MECHANORECEPTORS AND BEHAVIOR 343 



number of hair cells. Like that of other sensory receptor potentials, the 

 amplitude of the microphonic potential is linearly related (on a semilog 

 scale) to the amplitude of the stimulus (e.g. Furukawa, Ishii, and Matsuura 

 1972a). 



Because of its ease of detection many studies have been made with the 

 microphonic potential. For example, in bony fishes the recorded micro- 

 phonic has provided data on hearing thresholds, directional sensitivity, and 

 the audiogram range. Microphonics have also been recorded from lateral-line 

 organs where, interestingly, they occur at twice the frequency of the applied 

 stimulus (Jielof, Spoor, and de Vries 1952; Flock 1965), an explanation for 

 which was provided by Flock on the basis of the polarity of the hair cells as 

 seen with the electron microscope (Flock and Wersall 1962); the double 

 responses recorded from the goldfish sacculus (Furukawa and Ishii 1967a) 

 and from near the shark macula neglecta (Fay, Kendall, Popper, and Tester 

 1974) presumably have a similar basis. The only published record of micro- 

 phonics from an elasmobranch receptor is that given by Lowenstein and 

 Roberts (1951) although data from microphonic recordings were used by 

 Fay etal. (1974). 



It is worth emphasizing that if the microphonic does represent the receptor 

 potential of the hair cells it may provide data on hair-cell function but it 

 cannot be used to provide reliable data on the range or the threshold of the 

 sensation of hearing, because there is no certainty that the hair cell afferent- 

 fibre synapse will respond in a similar way to the microphonic nor, indeed, 

 that the higher sensory centres will not modify this signal. 



Receptor potentials — Intracellular recordings have now been success- 

 fully made for several different lateral-line organs: hair cells of Necturus 

 (Harris, Frishkopf and Flock 1970; Yanagisawa, Taglietti, and Katsuki 1974; 

 Sand, Ozawa, and Hagiwara 1975); Lota (Flock and Russell 1973a, b), 

 alligator-lizard ear (Weiss, Mulroy and Altmann 1974) and mammalian 

 cochlea (Russell and Sellick 1977). In these examples very sharp, dye-filled 

 electrodes were used so that the location of the electrode tip could be 

 confirmed by dye staining. The resting potentials of the hair cells appear to 

 be rather low (42 ± 13 mV (Yanagiswa et al. 1974); 33 ± 13 mV (Sand et al. 

 1975)) except for Szabo and Hagiwara's (1966) values of 40-70 mV for the 

 epithelial cells of Savi's vesicles in Torpedo, but the exact recording site was 

 not identified precisely. Intracellular recording in Necturus has also shown 

 that the resting potential level can be biased by sustained displacement of 

 the cupula, the effect being to depolarize or hyperpolarize the cell, according 

 to the direction of stimulation (Flock 1971). 



In their first recording, Harris et al. (1970) used computer averaging tech- 

 niques during applied repetitive stimulation and observed small oscillating 

 changes that did not exceed 800 juV in size. Such potentials (Figure 6b) 

 would certainly be too small to function as generator potentials in nerve 

 fibres, but it is possible that the hair-cell/afferent synapse is a sensitive 

 amplifier that responds to such small potentials. Recently Sand et al. (1975) 

 found that afferent nerve firing could be evoked in hair cells stimulated by 



