EXCITATION OF AUDITORY RECEPTORS 



579 



AUDITORY NERVE IMPULSES' 



Volleys and Latencies 



The auditory ner\e responds to a single click with a 

 sharp, well-synchronized volley of action potentials, 

 conventionally designated 'Ni'. If the click is of moder- 

 ate strength, Ni is usuallv followed about i msec, 

 later by a smaller second vva\e, 'N)'; and with still 

 stronger clicks, a third still smaller wave, 'Nj', may be 

 seen (fig. 17). N5 and N3 are due largely to repetitive 

 firing in some but not all of the responding fibers, the 

 interval corresponding to the refractory period of the 

 nerve fibers. 



The successive sound waves of a steady tone of 

 frequency below 4000 cps give rise to similar, although 

 smaller, volleys of action potentials. Between 4000 

 and 2000 cps the indi\iduai \'olleys are very small, 

 but the frequency of the tone is nevertheless clearly 

 reproduced in the action potential pattern, even 

 though no one fiber responds to every sound wave. 

 This pattern of occasional but synchronized response 

 to a regular but intermittent stimulus such as sound 

 waves is the basis of Wever's (23) 'vollev principle' 

 (fig. 18). 



At very low frequencies in the guinea pig, both Nj 

 and N2 may sometimes be seen in response to each 

 sound wave, but both are rather dispersed in time. 

 At 1000 cps and below, the sharp initial portion of Nj 

 is initiated in the lower turns of the cochlea in which 

 the partition moves almost in phase as a unit. The 

 sharp initial 'spike' is followed by a more diffuse 

 ' tail' of impulses from the inore apical regions. 



Not only do different fibers have different latencies 

 of response due to the travel time of the traveling 

 waves but, as shown by studies of individual fibers, 

 the latency of each varies from one response to the 

 next. This variability leads to a less and less perfect 

 synchronization of the impulses as the frequency is 

 raised, and above 4000 cps no synchronization is 

 visible on the oscilloscope or audible by ear. At the 

 onset of a high-frequency tone burst, however, there 

 is a very well synchronized Ni, N2 and perhaps N3 

 (figs. 16, 17). The latency of Ni, the sum of the whole 

 group of fibers, is very stable in spite of the variability 

 among individual fibers. The latency shortens from 2 

 msec, or more near threshold to about i msec, as 

 intensity is increased. The shortest latency reported is 

 0.55 msec. The latency is a function of rise-time as 



" See especially the papers of Davis (4) and of Tasaici 

 (14. 15)- 



\AA/I\AA/ 



FIG. 18. Single-fiber spikes in two different fibers of the 

 auditory nerve induced by 1000 cps pure tones about 55 db 

 above human threshold. Lower Iracitig is sound stimulus recorded 

 through a microphone. Exposure was about .015 sec. [From 

 Tasaki (14).] 



well as the intensity and perhaps also the frequency of 

 the acoustic signal. 



The latency of the action potentials is attributed 

 chiefly to conduction time in the nonmedullated 

 dendritic iiranches in the organ of Corti. It is meas- 

 ured from the beginning of the cochlear microphonic 

 to the foot of the action potential spike, recorded as 

 the volley passes through the modiolus. (No latency 

 can be seen between mechanical displacement of the 

 cochlear partition and the cochlear microphonic.) 



The response to a brief high-frequency transient 

 such as a click or the onset of a tone burst seems to be 

 determined by the wave-group as a whole as if a 

 rectifier-detector were operating in the ear. The 

 summating potential is proijably the electrical sign 

 of just such a mechanical detector action. 



At lower frequencies, below about 3000 cps, each 

 sound wave acts more and more like an individual 

 stimulus. Excitation apparently occurs during the 

 'falling phase' of the cochlear microphonic, i.e. while 

 the .scala media (and vestibuli) is ijecoming more 

 negative relative to the scala tympani. This corre- 

 sponds to the phase of outward movement of the 

 stapes. The latency of responses to individual waves 

 can be reckoned consistently and logically from the 

 positive peak (in the scala media) of the cochlear 

 microphonic; but latency measurements are compli- 

 cated at low frequencies because of the progressive 

 time delay of the travelinc: wave. 



Single Fiber Activity 



Many of the above statements concerning latency 

 of response, all-or-none activity, etc., of auditory 



