

H WDBOl IK (IF l'HYSll II i n.V 



NEUROPHYSIOLOGY III 



for. Moreover, input and output are not symmetrica] 

 in the nervous system, if only because of the great 

 convergence present, so feed-hack loops might have 

 different functions at various stages. Presumably 

 sonic control of volume and some gating of time must 

 be applied to every rivulet of information flow as well 

 as to the great through rivers. 



Moreover, this is true for the complex learned 

 behaviors no less than for the simple ones, e.g. eye 

 tracking (Whittridge), lens accommodation (Fry) 

 and cerebellar control (Paillard). Speech is moni- 

 tored by aural feed-back, and a delay in this of one 

 phoneme (about .05 sec.) can play havoc with 

 thinking as well as speaking (Zangwill). The question 

 ol positive feed-back loops certainly needs far more 

 Study. These are probably involved in attention 

 and vigilance (Bremer). [Warming the hypothalamus 

 decreases reticular arousal of the cortex (Strom). 

 Reticular formation stimulation can increase the cor- 

 tical response to light (French) or to thalamic stimu- 

 lation (Livingston); and a positive feed-back via epi- 

 nephrine liberation is seen for the reticular formation 

 (Ingram) and for receptors (Gray), including muscle 

 spindles, ] But little fact is available here. Even more 

 theoretical is the involvement of both positive and 

 negative feed-back loops in relation to learning. 

 1 hev should be involved, as will be seen, in the 

 progressive channelization of impulses in the useful 

 neuron path and the elimination ol useless irradiation 

 to others. 



FLOW PATHS. Still more generally, the over-all table ol 

 organization of synaptic patterns must account for 

 tin- inward (low of information and the outward How 

 of decisions. What are the channel capacities needed? 

 What are tolerable and optimal signal-to-noise 

 ratios. 1 wii.ii redundancy of information and equiva- 

 lence of channels is optimal.' What is the balance 

 between series and parallel channels.' Wh.it .ire the 

 is ol convergence and of divergence at various 

 decision points.' Above all, what is the command 

 hierarchy? What fraction and what kinds of de- 

 cisions c ,m Ik- handled at the local, peripheral level, 

 which must be relayed on to a more central command 

 Center; which ones must run the lull chain to the 



topmost centers? Ami where are these.' gathered 



I ilier in the cortex or cenlrcnccph.ilon (l'enlield) 



01 even in the reticulai formation see Jaspei '/ al. 

 (149) ,01 widelv scattered; anatomically constant or 

 shifting with conditions, so that the 'best -informed' 



ippropriateh; .novated neurons take command 



in each case? Livingston, and Ashby (13) carry this 



(lis! USsion lllltliei 



The same questions of efficient handling of 'intel- 

 ligence' and •command' face an army, a university, 



all operating institutions, and organisms and cells. 

 The optimal How pattern varies with the size of the 

 organization, required speed of response, tvpes of 

 input and many other factors. Students of the brain 

 can probably obtain valuable cues from the analvsis 

 of organizations (191 1 and from the theory of com- 

 puters (9, 22), themselves relatively simple organiza- 

 tions. Eventually, when the answers learned by the 

 nervous system in the course of organic evolution 

 are revealed, they will probably repay workers in the 

 other fields with fresh insights. How these patterns 

 are laid down in the nervous system is a separate prob- 

 lem, to be considered later. 



Cell Typi 1 inn/ Spei 1 /it ity 



structural CONSIDERATIONS. Besides the quantita- 

 tive differences in connections and locus just con- 

 sidered, neurons differ in kind. Subordinate to the 

 major dichotomy of cells in the nervous system be- 

 tween neurons and glia (and there is not vet clear 

 evidence that glia cells do or do not conduct im- 

 pulses; they contain pseudocholinesterase rather than 

 true cholinesterase according to Tower; but this is 

 also moot, especially in invertebrates), there are 

 such important secondary dichotomies as between 

 neurocrine neurons, central receptor neurons, central 

 autonomic neurons, long and short axon neurons, 

 and main more. In general, like neurons tend to be 

 clumped together; at least different anatomical 

 regions often possess cells with distinctive composi- 

 tion and appearance, are differentially sensitive to, 

 even destroyed by, different substances, occupy a 

 regular position on some quantitative scale, as 

 metabolic rate or glutamic acid concentration 

 ( lower, 88, 129, 139, 273), or differ in their physio 

 logical properties, such as ease of initiating and of 

 maintaining repetitive discharges. A characteristic 

 amino acid chromatogram constitutes a 'fingerprint' 

 ol each part of the nervous system (248). Synapses 

 differ fantastically in their detailed morphology, a 

 likelv guide to differences in functional properties 

 .mil mechanisms (82). They differ in raising or 



lowering the excitation level, or the threshold (e.xci- 

 tatory and inhibitory synapses); they operate with a 

 fat mi ol safety of less or of more than one, and thus 

 do or do not require some summative action to be 

 effective (98, 186); and they invest different neurons 

 to different extents and in differently patterned loci 

 (Chaugi. liber tvpes, different in structure, chemis- 

 try, timing, electrical pain ins and physiological 



