VIBRATING TELEPHONE DIAPHRAGMS. 



419 



record and tracing process, these four angles ordinarily are not quite 

 equal ; but their arithmetical mean may be taken as the observed 

 phase difference. Since in the photographic record, the oscillations 



Fig. 3. Lissajous' Diagram. 



of the telephone diaphragm produce the vertical component, and 

 the oscillations of the galvanometer mirror (in phase with the cur- 

 rent), produce the horizontal component, the lag of the diaphragm 

 vibration behind the current becomes determined. This lag angle 

 should agree with the lag computed by mechanical impedance (90" 

 — a^), except for the angle /?i°. It was found that by varying the 

 impressed A. C. frequency, the Lissajous ellipse could be made to 

 pass through all its forms, i. c, straight line, ellipse, circle, ellipse 

 and straight line. The particular form offering easiest recognition 

 and greatest precision of measurement, is the straight line, under 

 which condition the two displacements are in either cophase or op- 

 position. This means that the diaphragm displacement would be 

 in ± cophase with the current, or the diaphragm velocity in quadra- 

 ture with the current. Referring to Fig. 4, OA represents the 

 standard phase of A. C. vector current. OB, in quadrature there- 

 with, is the phase of the diaphragm's vibrational velocity when the 

 Lissajous figure is a straight line. By observing, on the Rayleigh 



