1897-98. Prof. M‘Kendrick on Theories of Voivel Sounds. 8,7 
Lord Rayleigh shows that even then it does exist in the external 
air. 
General Conclusion . — It would appear, on the whole, that the 
truth does not rest entirely on the side of one theory, but that 
both are partially true. The view of Donders, that each vowel 
has a mouth cavity of unchangeable and fixed pitch, is too exclusive, 
and, on the other hand, it cannot be denied (as is virtually done 
by those who uphold the relative pitch theory) that each vowel has 
a predominant partial or predominant partials which give it a 
definite character, and which must be produced by the mouth 
cavity as a whole, or by the double resonance of portions of the 
cavity, as contended by Lloyd. When a vowel is spoken either 
separately or as it exists in a word, the complex tone is produced 
by the vocal cords. The resonance chambers above assume a 
certain form, and one or more partials peculiar to the vowel are so 
strengthened as to give such character to the vowel-tone as to enable 
the ear at once to identify it. As the form of the mouth cavity 
varies in individuals and in different races, the partials may not be 
identical in each case, but they will be so near a common pitch as 
to enable the ear at once to identify the vowel. If, however, men 
of different nationalities attempt to fix the absolute pitch of the 
partial they will not agree, as has been found to be the case. 
Again, in singing a vowel in a scale, very slight alterations in the 
form of the resonance cavities take place. In passing from the 
lower to the higher notes the larynx rises, and more or less of a 
muscular strain acts on the walls of the resonance cavities. Thus 
there must be a slight change in the volume of these cavities lead- 
ing to the development of partials different from those formed by 
the cavity at rest or in speech, but still so sufficiently near as no 
materially to alter the quality of the tone. Consequently the ear 
still recognises the tone of the vowel, even when sung at a high 
pitch, and it may even recognise the special quality of a particular 
voice. Sometimes, though rarely, the quality may become richer 
as the voice rises in pitch ; usually for each voice there is a register 
in which the voice has its maximum of good quality. It is not 
surprising, therefore, that a singer almost instinctively chooses such 
vowels as best suit the resonating arrangements of his or her voice, 
and avoids tones, vowels, and words containing vowels that would 
force the production of notes of inferior quality. 
