neurophysiology: an integration- 



's 1 



be disposed in certain geometrical relationships, 

 with only so many layers of cells and certain dis- 

 tances of separation; with lateral positioning as well 

 as depth positioning playing an important role in 

 function (cf. Chang, Jasper, Bartley and Rose & 

 Mountcastle). Indeed, some sort of linear folding is 

 seen in a number of deep gray structures. 



Geometry has nothing to do with connectivity as 

 such; neuron A could connect with neuron B at any 

 distance of separation within the nervous system 

 and so preserve the topological network for the flow 

 of nerve impulses (201). Only what connections exist, 

 and their algebraic signs, matter in a 'graph' struc- 

 ture (131) or a sociometric one (218). The actual 

 geometry would influence the time taken for im- 

 pulses to pass between centers and so the temporal 

 pattern of arrival of impulses of mixed ancestry. 

 It would influence the number of connections to the 

 extent that axon collaterals or dendrite branches vary, 

 even on a statistical average, with distance from a 

 cell body. And it would influence the chemical ,mcl 

 electrical fields set up around active neurons and the 

 spread of such fields to other units, for example by 

 means of transmitters (von Euler). It is also, need- 

 less to say, essential to direct experimental manipula- 

 tion of the brain. 



The spread of activity waves and patterns through 

 a neuron mass is thus probably highly dependent on 

 the actual topographic relations, as is the passage of 

 light through plane or curved glass. Some kind of 

 focusing is indicated both by observation (179) and 

 theory (24), as will be discussed. A prototype is 

 perhaps to be seen in the experiment of compressing 

 a frog sartorius with a slightly tilted plane: conduc- 

 tion remains intact when approaching from the wide 

 to the narrow portion of the compressed region, is 

 blocked in the other direction (82). If, in the course 

 of evolution, neurons have tended to migrate to- 

 wards one another, centralization, and towards the 

 head end of the animal, cephalization, this presum- 

 ably has functional value and also indicates that 

 topology is not enough. 



Table of Organization 



decision points. Topology is, however, of the utmost 

 importance. Neurons are not randomly connected 

 throughout the nervous system; there are major 

 and minor flow channels in the grouped fiber tracts, 

 .is well as free seepage in the neuropil masses. An 

 enormous amount of detailed mapping has been 

 done, but very few general principles have yet 



emerged. Synapses, as compared to unbroken fibers, 

 are notoriously poor in transmitting information. 

 They are slow, nonquantized and variable, and the 

 information input and output are loosely linked. 

 In the nerve fiber, transmission is speedy, constant 

 and dependable, and there is normally a 1 : 1 rela- 

 tionship between input and output. Vet, although a 

 ^ins;le cell body can maintain an enormouslv long 

 nerve fiber, most cell processes are short, and the 

 conducting channels are chopped to bits with synap- 

 tic intersections and interpolations. Thev may even 

 conduct, by clectrotonic currents, in a graded fashion 

 and with a decrement (36). Clearly, the gain in 

 mixing different sorts of information, in modifying 

 current activity by past residues, in modulating func- 

 tional state in accord with wider body conditions, 

 as chemical milieu, and in integrating all these in- 

 fluences into a single output pattern, is of 1. 11 greater 

 value to the organism than are speed and precision 

 of transmission, as such. The synapse, or more broadly 

 the total pre- and postjunctional synaptic mechanism, 

 i-- tin- essential decision point in the nervous system; 

 and the number of synapses and their patterns of 

 interconnection are perhaps the most important 

 parameters of a nervous system. In no case are these 

 known or even scriousK guessed at. 



feed-back LOOPS. Some components of the connec- 

 tivity patterns are now recognized. Perhaps most 

 important is the feed-back loop, nearly always a 

 type of automatic volume control. At practically each 

 junction along a pathway, from receptor end organ 

 to effector element, including the synaptic links en 

 route, there is a negative feed-back control from 

 points downstream (Livingston, Neff, Paillard and 

 French). French, for example, specifies the existence 

 of feed-back control of ear, eye, olfactory bulb, 

 muscle spindles, skin, and sensory relax functions in 

 cord and brain stem, by cerebellar and cerebral 

 action on afferents, as well as of temperature regula- 

 tion, endocrine secretion, sensory gating, awareness 

 level, motor performance, etc. Adrian notes that 

 even the secondary controls on receptors feed in- 

 formation back to the nervous system. This negative 

 control is essential if explosive irradiation is to be 

 prevented, as will appear; but it is not immediately 

 clear why these controls are so numerous — a single 

 regulator at the input would suffice for the most 

 obvious control need. If, however, perceptions are 

 built, like motor acts, of learned components, and 

 higher level keyboards play upon lower level ones 

 (Paillard), clearly such multiple control is called 



