The Origin and Development of the Nervous System

It is in general true that in bilateral animals the chief receptors and the chief adjuster or central organ are situated in the head region, so that the functional direction of the reflex arcs is predominantly posterior, at least primarily. Attention has already been called to the fact that in various flatworms and annelids excitation is effective chiefly in the posterior direction. Stimulation of, or injury to, the head may produce effects throughout the whole length of the body, but stimulation of levels posterior to the head has relatively little effect anterior to the level of stimulation and much greater effect at more posterior levels. These facts indicate that the main conduction paths, or at least those which have to do with the body wall, function chiefly in the posterior direction, and that the head, so far as its relations to the rest of the body are concerned, is to a considerable extent autocratic. Unquestionably numerous local arcs exist in these forms, and it is by no means necessary to assume that the axial relations are the same for all organs of the body. As regards the alimentary tract of Planaria, for example, there is some evidence that the pharyngeal region, not the anterior end, is the dominant region, though whether nervous differentiation exists in the entoderm is not known. In most annelids the localization in early stages of a secondary growing region at the posterior end of the body determines a secondary gradient opposite in direction to the primary gradient and involving more or less of the posterior body region (Hyman, 1916; Child 191 7 d). In many of the aquatic oligochaetes the pos- terior end is more or less dominant over the levels of this secondary gradient. The dominance of this

posterior end over more anterior levels of the secondary gradient appears very clearly in the respiratory move- ment of the posterior body region when the animals are undisturbed. This movement of the body to and fro in the water is initiated at the posterior end, and that it is independent of the head is evident from the fact that it may occur in headless animals. But although this poste- rior region is ordinarily largely independent of the head, and the impulses in it apparently run predominantly in the anterior direction, a sufficient stimulation of the head, or in fact of any region of the body, brings this region more or less under the control of the head, the respira- tory movement is inhibited, and the animals may move in a more or less co-ordinate manner. In these forms, then, the functional directions of the chief arcs are apparently opposite in anterior and posterior regions under conditions of slight stimulation but with suffi- ciently intense excitation impulses may pass in either direction in all regions. In other words, the functional direction of the arcs which predominate under conditions of slight stimulation is down the gradients in both directions, but with strong stimulation impulses are conducted up as well as down. Paths upward are present, but a high degree of excitation is apparently necessary to make them available. To what extent or whether at all the upward and downward paths are iden- tical is not known, though it appears not improbable that they may be identical over some portions of their length.

While any extended survey of the conditions in the various groups is impossible at this time, it may be noted that with the progress of evolution the relation to the cephalic region of reflex arcs with receptors in postcephalic regions undergoes change. Apparently it becomes increasingly possible for excitation in such arcs to affect more anterior levels of the body and finally the cephalic region. In other words, the upward paths along the longitudinal axis apparently become more numerous and more readily traversed by impulses. It has been pointed out, for example (p. 99), that in the wormlike invertebrates even the excitation resulting from section of the body is much less effective at levels anterior than at levels posterior to the section, and less extreme excitations often produce no effect anterior but a marked effect posterior to the level of stimulation. The head region in such forms is to a considerable degree autocratic with respect to posterior levels. In the higher vertebrates, however, excitations from any level of the body may reach even the cerebral cortex, not directly, it is true, but only through the interme- diation of various correlation centers. In the higher vertebrates and man, as compared with other forms, the upward paths in the longitudinal axis have appar- ently become more numerous and more readily traversed by impulses. Even in these forms, however, the upward paths in their definitive condition differ in general from the downward paths in that they usually consist of several neurons, while the downward paths consist of a smaller number or of only one. Functionally the definitive upward paths appear to present obstacles to the passage of impulses, while the downward paths are more simple and more open.

Special attention must be called in this connection to two facts : first, while this difference between upward and downward is characteristic of the definitive paths in the higher vertebrates, it is much less distinct in the lower vertebrates. The condition in mammals and man evidently represents the most advanced stage of an evolutionary process as regards these paths. Second, although our knowledge of the developmental changes in the nervous system, which lead up to this definitive condition, is fragmentary, certain facts at hand show that, in the Amphibia at least, the condition in early developmental stages is not only very different from the definitive condition in that group but more or less the opposite of the definitive condition in the higher ver- tebrates. Since the facts as they stand appear highly significant in relation to the physiological gradients and to the changes which they undergo in the course of development, a brief survey of certain of them is essential.

