CAPABILITIES OF THE GOELENTERATE BEHAVIOR MACHINE

Abstract

Animals are essentially predatory behavior machines. So also are insectivorous plants which have developed raptorial feeding devices. Diploblastic and triploblastic animals meet the specification of such machines in different ways. In the Cnidaria both muscle and nerve-net seem organised on the basis of two-dimensional continuous sheets with local specialization. The condition is simplest in Anthozoa: in Scyphozoa, and still more in Hydrozoa, there are further complications. This simple picture of the Anthozoan nerve-net meets difficulties. Quick and slow contractions of the same muscle sheet are in fact operated by the same nerve-net. The slow contraction involves muscle-conduction and recruitment. A method of directly observing this is described. No complete explanation is yet forthcoming for reciprocal inhibition. The preservation of functionally significant shape seems to require proprioceptive machinery not yet discovered. It now seems well-established that both through-conduction and the original notion of interneural facilitation are valid elements in simple reflex responses. Knowledge of the importance of rhythmic phasic activity has, however, greatly increased in coelenterates generally. Many of these sequences of rhythmic activity seem to be based on modifications of the similar pattern sequences to meet different functional needs. Particularly in connection with these phasic activities, multiple action potentials both in response to stimuli and by spontaneous occurrence have been demonstrated. In Calliactis, 30% of records to threshold stimuli show evidence of multiple impulses. There is reason to associate such repetitive discharges with multipolar ganglion cells. The relation of these multiple discharges to the functional behavior is not always apparent. Complication of behavior in coelenterates is charactically on the motor side. Contrasted with triploblastic animals with probably the same order of number of nervecells (roughly 10⁵), there is a striking difference in the sensory equipment: exteroceptive information about the objects of the real world is lacking. A hunting-wasp with about that number of cells acts as though it had abstracted a world-model of objects, analogous to our own model, from the information received. But an anthozoan shows no evidence of that power. The importance of key-stimuli in anthozoan behavior is significant in connection with this. These deficiencies in complex behavior may be related to the topographical difficulty of complex correlation of sensory input in a two-dimensional net. The difficulty is easily overcome in the three-dimensional nets of triploblasts. Nevertheless, recent studies of conduction in the two-dimensional coelenterate net show striking ‘pre-adaptive’ features analogous to those of triploblast central nervous networks upon which sensory abstraction of information in these depends.