PRIMITIVE NERVOUS SYSTEMS

For nearly fifty years the dominant and accepted theory of the evolution of the nervous system has been that of G. H. Parker, developed over the course of the decade from 1909 to 1919 and summarized in his now classic book The Elementary Nervous System.' Like many successful theories, it was an oversimplification of contemporary knowledge presented in a forceful and stimulating manner. In essence, Parker viewed the initial evolution of the nervous system as involving three successive phylogenetic stages. At first there were only "independent effectors," represented today (in Parker's view) by sponge myocytes or coelenterate nematocysts. Secondly, receptor cells evolved from undifferentiated epithelium adjacent to the muscle cells. "The most primitive nerve cell from the standpoint of animal phylogeny is the sense-cell, or receptive cell, such as occurs in the sensory epithelium of the coelenterates."'1 Finally, "protoneurons" evolved between receptor and effector to give rise eventually to the reflex triad of receptor, adjustor, and effector. Parker's was not the first theory to be expounded. A generation before him Kleinenberg2 had used his discovery of the "neuromuscular cell" of Hydra to support the theory that the reflex triad originated from the division, in evolution, of what had once been a single cell. But the Hertwigs2 identified the "neuromuscular cell" as an epitheliomuscular cell, and, invoking the biogenetic law, suggested instead a simultaneous evolution of nerve and muscle cells from separate epithelial cells, as they are formed in coelenterate development. Until Parker's theory was advanced, the Hertwigs' theory was the one generally accepted; it was rapidly replaced, however, by the strength of Parker's concept of the independent effector stage, exemplified by the sponges. Since a great deal of what is known about nervous systems has been learned since the publication of Parker's book, in addition to a modest revival of interest in the nervous system of coelenterates, led by Pantin, it is noteworthy that the independent effector theory continues to be generally accepted. Neurophysiologists such as Bishop3 and Grundfest4 have emphasized that graded responses unknown in Parker's day certainly antedate the all-or-none action potential of the nerve cell axon, while the terminal secretory activity of the synaptic transmitter would also be expected to have evolved as a separate step. But neither question the fundamental sequence in Parker's theory that the independent effector evolved first and that then, in the next "brief and logical step,"3 a special sensory cell evolved to depolarize the adjacent contractile unit at the "first synapse."3 Pantin, whose approach has been that of the coelenterate biologist rather than that of the neurophysiologist, has been more critical5 of Parker's theory. Rather than being concerned with the evolution of certain physiological processes, he has considered the entire biology of the postulated ancestral forms, for "the metazoan behaviour machine from its origin... must have involved the structure of the whole animal, and it must have been complex enough and organized enough to meet all the varied requirements of behaviour."5 Just as in coelenterates today, the effector