ORGANIZATION OF PRIMITIVE CENTRAL NERVOUS SYSTEM 405 



efferent branches are jumbled together and consistent specific contacts have 

 been found only in a few instances where they are axon-axon synapses of 

 giant fibres to motor fibres, or in the optic ganglia of arthropods, where 

 spatial relationships must be maintained. In these few instances, as the giant 

 fibre-to-motor synapses of the crayfish (Johnson, 1924) or of the polychaete 

 nerve cord (Horridge, 1959) or the synapses between the giant interneuron 

 and the several giant motor neurons in the squid stellate ganglion (Young, 

 1939) there is a widespread activation which is anatomically addressed to 

 several motor neurons of a particular class. Here, the connexions, though still 

 widespread, show considerable differentiation of addressing, but these ex- 

 amples are drawn from giant neurons which are concerned with rapid and 

 stereotyped reflexes. 



The neuron patterns in the optic ganglia of insects offer examples of great 

 diversity of regional restriction and asymmetrical spread of histologically 

 known arborizations (Zawarzin, 1913; Cajal and Sanchez, 1915). The classes 

 of neurons that are localized radially below their corresponding ommatidia 

 are in a position to conserve the directional aspect of the excitation falling 

 on the eye, but many of these, having long arborizing branches, seem not 

 specific with respect to the depth to which they penetrate. On the other hand, 

 some interneurons spread laterally more widely, and lateral integration 

 between ommatidia could not occur unless some spread occurred. In contrast, 

 however, in all three optic ganglia there are classes of tangential or hori- 

 zontal neurons having processes which ramify throughout particular tangen- 

 tial layers of the ganglion. They are restricted to certain depths from the outer 

 surface of the ganglion but are in a position to take account of excitation 

 irrespective of the position on the eye of its ommatidia of origin. By its 

 laterally widespread connexions, this output from the optic gangha contrasts 

 strongly with the other, which has dendrites running more or less vertically 

 and axons which keep their ordered ranks as they run to the next proximal 

 ganghon. There are, therefore, two contrasting types of neurons to the brain, 

 those from particular directions of the visual field and those from various 

 definite depths in the optic lobes irrespective of the axes of the eye. The 

 functional significance of each individual neuron must be a consequence of 

 the directionality, extent and symmetry of the area over which its arboriza- 

 tions are spread in so far as it can have no connexions outside this area. 

 Within individual small areas, we meet the same problem as before in not 

 knowing how the axons are addressed. 



On the question of the origin of the differentiation of labelled lines, by what- 

 ever mechanism they function, there are hardly any relevant observations. 

 Neither anatomical nor physiological data, obtained by present techniques, 

 are a help in answering the questions raised. Certainly labelled lines can be 

 readily demonstrated physiologically but the mechanism of the channelling 

 of information as the physiological pathways pass through the neuropile is 



