156 On the Vehcity of' Sound. 



crepancies in their dimensions, and in the nnodes of operating peculiar 

 to different individuals. For, though Dr Moll says the guns were con- 

 stantly discharged, loaded, &c. by the same persons, yet he must mean 

 only the same rank or description of persons — not the same individuals; 

 because the guns were fully nine miles from each other, and discharged 

 nearly at the same time. Besides, if the sods, strongly rammed down 

 instead of wadding, presented the same resistance to the powder in 

 both guns, I should rather deem it an accidental coincidence. At the 

 same time, I am perfectly ready to admit, that I do not see how the 

 mode of experimenting with guns could have been better managed than 

 was done by Dr Moll and his associates; and I consider their results 

 among the most valuable we possess. 



In experimenting with cannon over great ranges, the intensity or 

 loudness of the sound must at first be very great, and then gradually 

 decrease toward the farther end of the range, where it has in some in- 

 stances been so faint as to be quite inaudible, when opposed by a veiy 

 slight wind. But since the results so obtained are only the mean of the 

 velocities over the whole range, they throw no light on the question, 

 whether, or how far, loudness affects the velocity. It is, besides, high- 

 ly probable, that such a mean velocity from cannon may often happen 

 nearly to agree with the mean from a bell, and yet, for all that, sound 

 be really moving with a retarded velocity, or slower as it gets fainter. 

 If, during experiments with cannon, additional observations were made 

 somewhere intermediate between the extremities of the range, possibly 

 a solution of the question might, to a certain extent, be obtained, by 

 comparing the times with the corresponding portions of the range. But 

 the difficulty of measuring these minute intervals of elapsed time with 

 sufficient exactness would here come into play, unless something like 

 the apparatus formerly suggested were adopted. By means of that me- 

 thod, the minute intervals may be ascertained with such facility, that 

 several observers could be ranged at various distances from the sonorous 

 body, which could scarcely fail to shew whether the velocity be uni- 

 form or retarded. 



Guns with percussion locks, it is true, could be set off in succession, 

 by means of clock-work ; but it would be nearly impossible to make 

 one and the same gun fire at sufficiently short equal intei-vals ; and I 

 i-ather doubt if two guns be often of precisely equal dimensions. To be 

 sure, this might be examined and rectified if necessary, though not 

 without considerable trouble. However, granting that any inequality 

 of size were obviated, or did not exist, a more insuperable uncertainty 

 remains ; for we cannot be sure that two charges, even those used sue- 



