H. V. Gill — Theory of Singing Flames. 



181 



these is evidently a pressure (5 — 2) = 3 acting on the gas 

 in the direction E. This pressure forces the gas back some 

 little distance into the gas pipe, and thus the flame is either 

 made very small, or forced back into the burner with the gas, 

 or extinguished altogether. But this state of things only lasts 

 a small fraction of a second. The condensation changes into 



2. 



tit 





R 



(5-2) 

 5 



A 



K 



(5+2) 

 7 



a rarefaction, and the air expands in the direction of the small 

 arrows (II) ; when the pressure is taken off the gas it rushes 

 forth. By the same reasoning as before we see the resultant 

 pressure is (5 + 2) = 7, and that the gas issues forth under this 

 pressure. This pressure, so much greater than the normal gas 

 pressure, causes the gas to escape with great rapidity, the name 

 lights up again with a slight shock (as one remarks when he 

 lights a gas jet), this shock gives an additional impulse to the 

 expanding air, but again comes the condensation, the flame is 

 again extinguished and so on. Thus we see how the note is 

 kept sounding, a very small periodic impulse being sufficient 

 to keep a note sounding once it has begun. 



There is a point to be noted. We said the gas issued forth 

 under a pressure of 7, but it is evident that the gas will come 

 out as soon as the rarefaction is so far developed that the pres- 

 sure of the air is a little less than 3. From this it is clear that 

 the gas issues forth under a pressure considerably greater than 

 its normal pressure, and that the lighting up of the flame 

 comes at such an instant that it assists the expansion during 

 the rarefaction. 



Am. Jour. Scl— Fourth Series, Vol. IV, No. 21. 

 13 



-Sept., 1897. 