In the course of anatomical and physiological studies of the development of the nervous system of salaman- ders, Coghill (1909, 1913, 1914) has found that in the earliest stages at which movement of the body in response to tactile stimulation is possible, a definite reflex mechanism has already been developed, which differs in certain ways, highly significant in the present connection, from the definitive mechanisms. This mechanism is discussed in a later paper by Herrick and Coghill (1915) and is indicated in Figure 70, taken from that paper. Certain large sensory cells in the cord (giant cells of Rohon-Beard) have given rise to outgrowths regarded by the authors as dendrites (see pp. 196-200), which reach both the skin and the muscles, and to other outgrowths (axons), which ascend the cord

all the way to the posterior end of the medulla oblongata. In this region they connect with the dendrites of neurons, whose axons form a ventral commissure across the cord FIG. 70. — Diagrammatic dorsal view of the upper end of the spinal cord of larval Amblystoma, illustrating the relations of the giant cells, commissural cells, and motor cells and the mechanism of the swimming reflex (based on the researches of Coghill, 1913, 1914; from Herrick and Coghill, 1915).

and connect with the motor neurons of the opposite side. These motor neurons send axons down the cord and collaterals to the muscles. At first the commissures occur only at the posterior end of the medulla, but in later stages the region of commissural connection extends both anteriorly and posteriorly, until commis- sures are present at all levels posterior to the optic region. The first motor response to tactile stimulation of regions posterior to the head is the contraction of muscles of the side opposite to that stimulated, first in the more anterior region just behind the head and later at more posterior levels, and the reflex mechanism indi- cated in Figure 70 constitutes the anatomical basis of this type of response. At this stage and in this the earliest reflex mechanism of the cord it is the upward path which is uninterrupted by synapses, while the downward path shows frequent interruptions of this kind and is apparently somewhat later in its develop- ment. In later stages the giant cells disappear, and other reflex mechanisms develop, until the definitive condition is attained.

These different characteristics of the upward and downward paths in the different vertebrate groups and in different stages of development of the individual must of course be dependent on differences in physiological conditions of some sort. The experimental data at hand concerning the physiological gradients in the vertebrates serve, I believe, to throw some light on these changes. It has already been pointed out (chap, viii) that in all segmented animals investigated, including annelids as well as vertebrates, a second region of high susceptibility and rapid growth arises early in develop- ment and gives rise to most or all of the postcephalic region. The presence of this second region of high metabolic rate has been demonstrated in fishes and in the chick by Hyman, whose work along this line is now

in press, and in the frog by Bellamy (1919). As the body develops, this region moves posteriorly and repre- sents a posterior growing region of the embryo. Hyman has found that in the fish its susceptibility is equal to, or higher than, that of the head and much higher than that of more anterior postcephalic levels. Moreover, it is electronegative externally to all levels of the body behind the head. Bellamy has found that the suscep- tibility relations of this region in the frog are essentially similar to those in the fish, and the newly hatched tad- pole shows a double gradient in susceptibility and electric potential, both susceptibility and external elec- tronegativity decreasing from the posterior tail region to the body and from the head to the posterior body region, the posterior tail region showing at this time the highest electronegativity of any part of the body (unpublished) . The reaction of the tadpole in a constant current, which consists in orientation with head toward the anode, instead of toward the cathode, as in most of the lower animals, is unquestionably dependent upon these physiological conditions. In still later stages of vertebrate development the posterior growing region disappears and with it usually the secondary gradient, though in fishes and in the tailed Amphibia it may persist to some extent in the tail region.

These facts show that there is a period of develop- ment in the vertebrates during which the longitudinal physiological gradient in the postcephalic region has its high end posteriorly and its low end anteriorly, i.e., just behind the head. Moreover, the presence of elec- trical differences associated with this gradient has been demonstrated in fishes and Amphibia. As regards the Amphibia, this is the period during which the reflex mechanism above described develops. In these experi- mentally demonstrated physiological axial relations we find, I believe, a simple basis for interpretation, not only of the remarkable reflex mechanism of early stages which Coghill has described, but of the progressive changes and the definitive condition. If the suggestions advanced in chapter xi concerning the significance of electrical polarization in determining neuron patterns are correct, there is no difficulty in understanding why axons growing down a general axial gradient are long and those growing in the opposite direction short. As I have suggested (p. 190), the growth direction of those growing up the general gradient is determined by local gradients arising from other neurons in the vicinity, or from other local conditions, and opposite in direction to the general gradient. The axon tip will remain posi- tively polarized and will therefore grow only as far as the electrical factors of this gradient over-compensate those of the general gradient.

