PREFACE. O 



way phonetics and philology must go. Automatic records, experimental 

 analysis, and careful measurement must be the foundation of phonetics 

 as a natural science. 



My investigations were begun at Yale University.* The special 

 apparatus devised for obtaining speech curves had furnished a large amount 

 of material for the study of which I had neither time nor assistance. The 

 work was inspected by a committee from the Carnegie Institution of Wash- 

 ington (Dr. S. Weir Mitchell and Dr. John S. Billings) and judged worthy 

 of support. In March the work was removed from Yale to Munich; here 

 a laboratory of six rooms was installed. Valuable assistance was received 

 from the Psychological Laboratory (Professor Th. Lipps, director) of 

 the University. During the six months in Munich many computers were 

 employed and the main conclusions were outhned.f In October, 1903, 

 the work was moved to Berlin, where a small private computing bureau 

 was opened. The tracing apparatus was set up in the Psychological 

 Institute (Professor C. Stumpf, director) of the University of Berlin; this 

 was made the occasion of a lecture before specially invited members of 

 the faculty and of the Academy of Sciences, in which the methods and 

 general conclusions were presented. J The apparatus was used by Professor 

 Brandl (English Department) for tracing the curves from a specially made 

 disc. The environment in Berlin was exceptionally favorable, and the work 

 was continued there for a year and a half. On March 1, 1905, a new labo- 

 ratory was opened in Zurich. 



This volume gives an account of the methods used in obtaining and 

 studying the curves. 



The first chapter gives a sketch of the way in which sounds are recorded 

 and reports some experiments on diaphragms. These experiments are, 

 I believe, sufficient to do away with the notion that the vibrations of 

 such diaphragms have any resemblance to those of plates that produce 

 the Chladni figures. The theses maintained are: (1) that the diaphragms 

 bend concentrically around the center; (2) that nodal fines are present only 

 as disturbing factors and this to a small degree or not at aU ; (3) that the 

 aim of the construction is to produce an air-tight, unbending piston which 

 will accurately follow the phases of condensation and rarefaction of the 

 air-wave. A further investigation might have important results in pro- 

 ducing better telephone diaphragms as well as better talking machines, 



♦Researches in Experimental Phonetics, Stud. Yale Psychol. Lab., 1899, vii, 14; 1902, x, 49; Speech 

 Curves, i. Mod. Lang. Notes, 1901, xvi, 73; A New Machine for Tracing Speech Curves, Amer. Jour. 

 Sci., 1903, XV, 147. 



tMechanics of the Voice, Carnegie Year Book No. 2, 243, Washington, 1903. 



JUeber das Studium der Sprachkurven, Annalen der Naturphilosophie, 1904-05, iv, 28- 



