030 feet in a second ; but 

 the velocity shewn by the pendulum was no more than 

 about 900 feet in a second. As the recoil was so nearly 

 equal to what it was in the y6th and 77 th experiments 

 before mentioned, when the velocities shewn by the recoil 

 and by the pendulum were almost exactly the same, I am 

 inclined to believe that there must have been some mis- 

 take in determining the velocities by the pendulum in 

 these last experiments, and that the velocity shewn by 

 the recoil is most to be depended on. 



With 290 grains, or half the weight of the bullet in 

 powder, in the 48th, 49th, and 5<Dth experiments, the 

 recoil was 22.58, 22.92, and 22.38 ; and the recoil with 

 the same charge of powder, without a bullet, at a mean 

 of the 6oth and 96th experiments, was 10.66. The 

 mean of the velocities of the bullets computed from 

 the recoil is therefore 1416 feet in a second, and the 

 velocity shewn by the pendulum was 1427 feet in a 

 second ; the difference is not considerable. The mean 

 of the velocities in the iO9th, noth, mth, H2th ex- 

 periments is by the recoil 1464, and by the pendulum 

 1444 feet in a second. 



With 330 grains of powder the velocities of the bul- 

 lets appear to have been as follows, viz. : 



Vent at o. Vent at 1.3. Vent at 2.6. 



By the Recoil, 1543 1620 1610 



By the Pendulum, 1594 1625 1528 



See the 62d, 63d, 64th, 65th, 66th, 67th, and i7th 

 experiments. 



The uniformity of the recoil was in all cases very re- 

 markable. Thus, in the first set of experiments with 



