WAVE ANALYSIS IN REFERENCE TO VOCAL ACTION. 115 



cavities were made by spreading meat over wire frames.. Innumerable 

 varieties of [u], [o], and [a] could be made with a wooden reed and soft 

 resonators. They were of a smoothness and beauty rarely attained by 

 the human voice. All attempts at vowels hke [e] and [i] failed, no matter 

 how small the resonators were. This led to the supposition that different 

 voM'els require different glottal action. Instead of reeds, rubber mem- 

 branes, rubber cushions, etc., were used, whose mode of vibration would 

 depend on the action of the air in the cavities above them. Even with 

 hard resonators it was possible to get almost any sort of tone out of the 

 membranes. Through a hole in the resonator they could be seen to 

 change their vibrations as the cavity was changed. All the vowels could 



be imitated by building up the resonators in 

 sets of two or three below and above the vibra- 

 ting membrane. This is a support for my sup- 

 position that the glottal lips vibrate differently 

 for the different vowels, but here the difference 

 was brought about by reaction of the rubber 

 glottis to the tones of hard cavities, whereas in 

 the larynx it presumably occurs by nervously 

 aroused contractions of the fibers of the muscles 



Fio. 105.— Top view of glottis. . i ,. i i- 



in the glottal hps. 



Let us now attempt a complete statement of the puff theory. 



Physiologically stated, the action for a vowel is as follows : Each 

 glottal lip consists mainly of a mass of muscle supported at the ends 

 and along the lateral side (figure 105). It bears no resemblance to a 

 membrane or a string. The two hps come together at their front ends, 

 but diverge to the rear. The rear ends are attached to the arytenoid 

 cartilages. When the ends are brought together by rotation of these car- 

 tilages, the medial surfaces touch. At the same time they are stretched 

 by the action of the cricothyroid muscles, which pull apart the points of 

 support at the ends. 



In this way the two masses of muscle close the air passage. To pro- 

 duce a vowel such a relation of air-pressure and glottal tension is arranged 

 that the air from the trachea bursts the muscles apart for a moment, after 

 which they close again; the release of the puff of air reduces the pressure 

 in the trachea and they remain closed until the pressure is again sufficient 

 to burst them apart. With appropriate adjustments of the laryngeal 

 muscles and air-pressure this is kept up indefinitely, and a series of puffs 

 from the larynx is produced. The glottal hps open partly by yielding 

 sidewise— that is,' they are compressed— and partly by being shoved 