In the early stages of salamander development the upward sensory paths in the cord are growing down the general axial gradient as it exists at that time in that region, and they grow approximately to the level at which this gradient is compensated by the cephalic gradient. The downward motor path, on the other hand, is growing against the general gradient, and the length of each axon probably represents approximately the length at that tune of the local gradient which determines it. In later stages, as this secondary developmental gradient becomes less marked, the sen- sory paths developing from the more posterior levels of

the cord do not extend so far anteriorly but end at lower levels. Still later, as the growth activity and the internal electropositivity decrease at one level after another in the posterior direction and the gradient of the head region in the primary direction extends farther and farther posteriorly, this reflex mechanism of the earlier stages disappears and others take its place, following the gradient changes, until the definitive condi- tion is attained. This case has been considered at some length because it seems highly significant. The appear- ance in the salamander of the early reflex mechanism so different in character from the definitive mechanisms and its complete disappearance later remain completely inex- plicable facts, except in terms of the gradients as experi- mentally demonstrated. Viewed in relation to the gradients, however, these facts receive a simple physio- logical interpretation and fall into line with many others. Various other features of nervous development will probably be found to be related to the general or to local gradients or both, as they exist at the various stages. In general a certain similarity exists between the structural and the functional characteristics of a path in relation to the gradients. This similarity appears in the differences between definitive upward and down- ward paths in man and other mammals. We may expect that a path which develops and functions up a general gradient will be made up of relatively short neurons and will show a high " resistance," while a path in the opposite direction will be relatively simple and open. Even the more primitive transmission paths exhibit a more or less definite functional direction with respect to physiological gradients present, excitation

being effectively transmitted to greater distances down than up a gradient. If the electric current produced at a region of excitation is the transmitting factor this difference is readily understood (p. 97). As structural differentiation of transmission paths progresses, the morphological features of this relation to the gradients become more and more evident, until the conditions found in the higher animals and man are attained. Moreover, as regards the functional relations of neurons, even if we admit that conduction may occur without decrement in the single axon, the probability exists of quantitative differences in excitation in different neu- rons and of the relation of these differences to levels of the gradients and to the changes in the gradients which occur during development and evolution.

These suggestions do not involve the assumption that the growth of the axons is determined by the passage of nervous impulses, but if the suggestions advanced in chapters iv and xi have any basis in fact, the growth of the axon and the transmission of excitation are both determined by the electrical factor. The growth of the axon is the physiological effect of increase in internal positivity produced by the electric currents in a physiological gradient, and transmission is the physiological effect of the electrical currents associated with a local excitation, this effect being primarily, according to Lillie, an increase in internal positivity. From this viewpoint the outgrowth of the axon and the transmission of excitation are simply reactions of protoplasm to essentially the same electrical conditions, and it is not strange that they should exhibit certain similarities.

Whether the paths from postcephalic regions upward are, in the higher vertebrates, parts of the main paths of the reflex arcs, as Pike (1909, 1912) maintains, or are lateral or accessory paths not essential to the functioning of the arcs, as has been generally believed, is of secondary importance in this connection. The upward paths exist, and their existence suggests that the general antero- posterior physiological gradient in the central nervous system has become in some way less effective in deter- mining functional direction in the higher than in the lower animals, and that the primitive relation of domi- nance and subordination with the more or less complete autocracy of the apical region or head in flatworms or. annelids is undergoing alteration toward a relation of true co-ordination in which many different arcs with receptors in various regions of the body are concerned in determining the final effect or action. The physio- logical dominance of the nervous system and of more anterior over more posterior levels is in fact departing from the primitive, autocratic relation and becoming more and more like that of a representative government. Such change does not mean a decrease in the degree of physiological integration but rather an increase in the complexity, flexibility, and potentialities of integration. It does not represent a functional degradation of the higher levels but a functional elevation of the lower levels of the gradients, and this constitutes a basis for relations and combinations of arcs and paths which are impossible in the lower forms.

Such a decrease in functional or apparent steepness of the physiological gradient must be associated with the development and evolution of the mechanism of excitation in protoplasm. Certainly the intensity of the excitation process increases in general in development and evolution, and such change undoubtedly determines more effective transmission or conduction, just as a rapid change in electric potential is more effective than a slow change in producing excitation. It is not improbable also that in the more highly specialized excitatory processes the electrical changes are greater than in the primitive processes, and this again must contribute to more effective transmission. Other changes may also be concerned, but the point which I wish to make is that the changes are such as to make it possible for .conduction of a local excitation to occur to a greater distance up a physiological gradient in the higher animals than in the lower. If the conception of exci- tation advanced in earlier chapters is correct, the change is probably primarily electrical.

