PHONO GRAMS. 1225 



phonograph, and the curves were submitted to harmonic analysis. This 

 was the beginning of the present discussion. These two observers 

 obtained good curves, even with the imperfect instrument, and the 

 curves were submitted to analysis, so as to determine the amplitude of 

 their constituents up to the sixth partial, and the process of measure- 

 ment and calculation was applied to more than one hundred curves. The 

 sixth partial was not a high limit to reach, but the phonograph, at that 

 date, did not record the higher partials. For example, it could not 

 reproduce the sound of I in machine, nor the French or German U. 

 Twelve values for y (the lengths of the co-ordinates for one period chosen) 

 gave the data for calculating the amplitude and phases of the first six 

 partials. Professor Tait supplied the authors with the solutions of the 

 simultaneous equations for twelve values of y, and the results are given 

 in a series of tables, in which the predominance of certain partials is 

 unmistakable. In this way each vowel was examined. The conclusions 

 arrived at were, on the whole, favourable to the constant-pitch theory, 

 and were thus summed up : " It is clear that the quality of a vowel 

 sound does not depend either on the absolute pitch of reinforcement of 

 the constituent tones alone, or on the simple grouping of relative partials 

 independently of pitch. Before the constituents of a vowel can be 

 assigned, the pitch of the prime must be given ; and, on the other hand, 

 the pitch of the most strongly reinforced partial is not alone sufficient 

 to allow us to name the vowel. To do this, we must also know the 

 relation of the constituent partials to one another." 1 Again, " The ear 

 is guided by two factors, one depending on the harmony or group of 

 relative partials, and the other on the absolute pitch of the reinforced 

 constituents. The ear recognises, as it were, when it hears a vowel sung, 

 at any pitch, the kind of oral cavity causing the reinforcement. As to 

 the question, " Is the resonance cavity for a vowel sound constant at all 

 pitches ? " the authors are cautious ; and their opinion appears to be that 

 whilst the resonance cavity for a vowel sound has an absolute pitch, it 

 may have a certain effect in reinforcing other subordinate tones or partials. 

 While experimentally it may be shown that a constant cavity may pro- 

 duce a vowel-like tone, say 0, over a wide range of pitch, it is probable 

 that the resonant cavities of the human being are slightly adjustable, so 

 as to be, as it were, tuned to the pitch on which the vowel is sung. In 

 other words, Fleeming Jenkin and Ewing hold that both the relative 

 and the absolute factors enter into the composition of a vowel, a 

 conclusion not far from the truth. 



The subject was taken up by Hermann 2 about 1890, and he used the 

 much improved wax-cylinder phonograph. He succeeded in obtaining 

 photographs of the curves on the wax cylinder, a beam of light reflected 

 from a small mirror attached to the vibrating disc of the phonograph 

 being allowed to fall on a sensitive plate while the phonograph was 

 slowly travelling. The curves thus obtained, representing the wave- 

 forms of the vowel tones (Fig. 438), 3 were submitted to analysis, with 

 the view of estimating the pitch of the mean partial or the formant, 

 as it is called by Hermann, according to the method already 

 described. Hermann also pointed out that the quality of a vowel tone 

 varies considerably, according to the rate at which the cylinder was 



1 Jenkin and Ewing, op. cit., p. 770. " Op. eft. 



3 The curves in Figs. 438 and 441 were kindly sent to Dr. M'Kendrick by Professor 

 Hermann. They are here introduced simply as examples. 



