188 M. F. Savart on some Acoustic Phenomena produced by 



in height and 0*041 metre in diameter, having an efflux tube 

 whose height and diameter were each 2'15 millims. On observing 

 tlie issuing jet during the several phases of the phfenomenon, it 

 was seen to swell out considerably whenever the tone attained its 

 maximum intensity, and to contract every time this intensity 

 decreased. In the first case the jet appeared to open, in the 

 second to close. 



As the charge diminished from 2'66 metres to 0'15 metre, 

 the tone lowered through two octaves; but this fall was not 

 gradual, it took place by fits, so that the whole series ccmstituted 

 a number of ventral segments, for each of which the tone was 

 nearly constant, except when the charge was feeble. These 

 ventral segments were separated by nodes, where the tone, with- 

 out disappearing altogether, was much less intense, and reduced 

 to a kind of rustling noise. At first this appeared to be inca- 

 pable of analysis by the ear ; nevertheless, on closer examination 

 the noise appeared to be composed of two tones, one being that 

 of the superior ventral segment, the other that of the inferior ; 

 from which it is evident that the nodes are the points where the 

 state of vibration which constitutes a certain ventral segment 

 becomes modified, in passing to that which constitutes the ven- 

 tral segment immediately succeeding. 



The number of such ventral segments deci'eases when the 

 aperture is enlarged. In the above case there were fifteen, and 

 the vessel was emptied in 315 seconds. In another experiment 

 with the same glass tube, but with an efflux tube whose height 

 was 5'456 millims., and diameter 5*4' millims., there were only 

 four such ventral segments, and the vessel was emptied in 

 47 seconds. In the former experiment, the tone correspond- 

 ing to the charge 2'66 metres was more than an octave higher 

 than in the latter ; the lowest audible tone in the latter experi- 

 ment corresponded to the charge 57 millim. When the vessel 

 emptied itself with great velocity, each ventral segment presented 

 this peculiarity, that the tone after reaching its maximum in- 

 tensity always rose a little higher in pitch, sometimes, indeed, a 

 semitone. 



By comparing the numbers of vibrations corresponding to each 

 pressure with the square roots of the pressures themselves, the 

 former were found to be in general proportional to the latter, or 

 to the velocity of efflux. It is clear that this law cannot be 

 rigorously exact for all pressures, because the tone remains 

 almost invariable throughout each ventral segment ; nevertheless 

 its existence was on the whole quite manifest. The degree ot 

 exactitude may be seen from the following table: — 



