(Return to index)

Our examination of the structural elements of the nervous system led us to conceive of the brain and myel, together with the nerves issuing from them, as a system of nerve-cells, inter-connected by their fibrillar runners either directly or through the contact of process with process. Our recent survey of the morphological development of the central organs lends support to this conception. We have found a series of cinereal formations which collect the fibres running centralward from the external organs and mediate their connexion with other, especially with more centrally situated grey masses. The paths of conduction that begin in the myelic columns pass upwards first in the crura and then in the corona until they penetrate the cerebral cortex. There we have the commissures, pointing to the inter-connexion of the central regions of the two halves of the brain, and the intergyral (arcuate) fibres, indicating the connexion of the various cortical zones of the same hemisphere. Hence from whichever point of view we consider the outward conformation of the central organs, we are presently met by the question as to the course taken by the various paths of nervous conduction. We know, of course, that the cell-territories stand, by virtue of the cell-processes, in the most manifold relations. We shall accordingly expect to find that the conduction-paths are nowhere strictly isolated from one another. We must suppose, in particular, that under altered functional conditions they may change their relative positions within very wide limits. But we may fully admit such a relative variability of functional co-ordination as is suggested by the neurone theory, and yet with justice raise the question of the preferred lines of conduction; -- of the lines which, under normal circumstances, are chiefly concerned to mediate determinate connexions, on the one hand, between the central regions themselves, and on the other, between the centre and the peripheral organs appended to the nervous system. This answered, we may in certain cases proceed to ask a second question, regarding the auxiliary paths or bypaths which can replace the regular lines of transmission in particular instances of interrupted conduction or of inhibition of function.[p. 151]

We distinguish two main kinds of conduction-path, according to the direction in which the processes of stimulation are transmitted: the centripetal and the centrifugal. In the former, the stimulation is set up at some point on the periphery of the body, and travels inwards, toward the central organ. In the latter, it issues from the central organ, and travels toward some region of the periphery. The physiological effects of a centripetally conducted stimulation, when they come to consciousness, are termed sensations. Frequently, however, this final effect is not produced; the excitation is reflected into a movement, without having exerted any influence upon consciousness. Nevertheless, the paths of conduction traversed in such a case are, at least in part, the same. We therefore give the name of 'sensory' to the centripetal conduction-paths at large. The physiological effects of centrifugally conducted stimulation are very various: it may find expression in movements of striated and non-striated muscles, in secretions, in heightened temperature, and in the excitation of peripheral sense-organs by internal stimuli. In what follows, we shall, however, confine our attention for the most part to the motor and the centrifugal-sensory paths, since these are the only parts of the centrifugal conduction-system that call for consideration in psychology. The muscular movements that result from the direct translation of sensory stimulation into motor excitation are termed reflex movements; those that have their proximate source in an internal stimulation within the motor spheres of the central organ we shall call automatic movements. In the reflex, i.e., centripetal is followed by centrifugal conduction; in the automatic movement, centrifugal conduction alone is directly involved.[1]

So long as the stimulation-process is confined within the continuity of determinate nerve-fibres; as occurs e.g. in the peripheral nerves, which often traverse considerable distances, it remains as a general rule isolated within each particular fibre, and does not spring across to neighbouring paths. This fact has been expressed in the law of isolated conduction. The law has usually been regarded as valid not only for the periphery, but for the conduction-paths within the central organs as well; on the ground that an external impression made upon some precisely localised part of a sensitive surface evokes a sharply defined sensation, and that a voluntary impulse directed upon a definite movement produces contraction of a circumscribed group of muscles. Really, however, these facts prove nothing more than that the processes in the principal paths are, as a rule and under normal conditions, separate and distinct. It has not been demonstrated with certainty that the stimulation is strictly confined to a single primitive fibril, even during the peripheral portion of its course. And in the central parts, any such restriction is entirely out of the question, as appears both from the general [p. 152] morphological features discussed in Chapter II., and from the phenomena of vicarious function which we shall speak of presently. The only principle that can be recognised here is a principle of preferential conduction. There is in every case a principal path, but this is supplemented by auxiliary or secondary paths.

We may avail ourselves of three distinct methods, in our examination of nervous conduction. Each one of them has certain imperfections, and must therefore be supplemented, where possible, by the other two. The first method is that of physiological experimentation; the second is that of anatomical investigation; and the third is that of pathological observation.

(I) Physiological experimentation attempts: to reach conclusions as to the course of the nervous conduction-paths in two ways: by stimulation-experiments, and by interruptions of conduction due to a division of the parts. In the former case, we look as a general rule for enhancement, in the latter for abrogation of function, in the organs connected with the stimulated or divided tissue. When we come to the investigation of the central paths, however, we find that both methods alike are attended by unusual difficulties and disadvantages. Even in the most favourable instances, when the stimulation or transsection has been entirely successful, we have established but one definite point upon a path of conduction; to ascertain its full extent, we should have to make a large number of similar experiments, from the terminal station in the brain to the point of issue of the appropriate nerves. Such a task holds out absolutely no hope of accomplishment, since the isolated stimulation or section of a conduction-path in the interior of the brain presents insuperable obstacles. There are, therefore, only two problems to which these methods can be applied with any prospect of success. We may use them to determine the course of conduction in the simplest of the central organs, the myel, and in the direct continuations of the myelic columns, the crura; and we may use them to discover the correlation of definite areas of the brain cortex with definite organs upon the periphery of the body. The answer to the former question has been attempted, for the most part, by isolated transsection of the various myelic columns; the answer to the second, by experiments upon the stimulation and extirpation of definitely limited cortical areas. Even with this limitation, however, it is difficult to secure valid results. A stimulation will almost inevitably spread from the point of attack to the surrounding part. This objection applies with especial force to the electric current, almost the only form of stimulus which fulfils the other requirements of physiological experiment, and a stimulus which the physiologist is therefore practically compelled to employ. The same thing is true of the disturbances consequent upon a division of [p. 153] substance. And if one is at last successful in securing the utmost degree of isolation of experimental interference, there will still be many cases in which the interpretation of the resulting phenomena is uncertain. The muscular contraction that follows upon a stimulation may, under certain circumstances, be due to a direct excitation of motor fibres, just as well as to a reaction upon the sense-impressions. And the derangements of function that appear as a result of transsections and extirpations always require a long period of observation before they can be accurately determined. This means that the certainty of the conclusions is, again, very largely impaired: the disturbances set up as the direct effect of operation for the most part disappear as time goes on, the explanation being that the principal path is functionally replaced by the secondary paths of which we spoke just now.

(2) The gaps left in our knowledge by the physiological experiment are largely filled out by anatomical investigation. The anatomist has followed two methods in the prosecution of his task: first, the macroscopic dissection of the hardened organ, and, later, its microscopic reduction to a series of thin sections. Of late years, the former of these two methods has fallen into disrepute, on the score that it runs the risk of substituting artificial products of the dissecting scalpel for real fibre-tracts. Carefully applied, however, it is a valuable means of orientation with regard to certain of the wider roads of brain-travel; while its critics are inclined, on their side, to underestimate the danger of error in the interpretation of microscopic appearances. And this danger is the more serious, the farther we are from the actual attainment of the ideal goal of a microscopic examination of the central organ, its complete reduction to an infinite series of sections of accurately known direction. For the rest, microscopical anatomy has been brought in recent times to a high degree of perfection by the application of the various methods of staining. The advantage of these is that they permit of the more certain differentiation of nerve-elements from the other elementary parts, and thus enable us to trace the interconnexion of the nerve-elements much farther than had before been possible.[2] Anatomical investigation is, further, very materially supplemented by embryological research. Embryology shows that the formation of the myelinic sheath in the various fibre-systems of the central organs occurs at different periods of foetal development, and thus puts it in our power to trace out separately certain paths of travel that in all likelihood are physiologically interconnected. This method, however, like the others, has its limits: the systems that develope simultaneously may [p. 154] still include numerous groups of fibres, possessing each a different functional significance.[3]

(3) Pathological observation is equally concerned with functional derangement find with anatomical change, and so in a certain measure combines the advantages of physiological and of anatomical investigation. The observations of pathological anatomy have been especially fruitful for the study of the nervous conduction-paths. Abrogation of function over a determinate functional area means that the fibres belonging to that area undergo secondary degeneration. The pathological anatomist can, therefore, appeal to a law very similar to that upon which embryological investigation is based. Unless there are extrinsic conditions present, which render an accidental concurrence of the degeneration probable, he can assume that all fibres which suffer pathological change at one and the same time are functionally related.[4] The observation of secondary degenerations is of especial value when conjoined with physiological experimentation. The joint method may follow either of two different paths. On the one hand, a severance of continuity may be effected at some point in the central or peripheral nervous system of an animal, and the consequent functional derangement observed. Then, after a considerable time has elapsed, the paths to which the secondary degeneration extended can be made out by anatomical means. On the other, a peripheral organ (eye, ear, etc.) may be extirpated in early life, and the influence observed which the abrogation of determinate functions exerts upon the development of the central nervous organs.[5] In the former case, the nerve-fibres evince the successive stages of degeneration represented in Fig. 23, p. 53. In the latter, the parts of the brain which serve as the centres for the abrogated functions sink in, and microscopic examination shows their nerve-cells in the various stages of atrophy that lead up, in the last resort, to complete disappearance (Fig. 22 B, p. 53).

