Mr. R. H. M. Bosanquet m? the Theory of Sound. 219 



dependent of the steady note ]iroduoed by any particular 

 niotle ot excitation is this. Take a hottle resonator \vh()se 

 })itch is about a minor third above that of a fork. Hold the 

 i'ork to the bottle, and cover tlie mouth partially with the 

 finoer till the pitch is reduced to that of the fork, which is 

 easily tested by blowing liohtly ; then by alternately raising 

 the finger and applying it to the hole, the pitch of the reso- 

 nator may be alternated between tliat of the fork and its own 

 a minor third higher. Now, if the fork be held steadily to the 

 month and the tinger moved in this manner, the note of the 

 fork is distinctly heard to be followed by the proper note of 

 the resonator for an instant, the excited yibrations dying away 

 in the proper note of the resonator. 



I cite two experiments only with tubes open at both ends. 

 These had their pitch so arranged as to fall in regions where 

 I had forks at intervals of four vibrations copied from Appun's 

 tonometer. The interval between two such forks is to my ear 

 a gross quantity ; and the division of the interval by estimation 

 is more feasible than with larger intervals. In some cases I 

 tuned a fork to the note to be determined, and then counted 

 the beats it made with the forks on each side of it. In others 

 I availed myself of the variation of temperature of the room, 

 noting at what temperatures the tubes coincided with any of 

 the standard forks. The velocity of sound at 0° C. was taken 



at 1087 feet per second (Note 1). ^^ is the ratio of the wave- 

 length to the diameter of the tube (approximate). The value 

 of the correction is the difference between the length deduced 

 from the observed note and the true length for one open end. 



=r. R. Leugtli. Con-ection. 



6 1 in. 5 in. -63511 



15 -97 in. U-1 in. -543^. 



I regard these two as probably the best determinations I have. 

 Both tubes were prepared of perfect form with great care ; 

 and tlie exact knowledge of the pitch, with the presence of 

 two open ends, enables me to say that I do not think it possible 

 that either determination can be in error by more than half 

 the ditference between the two. Under these circumstances 

 I state the following as an experimental law. The correction 

 for the open end of a tube increases and diminishes tcith the 

 ratio of diaDieter to wave-lenath. I need hardly say that this 

 does not rest only on these two observations ; but they illus- 

 trate it. 



I now come to tubes stopped at one end. These experi- 



