656 BRAIN MECHANISMS AND LEARNING 



with CS excitation being assumed to cause a general increase in excitability of the 

 neurones, which consequently can eventually be excited to discharge by the CS 

 alone, whereas initially CS was ineffective. Again there is no experimental evidence 



CS 



U^^,0 .R 



B 



us O *R 



Fig. I 

 Diagrams illustrating attempts to explain conditioned reflexes 

 by plastic changes in synaptic connections. Full description in 

 the text. CS and US show input into central nervous system of 

 conditioned and unconditioned stimuli respectively. In A the 

 arrows indicate nervous pathways, while B is a redrawing of a 

 highly simplified model in which Shimbel shows converging 

 synaptic connections of the CS and US lines, hi C nervous 

 pathways are drawn as broad bands along which conduction 

 occurs as described in the text, particularly in the neuronal 

 network, NN. The interruptions in the bands indicate synaptic 

 relays. Nerve centres containing large populations of neurones 

 are indicated by circles, while the neuronal network, NN, 

 would be an extremely complex neuronal system; for example, 

 an area or areas of the cerebral cortex (Eccles, 1953, reproduced 

 by permission of the publishers). 



for this non-specific increase in neuronal excitability, hicreascd activation ot 

 neurones changes only the synapses actually activated, i.e. it appears to be pre- 

 synaptic or synaptic not post-synaptic. Moreover, it gives rise to a serious problem 

 in attempting to extract from such a system an adec|uate degree of specificity when 