The first extensive collection of material for the investigation of the microscopical structure of the central organs was furnished by the researches of STILLING. The earliest attempts to construct a structural schema of the whole cerebrospinal system and its conduction-paths by STILLING'S method, i.e. by the microscopical examination of sections, date from MEYNERT and LUYS.[6] MEYNERT, especially, rendered great services to the science; he brought to his re-[p. 155]construction of brain-structure the results of a comprehensive series of original investigations and a rare power of synthetic imagination. It is true, of course, that the schema of conduction-paths which he published was largely hypothetical, and that it has already been proved erroneous in many details. Nevertheless, it formed the point of departure for further microscopical research; so that most of the later work takes up a definite attitude to MEYNERT'S structural schema, supplementing or amending. The application of the various methods of staining, and the consequent differentiation of the nervous elements, have played an important part in this chapter of scientific enquiry. Embryological investigation depends upon the fact that the myelinic sheath is formed in the different fibre-systems at different periods of embryonic development. It is this sheath which is responsible for the coloration of the alba, so that its appearance is easily recognisable. The signs of secondary degeneration consist on the other hand, in a gradual transformation of the myelinic sheath. The tissue becomes receptive of certain colour-stains, like carmine, by which it is unaffected in the normal state. Finally, the myelinic sheath disappears altogether. At the same time, the nerve-fibres proper (neurites) change to fibres of connective tissue, interrupted by fat granules. The value of these degenerative changes for the investigation of conduction-paths lies in the fact that the progressive transformation is always confined within an interconnected fibre-system, and that the direction which it takes corresponds in all fibres with the direction of conduction (WALLER'S law); so that the degeneration of motor fibres follows a centrifugal, that of sensory fibres a centripetal course. Nevertheless, this law of WALLER, like the law of isolated conduction, appears to be valid only as regards the principal direction of the progress of degeneration. In cases where the interruption of conduction has persisted for a considerable length of time, and more particularly in young animals, the deterioration of the fibres is always traceable, to some extent, in the opposite direction as well. We have, further, besides atrophy of the nerves separated from their centres, a similar though much slower degeneration of the nerve-cells which have been thrown out of function by transsection of the neurites issuing from them. This secondary atrophy of the central elements, the initial symptoms of which are the changes represented in Fig. 22, p. 53 above, is, again, especially likely to appear in young animals. It may, however, occur in the human adult, after long persistence of a defect. Thus it has been observed that loss of the eye is followed by atrophy of the quadrigemina; nay, more, in certain cases of the kind, a secondary atrophy of certain cerebral gyres has been demonstrated.

The nerve-roots leave the myel in two longitudinal series, a dorsal and a ventral. The dorsal nerve-roots, as a simple test of function by stimulation or transsection shows, are sensitive: their mechanical or electrical stimulation produces pain, and their transsection renders the corresponding cutaneous areas anaesthetic. The ventral nerve-roots are motor: their stimulation produces muscular contraction, and their transsection muscular [p. 156] disability. The fibres of the dorsal roots conduct centripetally; if they are transsected, stimulation of the central cut end will give rise to sensation, but not that of the peripheral. The fibres of the ventral roots conduct centrifugally; in their case, stimulation of the peripheral cut end will give rise to muscular contraction, but not that of the central. These facts were first discovered by CHARLES BELL, and their general statement is accordingly known as ' BELL'S law.' They prove that, at the place of origin of the nerves, the sensory and motor conduction-paths are entirely separate from each other. The same thing holds of the cranial nerves, with the addition that here, in most cases, the separation is not confined to a short distance in the neighbourhood of the place of origin, but persists either throughout the whole course of the nerves or at least over a considerable portion of their extent.[7] There can be no doubt that the union of the sensory and motor roots, to form mixed nerve-trunks, finds its explanation in the spatial distribution of the terminations of the nerve- fibres. The muscles and the overlying skin are supplied by common nerve-branches. While, therefore, the two sets of conduction-paths are functionally distinct, a spatial separation throughout their entire course occurs only in certain cranial nerves, where the terminations are comparatively near to the points of origin, but the points of origin themselves lie farther apart. Under these circumstances, a separate course involves simpler space-relations than an initial union of the sensory and motor fibres, such as we find in the trunks supplying adjacent parts of the body.

Not only the origin, but also the further peripheral course of the nerves is very largely determined by the conditions of distribution. Fibres that run to a functionally single muscle-group, or to adjacent parts of the skin, are collected into a single trunk. Hence it does not follow that the mixed nerve, formed by the junction of ventral and dorsal roots, always proceeds simply and by the shortest path to its zone of distribution. On the contrary, it frequently happens that there is an interchange of fibres between nerve and nerve, giving rise to what are called the nerve-plexuses. In explaining the occurrence of these plexuses, we must remember that the disposition of the nerve-fibres, as they issue from the central organ, meets the conditions of their peripheral distribution only in a rough and provisional manner; the arrangement is by no means perfect, and requires to he supplemented later on. The plexuses are most commonly formed, therefore. at places where there are parts of the body that need large nerve-trunks, e.g. the two pairs of limbs. Here it is evidently impossible, from the spatial [p. 157] conditions of their origin, that the nerves should leave the myel in precisely the order that is demanded by their subsequent peripheral distribution. But the plexus-formation is not only supplementary; it is, beyond question, compensative as well. The nerve-fibres that are nearest together as the nerves leave the central organs are those that are functionally related. Now functional relation does not always run parallel with spatial distribution. Thus the flexors of the upper and lower leg, e.g., are functionally related, and act in common; but those of the upper leg lie upon the ventral and those of the lower upon the dorsal side of the body, and consequently receive their nerves from different nerve-trunks, the crural and the sciatic respectively. If, then, the nerves for the flexors of the whole limb are in close proximity at their place of origin, there must he a rearrangement of fibres in the sacrolumbar plexus, in order that the two trunks may pass off in different directions. It is probable that the simpler connexions of the root-pairs are principally useful as supplementary mechanisms, while the more complicated plexus-formations are for the most part compensatory in function.

When BELL first established the law known by his name, he felt constrained by it to postulate a specific difference between sensory and motor nerves, -- a difference which found expression in this fact of the difference in direction of conduction, Physiologists for a long time afterwards gave in their adhesion to this hypothesis. There was, indeed, a prevailing tendency to refer all differences of function, e.g. those obtaining between the various sensory nerves, to conic unknown specific property of the nerve-fibres.[8] Later on, the belief gained ground that the nerves are simply in different conductors of the processes released in them by stimulation; though the only argument at first brought forward in support of it was the not very convincing external analogy of electrical conduction.[9] At the present time, we may say, with better reason, that BELL'S hypothesis of a specific conductive capacity of sensory and motor nerves is not tenable. The decisive evidence is drawn from two sources. On the one hand, the general mechanics of nerve-substance has thrown new light upon the processes of conduction in the peripheral nerve-fibre (pp. 80 ff. above). On the other, the morphological facts indicate that the difference in direction of conduction depends upon the mode of connexion of the nerve-fibres at centre and periphery. Figgs 20 and 27 (p. 50 above) gave a schematic representation of the structural relations involved in the two cases. Every [p. 158] motor fibre, as we said in describing them, is the neurite of a nerve-cell; and there is a certain principle of transmission of force which supposedly holds for all neurites alike. In accordance with this principle, the neurite is able to take up the stimulation-processes originated within the cell, or carried to it by its dendrites; but the excitatory processes of which the neurite is itself the sent, though they are conducted to the cell as a result of the general diffusion that every stimulation-process undergoes in the nerve fibre, are inhibited in the central substance of the cell (p. 99). The cells of origin of the sensory fibres always lie, on the contrary, outside of the central organ: in the invertebrates, for the most part at the periphery of the body; in the vertebrates, at any rate outside of the myel proper. Here, as we know, they form as it were little centres of their own; the spinal ganglia, situated in the intervertebral foramens. These ganglia are composed throughout of bipolar nerve-cells; that is to say, each cell sends out two, morphologically identical processes, which probably have the character of dendrites. In the lower vertebrates, the processes issue at different points -- in the fishes, at opposite sides -- of the cell. In man, the same conditions obtain in the early stages of development. As growth proceeds, however, the two processes fuse together at their point of origin, so that what were at first two distinct and separately originated processes now appear as the branches of one single process (Fig. 21, p. 50), which nevertheless retains the character of a protoplasmic process or dendrite. The two processes thus form a single neurone territory (N1), which divides into two halves. The one lies within the myel, and after giving off numerous collaterals penetrates with its terminal fibres into a second central neurone territory (N2). The other is continued in the sensory nerves, and is finally lost either in terminal arborisation among the epidermal cells, or in special end-organs, adapted for the support of the nerve-ramifications (H Fig. 21). We may therefore suppose, in agreement with what was said above regarding the diffusion of excitations carried by the dendrites (p. 42), that, where the peripheral and central processes issue separately from the cell-body, the process of stimulation is transmitted directly by the cell. When, as in man, the two processes unite to form one, the passage across, and then the transmission to higher neurones (N2), may actually take place within the fibre itself. The cell Z1 seems in this case to be cut out of the line of nervous conduction by a sort of short circuit; though this, of course, does not diminish its importance as a nutritive centre and storehouse of force. If, then, we regard the processes of conduction as conditioned in this way by the properties of the nerve-cells and the mode of termination of their processes within them, the principle of conduction in a single direction will hold only for the connexions of neurones, not for the nerves themselves; the stimulation of any nerve, at any point of its course, must, so long as its continuity [p. 159] is preserved, be followed by the production of a stimulus wave which spreads out centripetally and centrifugally at one and the same time. But, as a matter of fact, there is not the slightest reason why we should hesitate to adopt this hypothesis. It is obvious that, when a sensory nerve is cut across, the excitations carried to the peripheral cut end must disappear at the periphery without effect, i.e. without arousing any sensation in our mind, just precisely as the stimuli which act upon the central cut end of a transsected motor nerve are inhibited in the cell connected with the neurite. In both cases, it is not any property of the nerve-fibres, but the character of the nerve-cells, that is responsible for the result. We can see, more especially when we consider the different modes of origin of the cell-processes, that the nerve-cell is naturally qualified to determine the direction of conduction and to regulate the mode of transmission from one neurone territory to another. For the rest, we shall presently become acquainted with facts that speak definitely for a centrifugal conduction of certain sensory excitations (pp. 182 ff.).

