WAVE ANALYSIS IN REFERENCE TO VOCAL ACTION. 113 



in order to build up the vowels. One of the discs is shown in figure 103. 

 The form of the puff depended on the width and shape of the sUts; the 

 frequency was determined by the number of sHts that passed across the 

 blast. Resonators were held to the blast-pipe and the frequency was varied 

 by altering the speed of the motor. 



Four typical discs were used with openings of the forms indicated by 

 A, B, C, and D in figure 104. The puffs produced by them are of increas- 

 ing sharpness from A to D; their forms have not been mathematically 

 determined, but we may assume that they resemble somewhat those indi- 

 cated by the curves shown beside them in the figure. 



Various kinds of resonators were used, includ- 



fSh>rM\ '\C\f\(\ ^^S spherical brass resonators, cubical wooden 



1/1/PL/ \) \J \J\j resonators, resonators lined with felt, resonators 



'^ with walls of flesh, with walls of water, etc. 



The water resonators were made by spreading 



[][][][] '\PJ\r\j absorbent cotton inside or outside hemispheres 



B made of wire netting and uniting the two halves. 



Flesh resonators were made by spreading layers 



n n n n ^rAArA ^^ meat over cubical wire frames; the same 



U U U U ^^ frames could be used also for water resonators. 



^ For illustration a series of experiments with 



cubical wooden boxes, each having a circular 



{III Mr4-4- opening on one side, will be described. When 



D disc B was speeded up and down to produce puffs 



Fig. 101. -Siren holes and hy- of different frequencies, the resonators attached 



potheticai forms of puffs. ^^^^^ respoud loudly at a certain frequency; 



this frequency was different for each resonator. The loud response was 

 not sharply confined, but was spread over a considerable range of frequency. 

 With disc A (smoother puffs) the range of response became much smaller, 

 with disc C (sharper puffs) the range became greater, with disc D (very 

 sharp puffs) it was quite broad. At frequencies where there was no loud 

 response, there was nevertheless a peculiar quality added to the tone; 

 the sharper the puff the more strongly this quality appeared; it differed 

 for each resonator. With disc D this quality was distinctly vowel-like. 

 With a resonator of absorbent cotton soaked in water, Uttle or no special 

 resonance effect could be noticed with any of the discs, but the vocal 

 quality appeared at all times. 



It was observed that the audibiUty of the puff tone increased with 

 the sharpness of the puff. Smooth puffs sounded weak unless a resonator 

 was attached; then they became loud, with a soft vowel-Uke character. 

 With sharp puffs the puff tone appeared at all times. 



