82 



J. C. ECCLES 



-mV 

 -10 



-20 



A 



5inV 

 -30 MSEC 



-40 



JN^ 



-80—, E|P3P 



-90 



A_ 



-too- 



V. 



AK* 



D 



///k/l/// 



E ' 



zzzzzzzz 



INSIDE 



F 



7771 



77 



777 



INSIDE 



zzzzz 



Fig. 8. In a, the i.p.s.p.'s of Fig. 3a are shown arranged with their membrane 

 potentials on the scale indicated by short horizontal lines to the left of a, and 

 the equilibrium potential for the i.p.s.p. is shown by the broken line, b shows 

 the situation from 5 to 40 sec after the passage of a depolarizing current of 

 5 •; 10 * A for 90 sec through the microelectrode (filled with 0-6 m Na2S04); 

 the i.p.s.p.'s are shown similarly arranged on the same potential scale, the £'i.p.s.p. 

 being now — 35 mV. c shows, on the same scale, the i.p.s.p.'s obtained during 

 partial recovery at from 3 to 4 min after the electrophoretic injection with the 

 jE'i.p.s.p. at — 66mV. d and e represent the postulated fluxes of K+ ions in 

 conditions a and b, respectively, while f gives the postulated fluxes of K+ and Cl^ 

 ions normally occurring across activated inhibitory postsynaptic membranes. 



to recover to the normal value (cf. Fig. 8c). Hence Coombs et al. were led to 

 postulate that the flux of K+ ions was importantly concerned in the generation 

 of the i.p.s.p., for on this postulate K+ depletion would cause just such an 

 effect, which would have the slow time course of recovery that depended on 

 the linked Na-K. pump. 



But it now seems (Ito, 1960, personal communication) that the contri- 

 bution of CI ions to the more prolonged effect produced by Na+ injection 

 has been underestimated by Coombs et al. (1955b). After the K+ injection the 

 CI" could easily diffuse out across the membrane with the K+ ions. After the 

 Na+ injection and the consequent K.+ depletion, the membrane potential 

 was much diminished, in Fig. 8 from —75 mV to —57 mV in b; as a conse- 



