VIBRATING TELEPHONE DIAPHRAGMS. 425 



Fig. 4, already referred to in connection with the depression angles 

 )ffi° and /?o'', presents a smooth motional-impedance circle, all the 

 observed points falling nicely upon its circumference. Neverthe- 

 less, there is reason to believe that this circle contains a small inner 

 loop distortion between the origin o and the lowest indicated fre- 

 quency of 857 '—'. No measurements were made in this case below 

 857 '— ' ; but it is probable that, had they been made, the distortion 

 loop would have been too small to notice. The central amplitudes 

 of vibration were also measured by the explorer in the case of Fig. 

 4, and from these the vibrational velocities x of the diaphragm were 

 computed. They lie closely to the inner circle indicated in the figure. 

 This shows, moreover, as has been pointed out in earlier publica- 

 tions, that the motional-impedance circle may be regarded as a 

 velocity circle, taken to a suitable scale, disregarding the phase 

 lag ^/. 



When the diaphragm of the receiver B, considered in Fig. 4, was 

 loaded at the center by a small copper washer, about i cm. in di- 

 ameter and weighing 0.73 gm., a repetition of the test gave the dia- 

 gram shown in Fig. 5, the analysis of which appears in Table I. 

 This diagram shows two internal loops or abnormalities, one a little 

 loop near 691.9—', well defined however by numerous observations, 

 and the other a larger loop near 789.5 — '. The latter proved to be 

 a temporary visitor, and since it did not appear in subsequent tests, 

 it may be left out of consideration here. It suffices, therefore, to 

 notice that the loop at 691.9'—' had a length of about 2.5 ohms, and 

 occurs in the circle at an angular distance of about 30° from o, the 

 origin of the circle. The resonant frequency of the loaded dia- 

 phragm as a whole was apparently 794.5 ^~-'. 



The load on the diaphragm was then increased to a total of 0.975 

 gm. This had the effect of bringing the resonant frequency down 

 to 699.4 '—' ; or close to the frequency at which the abnormality (at 

 691.9'—') had appeared. The result of the test under this condition 

 is shown in Fig. 6. The distortion loop is central at 697.2 '— ', and 

 has a length of 42.5 ohms on the motional-impedance scale. It 

 is evident that when the diaphragm is brought nearly into reson- 

 ance with the frequency of the distortion loop, the magnitude of the 