We have now to investigate the farther course of the nerve-paths that lead into the myel, as they are continued in the interior of this central organ. We obtain information concerning them, in the first place, from physiological experiments upon the result of stimulation, and more especially upon the effect of transsection, of certain portions of the myel. We know that the motor roots enter the ventral, and the sensory roots the dorsal half of the myel. These experiments show that the principal lines of conduction retain the same arrangement, as they take their course upwards. The effects of outside interference with the ventral portion of the myel are predominantly motor; with the dorsal portion, predominantly sensory. At the same time, they show also that even in the myel the individual fibre-systems are interwoven in the most complicated fashion. The results of hemisection of the myel, e.g., prove that not all conduction- paths remain upon the same side of the body upon which the nerve-roots enter the myelic substance, but that some of them cross over within the myel from right to left and vice versâ. It is true that the statements of various observers as to the kind and extent of conductive disturbance after hemisection are not in complete agreement; and it is evident, also, that the relations of conduction are not identical throughout the animal kingdom. But experiments on animals and pathological observations on man have put it beyond question that the sensory fibres, at any rate, always undergo a partial decussation. Hemisection of the myel does not lead to a complete abrogation of sensation upon either half of the body. The motor paths [p. 160] appear be more variable in this respect. Experiments on animals again point to a partial decussation, though to one in which the greater part of the fibres remain upon the same side. Pathological observations, on the other hand, lead to the conclusion that in the myel of man the motor paths are uncrossed. We may, in particular, recall the well known fact that in unilateral apoplectic effusions in the brain, it is always only the one side of the body, viz. the side opposite to the apoplectic area, that is paralysed. Now there is, as we shall see presently, a complete decussation of the motor paths in the oblongata. If, then, decussation occurred to any considerable extent in the myel, the one arrangement would of necessity, so far as it went, compensate the other. We accordingly conclude that the principal motor path is situated in the ventral portion of the myel. And we are safe in affirming, similarly, that the principal sensory path lies in the dorsal portion,[sic] In the animals, it is true, we have a greater number of secondary paths, branching off to other parts of the myel, than we have in man, but even in the animals, there can be no doubt that the great majority of the fibres run their course without decussation. Impressions made upon the skin after hemisection of the myel upon the same side are not sensed, though stronger, painful stimuli will still evoke a reaction. Finally, in the lateral columns of the myel (m Fig. 66, p. 164) we have a combination of motor and sensory paths, drawn again for the most part from the fibre-systems of the same side. If the integrity of these columns be impaired, whether in man or in the animals, the resulting symptoms are in general of a mixed character.[10]

These experiments upon severance of continuity at various parts of the myel have brought to light a somewhat complicated interlacement of the fibre-systems. One of the chief factors in the resulting formation is, undoubtedly, the cinerea which surrounds the myelocele. The presence of this grey matter also explains the change of irritability brought about in the myelic fibres by stimulation-experiments. While the peripheral nerves may be readily excited by mechanical or electrical stimuli, this is so far from being the case with the myelic fibres that many of the earlier observers declared them to be wholly irresponsive to stimulation.[11] The statement, in its extreme form, undoubtedly overshoots the mark. Excitation can always be effected by summation of stimuli, or by help of poisons, like strychnine, which enhance the central irritability. At the same time, the marked change of behaviour points clearly to the intercalation of grey matter (cf. p. 86 above). If it be asked in what way this intrusion can materially affect the processes of conduction, we reply that a path coming in from [p. 161] the periphery will be brought into connexion, by the cinerea, not with one but with many paths of central conduction. These paths will, it is true, not be all equally permeable. Some will offer more resistance than others to the passage of excitation: in certain cases inhibitory effects, of the kind with which we have become familiar as the results of certain modes of connexion of the central elements (p. 99), may destroy or modify excitations already in progress. But there will always be various secondary paths available, over and above the principal path. The experiments upon severance of continuity call our attention chiefly to the principal path; but there are several ways -- increased intensity of stimulus, enhancement of irritability, destruction of the principal path -- in which the secondary paths may be thrown into function. If the white columns are entirely cut through, at any point upon the myel, so that only a narrow bridge of cinerea remains intact, sense-impressions and motor impulses may still be transmitted, provided that they are unusually intensive. And we find, similarly, that the phenomena of disability, which appear on transsection of a portion of the white columns, disappear again, after a short interval, although the cut has not healed.[12] The existence of these secondary paths or by-paths is attested, further and more particularly, by the phenomena of transference from one conduction-path to another, phenomena which prove the presence of a connecting path between different conduction-paths. They are of three kinds: the phenomena of concomitant movement, of concomitant sensation, and of reflex movement. As all alike are of importance fur a right understanding of the functions of the central organs, especially of the myel, we shall be occupied with them in the following Chapter. We are interested in them here only in so far as they bear witness to the existence of determinate conduction-paths, preformed in the myel, but functioning only under certain special conditions. The place of transference from motor to motor, sensory to sensory, or sensory to motor paths must be sought, again, in the cinereal structures. Complete severance of the cinerea, with retention of a portion of the dorsal and ventral columns of white matter, abrogates the phenomena in question. Transferences within the motor paths, manifesting themselves in concomitant movements, may be made, without any doubt, either on the same side of the myel or from the one side to the other. Thus, the innervation of a finger-phalanx is transmitted both to other fingers of the same hand and, under certain circumstances, to the skin of other parts of the body. Bilateral concomitant movements of this kind are especially observable in movements of locomotion and in pantomimic movements. It is clear, on the other hand, that excitatory innervation within a definite path may be connected with inhibitory co-excitation of the cells of origin of another motor path. We have an instance of this state of things in the relaxation [p. 162] of tonus in the extensors that goes with excitation of the flexors of a limb.[13] This illustration, like that of co-ordinated movements, shows further that co-excitations within the motor paths may establish themselves as regular functional connexions. Transferences within the sensory paths seem, contrariwise, to be confined almost exclusively to the same half of the myel. The concomitant sensations observed after stimulation of some part of the skin are nearly always referred to cutaneous regions on the same side of the body. They are brought out most clearly by painful stimuli and by the arousal of tickling: in the latter case, more particularly when the skin is rendered unusually sensitive by enhancement of irritability. Under these conditions, stimulation of certain regions of the skin usually evokes sensations in other regions. Certain sensory parts, e.g. the external auditory meatus and the larynx, are also pre-eminently disposed for concomitant sensation.[14] We can hardly explain these facts otherwise than by the hypothesis that, on the one hand, certain preferred paths of connexion exist within the sensory conduction-paths, and that, on the other, certain sensory areas (larynx, external auditory meatus) are peculiarly susceptible to co-excitation. As regards the conditions under which conduction takes place, it is clear that concomitant movements and concomitant sensations both alike depend upon cross-conductions, that may be effected at different heights in the myel, and that differ only in direction: the motor cross-conductions extending in all directions, while the sensory, so far as we can tell, are almost exclusively unilateral, and for the most part follow the direction from below upwards. For the rest, the concomitant sensations and concomitant movements that have their ground in connexions of the myelic conduction-paths can never be certainly discriminated from those mediated by transference within the higher centres.

This statement does not apply to transferences of the third kind, reflex connexions of sensory and motor paths. The myelic reflexes may be observed for themselves alone, after the myel has been separated from the higher central parts. The conclusion to be drawn from such observation is that branch-conduction of the reflexes is effected by a large number of conduction-paths, all of which are closely interconnected. Moderate stimulation of a circumscribed area of the skin is followed, at a certain mean degree of excitability, by a reflex contraction in the muscle-group, and in that only, which is supplied by motor roots arising at the same height and on the same [p. 163] side as the stimulated sensory fibres. If stimulus or irritability be increased, the excitation passes over, first of all, to the motor root fibres that leave the myel at the same height upon the opposite side of the body. Finally, if the increase be carried still farther, it spreads with growing intensity first upward and then downward; so that in the last resort it involves the muscles of all parts of the body which draw their nerve-supply from myel and oblongata. It follows, then, that every sensory fibre is connected by a branch-conduction of the first order with the motor fibres arising on the same side and at the same height; by one of the second order, with the fibres issuing at the same height upon the opposite side; by branch-conductions of the third order, with the fibres that leave the myel higher up; and, lastly, by branch-conductions of the fourth order, with those that emerge lower down.[15] This law of the diffusion of reflexes may, however, as we shall see in the following Chapter, be modified in two ways: by variation of the place of application of the reflex stimulus, and by the simultaneous application of other sensory stimuli (cf. Chap. VI. § 2).

