[ 132 ] 



XVI. 



ON THE CONTBOL SUPPLY-PIPES HAVE ON KEEDS. 

 By FEEDEBICK T. TKOUTON, B.E. 



[Read April 18, 1888.] 



If a free reed be supplied with, air through a long tube it will in 

 most cases be found difficult, often impossible, to get it to sound, 

 especially if the pitch be low. The introduction of a bulb on the 

 tube just before the reed generally prevents this effect to a large 

 extent, and allows the reed to sound more or less freely. It is 

 equally efficacious to have the bulb connected at right angles on 

 one side of the tube where the current of air does not pass through 

 it. The first explanation which suggested itself was, that to keep 

 tip the supply a reservoir of air near the reed was necessary, such 

 as the mouth presents when the reed is blown directly, or the air 

 chamber of the various musical instruments in which reeds are 

 employed. This idea was next found to be untenable, for a tube 

 of much larger volume than the bulb was attached to one side 

 without preventing the effect of the long tube. 



It was then suspected that the vibration period of the mass of 

 air attached in this way was important. To test this, close to the 

 reed a glass tube was attached at right angles, the effective length 

 of which could be varied by means of a piston. It was then found 

 that this tube had control over the vibrations of the reed ; for as the 

 piston was moved up and down the glass tube, at definite places, 

 of equal distances apart, the reed sounded quite freely, and at places 

 half-way between these it would not sound, while in intermediate 

 positions the note was more or less lowered according to the posi- 

 tion of the piston. The distance between these recurrent positions 

 was found to be always about half the wave length of the note of 

 the reed employed ; there was, however, some difficulty in deter- 

 mining the exact position of these points, for it was found that any 

 point within certain limits had pretty much the same effect. The 



