HALLOWELL DAVIS 



69 



impulses in the peripheral endings of the fibers. The latency of the 'foot' of the 

 action potential is so brief that it argues for rather than against direct electrical 

 excitation of the nerve by the microphonic (9). 



The present hypothesis, as we have said, designates the 'endolymphatic 

 potential' as the source of the energy of the cochlear microphonic. The endolym- 

 phatic potential was discovered by Bekesy in 1952 (2). A micropipette intro- 

 duced into the scala media reveals a resting DC potential of about +80 milli- 

 volts (in the guinea pig) relative to the perilymph of scala tympani or scala 

 vestibuli or to the tissues of the neck. The distribution of this positive j)otential 



Reissner's 

 membrane 



ENDOLYMPHATIC 

 SPACE 

 + 80mV 



STRIA 

 VASCULARIS 



CELLS OF THE 



ENDOLYMPHATIC.-/-; 

 WALL 



SPIRAL 

 LIGAMENT 



BONE 



Fig. 8. Endolymjjhatic space and distrrbution of the DC potentials of the cochlear parti- 

 tion (11). 



is shown in figure 8. In this figure is also indicated the familiar negative intra- 

 cellular potential, like that within nerve and muscle cells, that is encountered 

 when the tip of the electrode lies within cells of the cochlear partition. The 

 endolymphatic potential is of about the same or greater magnitude relative 

 to our reference electrode, but it is positive, not negative. There is, then, a 

 potential difference of about 150 millivolts across the 'membrane' of the hair 

 cells from which the hairs protrude. 



The cochlear microphonic and the endolymphatic potential are both im- 

 mediately dependent on o.xidative metabolism. They fall almost to zero if the 

 oxygen supply is cut off or if small quantities of cyanide or azide are injected 

 into the cochlea. The parallelism of these changes, which are reversible if not 