The conclusions which we have reached by way of physiological experimentation regarding the course of the conduction-paths in the myel are in complete agreement with the morphological facts revealed by histological examination of this organ. In particular, the arrangement of the nerve-cells and of the fibre-systems which take their origin from the cell-processes as shown in transverse and longitudinal sections, enables us to understand at once that every principal path is here accompanied by a large number of secondary paths, and that the most manifold connexions obtain between one line of conduction and another. We see, first of all, that the fibres of the ventral roots enter directly into the large nerve-cells of the ventral cornua, whose neurites they form; whereas the fibres of the dorsal roots, after their interruption by the nerve-cells of the spinal ganglia, divide upon entering the myel into ascending and descending systems, which there give off delicate branches at all points into the cinerea of the dorsal cornua. Here, therefore, as in the experiments with transsection, the white columns (l, m, n Fig. 66) appear in the role of principal paths: their ventral portions as motor, their dorsal as sensory. Secondary paths, for the conduction of unusually intensive excitations, or for the transferences required by concomitant movements, concomitant sensations and reflexes, can be mediated in a great variety of ways by the cellular and fibrillar system of the central cinerea (d, e). The interrelations of these different paths of conduction, and in particular of the two groups that in functional regard [p. 164] stand farthest apart, the motor and sensory, are then determined by their mode of connexion with their cells of origin, and with the processes which these cells give off. We thus find, in the properties of the neurone and its area of distribution as manifested within the myel, a continuation of the differences that we meet with in the primitive forms represented in Fig. 20, 21 (p. 50). Fig. 67 shows the various morphological elements in their natural connexion. Each of the large multipolar cells m of the ventral cornu has direct control of some peripheral region by means of its neurite n, which does not break up into its terminal arborisation until it reaches the terminal plate of a muscle-fibre (Fig. 20, p. 50). On the other side. the dendrites issuing from the same cell run a very short course, to enter at once into the cinerea of the ventral cornu. The dendritic reticulum stands in direct contact with the terminal fibrils of the neurite g of another nerve-cell, situated as a rule high up in the brain; so that the neurones of this motor conduction cover very extensive territories. Indeed, it is probable that in most instances the entire motor conduction involves only two neurones (N1, and N11, Fig. 20), the one of which extends from the cell n of the ventral cornu to the periphery of the body, while the other begins with some one of the fibres that run their course in the ventral or lateral column (l, m Fig. 66), and ends in a cell of the cerebral cortex. At the same time, still others of the dendrites belonging to the cells of the ventral cornua are in contact with the processes of the small cells s of the dorsal cornua, and with the small intercalatory or commissural cells c that lie scattered between ventral and dorsal cornu. In these latter connections we have, presumably, the substrate of reflex conduction. The sensory nerve-paths, on the other hand, follow a very different course. In their [p. 165] case, the spinal ganglion-cell sp forms the central point of a neurone territory, the one half of which extends by means of the peripherally directed processes h to the sensory termini of the organ of touch (Fig. 21, p. 50), while the other runs centralward in the central process f, which divides in the dorsal portion of the myel into ascending and descending branches (a, d). Both of these branches give off numerous collaterals, whose terminal ramifications stand in contact with the small cells of commissure and dorsal cornu. They themselves are finally resolved into fibrillar reticula, connected by contact with the dendrites of cells lying farther up and lower down. These structural relations seem to warrant the inference that the collaterals correspond to the various secondary paths by which transference, and especially reflex transference, is effected, and that the ascending and descending fibres constitute the principal path. The principal path of sensory conduction is, however, markedly different front the meter. As a general rule, there are several breaks in the line; the path consists of a number of neurone chains, arranged one above another. And this means, again, that the conditions of conduction in the principal path are less sharply distinguished from those in the secondary paths that begin in the collaterals. The whole morphological plan of the system of sensory conduction thus suggests a co-ordination of parts that is at once less strict and more widely variable than is the case on the motor side.

We have spoken so far only of the general properties of the myelic conductors, properties accruing to all nerve-fibres whose mode of origin [p. 166] and connexions conform to a certain type. In the higher regions of the myel, other conditions are at work, paving the way for that differentiation of the conduction-paths which characterises the higher central regions. Even as low down as the thoracic portion of the myel, certain funicles divide off from the three principal columns already named, the ventral, lateral, and dorsal columns (l, m, n Fig. 66). The principal paths, sensory and motor, that run their course within the length of the myel, are thus split up into several separate tracts. The significance of these new funicles can best be understood from their embryological connexions and from the course of the degenerations observed in pathological cases (pp. 154 f.). It can be shown, by both lines of evidence, that the motor division of the lateral columns ascends uncrossed in their dorsal half, in a funicle which, as seen in cross-sections, encroaches from the outside upon the cinerea of the dorsal cornu. Higher up, it passes over into the pyramids of the oblongata, and is accordingly known as the path of the pyramidal lateral column (Fig. 68). In the same way, the innermost division of the motor ventral columns, the part bordering directly upon the ventral sulcus, ascends uncrossed to the oblongata, where it too passes over into the pryamids [sic]. It is termed the path of the pyramidal ventral column, and is the only division of the pyramidal tracts to remain uncrossed in the oblongata. Of the more peripherally situated funicles of the ventral column, some take a straight course upwards, while others enter the ventral commissure and cross to the opposite side of the body. The division of the lateral column which overlies the pyramidal lateral column, at the periphery of the myel, is an uncrossed and, to judge from the conditions of its origin, a sensory path: it branches off to the cerebellum by way of the postpeduncles, and is termed the path of the cerebellar lateral column. The dorsal columns, which are exclusively sensory in function, and therefore receive from below the great majority of the fibres that enter the dorsal roots, divide in the cervical region into two funicles: the slender funicles or columns of Goll (fun. graciles), and the more outlying cuneate funicles (fun. cuneati, Fig. 68).[16][p. 167]

Oblongata and cerebellum, the parts of the brain stem that correspond developmentally to after brain and hind brain (p. 108), together with the pons that unites them, form in the brain of the higher mammals and of man a connected system of conduction paths. The system, as may be gathered from the general trend of the fibre-tracts that pass across it or decussate within it, is of importance in three principal directions. In the first place, this region furnishes the passage-way for the continuation of the sensory and motor conduction paths that come up from the myel. Secondly, it originates new nerves: the great majority of the crania nerves spring from separate grey nidi in the oblongata: and, in doing this, repeats, though in much more complicated fashion, the structural patterns which we have traced, in their comparatively simple form, in the lower central organ. Thirdly, it contains a great variety of connecting paths, them selves for the most part interrupted by deposits of nerve cells, between the various paths that lead across or arise within it; while, further, in the fibre tracts that run from the main conducting trunk to the cerebellum and back again, it possesses a secondary conduction path of very considerable extent that is interpolated in the course of the principal conduction path. It will be understood that, under these conditions, the lines of travel in the region we are now to consider, as well as in the adjoining regions of mid brain and 'tween brain, are extraordinarily complicated. A complete explication of them, in the present state of our knowledge, is altogether out of the question. But more than this: it is impossible, as things are, to put a physiological or psychological interpretation upon many of the structural features that have already been made out. The functional significance of some of the most prominent conduction paths, as e.g. the entire intercalatory system that runs to the cerebellum, is still wrapped in obscurity. Hence, in most cases, the tracing out of the fibre systems is a matter solely of anatomical interest. In physiological regard it is useful, at the best, merely as illustrating the extreme complexity of the conditions which here determine conduction. We shall therefore refer, in what follows, only to certain selected instances, adapted to give a general picture of the course of the paths of conduction in the gross; and we shall enter into some detail only in those cases which appear to be of importance for the physiological and psychophysical relations of the central processes. On the score of method, we must say also that the physiological expedient of isolating the paths by transsection of individual fibre tracts, which did good service in giving us the general bearings of the paths of conduction in the myel, can hardly come into consideration here,[p. 168] more especially in our study of the conditions of conduction in the hind and mid brain regions. Experiments of the kind are recorded not infrequently in the older physiology. The course of the paths is, however, too complicated and their origin too uncertain, to admit of any but an ambiguous result. The most that the method can give us is a point of view from which to appreciate the gross function of the organs or of certain of their parts; and we shall accordingly say nothing of the observations made by it until we reach the next Chapter. We may add that the method which has proved most fruitful for the problem of direction of conduction, apart from direct morphological analysis of the continuity of the individual fibre tracts, is the tracing of the course of degeneration in fibres separated from their centres of origin.

The simplest problem presented by our preset enquiry is that of the further course of the paths of motor and sensory conduction that come up from the myel. The two methods just mentioned furnish us with a fairly satisfactory solution, at any rate as regards the motor paths. The principal continuation of the main path of motor conduction that runs upward in the lateral and ventral columns of the myel is, as we already know: the pyramidal path (Fig. 68, p. 166; cf. Fig. 46, p. 118). The course of this path in detail has been made out, with some degree of completeness, by help of the descending degeneration which appears in it after destruction of its terminations in the brain. It is the continuation of that division of the motor principal path which lies in the myel in the dorsal portion of the lateral columns and along the inner margin of the ventral columns (Fig. 68 B). The branch of this path that belongs to the ventral columns decussates in the cervical region of the myel. Now the larger branch, from the lateral columns, also undergoes a complete decussation, clearly visible on the external surface of the oblongata (p, Fig. 47, p. 119). The central continuation of the path then runs to the cerebral cortex, without interruption by cinerea, Fig. 69 gives a schematic representation of the course of these paths, the longest and so far the best known of all lines of central conduction. After they have traversed the pons, the fibres of the pyramidal path enter the crusta (f, Fig. 56, p. 130) between lenticula and thalamus, and then trend upwards in the space between lenticula and caudatum to pass into the corona, where their principal branches constitute the fibre-masses that terminate in the region of the central gyrus and the surrounding area (VC, HC, Fig, 65, p. 145).[17] The path is thus fairly well defined. Part of it, as is proved by the paralyses following lesion of the pyramids [p. 169] and their continuation in the crus, undoubtedly subserves the conduction of voluntary impulses. In the animal kingdom, the pyramidal path affords a better measure than any other of the fibre systems collected in the brain stem of the general development of the higher central organs. In the lower vertebrates, the pyramids are altogether wanting. In the birds, they are but little developed. They steadily increase in importance in the mammalian series, up to man; while at the same time the tract from the lateral columns, which passes to the opposite side of the body in the pyramidal decussation, grows constantly larger as compared with the tract from the ventral columns, which decussates in the myel. A branch of the motor path which is forced inward by the pyramids, and which remains intact after removal of the pyramidal fibres, may be traced in part to the mesencephalon. It consists mainly of divisions of the ventral columns (mf, Fig. 70). Finally, certain of these remains of the ventral columns are collected in the interior of the rounded prominences to form the dorsolongitudinal bundle (hl, Fig. 72), which in its further course through the pons makes connexions with the pontal nidi and more especially, as it appears, with centres of origin of the oculomotor nerves and with the cerebellum.[18] We may accordingly suppose that these branches of motor conduction which run to the mesencephalon serve to mediate co-excitations in that region. The connexions of the dorsolongitudinal bundle, in particular, seem to point to connexions of the motor innervation of the eye and of the skeletal muscles, such as are involved in locomotion and in the orientation of the body in space.[p. 170]

