48 



J. C. ECCLES 



et al, 1952), it was concluded that the i.p.s.p. was the cause of the inhibition 

 of reflex discharge and consequently that its latency provided a direct and 

 accurate value for the latency of the inhibitory action. As so measured, the 

 inhibitory latency was almost a millisecond longer than the latency of the 

 monosynaptic e.p.s.p., yet at first an attempt was made to account for this 

 discrepancy merely by the longer conduction time in the longer central 

 inhibitory pathway, which extended for about 15 mm longitudinally in the 

 first investigation. However, systematic testing, as illustrated in Fig. 1, 



Fig. 1. Intracellular responses from motoneurons {lower traces) evoked by 

 various alTerent volleys which are recorded from the appropriate dorsal root as it 

 enters the spinal cord (upper traces). Upward intracellular deflections signal depo- 

 larization and downward hyperpolarization of the motoneuronal membrane. 

 All records are formed by the superposition of about twenty faint traces. The 

 arrows pointing to dorsal root records give times of entry of volleys into the 

 spinal cord, the arrows pointing to intracellular records mark the onsets of the 

 i.p.s.p. or e.p.s.p., i.e. intervals between arrows give the central latencies. With 

 c, D a deep peroneal motoneuron at Li level is inhibited by a gastrocnemius volley 

 (c) excited by a deep peroneal volley (d). There was virtually the same length of 

 central pathway yet the latency of the i.p.s.p. was 0-8 msec longer than for the 

 e.p.s.p. With A, the inhibition of the gracilis motoneurons in /.« by a quadriceps 

 volley had a shorter central pathway than its excitation by a biceps semiten- 

 dinosus volley, yet the central latency was 05 msec longer for the i.p.s.p. With 

 E, F the central inhibitory pathway was about 15 mm longer than the excitatory 

 and the central latency was 0-9 msec longer (Eccles, Fatt and Landgren, 1956). 



showed that the longitudinal conduction time could account for only a small 

 fraction of the discrepancy, and that about 0-8 msec could not be so explained 

 (Eccles et al., 1956). Moreover the discrepancy could not be due to a longer 

 delay in the actual synaptic mechanism concerned in producing the i.p.s.p. 

 because there was approximately the same interval (about 0-3 msec) between 

 the arrival of impulses in the respective presynaptic terminals and the onsets 

 of the e.p.s.p. and the i.p.s.p. (Eccles, Fatt and Landgren, 1956). This inference 

 has recently been strongly supported by the finding that stimulation by brief 

 electrical pulses through a microelectrode in the intermediate nucleus evokes 

 e.p.s.p.'s and i.p.s.p.'s of motoneurons with the same brief latencies of about 



