MECHANORECEPTORS AND BEHAVIOR 345 



potential (Figure 6) that has the same time course as the IPSP is recorded. 

 Second, if the sense organ is displaced mechanically at the same time as the 

 efferent nerves are stimulated, the microphonic potential becomes larger 

 (Figure 6). This is because efferent nerve stimulation causes a small increase 

 in hair-cell membrane conductance. 



Transmitter substances— Until quite recently it was acceptable to sug- 

 gest that the microphonic potential directly excited the nerve terminals, but 

 now all the electrical measurements, when taken together with the ultra- 

 structural data, indicate that the hair cells act on the afferent fibres via 

 chemical synapses. The electron microscope shows that both the hair cell 

 and the efferent ending contain many vesicles. In the efferent terminal these 

 are usually of two morphological types: a single population of lightly stain- 

 ing vesicles of spherical form, with a mean diameter of around 60 nm, and a 

 few larger vesicles with a densely staining core. The vesicles that cluster 

 around the hair-cell synaptic body constitute a single population with a 

 mean size of about 50 nm. 



The least confusing pharmacological data on the nature of the transmitter 

 substances are for the efferent endings, which are thought to be cholinergic 

 (Russell 1971b). The story for the afferent terminals is much less clear. 

 Recent studies by Flock and Lam (1974) on the bullfrog basilar papilla, 

 which has only an afferent innervation, showed that GABA was synthesised 

 in the neuromast, while in the lateral-line organ of the toadfish and the 

 ampulla of the skate (Raja), where both types of terminal are present, both 

 GABA and acetycholine were synthesised. On this basis, therefore, Flock 

 and Lam argued for GABA as a transmitter candidate at the afferent terminal, 

 and they went on to show that picrotoxin, which specifically blocks GABA- 

 transmitting synapses, stopped both spontaneous and evoked activity. How- 

 ever since then, Sand et al. (1975) have applied picrotoxin and bicuculline 

 to the hair cells of Necturus and found no change in sense-organ sensitivity. 



Other substances have been suggested as the transmitter at afferent 

 synapses. On the basis of the changed appearance of the afferent synaptic 

 body in the frog labyrinth after treatment with reserpine and guanethidine, 

 Osborne and Thornhill (1972) suggested that catecholamines must be pres- 

 ent at afferent terminals, while Steinbach and Bennett (1971) found that 

 glutamate had an excitatory action on the related hair cells of the ampullae 

 of Lorenzini. Clearly, the nature of the afferent transmitter substance re- 

 mains a problem. 



Construction of Hair-Cell Organs: The Acoustico-Lateralis System 



In nearly all cases the hair cell functions as a directionally sensitive displace- 

 ment detector. The "free" neuromasts found in some teleosts and aquatic 

 amphibia respond to water velocity (Harris and Milne 1966), but in the canal 

 system, where the sense organs are connected with the outside world by 

 special pore openings, the neuromasts register water displacement. Fluid 

 displacements set up by angular acceleration of the head are the primary 



