SYNAPTIC AND EPHAPTIC TRANSMISSION 



171 



properties of the axon contribute to the potential re- 

 corded from the soma. 



The nature and degree of excitability may be dif- 

 ferent in various parts of the soma and dendrites. 

 Thus the soma may be electrically inexcitable (ry, 

 33, 80, 186, 189, 190). The depolarizing p.s.p.'s 

 or generator potentials cv'oked at the soma excite 

 spikes at electrically excitable regions some distance 

 from the cell body. The superficial portions of apical 

 dendrites in the cat cortex are not electrically excit- 

 able (loy, 165), As mentioned earlier, the receptor 

 portions of various sensory cells are electrically in- 

 excitable and for this reason are capable of develop- 

 ing a sustained generator potential.^ 



Recent evidence (7, 17, 43, 82) also indicates that 

 different portions of electrically excitable mem- 

 branes of the cell body may have different thresh- 

 olds. The ■ initial segment' of the motoneuron (cf. 

 60) in the cat (82) and toad (7) responds first to an 

 electrical stimulus and gives rise to the early part of 

 the antidromic spike (fig. I2^'-F')- The spike of the 

 rest of the cell body (if the latter is electrically ex- 

 citable) occurs slightly later, the delay giving rise 

 to a slight break in the recorded response. 



In addition to these apparent inhomogeneities in 

 the excitability of different parts of the soma and 

 dendrites, slowed conductile spread, separate loci of 

 origin for spike and p.s.p.'s and different loci for 

 depolarizing and hyperpolarizing p.s.p.'s are all 

 factors that may contribute to variations in the 

 recorded response of the cell. Many variations can 

 be theoretically deduced, but their analysis is beyond 

 the present scope. 



'Retinal receptors in lisli (184) provide an interesting new 

 example (102). Their electrical response is probably generated 

 in cells other than the primary visual cells (cones). The re- 

 sponse is a sustained clectrogenesis. In some cells it is only 

 hyperpolarizing, in others depolarization is also developed, 

 depending upon the wavelength of the stimulating light. The 

 amplitudes of the responses are graded, not only with the in- 

 tensity of the light stimulus but also with its spectral composi- 

 tion. These characteristics of electrically inexcitable activity are 

 produced apparently in the absence of spikes, but the electro- 

 genesis, both hyperpolarizing and depolarizing, affects spike 

 production in other conductile elements. It has been suggested 

 (102) that these electrogenic cells (probably horizontal or bi- 

 polar cells or both) are excited by transmitter agents released 

 by photochemically activated cones. The clectrogenesis, in 

 which an electrically excitable component is lacking, is in turn 

 associated with secretory activity as in electrically inexcitable 

 gland cells. The secretory products acting upon the retinal 

 ganglion cells evoke neuronal activity of the latter, probably 

 including excitatory and inhibitory p.s.p.'s which lead to 

 patterns of spike activity seen in the optic nerve fibers. 



GENERAL AND COMPARATIVE PHYSIOLOGY OF SYNAPSES 



Forms and Magnitudes of Postsynaptic Potentials 



Viewed as the nonregenerative responses of elec- 

 trically inexcitable membrane, the forms and mag- 

 nitudes of the p.s.p.'s may be expected to have 

 rather simple relations to their excitants. The availa- 

 ble experimental data are as yet rather scanty, but 

 they do permit some general conclusions (cf. 60, 97). 



As a first approximation, the degree of synaptic 

 transducer action reflected in the rate and amount 

 of clectrogenesis may be considered to be roughly 

 proportional to the quantity of excitant. A brief jet 

 of labile transmitter or activating drug causes a 

 Ijrief response while the continued availability of 

 the excitant causes a sustained clectrogenesis. The 

 duration of the p.s.p. in the first case will be deter- 

 mined by the time course of the transducer action 

 initiated by the excitant (cf. also 53, 127, and papers 

 cited there). However, the responses will be dis- 

 torted by the electrical circuit properties of the 

 membranes. Thus, the rising and falling phases of 

 the p.s.p. may reflect this distortion which produces 

 a slowing such as occurs in electrotonic propagation 

 (fig. 2). The rise of the p.s.p. should be slowed less 

 than its fall since the former occurs when the mem- 

 brane resistance and time constant are relatively 

 low. This is the case experimentally as numerous 

 figures in this chapter indicate. The falling phase 

 probably bears some relation to the time constant 

 of the membrane (cf. 60), lasting longer when the 

 time constant is larger, like the ballistic response of a 

 slow galvanometer to a brief current. The relation, 

 however, does not appear to be a simple one (95, 

 97), and the duration of the p.s.p. probably reflects 

 importanth intrinsic time courses of transducer ac- 

 tions. 



The duration of the p.s.p. caused by a single 

 neural volley differs considerably in the various types 

 of cells. The p.s.p.'s of squid giant axons and of eel 

 electroplaques last only about 2 msec. (figs. 3, 19), 

 those of Aplysia giant neurons (fig. 7) or cat salivary 

 glands (fig. 20) may persist for nearly i sec. The 

 e.p.p.'s and p.s.p.'s of other neurons have inter- 

 mediate durations. In .some cases, physostigmine 

 and prostigmine both prolong the p.s.p., this effect 

 probably involving the prolongation of the life of the 

 transmitter, acetylcholine, by inactivation of cho- 

 linesterase (cf. 52, 53, 60, 68). Some of the quater- 

 nary ammonium compounds also prolong p.s.p.'s 

 (cf 52) and these actions may be caused by direct 



