CONDUCTION OF THE NERVE IMPULSE 



109 



0.2 M12) is connected between the two electrodes im- 

 mersed in the pools, this resistor serves to close the 

 external pathway of the local circuit and also to 

 measure the lons^itudinal current flowint^ through the 

 axis cylinder bridging the air gap. 



When the two pools are filled with normal Ringer's 

 solution, a familiar action current which we often 

 refer to as a ' binodal' action current is recorded. 

 Based upon the arguments described on earlier pages 

 (p. 88), this action current is explained as deriving 

 mainly from activity at the nodes (Ni and No in the 

 figure) in the immediate neighborhood of the record- 

 ing partition. The rapid rising phase of the action 

 potential at Ni develops a large gradient of potential 

 along the axis cylinder between node Ni and N2; the 

 phase of a strong (2 to 3 times lo"" amp.) current 

 flow in the binodal action current is the period during 

 which Ni is active but N2 is still inactive. When the 

 action potential starts also at node N>, the potential 

 gradient along the axis cylinder is greatly diminished, 

 resulting in a sudden fall in the longitudinal current 

 between Ni and N-j. At the end of the action potential 

 of a single node (fig. 16), the membrane potential 

 falls very rapidly. The abrupt end in the binodal 

 action current is related to the difference in the time 

 of termination of the action potential at Ni and N2. 

 Because of the capacities of the nodal membrane and 

 of the myelin sheath, the spread of current from No 

 to the internode between Ni and N2 prior to the 

 start of activity at Ni is very small. 



When a urethane-Ringer's solution barely strong 

 enough to block nervous conduction is introduced 

 into the proximal pool (in which Nj is immersed), the 

 upward deflection in the record (representing posi- 

 tivity of the right-hand electrode in the diagrain) 

 gradually decreases, indicating that the current arising 

 at Ni (partly from No) is reduced b)' narcosis. When 

 the upward deflection is reduced to one-fifth to one- 

 seventh of the original size, the downward deflection 

 which has gradually increased during narcosis sud- 

 denly drops out and, simultaneously, conduction 

 across the recording internode fails (the lowermost 

 record in fig. 28). From these observations, it is 

 found that the safety factor is between fi\e and seven 

 in large myelinated nerve fibers of the toad. 



The safety factor can be estimated from the meas- 

 urement of the threshold membrane potential and 

 the nodal action potential. It has been shown that 

 the action potential of a normal node is approximately 

 1 10 mv at the peak. When a membrane potential of 

 this size is developed at node Ni, the adjacent node 

 N2 is subjected to a strong outward current which 



would raise the membrane potential by 50 to 60 mv 

 if N2 had been made ine.xcitable (124). Since the 

 threshold depolarization of a fresh node is 10 to 15 

 mv, it is found that the safety factor estimated by this 

 method is about five. There are other methods of 

 estimating the safety factor (124). They all give a 

 figure between four and seven. 



As the result of this large safety factor in nervous 

 conduction, a nerve impulse can travel across one or 

 sometimes two completely narcotized nodes (124). 

 In the experiment of figure 26 it is often seen that 

 conduction across the middle pool remains unsus- 

 pended after introduction of a strong narcotic solu- 

 tion. A nerve impulse cannot travel across three 

 inexcitable nodes. 



Dots the .\ervc Imjnihc Jiiinji Jrom A'odr to Node? 



In 1925 Lillie (75) found that, when his iron wire 

 model of a nerve was covered with glass tubing broken 

 at regular intervals, the activation process jumped 

 from one break to the next. On the basis of this ob- 

 servation, he pointed out the possibility that the 

 nerve impul.se in the myelinated nerve fiber may 

 jump from node to node as in the model. This model 

 of 'saltatory conduction' has the following two fea- 

 tures: (a) the electrochemical changes underlying the 

 process of 'conduction' are localized at the 'nodes' 

 and (i) the time required for the conduction of the 

 impulse is determined solely by the rapidity of the 

 process at the node. In the model, therefore, the role 

 of the internodal segment is simply to provide an 

 ohmic conductance to the local circuit. 



We have described the main line of evidence indi- 

 cating that, in the vertebrate myelinated nerve fiber, 

 the physiological process responsible for producing 

 action potentials is localized at the nodes. We have 

 also seen that, although the d.c. resistance of the 

 myelin sheath is very high, the capacity of the myelin 

 sheath is large enough to have a marked effect upon 

 the threshold of the nerve fiber measured with short 

 current pulses (p. 99). This capacity of the myelin 

 sheath, therefore, sets a certain limitation to the 

 analogy between propagation of the activation wave 

 in the iron-wire model and the actual process of 

 ner\'ous conduction in the mvelinated nerse filjcr. 



The upper part of figure 29.I illustrates the arrange- 

 ment to demonstrate saltatory conduction in the 

 model nerve fiber. An iron wire covered \vith glass 

 tubings except at the ' nodes' is immersed in a bath 

 of nitric acid. When the wire in the passive state is 

 stimulated at one end, the process of activation as 



