ACOUSTIC IMPEDANCE. 25 



In other words, all motional-impedance circles of telephone receivers 

 are convertible into rectilinear mechanic impedance graphs of con- 

 stant mechanic resistance, neglecting the influences which are known 

 to distort the motional impedance graph from the purely circular 

 form.^^ 



If the motional-impedance circle of approximation O G E H of 

 Figure 12 is accepted for the receiver applied to the listener's ear; then 

 the dotted straight line g e h, Figure 13, would represent the corre- 

 sponding mechanic impedance graph, where the mechanic resistance 

 is 335 dynes per kine, and zero reactance is found near 1100<>^. The 

 curved line def corresponds more nearly to the actually recorded 

 motional-impedance graph O D E F, Figure 12. The particular shape 

 of this curved graph def may not, however, be relied upon. 



Figure 13 indicates therefore that, at least to a first approximation, 

 the acoustic resistance of the listener's ear in this measurement was 

 86 dynes per kine. At higher frequencies, this resistance appears to 

 increase. 



Experiments were also made with an artificial ear,^^ composed of 

 paraffin wax, and in form modelled closely after an anatomical model 

 human ear. A motional-impedance graph made for the same tele- 

 phone receiver, applied to this artificial ear, agreed fairly well in form 

 with the graph for the actual ear O D E F, Figure 12. The corre- 

 sponding approximate circle had a diameter of 98 X 44° ohms; or 

 was somewhat larger than the circle O G E H of Figure 12, cor- 

 responding to a somewhat lower mechanic resistance. It may be 

 noted that a receiver, when applied to a model ear in a solid substance 

 like paraffin wax, does not fit to it as closely as when applied to the 

 human ear. The imperfect fit would tend to lower the acoustic 

 resistance. 



With the receiver applied face downwards upon a flat slab of glass, 

 a motional impedance graph was a closed curve through the origin 

 roughly circular in form. The approximating eciuivalent circle had a 

 diameter and depression angle about the same as when the measure- 



is These deviating influences are the variation in the force factor A, with 

 impressed frequency, and the variation in r' the component of hysteretic 

 frictional resistance of the diaphragm, with change of impressed frequency; 

 see Bibhography 16. 



14 It is beUeved that an "artificial ear," consisting of a flat slab of some 

 suitable substance selected experimentally, on which a receiver is laid face 

 down, has been used for some years in the telephone laboratories of the West- 

 ern Electric Co. in New York. 



