348 15RAIN MECHANISMS AND LEARNING 



diminution in the size of the synaptic vesicles at the synapses made by both 

 rods and cones with the bipolar cells in the rabbit's retina (De Robcrtis and 

 Franchi, 1956). A further observation was that, after i day of darkness, the 

 synaptic vesicles tended to accumulate close to the presynaptic membrane 

 (De Robcrtis and Franchi, 1956). It would be of interest to sec whether, 

 correspondingly, there is a phase of increased synaptic efficacy after 

 comparable periods of disuse brought about by nerve section or by the 

 other procedures outlined above. 



The ultimate aim of these investigations on plasticity would be to 

 correlate the observed structural and functional changes and to under- 

 stand the way these changes are brought about by excess use and by disuse. 

 Inevitably, such a programme involves questions relating to the control 

 of the manufacture of transmitter substance and to its availability for 

 release by the activated synapses. It would seem more probable that use 

 gives increased function by enhancing the manufacture and availability of 

 transmitter substance, although enlargement of synaptic knobs and even 

 the sprouting of new knobs are alternative devices for securing an increased 

 synaptic action (Konorski, 1948, 1949; Hebb, 1949; Young, 1951; Eccles, 

 1953). Evidently, further investigation by electron microscopy is of the 

 greatest significance in providing evidence discriminating between these 

 alternatives. 



Perhaps the most unsatisfactory feature of the attempt to explain the 

 phenomena of learning and conditioning by the demonstrated changes 

 in synaptic efficacy is that long periods of excess use or disuse are required 

 in order to produce a detectable synaptic change. In contrast, conditioned 

 reflexes are established by relatively few presentations, and unique events 

 may be remembered for a life time. A probable explanation is that pro- 

 longed revcrberatory activity occurs in the neuronal network, so that a 

 single event may activate each synaptic link in a spatio-temporal pattern 

 thousands of times within a few seconds. Hebb (1949) makes a related 

 postulate when he supposes that 'a revcrberatory trace might co-operate 

 with the structural change, and carry the memory until the growth change 

 is made'. A similar suggestion has been made by Gerard (1949). Further- 

 more we may suppose that the plastic changes in synapses are susceptible 

 to reinforcement by the replaying of the specific spatio-temporal patterns 

 each time that the memory is recalled. It is also possible that synaptic 

 plasticity may be much more highly developed in the cerebral cortex. 