The course of the sensory path through the oblongata has not been made out as fully as that of the motor. The main reason for this defect in our knowledge lies in the difference of structure to which we referred above. It is characteristic of sensory conduction in the myel that the path does not pass upward in unbroken continuity, but consists of a chain of neurones. This structural complexity is not only continued but increased in the oblongata, where large numbers of cells, grouped together to form separate nidi, are interposed in the line of conduction. We may suppose that these nidi serve for the most part as transmitting stations -- points at which a path, whose course has so far been single, splits up into several branches that diverge in different directions. The main divisions of the sensory path pass in this way, within the oblongata, first of all into the grey masses deposited in the slender and cuneate columns (Fig. 68, A, and Fig. 46, p. 118). Further on, the sensory path continues in a bundle lying close under the pyramids (l, Fig. 70), which appears on the ventral surface of the oblongata directly above the pyramidal decussation (p, Fig. 47, p. 119) here in its turn suffers decussation, and then passes on in the lemniscus of the crus, a structure lying in the enter and upper portion of the tegmentum. The lemniscal decussation (formerly known as the superior pyramidal decussation) thus forms yet another continuation of the decussations of myelic fibres which begin within the myel itself. Other sensory fibres (ci, Fig. 70), down from the dorsal columns, pass into the tegmentum proper, which thus brings together portions of the motor (mf) and of the sensory path. All these sensory fibres terminate in the grey masses of the region of the quadrigemina and thalami, from which, finally, further continuations of the sensory path proceed to the cerebral cortex.

The general sketch of the course of the sensory and motor paths, given in the preceding paragraphs, makes matters much simpler than they really are. There are two facts, not yet mentioned, that are chiefly responsible for the complications actually found. The one of these consists in the origination of a large number of new sensory and motor paths, which are derived front the cranial nerves, and in their further course either join the paths formed by the myelic nerves or strike out special lines of their own, the other, in the appearance of large groups of central nerve cells, which serve either as transmitting stations for the conductions comings up to the cerebrum from below, or as junctions for the important branch-conduction to the cerebellum, here opened for travel. The difficult questions concerning the origin of the cranial nerves, questions that have not yet in every case received their final answer, are of interest for psychology [p. 171] only in so far as they involve that of the paths followed by the sensory nerves. Since these belong in large part to the mesencephalic region, we may postpone their consideration until later. It will suffice for our present purpose to refer to Fig. 72 (p. 176), as an illustration of the conditions of origin of the

We turn now to the grey nidi of this region of the brain. We have already mentioned the nidi of the dorsal columns, interposed directly in the sensory path. Very much more complicated are the functions of the largest nidi of the oblongata, the olives (Fig. 46 B, Fig. 47, pp. 118 f.) whose principal office seems to be the giving off of branch-conductions. On the one hand, the neurites of the cells give rise to a fibre system, the further course of which is uncertain: it is supposed to connect partly with the cerebellum, partly with the lateral columns of the myel. On the other hand, the dentata give rise to two fibre systems. The first of these covers the outer surface of the olivary nidus, in the form of zonal fibres (g Fig. 48, p. 120), and then bends round into the restes and their continuations, the cerebellar peduncles (cr Fig. 70). The second issues from the interior of the nidus and crosses the median line, to decussate with the corresponding fibre-masses of the opposite side. Other fibres from the olives enter the longitudinal fibre tract that lies between them, and then run within the pons to the lemniscus of the crus (l Fig. 70); they thus appear to join the sensory principal path to the cerebrum. Putting the facts together, we may say that the olives are structures which stand in intimate relation with the branching off of conduction paths towards the cerebellum Another ganglionic nidus, lying higher up -- in man concealed by the pons, in the lower mammals projecting on its posterior border -- the trapezium or superior olive, forms, as we shall see presently, a nodal point of great importance in the conduction of the acoustic nerve.

The conduction paths that branch off from the oblongata to the cerebellum, and there turn back again to join the caudex in its course through the pons, bear a striking external resemblance to a shunt interposed in the main current of an electrical conduction. And it seems, as a matter of fact, that this obvious comparison fairly represents the actual relations of the nerve paths, as they are shown schematically in Fig. 71. The sensory and motor principal paths, just described, have also been included in this diagram, in order that the reader may obtain a rough idea of their relation to the branch path leading to the cerebellum. The mammalian cerebellum contains, as we have already said, two formations of cinerea: the one appearing in the ganglionic nidi, the other in the cortical layer investing the entire surface of the organ (pp. 121 f.). Our present knowledge of the relations between the fibres that enter into and issue from the cerebellum and these grey masses may be summarised as follows (cf. Fig. 48, p. 120). The fibres of the restes are deflected round the dentatum, more especially over its anterior margin. They do not appear to connect with the cinerea of the nidus, but radiate from its upper surface towards the cortex, where [p. 173] they terminate and are lost. From the cortex itself comes a system of transverse fibres; which cut across the more longitudinal radiations of the restes, and draw together in stout fascicles to form the medipeduncles (brachia of the pons). The interior of the dentata gives rise, further, to the funicles which pass into the prepeduncles (crura ad cerebrum). And, finally, there is a connexion between the dentata and the cerebellar cortex. This path, together with the radiation of the restis and the medipeduncle, occupies the outer division of the alba, while the innermost portion is constituted by the prepeduncle. It is therefore probable that all the fibres running through the postpeduncles of the cerebellum from the oblongata have their termination in the cortex. The cortex itself gives rise to two fibre systems: the one passes directly over into the medipeduncles, the other appears first of all to connect the cortex with the dentatum, which then gives off the vertically ascending fibres of the prepeduncles. These run upwards, with the continuations of the myelic columns, converging as they proceed; just anteriorly to the upper end of the pons they reach the middle line, and undergo decussation. Besides the two divisions of this system of ascending fibres, we find, lastly, further radiations, whose fibres subserve the interconnexion of more or less remote cortical areas. Some of the longer lines cross from the one, side to the other in the vermis.

The further course of the paths leading from the cerebellum to the cerebrum is as follows. The path which is continued in the medipeduncles appears, first of all, to terminate in grey masses in the anterior region of the pons. From these masses arise new, vertically ascending fibre, some of which can be traced to the anterior brain ganglia, the lenticula and striatum, while others proceed directly to the anterior regions of the cerebral cortex. The fibres collected in the prepeduncles find their proximate termination in the rubrum of the lemniscus (hb Fig. 56, p. 130). A small number of the fibres issuing from this point probably enter the thalami; but the greater portion pass to the internal capsule of the lenticula, and thence in the corona to the cerebral cortex, ending in the regions posterior to the central gyre, and more especially in the precuneus. The valvula (vm Fig. 48, p. 120), which joins the prepeduncles at the beginning of their course, serves in all probability to supplement the connexions of the cerebellum with the brain ganglia, by mediating a conduction to the quadrigemina.

We must believe, in view of these results of anatomical investigation, that the concurrence of conduction paths in the cerebellum is extremely complicated. Let us consider these paths as a branch conduction, interposed in the course of the direct conduction from myel to cerebrum as mediated by oblongata and pons. We have two divisions, a lower and an upper. The lower division of the branch conduction carries sensory fibres from [p. 174] the dorsal and ventral columns (olivary path of the dorsal columns, and cerebellar path of the lateral columns), which connect the myel with the cerebellum; and motor funicles, which branch within the pons to enter the restes. The upper division makes two principal connexions, by way of the medipeduncles: the one with the cerebral cortex direct, the other with the anterior brain ganglia (lenticula and striatum). At the same time, there is a connexion, mediated by prepeduncles and valvula, with the posterior brain ganglia (thalami and quadrigemina). The most extensive of these conductions, that to the cerebral cortex effected by the medipeduncles, radiates out to all parts of this organ, but is principally directed forwards to the frontal brain and the adjacent regions.