It may be also that the gradients themselves are less steep in the higher than in the lower forms, but as yet no adequate basis exists for comparison of steepness in different forms, though the electrical data will probably give some evidence along this line. Finally, the fact that the higher animals consist of segments and in con- nection with segmentation develop a secondary posterior growing region and a secondary longitudinal gradient hi the opposite direction to the primary gradient is unquestionably a factor of great importance in deter- mining that the more posterior levels of the body are able to send impulses all the way to the head. This gradient determines the primary reflex mechanism in Amphibia, as already pointed out, and the definitive relations arise later after the gradients have undergone

changes. The segmented animal does not represent the simple axial gradient of the lower forms, but is the result of a repetitive process of reproduction (p. 131). In early development a partial breaking up of the primary individual occurs and reintegration takes place only gradually as the segments successively become less active and the head region regains some degree of dominance. Under such conditions the relations of the different axial regions and of the different segments become much more complex than in the simple indi- vidual, and the possibility of impulses from posterior regions reaching the head is much greater. Segmenta- tion of the body is then also a factor in the development of the more complex and more flexible type of physio- logical integration characteristic of the vertebrates.

It appears probable that the conditions determining the increase in number and functional significance of the upward paths in the central nervous system in the higher animals are in some way and to some degree concerned in the origin and development of the cerebral cortex. The cortex is not simply a central or adjuster organ for the receptors of the head region, as the cephalic ganglia of the invertebrates appear in large measure to be. It does not properly belong to the original reflex arcs at all, but represents functionally an additional arc superimposed upon the adjuster portion of the original arc. It may be termed a super-adjustor, but it is not simply that, for super-adjustors exist in the subcortical correlation centers. The cortex is in fact the supreme

adjuster. This, however, does not mean that it is physio- logically dominant in the more or less autocratic manner in which the head or the apical end of the physiological gradient is dominant. The cortex resembles rather a deliberative assembly to which reports of matters requiring consideration come in from the various con- stituent groups or bureaus and. in which they are considered and action finally taken through the proper channels (cf. Herrick, 1913, p. 230). Not every afferent impulse reaches the cortex. In many, and particularly postcephalic arcs, as long as the reflex mechanism of lower centers gives satisfactory results the cortex does not become involved ; only when the stimulus continues and summation, or some other effect, makes it possible is the impulse able to reach the cortex. As Herrick has so clearly pointed out:

There are no afferent tracts leading to the cerebral cortex directly from any peripheral sense organ or from any center within the brain which is "pure," i.e., devoted to a single sensory function. In other words, no simple sensory impulses ordinarily reach the cortex, but only nervous impulses arising from lower correlation centers, where complex reflex combinations of various sensory systems are possible. The optic impulses reach the cortex most nearly pure, i.e., with less subcortical associational relation with other sensory systems (it is no accident that the visual sense plays a dominant role in human cortical function) ; but even here the optic centers in the thalamus from which the optic projection fibres arise are intimately related with acoustic, tactile and other important sensory centers. And in the case of all the other sensory systems, the projection fibers which enter the cortex come from centers which are separated from their respective sense organs by two or more association centers of a high order of com- plexity. Each of these subcortical associational centers may be dominated physiologically by a single sensory system, but it is structurally adapted for bringing that system into relation with

several others, so that the nervous discharge which emanates from it may be the efferent link in a very complex reflex arc. This efferent discharge may descend to the appropriate motor center, or it may ascend to enter a still higher association center, all of whose afferent tracts come from similar lower centers and therefore carry nervous impulses which represent a sort of physi- ological resultant of the functional factors there interacting [Herrick, 1913, pp. 223-24].

The cortex represents in fact a region into which impulses originating in any receptor may under proper conditions enter, but only after passing through one or more correlation centers. Physiologically speaking, it is the "central" nervous system par excellence in that it is hi a sense the meeting place of impulses from the most various sources. The physiological conditions concerned in the origin and development of the cortex must be conditions which make possible or determine the localization of such a truly central nervous region. Among these conditions, those which have led to the development of the upward paths discussed in the preceding section must have been important factors. It has been suggested in chapter xi that the primary directive factors in development of the nervous system are electrical, arising in connection with the physiological gradients and their modifications and producing physiological effects through polarization, and if that is true it is allowable to suggest that similar electrical factors acting in the same way may be the chief or the primary physiological factors concerned in the ontogenetic localization and development of the cortex.

