356 MECHANICAL AND ACOUSTICAL SENSES 



With rotation at constant velocity (Figure 13C) there was an initial abrupt 

 increase or decrease in discharge rate, followed by a slow return to the 

 resting frequency. Lowenstein and Sand (1940a) found a threshold effect 

 somewhere around 3°/s 2 , which ten Kate (1973) calculates as being a devia- 

 tion of the hair cell cilia of about 0.1 nm. 



The experiments also showed that in Scyliorhinus the horizontal canal 

 responded only to horizontal rotation, whereas the vertical canals could 

 register rotation in all three axes. These out-of-plane responses are much less 

 pronounced in Raja (Lowenstein 1974). O'Leary and his collaborators 

 (O'Leary, Dunn, and Honrubia 1974; O'Leary and Honrubia 1976), working 

 with the horizontal canal of Rhinobatus, suggested that the sense organs 

 receive a systematic projection of nerve fibres in individual bundles that 

 behave quantitatively differently from each other when exposed to rota- 

 tions. From this finding they developed a theory that each afferent fibre is 

 'tuned' to a particular range of head accelerations. 



Groen, Lowenstein, and Vandrik (1952) used the elasmobranch canal 

 system to examine the torsion-pendulum theory closely by recording from 

 horizontal canal afferent fibres of Raja during sinusoidal movement and in 

 response to sudden changes in the velocity of a rotating turntable. They 

 confirmed the presence of a resting discharge of 6-100 imp/s (mean 26.5), 

 which they found to be very constant, deviating by only about 4% from the 

 mean, and went on to show that during a sinusoidal movement the time the 

 impulse frequency was at the "resting" value did not coincide with zero 

 position (i.e., there was a phase difference) and that it took at least 55-100 

 ms for the impulse discharge to return to its resting value after a sudden 

 acceleration. They assumed that this long delay was a reflection of the slow 

 return of the displaced cupula to its resting position. 



Groen et al. (1952) believed that their data conformed satisfactorily to 

 the equations that describe a torsion pendulum, but not all authors would 

 agree with the values they obtained (see Money et al. 1971). Quite recently 

 in electrophysiological recordings from the squirrel labyrinth, Fernandez and 

 Goldberg (1971) encountered deviations from the model which they at- 

 tribute to adaptation of the sense organ and to a response to the velocity as 

 well as to the displacement of the cupula. 



Jones and Spells (1963) point out that the labyrinth of a fish is about 

 twice the size of the labyrinth of a mammal of similar body size and, because 

 the sensitivity is dimensionally dependent, they attempted to explain this 

 significant size difference in terms of the type of head movements made by 

 fishes when swimming. More recently, however, ten Kate, Van Barneveld, 

 and Kuiper (1970) have shown that the large labyrinth size in fishes is a 

 result of the way they grow and that canal sensitivities are very similar in all 

 vertebrates. 



The utriculus and the sacculus— Because the otoconia and the cupula 

 membrane are about twice as dense as the endolymph, the otolith organ 

 functions as a differential density accelerometer in responding to linear ac- 

 celerations (Trincker 1962). Once again, elasmobranch preparations have 