The schema given in Fig. 71 shows the main features of this conduction system. The reader will recognise, first of all, the pyramidal path, with its crossed branch from the lateral and its uncrossed branch from the ventral columns, running directly between myel and cerebral cortex (p1p2, p). He will next notice the other motor paths, derived from the ventral columns, and interrupted in the mesencephalic region by masses of cinerea. Some of these paths are continued in a new neurone chain, and extend to the cerebral cortex; other fibres of the same system probably terminate in the mesencephalic region itself (vv'). A considerable division of the sensory path (gg') drawn from the dorsal columns, passes in the lemniscal decussation (k2) to the opposite side: part of it is lost in the [p. 175] grey masses of the pons, part continues in fibre tracts which, interrupted by grey nidi, run to the anterior brain regions and so finally to the cortex. There is also an uncrossed sensory path (cc), derived from the dorsal and lateral columns, which passes into the tegmentum of the crus and finds its proximate terminus in the tegmental grey nidi. Another path, also sensory in origin, is the uncrossed branch conduction (cs) carried to the cerebellum from the restes in the postpeduncles; it terminates in the cerebellar cortex, for the most part in the vermis. Finally, there is a crossed conduction (f), issuing from the grey nidi of the olives, which, unlike the former, enters into the nidal structures (N) of the cerebellum. These are all incoming paths. The outgoing lines, leading to the cerebrum, are two in number: the prepeduncles, which start from the cerebellar nidus, and may be traced partly into the prosencephalic ganglia, partly to the cerebral cortex (e'); and the fibres of the medipeduncles (bb'), which run direct from the cerebellar cortex to the cerebrum. These latter enter, first, into the grey nidi of the pons, and are by them brought into connexion, in some measure, with the brain ganglia, but most extensively with the cerebral cortex, and in that principally with the frontal region. The system is completed by the paths of connexion between nidal structures and cortex (rr) which belong exclusively to the cerebellum.

The general relations of these incoming and outgoing paths suggest that the cerebellum brings into connexion with one another conductions of different functional significance, This inference finds further support in the peculiar structure of the cerebellar cortex. The characteristic constituents of this region are, as we saw above (Fig. 15, p. 44) the cells of PURKINJE, easily distinguished by their large size and the manifold arborisation and reticulation of their protoplasmic processes. If, now, the cerebellar cortex serves to connect fibres of different function, sensory and motor, as is suggested by the relations of the incoming and outgoing paths, it is clear that we may look upon these cells of PURKINJE as elementary centres of connexion between functionally different fibre elements. We should then have to assume, on the analogy of the large cells in the ventral cornua of the myel, that the dendrites mediate centripetal, the neurites centrifugal conductions: in other words, that the chief office of the former is to take up the excitations carried in the postpeduncles, while the latter collect to form the paths of conduction that continue in the medipeduncles to the cerebrum and there, as it appears, are chiefly connected with the centres of innervation of the prosencephalon.

The pons is chiefly important as receiving the paths to be carried up from cerebellum to cerebrum, and associating them to the vertical ascending fibres of the crus. Its development in the animal kingdom thus keeps even pace with the development of all these paths of conduction, and [p. 176] especially of the pyramids and medipeduncles. The fibres that cross over from the one side to the other in the median line of the pons (at R Fig. 72) are decussating fibres belonging in part to the direct continuations of the myelic columns through the pons, in part to the medipeduncles of the cerebellum. The decussation of these latter has been established by pathological observations: atrophy of a cerebral lobe is ordinarily attended or followed by a wasting away of the opposite half of the cerebellum. The fibres of the medipeduncles, probably without exception, pass through internodes of grey matter before they are deflected into the vertical paths; and small grey nidi are also strewn in the path of the directly ascending prepeduncles (ba Fig. 72). These presently decussate, and come to art end in the rubrum of the tegmentum. In this way, by collection of the myelic columns that come up from below, and of the continuations from the cerebellum that join them front above and from the side, there forms within the pons that entire fibre tract which connects the lower-lying nerve centres with the structures of the cerebrum, -- the crus. At the same time, the pons is broken root bundles of certain cranial nerves, which take their origin higher up. The nidi of origin of these nerves are situated partly upon the cinereal floor of the highest portion of the fossa rhomboidalis (metacele), partly in the neighbourhood of the Sylvian aqueduct (mesocele), which forms a continuation of the central canal.[p. 177]

As a result of its cleavage by cinerea and by the cross fibres of the medipeduncles, the crus divides into two parts, distinguishable in the gross anatomy of the brain, and known as crusta and tegmentum. A third division, the lemniscus, belongs to the tegmentum so far as regards the direction of its course, but in all other respects is clearly differentiated from it. Neither of the two principal parts constitutes a complete functional unit; on the contrary, each of them includes conduction paths of very diverse character. Nevertheless, the twofold division of the crus seems to represent a first, even if a rough classification of the numerous paths of conduction to the cerebrum. Thus the inferior portion or crusta (p--p' Fig. 72) is principally made up of the continuations of pyramidi, remains of the dorsal columns, and medipeduncles. Its outermost portion carries that continuation of the dorsal columns which passes in the lemniscal decussation to the opposite side of the body (k2 Fig. 71). The intercalatum (substantia nigra of SÖMMERING: Sn Fig. 73) is a ganglionic nidus, belonging to the conduction paths of the crusta, which separates crusta from tegmentum. The portion of the crus which lies above the intercalatum, the tegmentum (v'--hl Fig. 72), is at first composed of the remains of the lateral and dorsal columns, and of a part of the remains of the ventral columns. In its further course, beyond the point at which the rubrum appears in cross sections of the tegmentum (R Fig. 73), these are reinforced by the prepeduncles (mf, hi, cr Fig. 70). Finally, the lemniscus, which we have recognised as a separate subdivision of the tegmentum (sl--sl' Fig. 72), also carries fibres from the dorsal columns, as well as fibres from the ventral columns and the cerebellum. Taking the origin of all these tracts into consideration, we may designate the crusta as that part of the crus which, so far as it derives directly from the myel, is especially devoted to the conveyance of motor paths; the tegmentum and lemniscus are of mixed, and mainly, as it seems, of sensory origin. At every point, however, these direct continuations of the myelic systems are augmented by intercentral paths, the conductions from the cerebellum. In this way, as may be seen from Fig. 72, which shows a cross section taken approximately through the middle of the organ, the structure of the pons becomes extraordinarily complex. We may add that it contains, crowded together in a comparatively small space, the whole number of conduction paths, many of which in their later course are widely divergent. It is, therefore, a remarkable coincidence that, besides the epiphysis, which is not a nervous centre at all (see p, 124, above), the pons should have been regarded with especial favour by the metaphysical psychology of past times as the probable 'seat of the mind.' HERBART himself accepts this view. If, on the contrary, one were asked to lay one's finger upon a part of the brain that by its complexity of structure and the number of elements it compresses into a small space [p. 178] should illustrate the composite character of the physical substrate of the mental life, and therewith show the absurdity of any attempt to discover a simple seat of mind, one could hardly hope to make a happier choice.

If we look at the series of cerebral ganglia, we see at once that those of mesencephalon and diencephalon, the quadrigemina and the thalami, serve as intermediate stations on the line of conduction: peripherally, they receive sensory and motor fibres; centralwards, they stand in connexion with the cerebral cortex. They lie, as their function requires, directly upon the crura, whose fibre masses partly run beneath them straight to the prosencephalon, partly curve upwards to enter into the grey nidi of the ganglia. There is a difference, however: the thalamus takes up comparatively few fibres from below, and sends out very considerable bundles to the cerebral cortex; the quadrigemina do just the reverse. Both ganglia, as we shall see in detail later, are of especial importance as nodal points in the optic conduction. Fig. 73 shows a section taken through the middle region of this whole area, and will assist in some degree towards an understanding of the structural relations.

The position of the prosencephalic ganglia, the striata with their two subdivisions, caudatum and lenticula, is more obscure. The incoming and outgoing fibres tell us but little of their function. Both divisions receive fibres front the periphery, derived for the most part from the diencephalic and mesencephalic ganglia. The crural fibres, on the other hand, pass below and between the prosencephalic ganglia, without entering them (Fig. 74). The grey masses of the ganglia send no further reinforcements [p. 179] to the coronal radiation. It would appear, then, that these structures are terminal stations of conduction, analogous to the cerebral cortex, and not intermediate stations like the thalami and quadrigemina.[20]

An important place is filled in the system of conductions that falls within the region we are now considering (prosencephalon, diencephalon, mesencephalon) by the paths of the sensory nerves. Fortunately, these are among the conductions that have so far been most fully investigated, and whose functional significance is at the same time relatively easiest of interpretation. In view

There is, however, a further point, in which the path of the gustatory nerves differs from those of the other nerves of special sense. The gustatory fibres, in consequence, we may suppose, of their distribution over a functional area of some considerable extent, run their course in two distinct nerve trunks: those destined for the anterior portion of this area in the lingualis (L Fig. 75), and those intended for the posterior portion in the glosso-[p. 181]pharyngeus (G). This division appears, however, to be simply external. Both of the gustatory nerve paths take their origin from the same masses of nidal cinerea on the floor of the metacele. At first, however, the gustatory fibres that run to the anterior portion of the tongue join the facialis, at the genu of which (F) they pass through the cells of a small special ganglion. Thenceforward they are continued in the chorda tympani (Ch) side by side with the lingual branch of the trigeminus. The glossopharyngeus, on the other hand, which supplies the posterior portion of the tongue, passes through its own ganglion. At the periphery, as we shall see when we come to consider the peripheral sense apparatus the two nerves break up into terminal fibrils, which end in and among the taste breakers, without, as it appears, coming into contact with other than epithelial terminal structures. The course of the fibres, as shown synoptically in Fig. 75, accordingly corresponds in all details with a general sensory conduction, such as is represented in Figg. 21 and 67 (pp. 50, 165) for the myelic nerves: the ganglia VII. and IX. may be regarded as analogues of the spinal ganglia.

