Barrett — Note on Combination- Tones. 371 



As regards the difference-tone emitted by the sensitive flame, 

 we have here an audible vibration set up in the flame itself, the 

 greater amplitude of the waves recurring at the beats exciting a 

 more vigorous disturbance of the flame. 1 The fact of the flame 

 speaking, i.e., giving out a note of its own corresponding to the 

 impressed disturbance, indicates that the issue of the stream of 

 gas through the orifice is rendered periodically variable. 



Now, a jet of air or a flame can act as a resonant jar ; " a 

 tuning-fork vibrating feebly, and presented to the jet, is loudly 

 heard," or the ticking of a watch is reproduced by a sensitive 

 flame so loudly as to be heard all over a large lecture-room. 



How the disintegration of the flame occurs under the influence 

 of sound is not yet clearly ascertained. The disturbances by 

 which the equilibrium of a sensitive jet is upset are certainly 

 impressed on the stream of gas or liquid as it leaves the orifice ; 

 hence the root of the flame is the seat of its sensitiveness. Lord 

 Eayleigh's experiments show that a sensitive flame is affected at 

 the loops of a sound-wave, where the variation in motion of the 

 air-particles is greatest, and not at the nodes, where the variation 

 in pressure is greatest ; and somewhat doubtfully he was led to the 

 conclusion that the disintegration of a sensitive flame was due to 

 an increased sinuosity given to the issuing stream of gas by the 

 sonorous disturbance ; " the necessity, as remarked by Barrett, 

 for an unsymmetrical orifice points strongly in this direction." 2 

 For my own part, I am disposed to think we have not yet arrived 

 at any adequate explanation of the phenomenon. Some experi- 

 ments indicate that the effect is due to symmetrical swellings 

 about an axis on the issuing stream — varicosity, as Lord Rayleigh 



doubt here — that if there had been two 256 puffs per second, they might have been 

 arranged to give equally -spaced puffs 512 per second, and so the summation-tone 

 would have been heard. Just as with the water-jet difference-tone, where the 

 difference-tone alone is heard, and not the primaries, here also, I imagine, the 

 summation-tone alone would be heard, and not the primaries." 



1 So far as regards notes nearly in unison, this was pointed out in my papers, and 

 shown in my lecture to this Society thirty -six years ago — a sensitive flame visibly 

 throbbing to the beats of two tuning-forks nearly in unison, even when the sound of 

 the forks was barely audible. This is of course due to the fact that the energy of the 

 sonorous vibrations which affect the flame is proportional to the square of their 

 amplitudes. 



2 " Theory of Sound," vol. ii., p. 402. 



