DYNAMICAL STUDY OF THE VOWEL SOUNDS 



101 



Fig. la is a diagram of a double resonator. The volumes of the 

 chambers are respectively Vi and Vi; the conductivities^ of the 

 orifices are Ki and K2. In this structure the outer orifice corresponds 

 to the mouth (see Fig. \b), the outer cavity to the buccal cavity, the 



la \b 



Fig. la and \b — Diagram of the mouth-pharynx system 



inner orifice to the constriction between the soft palate and the back 

 of the tongue, and the inner cavity to the pharynx. The source of 

 sound in the back of the inner chamber is of course the glottis, or 

 rather the periodic pufifs of air to which the glottis gives rise, and we 

 may remark that at resonance the apparatus is driven at a node (or 

 pressure maximum) which is a condition for maximum efificiency. In 

 Paget's models, a small opening was made at the back for the source 

 of sound, which was a loosely stretched strip of rubber, mounted in a 

 slit, and blown by an air stream. To be successful, in connection with 

 the resonator model, in producing a vowel sound artificially, such a 

 source must of course generate a sound whose fundamental is some- 

 where near that of the human larynx, and which has in addition a very 

 extended range of harmonics; that is, for a bass voice, we should need 

 a fundamental frequency of about 100, and additional sound energy 

 scattered through the frequency range up to 4,000 or 5,000 cycles.'* 



^ The average mass of air which surges to and fro in the orifice of a resonator is 

 pS'/K, in which p is the density of air, S the area of the orifice, and K the conductivity. 

 K" is a linear quantity, proportional to the width of the orifice, and is a measure of the 

 ease of flow of fluid through it. It may be defined as the ratio of the (velocity) 

 potential difference, between the two ends of the orifice, and the flux or current (S^) 

 flowing through the orifice. 



^ Sacia suggests that a source of sound giving a saw-toothed wave (rip saw tooth: 

 one slanting and one vertical side) should be ideal for driving vowel resonators- 

 (An experiment with such a device will be described later.) This wave shape 

 corresponds to a fundamental and full retinue of harmonic tones, and should be of 

 service in many ways in acoustic experiments. 



