I.—PHYSIOLOGY 169 
In reactions where there is no evidence of localisation (e.g. the learning 
of maze habits in the rat), Lashley finds that the important factor is the 
total mass of the cortex and not the presence of particular regions. ‘The 
effect of an injury depends on its extent and not on its situation. It 
depends, too, on the amount of grey matter (nerve cells and dendrites) 
destroyed, and not on the cutting of connections between the different 
parts of the cortex. ‘Thus the ability of the brain to form new associations, 
and generally to control the behaviour of the animal, depends primarily 
on the total area covered by the nerve cells of the cortex and their inter- 
lacing dendrites. For certain reactions it depends to some extent on 
the arrangement of pathways, but this arrangement is not essential. 
There is more localisation of function in the large brain of man than in 
the very small brain of the rat, for different cortical regions may be 
completely equivalent when they are separated by 5 mm., but not when 
they are separated by 100. But apart from this difference in scale it is 
likely that the human cortex has the same mass effect and plasticity of 
function. 
How do the individual neurones combine to produce a system which 
can recognise a triangle or direct the movements of the organism with 
such disregard of detailed structure? If particular neurones or path- 
ways are not tuned to triangularity, how can the whole mass be tuned to 
it, and why should the tuning be more certain when the mass is greater ? 
Our data may be at fault and the mass effect an illusion, but there is 
certainly enough evidence for it to be taken seriously. ‘Though there 
is no solution at the moment, I cannot believe that one will not be 
found—a solution which need not go outside the conceptions of physiology. 
It should be possible, for instance, to find out how many neurones must 
be combined to give a system which reacts in this way and what kind of 
structure they must form. ‘The nervous systems of insects may provide 
the clue, for these may contain a few thousand nerve cells in place of the 
ten thousand million in the human brain. It is possible also to study 
the reactions of isolated parts of the central nervous system, to see how 
far their behaviour can be explained in terms of the units which compose 
them. The retina is an interesting example of this kind, for it contains 
an elaborate structure of nerve cells and dendrite connections, and has 
some of the reactions which we might expect from a mosaic of sensory 
endings, and some which depend on interaction between the different 
neurones. But even now we can form some idea of the way in which the 
grey matter ‘can act as a whole. The electric oscillations in the cortex 
and in the grey matter generally are often due to a large number of units 
pulsating in unison. Sometimes there are several competing rhythms, 
and sometimes the collective action breaks down altogether, to reappear 
from time to time when some part of the system is stimulated to greater 
activity. When these collective rhythms appear the neurones are already 
acting as though they formed one unit. There is no need to regard the 
dendrites as forming a continuous network, electric forces may well 
bridge the gaps between them, but they may form a system in which 
activity can -be transmitted more or less freely in all directions. The 
patterns of activity in a system of this kind would be like the ripples on 
G2 
