78 J. C. ECCLES 



(Araki et ciL. 1960). Thus there is again a dual inhibitory action, which is 

 hkewise attributable to the inhibitory current and to the hyperpolarization of 

 the i.p.s.p. When the rectangular pulse was much briefer (from OT to 0-2 

 msec), it had correspondingly to be much more intense, and the initial phase 

 of inhibition was much less prominent (Araki et al., 1960). This is to be 

 expected because the inhibitory current would be much less effective in 

 counteracting the much more intense depolarizing current. 



In conclusion it can be stated that the inhibitory synaptic mechanism shown 

 diagrammatically in Fig. 2b, c, d satisfactorily accounts for the time course 

 that is exhibited for inhibition of motoneurons, both of their reflex discharges 

 and of their responses evoked by direct stimulation. Thus these investigations 

 conform with those described above on inhibitory latency in showing that 

 there is no justification for postulating (cf. Lloyd and Wilson, 1959; Lloyd, 

 1960) that inhibition is due to some process in addition to the potential change 

 (the i.p.s.p.) revealed by intracellular recording and the inhibitory currents 

 that cause that potential change. 



As mentioned above, the inhibitory curve has occasionally been found by 

 Jack et al. (1959, and personal communication) to exhibit only the brief 

 phase, i.e. to have no phase attributable to the hyperpolarization of the 

 i.p.s.p. Further support for this explanation was derived from the observation 

 that in contrast to e.p.s.p.'s, no i.p.s.p. 's could be recorded as a result of 

 electrotonic transmission to the ventral root as it emerged from the spinal 

 cord (Jack et al., 1959; Lloyd and Wilson, 1959). It was therefore postulated 

 that the intracellularly recorded i.p.s.p. resulted from the lowering of mem- 

 brane potential due to impalement by the microelectrode. However, a re- 

 investigation (Araki et al., 1960) has shown that i.p.s.p.'s can always be 

 recorded in ventral roots provided that the experimental situation is designed 

 so that it is particularly favourable for the production of i.p.s.p.'s and there 

 is a minimum of comphcation by superposition of e.p.s.p.'s. For example, 

 in Fig. 6a a quadriceps la afferent volley produced the i.p.s.p. electrotonically 

 transmitted from biceps-semitendinosus motoneurons to a Si ventral rootlet. 

 The central latencies of the i.p.s.p.'s recorded intracellularly (b) and electro- 

 tonically are virtually identical, but, as would be expected, the i.p.s.p. recorded 

 from the ventral root has a slower time course. The relative sizes of the 

 e.p.s.p.'s (c) and i.p.s.p.'s recorded from the ventral root are not at variance 

 with what would be expected from the mean of the intracellular i.p.s.p.'s 

 and e.p.s.p.'s produced similarly in motoneurons (Araki et al., 1960). The 

 mean sizes of the i.p.s.p.'s and e.p.s.p.'s produced by quadriceps and pos- 

 terior biceps-semitendinosus (PBST) volleys on BST motoneurons in the 

 lower Lt or upper 5"! segments were 10 and 90 juV, respectively. Similarly, 

 with stimulation of the S3 dorsal root, i.p.s.p.'s of appropriate size were 

 regularly recorded from the contralateral S3 ventral rootlets, the mean size 

 being 28 fiW. 



