394 
NATURE 
[ FEBRUARY 21, 1907 
By a special method Dr. Scripture plots a melody 
curve from a transcribed record, showing, for 
example, the curve when ‘‘ Oh”? is uttered ‘‘ sorrow- 
fully,”’ or ‘‘ admiringly,’’ or ‘‘ questioningly,’’ &c. He 
works out the ‘‘ melody curves ’’ in Depew’s speech, 
and then writes the melody in musical notation. With 
regard to the emphasis of speech as indicating the 
emotional condition of the speaker’s mind, we must, 
however, take into account not only melody, or the 
sequence of tones of different pitch, but also the 
intensity, the passing from diminuendo to crescendo, 
or vice versa. Dr. Scripture has not attributed suf- 
ficient importance to this element in the analysis. 
The amplitudes of the wave forms increase or 
diminish according to the intensity. 
Dr. Scripture expounds the principles of harmonic 
and inharmonic analysis in two chapters at great 
length and with much clearness. Nowhere have I 
met with a fuller exposition of Fourier’s theorem 
and its application to acoustical problems. He does 
not hold, however, that a vowel curve is produced by 
combining simple sinusoid vibrations in a harmonic 
series, and he concludes that 
“the sounds from the musical instruments are presumably 
produced in this way, but we dare not assume that the 
vowels are so produced until the fact has been proven ”’ 
(p. 78). 
He shows how to separate, by the rules of Her- 
mann, harmonic and inharmoniec sinusoids from the 
mixed results of a harmonic analysis. How is one 
vowel distinguished from another? Are the differ- 
ences due to the presence of certain tones of definite 
pitch, as held by both Helmholtz and Hermann? If 
so, are we to hold with Helmholtz that these tones 
are harmonic overtones of the glottal tone or that 
they are inharmonic to it, as stated by Hermann? 
Dr. Scripture holds that Hermann has completely 
disproved the theory of Helmholtz. After discussing 
the method of analysis with frictional sinusoids, as 
distinguished from simple sinusoids, he states :— 
“The vibrations of the voice in speech are... com- 
posed exclusively of frictional sinusoids and not of simple 
sinusoids, as has hitherto been assumed. Can a method 
of analysis into frictional sinusoids be found? Does an 
analysis into simple sinusoids give false results for the 
vowel curves? ”’ (p. 101). 
He answers the question thus :— 
“The treatment of the curves by simple harmonic 
analysis—the only method that has hitherto been tried— 
furnishes results that are so wrong as to be utterly mis- 
leading when used to indicate the manner in which the 
vibrations were produced.”’ 
I observe that Dr. Scripture states that Prof. 
Weber, of the Swiss Polytechnicum, along with 
Schneebeli, was the first to apply the Fourierian 
analysis to a vowel curve, but he does not give the 
date when this was done. We must not forget that 
such an analysis was made by Fleeming Jenkin and 
Ewing in 1878 (“On the Harmonic Analysis of 
certain Vowel Sounds,” Trans. Roy. Soc. Edin., vol. 
XXVH1., Pp. 745). 
As to the mode of production of vowel tones, Dr. 
Scripture discards the views of Wheatstone, Grass- 
mann, and Helmholtz that the glottal lips vibrate 
after the manner of strings or the borders of a mem- 
brane on each side of a narrow opening, and he fully 
adopts the ‘‘ puff ’’ theory of Willis and Hermann, 
according to which 
“the glottis emits a series of more or less sharp puffs; 
each puff, striking a vocal cavity, produces a vibration 
whose period is that of the cavity; a single wave-group 
shows the sum of these vibrations from all the cavities ; 
the periods of these vibrations may stand in any relation 
to the interval at which the pulfs come, that is, to the 
fundamental.”’ 
NO. 1947, VOL. 75] 
There can be little doubt that, at all events in his 
later days, Helmholtz saw the analogy between the 
action of the glottis and the ‘‘ puffing ’’ sounds of a 
syren, but he undoubtedly held that the overtones 
were harmonics of the glottal tone. Hermann, how- 
ever, has conclusively shown that at least some of the 
tones of the cavities may be inharmonic to the glottal 
tone, and Dr. Scripture supports this view by many 
ingenious experiments. His description, however, of 
the glottis is not either anatomically or physiologically: 
quite satisfactory. It is not in accordance with 
anatomical detail to write, ‘‘ Each glottal lip consists 
mainly of a mass of muscle supported at the ends 
and along the lateral side,’’ or that ‘‘ the two masses 
of muscle close the air passage,’’ or that the air from 
the trachea ‘‘ bursts the muscles apart.’’ The glottis 
is a much more delicate structure than these words 
would imply. It contains much elastic tissue at the 
borders which come together, according to the “ puff ”’ 
theory, and the muscular structures are devoted to 
placing strains on this tissue and to separating or 
approximating the lips of the glottis. Dr. Scripture’s 
view is that 
“the effect of each puff on each element of the vocal 
cavity is double: first, to arouse in it a vibration of a 
period depending on the cavity; second, to force on it a 
vibration of the same period as that of the set of puffs.” 
The glottal puff produces a frictional sinusoid with 
large amplitude and a very large coefficient of fric- 
tion, and the cavity vibrations are also of the fric- 
tional sinusoid form. This may explain the failure of 
a simple harmonic analysis to reveal the real elements 
of the vowel curve. 
In chapter ix. Dr. Scripture gives his views as to 
the action of the organ of Corti in relation to wave 
analysis, and he conjectures that portions of it are 
affected by ‘* groups of stimuli,’’ when complex wave 
forms reach it. This does not seem very conclusive, 
and in my judgment the theory of Helmholtz, by 
which he explained the action of the organ by adopt- 
ing the principle of resonance, still holds the field. 
Dr. Scripture has also attempted a synthesis of 
vibrations by ingenious mechanisms, by which he 
obtained curves somewhat similar to speech curves. 
There is no doubt a great future for this line of ex- 
perimental research. After fully worked out examples 
of vowel analysis, with all arithmetical details, Dr. 
Scripture appends to the end of the monograph a 
number of elaborate schedules to assist in the 
Fourierian analysis, namely, schedules of 12 ordi- 
nates, 24 ordinates, 36 ordinates, and 72 ordinates. 
The preparation of these schedules was a_ very 
laborious task, and the work will be much appre- 
ciated. 
We congratulate Dr. Scripture on the production 
of a splendid monograph. It might have been im- 
proved by fuller bibliographical details, and perhaps 
by a more adequate recognition of the work of others. 
Joun G. McKenprick. 
AGRICULTURAL EDUCATION AND 
RESEARCH. 
NE of the functions of the Board of Agriculture 
is the administration of a Treasury grant for 
the purposes of agricultural education, and though 
the total distributed is not large it has been a potent 
factor in stimulating the development of the higher 
forms of agricultural education during the last fifteen 
years. It is certain that many of the county councils 
1 The Annual Report of the Board of Agriculture and Fisheries on the 
Distribution of Grants for Agricultural Education and Research in the 
Year 1905-6. 
