MECHANICS OF THE HUMAN VOICE ' 259 



Except in very long vowels, the ear cannot hear any change, but 

 gets a general impression. This is necessarily so crude that trained 

 phoneticians will assume utterly different vowels in such a word as 

 "not," even when spoken on the same occasion by the same 

 speaker. The changes within the vowel lie, of course, beyond the 

 grasp of the ear. This new machine will furnish the means of an 

 acoustic analysis of the vowels in the following manner: The first 

 wave of the vowel, as it appears in the tracing, is engraved on the 

 wooden cylinder in a continuous repetition. The reproduction on 

 the wax cylinder then gives continuously the sound for the first 

 wave of the vowel. This is done for each wave in succession. In 

 this way the acoustic vowel elements are determined throughout a 

 vowel. The cost of this apparatus can be placed at $300. 



A special measuring machine should be constructed. At present 

 the curves are enlarged by photography and measured directly in 

 tenths of a millimeter. To avoid the enlargement, an apparatus 

 was designed in which a magnifying glass can be moved by milli- 

 meter screws in two directions above the curve without touching it, 

 the readings being in hundredths of a millimeter. The construction 

 was not begun, as the estimated cost (mainly for the fine screws) 

 was $200. It should be made at once, however, as the photography 

 is a heavy running expense, which can be avoided once for all by 

 the apparatus. 



6. A mathematical treatment of the phenomena in resonators with 

 soft walls and of the vibrations of soft bodies should be sought. 

 In the works of Helmholtz and Rayleigh these problems are quite 

 overlooked, and yet the human voice — whose vibrations are pro- 

 duced in this way — is the most important of musical instruments. 

 L,ord Rayleigh, in the second edition of his Theory of Sound, has 

 shown great interest in the study of vowel sounds. I suggest that 

 he be requested by the Carnegie Institution directly to treat the 

 subject. I would take great pleasure in laying before him any 

 desired curves of speech or results and in carrying out any experi- 

 ments, measurements, or computations he may wish. If he is not 

 willing to do the work I shall be pleased to cooperate with any one 

 whom the Institution may suggest. Presumably, no cost will be 

 involved unless Lord Rayleigh desires a personal interview in 

 England. In this case traveling expenses will have to be covered. 