Gradients in electric potential are characteristic features of the primary physiological gradients and undergo modification with the other features of the gradients. The cerebral cortex apparently arises in organisms in which the primary gradients become so modified in the course of development that a truly central region is localized, which may be affected to some extent by changes of potential in any part of the organism. This being the case, the electrical situation in the region from which the cortex develops must be characteristically different in some way from that in other regions. More specifically stated, this region may be conceived as a region which represents not merely a certain level in the general axial gradients but which is affected, for example, by the electrical changes con- nected with localized developmental activity in various organ primordia, first of all the receptors of the head. Later, as the changes which have been briefly discussed in the preceding section occur, as various parts of the nervous system successively become functional in a specifically nervous way and as the upward paths in the cord develop, progressive modification and complica- tion must occur.

The cortex represents a region which differs from others, not in kind, but in the variety and complexity of the structural and functional relations between its parts, and these relations must be determined by physiological factors of the same sort as those concerned in other parts of the nervous system, but acting under less simple conditions. If the conclusions and sugges- tions of earlier chapters concerning the part played in the development of the nervous system by electrical factors through their effects on physiological condition are correct, the physiological factors concerned in the

development of the cortex must be in large part electri- cal in origin. The anatomical and functional relations of the cortex paths indicate clearly enough that condi- tions there are unique as regards complexity, though not different in kind from those in other parts, and it is dim- cult to conceive any other than electrical factors which could determine these relations. If we could trace out the electrical gradients and their changes and the sequence in which different body regions and organ primordia affect the cortex region electrically, either through developmental activity or functional excitation and so bring about physiological changes in this region, we might attain some conception of the basis for its remarkable structure and function.

The evolutionary aspect of the origin and develop- ment of the upward paths and of the cortex is of course another matter, involving, as I have repeatedly pointed out, the hereditary mechanism of the protoplasm con- cerned. In other words, the gradient situation works out differently in different protoplasms, and in those in which it works out in certain ways a cerebral cortex and the other features associated with it appear. Whether the evolutionary process requires Lamarckian assump- tions or not does not primarily concern us here, but the gradient conception does not make such .assumptions necessary. However the necessary mechanism of proto- plasm is established, the development of the individual is a physiological process, and real advance in our con- ceptions of heredity is possible only as our knowledge of the physiology of development increases. The con- ception of the physiological gradient and of its relation to development and integration places these problems

of developmental physiology in a new light, and I have ventured to suggest that even the more complex aspects of nervous development and integration are related to the fundamental physiological features of organismic pattern, the gradients. From this point the attempt might be made to push suggestion and speculation along these lines into the field of psychology, but we are concerned primarily with the origin of the nervous system rather than with its relations to psychological phenomena. It seems allowable, however, to point out the possibility that the field or content of consciousness at any given moment may be associated with the determination of a field or area of relatively intense physiological activity in a particular region of the cortex as the resultant of the incoming impulses at that moment. If this activity is sufficiently intense, this area becomes for the time being the dominant region and its connections with other parts, together with the physiological conditions in other parts, determine the course of action. Hesita- tion, deliberation, delays, so characteristic of cortical reaction and consciousness, represent the period before complete dominance is attained and can be interpreted in terms of physiological factors. I do not mean to imply that each particular content of consciousness is localized in a particular cortical area, but merely that increased activity in a particular cortical region or regions is itself a resultant of a particular functional complex in the nervous system and the body as a whole, and that if such activity becomes sufficiently intense it determines a certain reaction complex. The structure of the cortex indicates that a general physiological

axiation of the cortex as a whole is absent or practically absent, although a very definite surface-interior pattern does exist. The dominant region of the cortex may therefore shift from moment to moment and is purely physiological and transitory rather than morphologically persistent. That there is anything essentially new in these suggestions I should be the last to claim. They are brought forward here merely as the final point in the discussion of functional direction in relation to the physio- logical gradients.

In the preceding chapters an attempt has been made to show the physiological relation and continuity between the structural and functional development of the nervous system and the primary factors in organismic pattern, the physiological gradients. From the viewpoint which underlies the book the organism is regarded as represent- ing essentially a behavior pattern, arising in protoplasm in response to its relations with the external world, and particularly the relations concerned in excitation and its effects. When we take all the facts into consideration, it is impossible, I believe, to reach any other conclusion than that the organism is such a pattern. The organ- ism is without significance or meaning of any kind, except in relation to an external world, and particularly to the factors in that world which bring about excitation. I have endeavored to show that organismic pattern is primarily excitatory and transmissive in character, and that the material exchange between protoplasm and environment is of secondary importance, so far as this pattern is concerned, though necessary of course for the continued existence of protoplasm.