■938 



II WMiCKJK 111- IMIYSIOI ( ><;Y 



NF.L'Knl'IIYSloI.OUY 111 



The nerve fiber is essentially all-or-none or digital 

 and, although modifiable over a moderate range, it 

 delivers quanta of excitation at a synaptic ending. The 



nerve cell, and probably its short processes, in con- 

 trast, is essentially continuous or analogical in its 

 properties, and can show great and graded changes 

 in its metabolism, membrane potential, threshold, 

 somatic potential and the like (98). A time modula- 

 tion is important; further, the spontaneous rhythm 

 that obtains for many, if not most, neurons may con- 

 tribute to the time sense and to the discontinuity of 

 experience. Not only does the size of an evoked visual 

 potential of the occipital cortex depend on the phase 

 of the alpha rhythm at which it arrives (Bartley), a 

 similar fluctuation in reaction time has also been re- 

 lated to the phase of the alpha rhythm at which it is 

 measured (17, 180). Moreover, experience seems to 

 advance in a series of time frames, of about a tenth of 

 ,1 second in man, much as do moving pictures (206, 

 272). And spontaneous rhythmic beats of single 

 neurons, as well as circuits between cortex and 

 diencephalon, are involved in the rhythms of the 

 mammalian cortex [but see Burns (39)] as well as 

 of the amphibian telencephalon (87, 174). 



fields. The field effects can be general, depending on 

 the over-all chemical environment as determined by 

 blood and intercellular fluid exchange, or by widely 

 separated sources and sinks of polarizing currents, 

 or they may be localized, depending on the diffusion 

 of metabolites ,md flow of currents between active 

 and inactive neurons or small neuron groups. The 

 vasodilation restricted to active neural regions (85; 

 Kety) is an example of such a local effect, and the 

 relation of activity to polarization, which in one 

 direction increases and in the other decreases sponta- 

 neous rhythms and actual discharges, has been 

 extensively demonstrated (39, 60, 71, 256; O'Learx & 

 Goldring). Activity, in turn, alters the steady poten- 

 tial (10, 17b). Intercellular currents undoubtedly 

 also contribute to the synchronization of cell masses, 

 most directly shown in Nitella (6, 7-', 138) and in the 

 frog brain (175), and probably to the spread <>l 

 inhibitory waves I |8, 167, 168, 283) and to the 

 physiological state in general, including dendrite 

 potential 19 Strom <ites steady potentials gene- 

 rated b\ blood changes temperature in the supra- 

 optic nucleus, carbon dioxide in tin- medulla thai 

 altei the I EG, muscle tension and pituitary activity, 

 ,,i otherwise modulate the level oi physiological 

 i eadini 



Such held 01 mass actions allow great freedom as to 



the actual neural units entering into a functioning 

 s\stem. The outcome can depend on the topography 

 of the steady potential held rather than on the specific 

 neurons that are active, just as cells in a cut flatworm 

 regenerate in terms of their position on a metabolic 

 gradient rather than in prespecified directions. Mass 

 effects also come closer to making understandable the 

 unitary whole of consciousness and behavior than 

 does a picture of innumerable impulses racing along 

 their separate nerve pathways. Yet this necessarily 

 sacrifices information and some experimental evidence 

 presents problems to field interpretation (268). 



Synchrony and hypersynchrony of neuron masses 

 can also well explain the phenomena of causalgia 

 and other types of 'physiological inflammation' ( 101 ) 

 associated with a continued excess of one modality of 

 input, and a deficiency in other modalities (but see 

 Sweet). Similar mechanisms may well underlie the 

 build-up of convulsive discharges, motion sickness, tin- 

 sex climax, certain neuroses and even the hallucina- 

 tory experiences of sensory deprivation. 



nets. Whereas field effects tend to be general and 

 graded, synaptic action and nerve nets tend to be 

 specifically patterned and quantized. As will be seen, 

 the mechanisms grade into one another for large 

 groups of neurons, in sheets or masses; yet the field 

 and net machines arc essentially polar to each other. 

 The influence of incoming impulses on a neuron is 

 determined by simple parameters of the synapses. 

 Endings from one or several fibers can differ in num- 

 ber or intensity (two endings close together would be 

 the equivalent of a single doubly strong one), locus 

 (two like endings far apart may not sum their effects 

 or may even have opposed effects on the cell activity >, 

 timing (altered frequencies or phases of incoming 

 impulse trains can change responses from positive to 

 zero to negative ones), and kind (where excitation and 

 inhibition depend on qualitative differences in termi- 

 nals rather than on their position on the cell body, or 

 on comparable differentials). 



Impulses lend to How forward, from input to out- 

 put, in a neuron chain or net, but they also reverberate 



to an important extent within dosed loops or as- 

 semblies Simple feed-back loops can give the normali- 

 zation oi perception and of action, already described, 

 and afford the underlying mechanisms proposed for 

 purposeful or servomechanistic behavior (20a). They 

 also would give the needed freedom in time, so that 

 the response is not necessarily linked immediately to 

 the siimulus K|), 97), and oiler .1 mechanism whereby 

 classes or universals c.m be generated from a succes- 



