278 PROCEEDINGS OF THE AMERICAN ACADEMY 



In like manner the sentence "This was the noblest Roman of 

 them all " was impressed upon the cylinder at a speed of one revo- 

 lution. Reproduced at two revolutions, the o's had a sound which 

 closely approached u. Lowering the speed slightly changed it to 

 a very short 5 ; and raising the speed slightly, to a clear u. At 

 three revolutions the quality of all the vowels was so completely 

 changed that the sentence was entirely unintelligible even to one 

 knowing what it really was, the rapidity of the speech, as repro- 

 duced, however, contributing largely to this lack of intelligibility. 

 To avoid this latter source of difficulty, the same sentence was 

 spoken into the instrument at a speed of two revolutions, and re- 

 produced at one revolution and also at three revolutions. At the 

 lower speed the i in "this " changed to a, so that, the s in "this " 

 being rather indistinct, the word seemed almost to have changed 

 to "that." But the o's, while deepened in quality, did not appar- 

 ently actually change to o. Lowering the speed still more caused 

 a closer approach to this, but the o quality still seemed to persist. 

 This was doubtless due in part to the association of the sound with 

 the word, and in part to the acoustic effects due to logographic 

 pressure, and to the connection of the vowel with the preceding 

 consonants. To study these, the same sentence, preceded by four 

 o's and followed by the same number, was spoken at two revolu- 

 tions and reproduced at one revolution. The separate o's seemed 

 to fall to o, and there was likewise a perceptible fall in the same 

 sounds in the body of the sentence, but these seemed still to retain 

 their o sound. To test the matter still further, the same sentence 

 with the o's mispronounced was spoken into the phonograph, at a 

 speed of one revolution, as follows: "This was the noblest Roman 

 of them all." On reproducing it at two revolutions the o's changed 

 to o, and the sentence was clearly heard with the mispronounced 

 words rectified, and as distinct as if they had been properly pro- 

 nounced when impressed upon the cylinder, and reproduced with 

 the same speed of the cylinder as when uttered. Lowering the 

 speed below one revolution, the 6" sounds became still deeper and 

 clearer. 



As the vowels o and o are more persistent in their character than 

 some others, a similar experiment to those already described was 

 tried with the line, " Though the harbor bar be moaning." This 

 was impressed upon the cylinder at a speed of one revolution per 

 second. Reproduced at two revolutions, the words " harbor bar " be- 

 came " harbor bar " (a as in hat), and at three revolutions the a's 



