52 THE STUDY OF SPEECH CURVES. 



dinate vibrations is one wave-group representing one puff from the glottis 

 and that the systematic changes in amplitude have some special cause. I 

 have no final explanation to offer for this phenomenon, but I venture the 

 suggestion that the smaller groups arise from vibrations of the muscular 

 glottis, while the larger groups come from vibrations of the whole mass, 

 including the vocal muscles and the arytenoid cartilages; that is, of the 

 entire glottis, cartilaginous as well as muscular. In an ordinary vowel 

 the whole glottis is supposed to vibrate in the chest register and only the 

 muscular glottis in the head register. Here, although the whole glottis 

 may vibrate with a longer period, the muscular glottis vibrates in addition 

 with a shorter period. Such a combination of two vibrations is not diffi- 

 cult to imitate mechanically. This action may perhaps arise on account 

 of the preceding [h], which may be sonant in its latter portion. If this 

 sonant portion is produced by the muscular glottis vibrating while the 

 cartilaginous glottis is open for the air to pass with aspiration, the vowel 

 might be produced by closing the latter somewhat more, whereby the 

 entire mass might come into vibration without stopping the vibrations 

 of the muscular glottis. In the last part of the line the vibrations die 

 down rapidly; the pitch also falls. The smaller vibrations toward the end 

 of the first line and in the second line show signs of conflicting grouping 

 suggestive of conditions similar to those at the beginning; there is prob- 

 ably present some such sound as sonant [h], although the ear is quite 

 unable to decide what the sound is. The curve of the second laugh shows 

 vibrations in many respects similar; there is, however, a tendency to fall 

 irregularly into groups of three instead of two, or even into groups of 

 four, throughout the whole curve. We note far less change from wave to 

 wave; in fact, not more than in vowels of ordinary speech. The unusual 

 curves found for these laughs may have some other explanation than the 

 one I have suggested; in particular, we may note the possibiUty of vibra- 

 tions or flappings of other parts of the vocal apparatus, namely, the 

 ventricular bands or the epiglottis. A somewhat similar phenomenon 

 has been observed in the curve of " So," spoken by Jefferson (plate vii of 

 "Elements of Experimental Phonetics"). This word was spoken after 

 a toast and the sound is hke that produced when a particle of hquid 

 remains in the throat. It may have been produced in this case by 

 some flapping of the ventricular bands or the epiglottis. Still another 

 example of tliis grouping of waves occurs in the curve of the trill as 

 already noted on p. 36. 



The quaUtative analysis in all these examples has been a purely 

 descriptive one made by the eye, yet it advances far beyond the older 