The paths of the olfactory nerves follow a radically different course. Their point of origin lies furthest forward of all the sensory nerves, so that they border directly upon certain cortical regions of the cerebrum. This is the reason that the olfactorius, from the outset, is not a single nerve, but appears in the form of numerous delicate threads, which issue direct from a part of the brain that belongs to the cortex, the olfactory bulb (Fig. 52, p. 125). Conduction begins at the periphery in the cells of [p. 182] the olfactory mucous membrane (A Fig. 76), which are set between epithelial cells, and have themselves the character of nerve cells that send their neurites centralward. These neurites, in their course to the olfactory bulb, break up into delicate fibrils, which for the most part come into contact with the dendrites of small nerve cells: the two sets of processes together forming a compact ball of tissue (a, b). Each of these cells, in its turn, sends out a principal process, which passes into one of the large nerve cells of the bulb (C). Here we must place the proximate cortical station of the olfactory path. The dendrites issuing laterally from the cell bodies represent, in all probability, ramose secondary conductions; while the main path of centripetal conduction is continued, in the direction of the arrows, in the neurites, which leave the cell upon the opposite side, and pass into the olfactory tract. This, the principal path, accordingly extends over a peripheral and a central neurone territory. There is, now, a second group of central olfactory cells which, if we may judge from their connexions and the direction of their processes, are probably to be regarded as nodal points of a system of centrifugal conduction. These cells (D) send out a single peripherally directed neurite, which breaks up within the glomeruli in a delicate reticulum of terminal fibrils (c). The olfactory path thus shows a marked divergence from the type of sensory conduction represented by the cutaneous nerves. The peripheral organ itself appears as a peripherally situated portion of the cerebral cortex, and the olfactory fibres, by a natural consequence, resemble central rather than peripheral nerve fibres. Another novel feature is introduced in the probable existence of a secondary path of centrifugal conduction,. And, lastly, we must note the central connexion of the olfactory region; of the two sides by the precommissure (ca Fig. 53, p. 127). The connexion is presumably to be interpreted as an olfactory decussation, by which the centripetal paths are carried to the opposite hemisphere, and the neurones D c are also enabled to mediate co-excitation, in the centrifugal direction, of the peripheral cells A of the opposite side of the body.

In man and the higher vertebrates, the cochlea of the auditory organ is, in all probability, the only part of the labyrinth of the ear that subserves auditory sensation. If we may judge from the character of the cochlear nerve terminations, the peripheral starting-point of the acoustic conduction conforms in all respects to the conduction type represented by the cutaneous nerves. The terminal fibrils of the acusticus extend among the epithelial and connective tissue structures of the basilar membrane (cf. Ch. VIII., § 4, below), and then, in the auditory canal of the cochlea (S Fig. 77) traverse groups of bipolar ganglion cells (g) which resemble the cells of the spinal ganglia [p. 183] and are termed in common the spiral ganglion. The cells of this ganglion, which accordingly corresponds to an outlying spinal ganglion, give off neurites which run centralward, and finally break up into

This list shows us how extraordinarily complex is the network of relations into which the auditory organ is brought by its central paths. Apart from its twofold crossed and uncrossed connexion with the cerebral cortex, the following facts should be noted as of especial significance. First, there is a reflex path connecting the acoustic centres with the points of origin of muscular nerves, and among them with the centres for the movements of articulation and for the movements of the eyes, which latter are extremely important in the spatial orientation of the body. Secondly, we find that the conduction system, like that of the olfactory nerve, includes centrifugal paths, whose office is, perhaps, to transmit the excitations of the auditory organ of the opposite side, or other sensory excitations that find their nodal points in the mesencephalic region, in the form of concomitant sensation.

We remark, in conclusion, that the acoustic nerve proper, which comes from the cochlea, is connected over a part of its peripheral course with the nerve that comes from the vestibule and canals. This, the vestibular nerve, is a branch of the eighth cranial, and is commonly accounted, like the cochlear, to the acoustic nerve. In its central course, however, it appears to follow a different road. It passes through special nidal structures, and finally, as its secondary degenerations prove, terminates in separate areas of the cerebral cortex. [22]

The principal difference between the optic and acoustic conductions is that the optic surface itself, like the olfactory surface, is an outlying portion of the central organ, displaced to the periphery of the body. It is natural, therefore, that the optic fibres too, when they emerge from the retina, should at once appear, as by far the great majority of them do, in the character of central nerve fibres. The cells that give visual sensation its specific quality, the rods and cones (S and Z Fig. 78) -- usually termed, on this account, visual cells -- are sensory epithelia which, like the gustatory cells, are connected only by contact with the terminal fibrils of the optic conduction. In the retinal layers that cover them are several strata of nerve cells, easily divisible by their marked differences of form into two main groups: the large multipolar ganglion cells (G2), which my be regarded, from the relations of their neurites and dendrites, as proximate points of departure for the optic conduction running centripetally from the retina to the brain; and bipolar ganglion cells (G1) to which may be added stellate intercalary cells, found far forward in the neighbourhood of [p. 186] the elements S and Z, and not represented in the Figure. These last two classes constitute together a neurone territory, intervening between the last termina: fibrils of the peripheral optic conduction and the large ganglion cells (G2), which may be considered as the extreme peripheral member of the centripetal optic conduction. Between its limits we find, further, terminal arborisations of neurites (e), derived not from cells of the retina itself but from more central regions, -- probably from the pregemina, since these, as we shall see in a moment, form important nodal points in the optic conduction at large. There is thus a further point of resemblance between the outlying central area represented in the retina and the olfactory surface: here as there, the structural relations indicate the existence of a centrifugal secondary path, running alongside of the centripetal running alongside of the centripetal principle path.[23][p. 187]

The fibres collected in the optic nerve conduct, then, for the most part centripetally; though there is, in all probability, a small admixture of centrifugal conductors. Following its course, we come upon the decussation of the optic nerve, the chiasma, where a distribution is made of the optic fibres, to the paths running further towards the central organ, that is obviously of extreme importance for the co-operation of the two eyes in binocular vision. It is instructive, in this regard, to trace the phylogenetic stages through which the mode of distribution in the human chiasma has gradually been attained. In the lower vertebrates, up to the birds, there is a complete decussation of the paths, the right half of the brain receiving only the left optic path, and conversely. In the mammalian series, from the lower orders onwards, direct paths play a larger and larger part alongside of the crossed fibres; until finally, in man, the distribution has become practically equal; so that the one (the temporal) half of the retina passes into the optic tract of the opposite side, and the other (the nasal) into the tract of the same side.[24] We owe our knowledge of this fact less to direct anatomical investigation, which finds great difficulty in the tracing of the detailed course of the optic nerve, than to pathological observations of the partial loss of sight resulting from destruction of the visual centre of one hemisphere or from the pressure of tumours upon the optic tract of one side. The main results of these observations are brought together, in schematic form, in Fig. 79. It will be seen that the corresponding halves of retina and optic nerve are cross-hatched in the same direction. The temporal halves of both retinas have a crossed (tt), the nasal halves a direct path (nn). Before decussation, the crossed path lies on the outside, the uncrossed on the inside of the optic nerve; after decussation, the crossed changes to the inside, the uncrossed to the outside of the optic tract. In contradistinction to the retinal halves which thus receive only a one-sided representation in the brain, the central area of the visual surface, or macula lutea, where the retinal elements are set most thickly, is favoured with a bilateral representation. Destruction of the central optic fibres of one side is accordingly followed by half-blindness (hemianopsia) or limitation of the field of vision to one-half of each retina (hemiopia), with the exception of the area of direct vision around the fixation point, which becomes blind only when the central disturbance affects both sides of the brain.[25] We return later (pp. 229 ff.) to the relations which this peculiar mode of distribution sustains to the function of vision.

The conduction of the optic tract of either half of the brain is thus composed of temporal paths from the opposite retina, nasal paths from the [p. 188] retina of the same side, and macular paths from both retinas. It divides again, on both sides -- as is shown schematically in Fig. 78, where abstraction is made from the decussations which we have just been discussing -- into

We have now traced, as accurately as may be, the course of the fibre systems that run to the cerebral cortex, whether directly from the crura, or indirectly from the cerebellum and the brain ganglia. We have made use, in this enquiry, both of the results of anatomical investigation and of the degenerations set up by severance of the fibres from their centres of function. But we have not been able to say anything at all definite of the final distribution of the central fibre systems in the cortex itself. As a matter of fact, there are still certain mazes of interlacing fibres to which anatomists have not yet found the clue; and our two methods fail us, when we seek to determine by their aid the precise relations in which the various regions of the cerebral cortex stand to the deeper lying nerve centres and to the peripheral parts of the body. We therefore ask assistance at this point from two other sources, physiological experiment and pathological observation. The former supplies us with a certain correlation, in the animal brain, between definite cortical areas and the various motor and sensory functions of the peripheral organs. The latter attempts the same problem for the human brain, by a comparison of the functional derangements recorded during life with the results of post-mortem examination. The conclusions drawn from experiments on animals may be transferred to man only, of course, in so far as they answer the general question [p. 191] of the representation of the bodily organs in the cerebral cortex. When we attempt to map out, on the human brain, the terminal areas of the various paths of conduction, we have to rely solely upon pathological observations. These possess the further advantage that they allow us to make more certain tests of the behaviour of sensation than do the experiments, upon animals. On the other hand, they have the disadvantage that circumscribed lesions of the cortex and pallium are of comparatively rare occurrence, so that the collection of data proceeds but slowly.

