ORGANIZATION OF PRIMITIVE CENTRAL NERVOUS SYSTEM 399 



animals such as sea anemones, corals and hydroid polyps, the nervous 

 system consists of a net of neurons in which every neuron appears to be 

 equivalent to every other, and to be connected by synapses or fusions with 

 any other neurons with which it comes in contact. Considered as a problem 

 in morphogenesis the neurons are merely scattered about and they need have 

 no morphogenetical specifications for forming physiologically significant 

 contacts with only certain other neurons; in fact the net seems to be what it is 

 because synapses are formed at random. Apparently a pattern of only this 

 degree of complexity serves adequately as a nervous system for most polyps, 

 and all known types of their rapid behaviour can be paralleled in a model by 

 changing the parameters of probabiUty of transmission, density of connexions 

 and so forth, in a randomly connected non-addressed net (Horridge, 1957). 



A similar set of apparently random connexions between neurons occurs in 

 the sub-retinal layer of the lateral eye of Limulus. There are two classes of 

 neurons, retinula cells and eccentric cells, but the branches of both have 

 close relations with their neighbours in a neuropile of only these two compo- 

 nents. Their effectiveness in inhibiting their neighbours falls off with distance 

 and only parameters which define this spatial relationship are required to 

 describe the mutual inhibition between the ommatidia (Ratliff and Hartline, 

 1959; Hartline f/ a/., 1961). 



The next grade of complexity is found where there are two overlying nerve 

 nets over much of the animal, and, as compared with the single nerve net, 

 these nets must be kept separate by an additional degree of specificity during 

 their growth. For example, Heteroxenia, Fig. 1, is an Indo-Pacific tropical 

 soft coral found in small clumps a few centimetres broad, from which spring 

 a few hundred polyps several centimetres long. The polyps continually beat 

 out of phase and by cutting them into pieces it can be shown that each of their 

 eight arms contains a spontaneously active centre which is apparently nervous. 

 Stimulation anywhere on the colony causes all polyps to stop at once and 

 analysis shows that this is co-ordinated by a distinct inhibitory nerve net 

 which must run to all the pacemakers of the colony (Horridge, 1956b). A 

 second example of widespread inhibition of pacemakers is provided by the 

 tropical anemone Boloceroides which swims by simultaneously twitching all 

 its tentacles downwards and outwards. The tentacles are co-ordinated by a 

 through-conducting pathway round the oral disk, and there are many 

 spontaneously active centres or pacemakers. Stimulation anywhere on the 

 anemone by a single shock stops the spontaneous twitching. Again it must be 

 inferred that an inhibitory net spreads to all pacemakers. In these examples 

 the inhibitory pathway must be widespread, anatomically irregular, and 

 addressed in only one respect, that is to seek out the net containing the 

 pacemakers. 



In the jellyfish the situation is known in more detail (Horridge, 1956a). 

 There are many spontaneous centres arranged round the margin of the bell, 



27 



