WAVE ANALYSIS IN REFERENCE TO VOCAL ACTION. 109 



According to ^Vheatstone, Grassmann and Helmholtz the glottal 

 lips vibrate after the manner of strings and produce a series of partials 

 of which the first (or fundamental) — that is, the tone of the voice — is 

 the strongest. The series of cavities above the glottis reinforce certain 

 of these partials; for each vowel the cavities are readjusted and a differ- 

 ent set of partials is reinforced. The curve of a vowel is therefore the 

 sum of the curves for the series of the partials of the voice, some of these 

 partials being reinforced by the resonance. The curve will contain the 

 first partial (fundamental) as the strongest element, because the glottal 

 tone is in every case by far the loudest; whatever we may lose of a 

 sung or a spoken vowel by deafness or distance, we always hear the glottal 

 tone if we hear anything. For example, let us suppose the glottis to 

 emit a tone whose partials are in the relations of strength indicated by the 

 sizes of the numbers in figure 98. A vowel would increase the strength 

 of some of the partials, for example, as indicated in figure 99. 



9 lo 



12 3456789. 12345678 



Fig. 98.-Scheme of partial tones. Fi«- 99 -Scheme of reinforced 



partials. 



The overtone theory of the vowels can not be correct. The following 

 reasons seem conclusive. 



If the theory were correct, the harmonic analysis into a series of 

 simple sinusoids would be the proper one, and its results would give the 

 series of partials with their proper ampUtudes. The analyses made by 

 various investigators show that in most cases the first partial is very weak 

 or entirely lacking, that is, it gives a result that is known to be false 

 because the ear hears just this partial as by far the strongest of all. Either 

 the theory is incorrect or the method of obtaining the curve introduces 

 a falsification. As long as the curves were obtainable only by the phon- 

 autograph, the latter might be supposed to be the case. With the curves 

 obtained by the tracing methods from phonograph and gramophone 

 records a new test can be made. The chief tone of the voice, the glottal 

 tone or first partial, is present in the records and can be heard as distinctly 

 as in the original voice; in fact, as long as anything can be heard from 

 a record — whether weakened by distance or by wear — just this tone 

 remains. Yet a harmonic analysis of the curve of such a record usually 

 shows the absence of this tone. One investigator was led to the remark 

 that "the phonograph must be deaf to the glottal tone" and failed 

 to see that this must be absurd because as long as the phonograph 



