
( 745 ) 
XX XII.—On the Harmonic Analysis of certain Vowel Sounds. By Professor 
FLEEMING JENKIN, F.R.SS.L. & E., and J. A. Ewrne, B.Sc., F.R.S.E. 
(Plates XXXIV. to XL.) 
(Communicated June 3 and July 1, 1878. Now published, with additions down to July 19, 1878.) 
The permanent record obtained with Mr T. A. Epison’s phonograph has 
afforded a new opportunity of investigating the nature of spoken sounds, and 
the method which this invention placed at our disposal appears to us to possess 
several important advantages. 
This method consists in obtaining a magnified transcript on paper of the 
indentations impressed by spoken vowels on the tinfoil of the phonograph, and 
then subjecting the periodic wave-forms thus obtained to harmonic analysis. 
The curves as drawn in ink on paper represent to a large scale the surface 
of a longitudinal section of the tinfoil made along the centre of the furrow im- 
pressed by the pricker of the phonograph. The forms impressed on the tinfoil 
depend essentially on the movement which a particle of air performs when the 
given sound is being uttered, and the harmonic constituents of each period of 
the continuous wave-form indicate the relative proportions in which the prime 
tone and its harmonics are present in the sound, 
We thus obtain what may be called a harmonic analysis of the vowel 
sounds. 
The curves obtained are precisely those which the phonautograph is intended 
to give, and our experiments confirm the accuracy of some of the records already 
obtained by DonpErs with that instrument. We consider, however, that our 
method has a double advantage over the older plan; for, in the first place, we 
can ascertain whether the curve produced on the tinfoil really does correspond 
accurately with the sound in question, by making this very curve reproduce the 
sound ; we can thus assure ourselves that the proper periods of the disc and 
its connected parts have had no serious influence on the form of the curve; in 
the second place, the risk of this influence is much less with our arrangement 
than with the phonautograph, because the magnification of the curve which is 
necessary to allow of its examination by the eye, can with the phonographic 
record be performed at leisure, so that the motions of the marking points and 
multiplying levers can be reduced in speed until their inertia is without sensible 
influence on the curve finally registered. It is difficult to secure this condition 
with the phonautograph, especially upon high notes. 
VOL. XXVIII. PART III. 91 
