22 
NALTORE ~ 
[January 6, 1923 

not all in the same direction. From a (calm) onward 
the resonances go down, as it were, hand-in-hand, 
keeping an equal distance of about 8 semitones apart ; 
and it is possible, owing to this fact, that they have 
not been generally recognised as separate resonances. 
The double resonance of the oral cavity when forming 
the vowel sound u (who) may be demonstrated by the 
clapping method (see Nature, March 16, vol. 109, 
p- 341) ; also the possibility of varying both resonances 
independently at the same time. Similarly, the inde- 
pendence of the larynx note and the front resonance 
may be illustrated by simultaneously humming and 
whistling a convergent scale. 
Having identified the various resonances on which 





Seeesearc 
eee 
Pa a 






ro 
NO 
ON |S 






SUETSUONTRRRNON 
PEEPEEEEEEEEEE ET 















| ment. 















Fic. 2.—Vowel resonance chart. 
lines denote the actual resonances of the models ; the numerals i, ii denote 
successive models of the same vowel, 
the production of the breathed vowels appeared to 
depend, the attempt was made to reproduce these 
vowel sounds by constructing some form of resonator 
which had resonances identical with those of the 
human mouth when a stream of air was blown through 
it. It seemed reasonable to expect that, if this could 
be done, the breathed vowel sounds would be reproduced. 
Models in plasticine were therefore made, the internal 
form of which very roughly imitated that of the human 
mouth and throat, except that the back portion corre- 
sponding to the pharynx was, for convenience, shortened 
and made more bulbous. 
With this and similar models a number of experiments 
were made to test the effects of various alterations of 
the internal form—such as are actually made in the 
human mouth by the movement of the tongue, lips 
etc—and to discover the rules for tuning the instru- 
NO. 2775, VOL. 111] 
The horizontal dashes on the thick: vertical 
An artificial larynx was made of a rubber strip 
lying across a flattened air passage—on the principle 
of the reed instrument which boys make with a blade 
of grass held between their two thumbs. When this 
reed or larynx was fitted to the back orifice of the 
model and blown, the model gave a voiced vowel. 
The rules for tuning these models may be shortly 
summarised as follows : Enlarging the mouth generally 
raises both resonances. Increasing the projection of 
the lips or reducing the size of the mouth lowers both 
resonances. Raising the front of the tongue upwards 
or forwards raises the upper resonance but lowers the 
lower resonance. Pressing the back portion of the 
tongue backwards—so as to reduce the capacity of 
the back cavity corresponding to the human pharynx 
and to prolong the passage between the front and 
back cavities of the mouth—raises the lower resonance 
but lowers the upper resonance. 
The experiments in tuning the plasticine cavities 
eventually made it clear that the human mouth, when 
making vowel sounds, always acts as two separate 
“ Helmholtz” resonators connected in series—one behind 
the other—the back resonator being formed by the 
pharynx, the back of the tongue, and soft palate; the 
front resonator being formed by the front of the tongue, 
the hard palate, and lips; and the passage between 
the two resonators being formed by a hump in the 
middle of the tongue which approaches the roof of 
the mouth. By humping the tongue in different 
positions—forward or backward—the relative sizes of 
the front and back resonators can be altered at will, 
while the tuning can further be modified over a wide 
range by varying the opening of the mouth. 
The resonant note of a cavity with an orifice to the 
open air depends, as is well known, on the relation 
between the volume of the cavity and the size of its 
orifices. The larger the cavity the lower the note ; 
the larger the orifice the higher the note. With a 
resonant cavity having a neck—such as the neck of 
a bottle—the resonant pitch also depends on the length 
of the neck, being lower as the neck is made longer, 
and higher as the neck is shortened. 
It follows from this that when two such resonant 
cavities are joined together, each one becomes, as it 
were, a neck to the other, and therefore influences its 
pitch. The effect is always to lower more or less the 
resonance of the neighbouring resonator according to 
the relation of the relative sizes of the two, and of the 
relative sizes of the connecting opening between the 
resonators and the opening to the air of the front 
resonator. The pitch of the resonators was ascertained 
by tapping them and listening to the resonant notes, 
or by blowing across the open mouth. 
Each of the plasticine models (Fig. 3) made on this 
principle gives two resonances corresponding to a 
separate vowel sound. When the various models are 
blown in succession, first by mouth and afterwards 
for @ (earth) and 9 (all) by bellows, the vowel characters 
are made more recognisable by covering and uncovering 
the mouth of the model by hand during blowing, so 
as to give an associated consonant (m or w). It was 
thus demonstrated that the vowel sound remains 
appreciably constant however much the pitch of the 
larynx note is altered by varying the air pressure. 
Instead of putting the two resonators in series, as 
