ACOUSTICS. INVESTIGATION OF THE JAPANESE BAMBOO JMPE. 27 



This relation has been obtained on the supposition tliat tlic 

 velocity of flow in every transverse section is uniform. Tlie as- 

 sumption does not hold in the cases investigated, since the abrupt 

 change of the section at both ends of the diaphragm produces 

 an irregular flow of the air. It will be interesting, however, to 

 test the above relations for the present case. 



Let JS = const, between .r=^ + (i and c — (^, and elsewhere = 

 Then, we have 



J' — -:r— CO^i 



For J/ = 0, we obtain ç = | /. This is nearly the ease in Figs. 

 and 10. 

 If ^ = /, 



,j I JS . -l-a . 



J' —— —z SDl 



Therefore, if a = const., 



J/ oc J s or oc {li — r)'- 



where K is the radius of the cylinder and r that of the channel. 

 This is verified qualitatively in the table given in p. 24. 



Comparing the above formulae Avitli the results given in 

 Figs. (S, V), and 10, Ave see generally that the actual values of 

 ^l are far greater than the theoretical values, the discrepancy in- 

 creasing with the decrease of the radius of the passage. This 

 shows that the disturbance due to the end of the dia2>hragm is 

 considerable. 



APPLICATIONS OF THE RESULTS OF THE EXPERIMENTS 



TO THE SYAKUHATI. 



The results of the present ex|ieriments may be applied to 

 the case of the sjjakuhaii in the following ways : — 



