EXPERIMENTAL PHONETICS. 257 



each coMsistin^^ of two slits; and for i and ou there are many slits, 

 without these being arranged in groups. The slits are very large for 

 o and narrow for k, and large for ou and narrow for i. He then 

 molded a series of casts of the interior of the oral and pharyngeal 

 cayities of a human subject, as these were adapted for the singing of 

 the different vowels, and from them constructed masks or headpieces 

 which could l)e placed over the syren so that the air escaping from it 

 passed through the cavities of the mask. He found that if air was 

 driven through the masks under a pressure of only 7 centimeters of 

 water the timlire of the corresponding vowel was at once perceived, 

 as in whispering. Marage's view is that to form a vowel the true 

 vocal cords vibrate in a horizontal plane, in such a way as to influence 

 by their greater or less degree of approximation the escape of air. 



If the air escapes in three little putt's as it were (the cords vi])i'ating 

 during each putf a number of times equal to the pitch of the note on 

 which the vowel is spoken or sung), so that there are inttn-vals between 

 the groups of putfs, then the vowel a is the result. The oral resonater 

 is in unison with the sum of the vibrations and the vowel is emitted. 

 If the resonater (eithiM- artiticial or the oral cavit}', as in life) is turned 

 to the third harmonic of this note, then the vowel a is uioditied; the 

 same applies to e and o, which have the second harmonic, and in 

 passing from the one vowel to the other it is sufficient to change the 

 aperature of the glottic opening. Thus for a, if the fundamental note 

 is J), the oral resonator nuist l)e tuned to 3;/; for Vj and o, if the funda- 

 mental is /)', the oral resonator gives 2//; and for land ou the resonater 

 is in unison. If this is not so, then the (juality of the vowel is nmch 

 altered. Thus if the syren gives a, and the plate used is that for ou, 

 then the sound is a moditied. This agrees with the experience of 

 teachers of singing, who hold that a badly sung vowel is a vowel 

 sound emitted into a cavity adjusted for another vowel. Marage has 

 also found that wdien the sounds of his syren, aid<Kl I)y the masks, are 

 examined l)y the manometric method, the flamc^ pictui-es ajipear as 

 they may be expected to do — that is, groups of three flames for a, of 

 two for E, EU, and o, and of one for i, u, and ou. Vowels then, 

 according to him. are due to an intermittent aero-laryngeal vibration, 

 strengthened by the oral cavity and producing ou, o, a, p:, and i, 

 when it is in unison with the sum of the \'il)i'ations; transformed by 

 it, and giving origin to other vowels, when there is no unison; and 

 the number of intermittences gives the fundanuMital notc^ on which the 

 vowel is emitted. If the oi'al cavity acts alone, the Nowel is whispered; 

 if the larynx acts alone, the \owel is sung; and if the two act, the vowel 

 is spoken. Marage has applied his method with much success in test- 

 ing the ear and in the treatment of unites who are not absolutely deaf. 

 His memoir is characterized l)y great simplicity and at the same time 

 by thoroughness. ^ 



SM 1902 17 



