40 Dr. Schroder van der Kolk on the Velocity of Sound, 

 we can then get / by integration, 



dp r-R XR, X(r-R) 

 %tan — 



-i 



,( 1+ ,E) 



Hence the time required by sound to traverse the distance 

 r — R, when it is propagated in a tube, is 



- r ~ R 



and when it is propagated in space, 



r— R XR, . X(r-R) 



Let us now test this formula by applying it to a particular 

 case. 



For fi = \/ ¥-j--y, I will adopt the mean value of 331 metres 



deduced from the experiments of Wertheim and of Masson; 

 this value is arrived at by means of organ-pipes, in which the 

 intensity of the sound is not great. Hence we get, by means 

 of the ordinary formula, y = 1'398, and 



AV 

 X 2 =Q-&857^-- 



v o 

 AV 

 The value -~ can only be found approximately; I have made 



the attempt which follows to apply the formula to Moll and 

 Yan Beek's experiments. 



Following the statements of Regnault (Cours de Chimie, vol. ii. 

 p. 291), I assume for the volume of the gases generated by the 

 combustion of gunpowder, 329 times the bulk of the powder 

 itself, at 0° C. and under a pressure of 0-760 metre of mercury; 

 but as the temperature is in reality much higher than this, we 

 may certainly assume that we have three times this volume. If, 

 after Regnault, we take the specific gravity of gunpowder as 1, 

 the volume of a kilogramme of the gases will be =329 X 0'003 

 = 0-987 cubic metre. In the experiments above mentioned, the 

 quantity of powder expended was always 3 kilogrammes, and 

 consequently the volume of gases formed was 2*961 cubic 

 metres. 



The wave-length now remains to be determined also. As I 

 had no means of ascertaining this directly, I adopted for the 

 pitch of the report of a cannon the tone AAA, making 27| vibra- 

 tions in a second, the lowest orchestral tone. The wave-length 



