282 



HANDBOOK OF PHYSIOLOGY 



NEUROPHYSIOLOGY I 



be a series of autosiphons, each emptying into the one 

 beside it so that should any one element operate, the 

 whole chain would be initiated. This model would not 

 work, however, for it requires that each element be 

 above all the others which is absurd. In the electric 

 model this difficulty is overcome by providing capaci- 

 tors which direct a proportion of the discharge cur- 

 rent of each element to the trigger tubes of the ad- 

 jacent ones so as momentarily to raise the potential 

 differences across them to their discharge threshold. 

 This is an important feature, since it suggests that in 

 a living nerve the current generated at any active 

 region, the action current, may produce fields of po- 

 tential difference great enough to initiate at a distance 

 the electrochemical process responsible for the char- 

 acteristic depolarization and discharge. This effect is 

 of course demonstrated ijy the circumvention of a 

 blocked region in a nerve by leading the action cur- 

 rent through an inert conducting bridge (30). 



SPONTANEOUS RHYTH.MIC ACTIVITY IN 

 EXCITABLE TISSUES 



Rhvthmic Activity in Singh' Units 



With these properties in mind we may revert to 

 the question of spontaneous rhythmicity. In a single 

 nerve cell, as in a single clement of the nerve model, 

 spontaneous rhythmic activity will tend to occur 

 whenever the discharge threshold is at or below the 

 polarization potential. As is well known, depression of 

 the threshold of excitation, or lengthening of the ac- 

 commodation constant by the action of drugs, does 

 induce spontaneous rhythmic activity, even in 

 normally passive peripheral nerve fibers (19). In gen- 

 eral, the amplitude and rate of .such a discharge de- 

 pends on two factors: the rate of charge or polariza- 

 tion, and the rate of discharge or depolarization. 

 These two time constants are independent variables 

 to a first order of approximation and may be analo- 

 gous to the two resistors in a sawtooth relaxation 

 oscillator which control the sweep speed and fly-back 

 speed, respectively. In this comparison the flyback is 

 equivalent to the action potential or spike discharge, 

 which need not be numerically equal to the total 

 available polarization potential. 



Rliytlimic Activity in .Netwurks 



Now, when several such elements can interact wiih 

 one another by their electric fields, the aggregate 



system will tend to exhibit generalized rhythmic ac- 

 tivity at a frequency very much lower than that sug- 

 gested by the time constants of the single elements. 

 The repolarization time of neurons in the central 

 nervous system is probably equivalent to their re- 

 fractory period and lasts about i msec. The maximum 

 frequency of spontaneous discharge for such a neuron 

 is therefore of the order of 1000 pulses per sec, but 

 the lowest rate depends on the relation of the degree 

 of depolarization to the threshold. In effect this im- 

 plies that there should be an inverse relation between 

 the amplitude and the frequency of a spontaneous 

 rhythm. 



The i^asic waveform of a discharge determined in 

 this way should be of an asymmetrical sawtooth 

 variety, the asymmetry being more apparent at 

 higher amplitude, though the proportions are actually 

 constant. It can be shown, however, that in the case 

 of a large population of mutually interacting unstable 

 elements the waveform of the aggregate discharge 

 may be so smoothed as to lose all traces of its angu- 

 larity and come to lie indistinguishable from a 

 sinusoidal rhythm. This principle is actually applied 

 in electronic circuit design to obtain a sine wave sig- 

 nal from a square or triangular source which can be 

 activated or synchronized without the inertia of a 

 conventional sine wave oscillator. 



The conclusion to be drawn is that the wav-e form 

 of a spontaneous rhythm originating in a population 

 of active elements is of limited assistance in determin- 

 ing the mechanism of its .source; a pure sine wave may 

 originate in an assembly of relaxation oscillators, but 

 a relaxation wave form is less likely to be the output 

 of a single harmonic source. 



Rhytlimic Activity in Piiinitive Organs 



Having now considered the basic properties of 

 rhythmic generators in general, we may turn to the 

 specific features of this class of activity in the brain. 

 At the very outset it must be admitted that no con- 

 venient generalization is possible. Rhythmic dis- 

 charges are common in the nervous and muscular 

 systems of nearly all animals, but there is as yet no 

 proof that they can all be attributed to the same 

 mechanism. 



For example, details of the intrinsic rhythms of the 

 cardiac ganglion cells in Crustacea have been de- 

 scribed by Hagiwara & Bullock {26) and Bullock & 

 Terzuolo (14). The wave form of the.se rhythms 

 seems to be typical of the relaxation oscillator type, 

 as shown in figure i. Harris & Whiting {27) hax'e 



