1940 



II WI1BOOK OF I'HVSKH uov 



\i i kui'HYsioi.ooY in 



and intervals uot unlike the Morse code of FM — as 

 well (Neff). Filtering out sound frequencies under 

 i ,i pi mi cps eliminates 90 per cent of the power but only 

 in per cent of the information in speech (Zangwill). 

 Complexity also exists foi spatial patterns which 

 determine convergence and summation, reverbera- 

 tion .md feed-back, synchrony and potential fields, 

 and the like (Gray). Even at the simple quantitative 

 level, synapses m.i\ reduce an input frequency of 

 several hundred per second to an output frequency 

 under 10 (Chang); and one input may lead to one or 

 to two do/en discharges (Davis). The interaction of 

 different impulse trains has been shown to alter Un- 

 experienced modality (synesthesia) in temperature 

 (Zottermann), touch and kinesthesia (Rose & Mount- 

 castle), taste (Pfaffman), etc. A io-jisec. time differ- 

 ence in the arrival of air waves at the two ears suffices 

 to give direction to heard sound (Davis); a shift in 

 timing of stimuli to right and left afferentscan start or 

 stop swallowing (56); impulse frequencies relate to cell 

 discharges in the lobster cardiac ganglion (37). 



As discussed, different frequencies of stimulation, 

 or different initial states of a neuron pool, can reverse 

 the response elicited (see also Patton & Amassian). A 

 single neuron can similarly be excited or inhibited by 

 different frequencies of stimulation, from other neu- 

 rons (e.g. French) or receptors even depending on 

 previous positive or negative conditioning (148). And 

 the number, frequency and pattern of discharges 

 from a single neuron can be altered over wide ranges 

 with different inputs. A single chemoreceptor unit can 

 lire at different frequencies to different taste stimuli; 

 and a single photoreceptor, to different colors (e.g. 

 Neff). Two axons of a single neuron can even fire 

 simultaneously at separate rates 1 [6). 



The ideal approach to the problem of neural coding 

 is to obtain the output for all ranges and kinds of 

 input, and to do this for single junctions, single 

 neurons activated through all combinations of junc- 

 tions and neuron pools receiving impulses from 

 various groups ot allerenl libers. This program is well 

 under wa\ for single units in mam laboratories, and 

 a promising start has been made even for whole 

 nglia. By autocorrelation (16) applied to post- 

 ■Iiimiii nerve trunks while preganglionic ones are 



tlated at differem strengths and frequencies, 



( :.isb\ (personal communication) has been able to 

 identify the impulses in fibers conducting al eat h 

 velocity or similar characteristic. Such data will m 

 time reduce the behavior o! neurons and neuron nets 

 and pools to the same quantitative rigor ol description 

 and prediction as has been achieved for activation ol 

 the nerve fiber 



Majm S) rt< ms 



The properties of neuron nets and masses have been 

 considered so far al the most general level; later, the 

 formation of highly specific p. uterus will receive 

 attention. Between these, there exist in the nervous 

 system major systems or organization schemes, prob- 

 ably including many not yet recognized. The more 

 particular nuclei and paths, which constitute the 

 subject matter of physiological anatomy, are exten- 

 sively treated in the Handbook. It will not be possible 

 here to do much by way of summary. Clearly, most 

 structures connect with most others, and often by- 

 direct paths. And, equally clearly, each author finds 

 the neural region that has won his devotion to be 

 responsible for great physiological deeds. The compe- 

 tition is especially keen between the reticular forma- 

 tion (French), hypothalamus (Ingram), cingulate 

 (Kaadaj, hippocampus (Green), amygdala (Gloor) 

 and the whole limbic system (MacLean). At least the 

 amygdala facilitates and the hippocampus inhibits the 

 hypothalamus (195); this controls pituitary hormones 

 in a highly patterned way ( 143), and influences drives 

 and vigilance 1 Stellar 1, partly via the reticular 

 G 11 mation (French 1. 



LOCAL and RELAY. Perhaps the most important 

 dichotomy, certainly the Longest and best known one, 

 is that between short-circuit and long-circuit re- 

 sponses. Most responses were primitively at the 

 ganglionic or segmental levels, with minimal spread 

 to adjacent aggregates. Later, with anatomical cen- 

 tralization and cephalization, then- was aho spread ol 

 physiological activity, involving especially long- 

 circuiting towards the head end. Then were de- 

 veloped the powerful distance receptors feeding 

 increased information to new superposed nuclei, and 

 new unbroken nerve tracts to handle the information 

 How between these new centers and the main body 



The central gray columns and the fasciculi proprii, 

 with their rich neuropil, constitute a widespread 

 mechanism for local traffic and peripheral decisions. 

 As pointed out earlier, a major challenge is to rational- 

 ize the nerve impulse traffic and the associated 

 information How. Presumably the besi established, 

 most regular .un\ phylogenetically oldest responses 

 can most safely be entrusted to local autonomy. This 

 is well exemplified in the axon reflexes and the 

 ganglionic reflexes of the autonomic nervous system, 

 which continue with little disturbance when discon- 

 nected from the central neur.ixis, and only less well in 

 the basic spinal reflexes which, although clearly 

 modulated from higher up, continue to function after 



