24 ILLINOIS STATE ACADEMY OF .SCIENCE 



the momentum of the air jet at the center of the tube, vi- 

 brations of the column would be discouraged. 



We can communicate momentum to a vertical open air 

 column by heating it. If we heat the air in a tube, 2, by 

 a wire net placed in the lower half of the tube, we will 

 obtain a uniform current of air upwards. If the air is 

 vibrating then as it is moving inward its vibration is 

 increased by the momentum of the upward stream of air, 

 but not increased by moving downward. When we place 

 the hot wire net in the middle of the tube it will tend to 

 increase the pressure of the gas when it is a minimum, 

 i. e., it will discourage oscillations. The same will hap- 

 pen when we place the hot net above the middle. In order 

 to encourage oscillations we have to add momentum in a 

 position and at a moment such that the pressure in the 

 node increases more than it would do because of the oscil- 

 lations alone. If we put the hot wire net at the lower end 

 of the tube, i. e., in a loop, the effect will be very small, 

 or zero. The transfer of heat will depend upon the tem- 

 perature of the air in contact with the wire net, being 

 greatest when the temperature is lowest. But the tem- 

 perature in the loop at the lower end does not vary; 

 therefore, the transfer of heat in this position of the 

 gauze does not give rise to oscillations. It tends only to 

 raise the temperature of the gas uniformlj^ Heat must 

 therefore he applied between a loop and a node. 



If we cover the upper end of the tube. Fig. 3, with the 

 hot net in the most favorable position, the sound ceases, 

 and if we heat by means of a Bunsen burner the outside 

 at the top, as in 4, we get no sound. This was considered 

 by Kayleigh** as possible. But if we change the cross 

 section of the tube as in 5, and heat at p, then the tube 

 will emit a sound. The pressure in the upper half of the 

 tube will increase, partly because the air is heated, partly 

 because of the condensations of the air in the node on the 

 top. The air will expand, and now the expansion in the 

 narrow neck is aided by the air being heated by the wall. 

 Here the oscillations are encouraged because each time 

 when the air is expanding by the oscillation the expansion 

 is increased by the heat. In each cycle the vibrating 



** Theory of Sound. Vol. II, v>. 231. 



