336 BRAIN MECHANISMS AND LEARNING 



that changes in excitabihty of neurones may be produced by the electrical 

 fields generated by activity of adjacent neurones. 



But even such conditions of electrical field interaction would merely 

 have the effect of distorting the finely patterned influences that arise 

 through synaptic action. There seems to be no alternative to the highly 

 selective synaptic actions when attempting to account for the enormous 

 wealth of information that is preserved during transmission through the 

 central nervous system. It may be stated as a general rule, that the number 

 of nerve fibres in any pathway is related to the wealth of information 

 that has to be transmitted along that pathway. It seems inconceivable, for 

 example, that all the fine grain of information transmitted by the optic 

 nerve and tract could be dissipated in the mere generation of electric 

 fields in the occipital cortex. In contrast it should be mentioned that some 

 synapses in Crustacea do operate by electrical transmission (Furshpan and 

 Potter, 1959). But such a mechanism depends on special permeability and 

 rectification properties of the apposed synaptic membranes, and is thus 

 just as unique and localized in its action as is chemical synaptic trans- 

 mission. Moreover such an electrical synaptic mechanism would have 

 been detected if it were opera.tive at any of the vertebrate central synapses 

 that have been systematically investigated. 



Two explanations, not mutually exclusive, have been proposed for the 

 neurological basis of learning and 'conditioning (Hebb, 1949; Eccles, 

 1953; Young, 1951; Thorpe, 1956). According to one, learning is a 

 dynamic process, due to continuously circulating patterns of impulses in 

 closed neural chains (Rashevsky, 1938 ; Young, 1938 ; Hilgard and Marquis, 

 1940; Householder and Landahl, 1945). As a consequence, the reaction of 

 the nervous system to any particular sensory input is changed in a unique 

 way so long as this circulation of impulses continues. Conceivably, this 

 explanation could apply at brief intervals — seconds or minutes — after 

 some initial conditioning stimulus. It certainly cannot account for 

 memories or conditioned behaviours that survive either a virtual suppres- 

 sion of all activity in the cerebral cortex — e.g. deep anaesthesia, concus- 

 sion, coma, extreme cold, or even deep sleep — or the converse, convulsive 

 seizures of the whole cortex. The alternative explanation is that activation 

 of synapses increases their efficacy by some enduring change in their fine 

 structure (Tanzi, 1893; Cajal, 191 1; Hebb, 1949; Toennies, 1949; Eccles, 

 1953 ; jLiiig, 1953 ; Mclntyre, 1953 ; Thorpe, 1956). We may assume that a 

 given sensory input results in a uniquely patterned activation of central 

 neurones, and, according to this explanation, a subsequent re-presentation 

 of this input would tend to be channelled along the same pathways 



