18 



THE STUDY OF SPEECH CURVES. 



stresses and strains favor the components of greater frequency. The 

 problem of how this occurs still remains uninvestigated. The experi- 

 ments of Bourget were confined to paper membranes glued to wooden 

 frames; the analytical treatment by Rayleigh is confined to firm edge- 

 fixation; the results are not applicable to the present case. It is natural 

 to suppose that the change of the simple fork tone into a sharp tone occurs 

 by vibration of the diaphragm in parts; as far as I am aware this is the 

 thought in the minds of all who have discussed vibrating membranes. In 

 the case of sound boxes from talking machines this view is erroneous, as 

 the following experiments show. 



To test the effect of the entire removal of the edge-fixation a repro- 

 ducer was employed in which the diaphragm was not fastened at the edge 

 at all, being supported only by the arm at the center. When set in vibra- 



FiG. 9. — Nodal figure from a 

 diaphragm with free edge. 



Fio. 10. — Circular distribution of light 

 from a diaphragm with fixed edge. 



tion by a gramophone record, the box gave forth only a very weak tone 

 with occasional rattling. The edges of the diaphragm could be seen to 

 be in strong vibration; this occasionally became so strong that the glass 

 struck the metal. When water was placed round the edge, the result 

 was not greatly altered, as the water was quickly blown awaJ^ As soon, 

 however, as vaseUne was appHed, the sound came out strongly and clearly. 

 We can thus conclude that the fixation at the edge has at least two pur- 

 poses: (1) To produce an air-tight closure around the diaphragm, and 

 (2) to hinder the edge from striking against the metal. 



A third purpose of the fixation appears in the following experiments : 



The light from an arc lamp was allowed to fall upon the diaphragm 



of a reproducer; the reflection was caught upon a white screen. When 



the diaphragm had been so inserted in the reproducer that its surface was 



flat, the reflection appeared as a white circle. 



