336 HARRY GRUNDFEST 



potentials. This is only a partially useful definition and justifiable only if the 

 definition also recognizes that the potentials may be absent or may be of 

 either sign and that in any case they need have nothing to do with the process 

 of transmission of the information which is analogous to neurosecretory 

 activity. 



Information about the functioning of auditory receptors and equihbrium 

 receptors of vertebrates and, indeed, of sensory receptors in most forms of 

 vertebrates and invertebrates is at present known only indirectly from the 

 recording of the impulses in the nerves. However, there is now ample evidence 

 that the relatively simple receptor systems like that of the Limulus eye have 

 complex excitatory and inhibitory interactions (Ratlifif and Harthne, 1959) 

 and this is probably even more elaborate in the vertebrate systems (cf. 

 Kuffler, 1953). Accordingly, the views on the mode of function of the auditory 

 and vestibular receptors are as yet speculative. Davis (1959) considers that the 

 hair cells of the auditory receptors are mechanosensory transducers which 

 change their resistance during rarefaction and condensation of the air. There 

 is a potential difference of about 80 mV between the scala media and scala 

 tympani. If the hair cells act as resistance microphones they might thus 

 impress a change of potential on the terminals of the auditory nerve fibers. 

 In that case the terminals ought to be electrically excitable and might perhaps 

 be related in evolution to the electrical receptors in electric fishes (Grundfest, 

 1961). 



Another possibihty is that the hair cells are in reahty secretory cells pro- 

 ducing and emitting a transmitter agent. At first glance this seems unfikely. 

 The cells and nerve fibers must follow relatively high-frequency changes in 

 pressure. However, with the discovery that the continuous liigh-frequency 

 discharges of electric fish are mediated by synaptic means (Bennett and Grund- 

 fest, 1961, and unpublished), this is no longer a stumbling block. Some of the 

 fish discharge at frequencies as high as 1600/sec (Lissmann, 1958), wliile in 

 the mammahan nervous system impulse production at such high rates is 

 possible for only a brief time (Gasser and Grundfest, 1939; Grundfest and 

 Campbell, 1942). The frequencies of the discharges are remarkably constant 

 in many gymnotids and are set by a command nucleus (Szabo, 1961 ; Ben- 

 nett and Grundfest, 1961, and unpublished). Every discharge of the organ 

 is associated with a discharge of a spinal neuron (Fig. 9). That discharge is 

 initiated by synaptic excitation wliich accordingly must arise from a spike 

 and a pulse of secretory activity in the nerve fibers descending from the 

 nucleus. As noted above, auditory nerve fibers probably do not conduct 

 impulses at rates higher than 1000-1 500/sec. Thus, it is entirely possible that 

 the auditory receptors are also secretory cells converting mechanical to 

 chemical energy. The same considerations apply to the vestibular receptors. 



The frankly neurosecretory cells which are found in lower forms as well as 

 higher (cf. Scharrer and Scharrer, 1954; Ortmann, 1960) provide a fink 



