1 2 2 o ON VO CA L SO UNDS. 



Each of these pictures is composed of two groups of teeth, reproduced 

 again and again, and thus showing a regular periodic vibratory move- 

 ment. The three pictures of ut l contain eight groups of teeth, while 

 those of soZ x contain twelve, and those of ut., sixteen. As the length of 

 all these pictures corresponds to the same duration of a vibratory state, 

 it is evident that the period of the vibratory movement, or the length 

 of the wave characteristic of a tone of determinate pitch, is independent 

 of the vowel upon which the sound is emitted. But the form of the wave 

 characteristic of the tone of given pitch varies much with the vowel upon 

 which it is sung. This alteration of the form of the wave, while the period 

 is constant, must be due to the superposition of a tone developed in 

 the mouth, characteristic of the vowel, upon the tone emitted by the 

 larynx. 1 



According to Konig, OU is always easily emitted with st 2 f>, and 

 the neighbouring tones la.,, sol 2 with the lower partials, si^, mi x \). 

 is emitted with si 3 r>, the partials la 3 , sol 3 , and the lower partials 

 si 2 \), mi 2 \f, s\\>. A comes with s%\), with la it with sol^, along with the 

 lower partials, si 3 \>, mi 3 \), si 2 \>. E and / are not easily sounded on low 

 tones, as their characteristic partials are very high. On tones lower than 

 ut 3 , the female voice turns involuntarily to or OU, which have sijp and 

 si 2 \} as their characteristic partials. Above /a 4 it is A, characterised 

 by s%\}, which is most readily given. Above si 4 the voice passes into E 

 and I. These observations of Konig are harmonious with the conclu- 

 sions of Helmholtz, and favour the fixed pitch theory. 



Hallock 2 has recently employed a device founded on that of Konig. 

 Eight resonators in harmonic series were each connected with a mano- 

 metric capsule, and photographs were taken of the eight bands of flame 

 pictures, reflected in a mirror, when the vowels were sung before the 

 resonators. 3 From these photographs, the partials present in any vowel 

 tone within the range of the resonators could be determined. Of course 

 the higher partials, on which, as pointed out by Helmholtz, much 

 depends, were not detected. 



Wave-forms of vowel tones.— We must now turn to the evidence 

 adduced by an analysis of the wave-forms of vowel tones. To appreciate 

 this evidence, the following statements must be kept in view : — 



1. Pitch depends on the length of time in which each single vibration 

 is executed, or, in other words, on the vibrational number of the tone. 



2. Musical tones are higher, the greater their vibrational number ; 

 that is to say, the shorter the vibrational period. 



3. The sensation excited by a periodic vibration is a musical tone. 

 This tone is usually compound, the constituents being partial tones. 



4. Only one form of vibration, like that of a pendulum or the limbs 

 of a tuning-fork, can give rise to a simple tone, destitute of partials. 

 This is a simple pendular vibration, and the sensation is a simple 

 tone. 



5. A compound tone is the sensation produced by the simultaneous 

 action of several simple tones, with definite ratio of pitch. Such a 

 compound tone corresponds physically to a wave of more or less 

 complex form. 



1 Gavarret, " PWnomenes physiques de la phnnation et de l'audition," Paris, 1877, 

 p. 394. 



2 Am. Ann. Photo., 1896, p. 21. 



3 For a figure of such an analyser, see M'Kendrick's " Physiology," fig. 418, vol. ii. p. 686. 



