46 THE STUDY OF SPEECH CURVES. 



otherwise there would have been a break in the even course of the pitch 

 of the glottal tone. 



The short pause after "eye" is followed by an emphasized "I" which 

 has all the characteristics of the initial vowel (see above) ; its latter portion 

 has the same form of curve, but is louder and much shorter than that 

 for the latter portion of "my." The sharp, quick fall at the end of " I" 

 is a means of emphasis. There is a short pause after "I," followed by 

 [s] of "saw him." The vibrations for [o] of "saw him" show that the 

 vowel action is similar to that for [a]; the curve passes into a curve which 

 resembles something between that of the latter half of "m}'" and that of 

 "die"; in fact we have here apparently the diphthong [oi] constructed 

 like [ai]. 



The rather weak vibrations for [i] and the still weaker ones for [m] are 

 distinctly seen in line 9. Is the breath-sound Qi] lacking? We might 

 suppose that the [h] was omitted, especially as this is generally the case 

 in such phrases in England and as the same has been asserted to be true 



Fig. 44. — Sonant [h] in the Jefferson record. 



for America. But the record can be heard to distinctly pronounce " saw 

 him" and not "saw 'im." The [h] is therefore present between [o] and 

 the following [i], although the vibrations of the glottis do not cease for an 

 instant. We thus have here distinctly a sonant [li] — prescribed by the 

 Sanskrit grammarians, long supposed to be non-existent in modern 

 languages, and pronounced an impossibility by some phoneticians. The 

 curve for this sonant \h] can be seen in the weakened vibrations after [a] 

 in line 9. Another case of sonant [h] was found in the other curve of " saw 

 him" ("Who saw him die?") in the Cock Robin record. Still another 

 was found in the Jefferson record (figure 44). The proof is the same 

 in each case; the vibrations are continued without interruption between 

 the two vowels, and yet the gramophone disc speaks a distinct [h]. 



How the breathy [li] is produced while the glottis is vibrating is a 

 question to which a decisive answer can not be given at present. One view 

 of the mechanism of sonant Qi] is that the glottis opens while the glottal 

 lips are vibrating and that this permits an escape of air with a rushing 

 noise (this does not appear in my curves) while the vibration continues. 

 This involves a fairly constant tension in the glottal lips, in spite of the 

 opening of the glottis. The action is presumably like that of the sonant 



