VIBRATING TELEPHONE DIAPHRAGMS. 451 



first by raising the frequency slightly above resonance, and then 

 lowering it. These measured angular velocities (Oo and oj^ cor- 

 respond to the quadrantal points on the motional-impedance circle, 

 and supply A by formula (lo). Finally, the continuous-current 

 strength,. /s abamperes, necessary to produce a steady mirror de- 

 flection 9s radians, is measured as in (14). These four equations 

 suffice to evaluate A, m, r and .y for the instrument. An oscillo- 

 graphic natural decay curve, corresponding to (16), may also be 

 taken as a check on the results. 



The following are the results of a series of observations made 

 on an experimental bifilar oscillograph^" with two strips, each of 

 active length 3.5 cm. in a magnetic field of approximately 16 kilo- 

 gausses. The strips were of phosphor bronze, 0.366 mm. wide (15 

 mils), and 0.013 ^'^^- thick (0.5 mil), each under a tension of ap- 

 proximately 30 gm. weight, spaced 1.5 mm. on centers, and having 

 a mirror fastened to and across them, about i mm. X 0.5 mm., near 

 the middle of their active length. The vibrator was air damped, 

 /. e., it did not work in oil. 



^=3,750 dyne perp. cm. per abampere, 

 ^'0 = 2,530.5 ^, coo = 1.59 X 10* radians/sec, 

 ?n= 1.322 X lO"^ gm.-cm.-, 



^ = 2.78 X io~^ dyne perp. cm. per radian per sec, 

 ^ = 3,360 dyne perp. cm. per radian, 

 «., = 2,547^, «.,= 2,514^, 

 A= 103.7 hyps, per sec, 

 A = 76.7, 

 M2 — "i = 33'-"> 



^s//s = 1-11 5 radians per abampere, 

 Im == 0.002 absampere, 

 Z,„ = 5.075 X 10^ absohms =5.075 ohms. 



By plotting the deduced angular velocities 6 at different impressed 

 frequencies close to resonance, a fairly good circular locus was ob- 

 tained. The diagram is the same as that of Fig. 19, except as to the 

 scales of magnitude and numerical values. 



12 Bibliography, 11. 



