\V. R. ADEY 587 



to lie at the basis of the neural process ot learning, rather than, perhaps, 

 cybernetic notions of reverberant circuits, or even more convenient neural 

 models in which every brain neurone is connected to every other, we might 

 speculate about the parts ot the neurone in which training might produce 

 altered conduction and excitability characteristics, responsible in turn tor 

 altered patterns of phase relations. It is probably siguiticant that the 

 regular rhythmic hippocampal slow waves seen here in association with 

 the motor pcrtormance arc maximally distributed in zone CA4, a region 

 particularly rich in both dendritic arborizations and tnie presynaptic 

 terminals. 



There is much evidence that the rhythmic hippocampal slow waves 

 arise in this dendritic layer, and that this activity is not associated with a 

 massive tu'ing of the cells themselves (Green and Machiie, 1955). Such 

 waves have been attributed to ephaptic activity between adjacent dendritic 

 trees in the production of the cortical alpha rhythm (Li, McLennan and 

 Jasper, 1952). These findings may indicate the importance of neural 

 integration ot considerable complexity occurring in dendritic and 

 presynaptic mechanisms. 



It is not improbable that neural integration ot this type may occur in the 

 vertebrate central nervous system only in the brain and not in the spinal 

 cord, and that it may be dependent on functional and structural arrange- 

 ments of dendritic and associated presynaptic structures not seen outside 

 the brain itselt. It should be a continuing and salutary reminder to those 

 disposed to extrapolate directly from spinal to cerebral mechanisms ni 

 processes of learning that, at the physiological level, singular differences 

 exist between spinal and cerebral synaptic functions on even such a simple 

 basis as the activity of strychnine. This evidence has been reviewed else- 

 where (Adey, 1959). 



At the behavie^ural level, it is a matter ot concern that the brain ot many 

 invertebrates, including arachnids and earthworms, with a complexity far 

 less than that of the human spinal cord (Adey, 195 1), can be readily 

 conditioned, whereas conditioning of the human spinal cord has remained 

 at best a most difficult procedure. We might inquire as to whether it is 

 reasonable to interpret all aspects of neuronal excitability solely in terms 

 ot the membrane potential and trans-synaptic events impinging on it. It is 

 difficult to avoid the conclusion that such a point ot view would reduce 

 the role of intracellular structures to that of air in a balloon. An embracing 

 view of the neurone in the learning process must surely take account, not 

 only of the cell membrane, but also of intra- and extracellular relations, 

 particularly as the so-called 'extracellular' space has been shown in both 



