MECHANICS OF THE HUMAN VOICE 253 



only the former palate of hard rubber, but also a specially thin one 

 of aluminum. I have also lately devised a palate of paper soaked 

 in cobalt chloride, which registers automatically the regions of con- 

 tact by changing the color of its surface. 



2. The Melody of Speech. 



Under melody we understand the rise and fall in pitch of the tone 

 from the vocal cords. The pitch of the cord tone at any instant can 

 be obtained from the speech curve by measuring the length of the 

 group of waves corresponding to one vibration of the cords. An 

 extensive study of English melody is of the highest importance for 

 oratory, for the history of the language, for the psychology of the 

 emotions, and, according to the recent discovery of Professor Sievers 

 (L,eipzig), as a means of textual criticism. Only three experi- 

 mental studies of English melody have yet been made, all in my 

 previous investigations. In the present investigation part of a 

 speech by Senator Depew was studied ; it furnished data concerning 

 American melody in a speech without oratorical exaggeration. A 

 record by another speaker furnished data for American speech with 

 satirical expression. The melody of Rip Van Winkle's toast, by 

 Joseph Jefferson (previously published), was recalculated by more 

 accurate methods ; it furnished data for emotional expression. In 

 this way the characteristics of American melody for oratorical speech 

 were obtained. When compared with the previous studies of the 

 melody of American sentences and of the I^ord's Prayer, the dif- 

 ferences from conversational and religious melody became apparent. 

 The work on melody should be extended to more persons, to dif- 

 ferent subjects, and to different languages. Problems of the fol- 

 lowing kind would be answered : 



(a) Has each piece of prose or verse a charcteristic melody of its 

 own that appears in spite of the individual differences among the 

 speakers ? If so, is it possible on this principle to pick out collabora- 

 tions and insertions from a text ? Professor Sievers asserts this to 

 be true, and is just issuing an edition of the Hebrew bible with text 

 criticism on this basis. His judgments are entirely by the ear as he 

 reads the passages himself. The fundamental principles, however, 

 should be established by experimental records. I discussed the 

 matter last spring with Professors Wundt and Sievers in Leipzig, 

 and the Psychological Laboratory there has, as a result, undertaken 

 of its own accord some researches on German melody. At the 



