ACOUSTICAL IXVESÏRIATIOX OF THE JAPANESE 15AM1500 IMl'K. l'.) 



however, we confine our study to the former mode exchisively. 

 Tlie latter mode will, T hoi)e, form the subject of a future com- 

 munication. 



The results wlien a conical passage was used, are given in 

 Figs 8, Î), and 10, in broken lines ; the diameter given there is that 

 of the narrower end, the wider one being always 4 cm. i.e. equal 

 to the inner diameter of the pipe. Here, again, the abscissa is 

 the distance between the open end and the middle section of 

 the diajDhragm. The results are quite unaltered if the direction 

 of the conical channel be inverted. It will be seen from the 

 figures : 



a) That the correction due to the diaphragm is about 0.4 

 times less than that of the corresponding cylindrical channel 

 whose sectional diameter is equal to the diameter of the narrower 

 end of the conical channel. 



b) That the full line and the corresponding broken line 

 intersect each other at a point near the corresponding horizontal 

 dotted line. This shows that, where the eftect of the diaphragm 

 is null, its form and size are immaterial. 



c) That the straight line joining these points of intersection 

 for ditterent notes, is directed toward the origin. This shows 

 that the position at which the effect of the obstacle is null, 

 is given by the ratio of the distance from the piston or the open 

 end to the wave length of the note. 



Lastly, the efîëct of the length of the diaphragm was studied. 

 The result is shown in Fig. 11. It will be noticed that the 

 correction increases with the length of the channel. Though in 

 this case the length of the diaphragm ranges from 5 to 1 cm., 

 the different curves are nearly concurrent to a point on the 

 corresponding dotted line, if the abscissa be taken proportional 



