42 Experiments upon Gunpowder. 



ence of those velocities with all the charges except the 

 first, thus : 



Sum of the Velocities 1764. 



Computed. Actual. Difference. No. of Exp. 



9864 9854 IO 23 



So that it appears that the difference or the actual 

 velocity was smaller than the computed by ^|- part 

 only, at a mean of 23 experiments. 



But as by far the greater number of the experiments 

 were made with the following charges, viz., 290, 218, 

 208, 165, and 145 grains of powder, let us take the 

 sum and difference of the computed and actual velocities 

 of those charges, thus : 



Sum of the Velocities. 

 Computed. Actual. Difference. No. of Exp. 



5985 6044 -f- 59 1 8 



Here the agreement of the theory with the experi- 

 ments is so very remarkable that we must suppose it 

 was in some measure accidental; for the difference of the 

 velocities in repeating the same experiment is, in gen- 

 eral, much greater than the difference of the computed 

 and actual velocities in this instance; but I think we 

 may fairly conclude, from the result of all these trials, 

 that the velocities of like musket-bullets, when they are 

 discharged from the same piece by different quantities 

 of the same kind of powder, are very nearly in the 

 sub-duplicate ratio of the weights of the charges. 

 Whether this law will hold good when applied to 

 cannon-balls, and bomb-shells of large dimensions, I dare 

 not at present take upon me to decide ; but for several 

 reasons that might be mentioned, I am rather of opin- 

 ion that it will not ; at least not with that degree of 

 accuracy which obtained in these experiments. 