Experiments on animals fall into two main classes: stimulation experiments and abrogation experiments. Under the latter heading, we include all experiments which are intended to abrogate, temporarily or permanently, the function of some cortical area. In stimulation experiments, the symptoms to be observed are phenomena of movement, twitches or contractures in the muscles; abrogation experiments bring about abrogation or disturbance of movements or sensations. Both forms of experiment are of value for the definition of the terminal areas of the motor paths; for the sensory areas, we must have recourse in most cases to abrogation experiments. There are, however, many regions of the cerebral cortex which form the terminal areas of intercentral paths from the cerebellum and brain ganglia, paths which are connected only in a very complicated and roundabout way with the lines of sensory or motor conduction, or with both. We shall, therefore, expect a priori that not every experimental or pathological change, induced over a limited area, will be followed by noticeable symptoms; and that, even where such symptoms appear, they will not, as a rule, consist in simple phenomena of irritability and disability such as arise from the excitation or transsection of a peripheral nerve. This expectation is amply confirmed by experience. At many points, stimulation may be applied without producing any symptoms whatsoever. Where it does produce a result, the muscular excitations often have the character of co-ordinated movements. The symptoms of abrogation, on the other hand, are for the most part simple disturbances of movement or impairments of sense perception; it occurs but seldom, and in general only where the lesion is of considerable extent, that there is complete abrogation of function, sensory or motor. It is well, therefore, in speaking of experiments upon the cerebral cortex, to use expressions, that in some way indicate this ambiguity of result. We shall accordingly distinguish between centromotor cortical areas, whose stimulation produces movements of certain muscles or muscle groups, and whose extirpation is followed by a derangement of these movements, and centrosensory areas, whose removal brings in its train symptoms of loss or defect upon the sensory side.[27] These terms must [p. 192] not, however, be interpreted at the present stage of the enquiry as implying any hypothesis whether of the significance of the phenomena of stimulation and abrogation or of the function of the cortical areas to which they are applied. The only question to be discussed here is that of the termination of the paths of conduction in the cerebral cortex; and all that we require to know, in order to answer it, is the functional relation obtaining between the various regions of the cortex and the peripheral organs. How these functional relations are to be conceived, and in what manner the different cortical areas co-operate with one another and with the lower central parts, -- these are questions that we do not yet need to consider. There is, however, one point, so important for the right understanding of the conditions of conduction that it should, perhaps, be expressly mentioned in this place; a point that follows directly from the extreme complexity of interrelation which we have found to prevail in the central parts. It is this: that, for anything we know, there may exist several centromotor areas for one and the same movement, and several centrosensory areas for one and the same sense organ; and that there may quite well be parts of the cortex which unite in themselves centromotor and centrosensory functions. Suppose, then, that we are able to demonstrate certain results of stimulation and abrogation. They will simply indicate that the particular area of the cortex stands in some sort of relation with the conduction paths of the corresponding muscular or sensory region. The nature of the relation can be conjectured only after a comprehensive survey has been made of the whole body of central functions. All questions of this kind must, therefore, be postponed until the following Chapter.

The extreme complication of the course of the conduction paths, and the unusually complex conditions that govern the central functions, in face of which the formation of a critical judgment becomes a matter of serious difficulty, make us realise all the more keenly the comparative crudeness and inadequacy of all, even the most careful, experimental methods. In stimulation experiments, it is never possible to confine the stimulus effect within such narrow limits as is desirable, if we are to establish the relations of conduction obtaining between distinct cortical areas. Moreover, the central substance, as we have seen, has its own peculiar laws of excitability, which make negative results practically worthless as data from which to draw conclusions. Physiologists are therefore inclining more and more to attribute the higher value to abrogation experiments. But here, again [p. 193] there are difficulties, as regards both the performance of the experiments and the interpretation of their results. The shock given by the operation to the whole central organ is usually so violent, that the immediate symptoms cannot be referred to any definite cause; they may be due to functional disturbance in parts of the brain widely remote from the point of injury. Hence almost all observers have gradually been led to agree that the animals must be kept alive for a considerable period of time, and that only the later, and more especially the chronic symptoms may be made the basis of inference. Even so, however, various sources of error are still possible. Thus, as GOLTZ pointed out, inhibitory influences may continue to be exerted, either upon the entire central organ or upon distant regions, particularly if but a short interval has elapsed after the operation. Or, if a longer time has passed, the injured part may have been functionally replaced by other cortical areas: numerous pathological observations on man have put the efficacy of such vicarious function beyond the reach of doubt. Or, finally, as LUCIANI remarked, the cortical lesion may, on the contrary, set up a secondary degeneration of deeper lying brain centres, so that the abrogation of functions may be extended far beyond its original scope. In view of these difficulties, which mean that the experimental result may be obscured by sources of error of the most various kinds and of opposed directions, it is obvious that conclusions in which we are to place any measure of confidence, must be drawn without exception from a large number of accordant observations, made with due regard to all the factors that might affect the issue. And when these precautions have been taken, it is still inevitable that the conclusions, in many cases, attain to nothing higher than a certain degree of probability. In particular, they will as a general rule fail to carry conviction, until they are confirmed by pathological observations upon the human subject.

Centromotor areas in the cerebral cortex may easily be demonstrated, as HITZIG and FRITSCH were the first to show, by experiments with electrical or mechanical stimuli. The simplicity of the structural plan of the carnivore brain (Fig. 61, p. 138) makes it comparatively easy to rediscover the irritable points, when they have once been found. In Fig. 80 there are marked upon the brain of the dog the principal points about which the statements of the different observers are in general agreement.[28] Besides these superficial areas, there appear to be other cortical regions in the same neighbourhood, lying concealed in the depth of the crucial fissure, which [p. 194] are mechanically excitable: their exact localisation is, however, impossible, owing to their inacessible [sic] position.[29] The motor areas are all situated over the anterior portion of the brain, between the olfactory gyre and the Sylvian fissure. With stimuli of moderate intensity, the effect of stimulation is produced on the opposite side; bilateral symptoms are observed only in the case of movements in which there is a regular functional connexion of the two halves of the body, e. g. in ocular movements, movements of chewing, etc. With stronger stimuli, the effect is confined as a rule to the muscles of the same side of the body. The stimulable areas are seldom more than a few millimetres in extent, and excitation of points lying between them is, if the stimuli are weak, unaccompanied by any visible effect. If the stimulus is made more intensive, or is frequently repeated, contractions may, it is true, be set up from these originally indifferent points; but it is possible that such results are due to diffusion of currents (in electrical stimulation) or to an enhancement of excitability brought about by the preceding stimulation. There can, indeed, be no doubt that repetition of stimulus is able to induce this enhancement for it is often found that, under such conditions, the excitation spreads to other motor areas, so that the animal is finally thrown into general spasms, the phenomena of what is called cortical epilepsy.[30] For the rest, the contractions set up by cortical stimulation are always distinguished from those released by electrical stimulation of the coronal fibres by a much longer duration of their latent period, the expression of that retardation of the stimulation processes which is of universal occurrence in the central elements.[31]

The phenomena of abrogation, observed after extirpation of definite portions of the cerebral cortex of the dog, differ in two respects from the [p. 195] results of the experiments with stimulation. In the first place, they show that the removal of a stimulable area is usually followed by disturbances of movement in other groups of muscles, which were not excited by stimulation of the same area. Thus, extirpation of the area d in Fig. 80 is likely to produce paralytic symptoms in the fore leg as well as paralysis of the hind leg, and, conversely, extirpation of the area c is attended by a partial paralysis of the hind leg; again, destruction of the centres of neck and trunk aa' involves both the extremities; and so on. At the same time, the paralysis of the stimulable areas is always more complete than that of the areas sympathetically affected. In the second place, the extirpation of parts of the cortex that are irresponsive to stimulation may also give rise to phenomena of paralysis; and this statement holds not only of points of the cortex lying between the stimulable areas, within the zone of excitability, but also of more remote regions. It can thus be demonstrated that the entire anterior portion of the parietal lobe, and even the superior portion of the temporal region as well, are in the dog centromotor in function. Only the occipital and the larger, inferior portion of the temporal region can be removed without producing symptoms of abrogation on the motor side. Fig. 81 gives a graphic representation of these facts. The sphere of centromotor abrogation is dotted over; the size and number of the points in any given area indicate the intensity of the phenomena of abrogation appearing (always on the opposite side of the body) after extirpation of that particular zone.[32] The character of these disturbances, and more especially the regularity with which definite muscle groups are affected by the extirpation of definite parts of the cortex, render it improbable that the results obtained from non-stimulable areas are the outcome of transitory inhibitions, propagated as simple sequelæ of the operation from the point of injury to other, uninjured parts. We may more reasonably explain the differences between the phenomena of stimulation and those of abrogation [p. 196] by supposing that the excitable zones stand in closer relation to the peripheral conduction paths than do the others, whose centromotor influence can be demonstrated only by way of the inhibition of function which follows upon their removal. For the rest, it is a significant fact for the theory of these phenomena of centromotor abrogation that they do not consist by any means in complete muscular paralyses. In general, there is inhibition of voluntary movement only: the muscles involved will still contract reflexwise upon stimulation of the appropriate points upon the skin, and may be thrown into sympathetic activity by the movement of other muscle groups. Further, all symptoms of abrogation, save where very considerable portions of the cortical investment of both hemispheres have been removed, are impermanent and transitory; the animals will, as a rule, behave, after the lapse of days or months, in a perfectly normal way, and the restoration occurs the more quickly, the smaller the extent of the cortical area destroyed.[33]

The demonstration of the centrosensory areas, if it is to be accurate and reliable, must, as we said above, be undertaken by help of the phenomena of abrogation. This limitation of method, and more especially the uncertainty which attaches to sensory symptoms, place serious obstacles in the path of investigation. There are, however, two points in which the disturbances of sensation set up by extirpations of the cortex in the dog appear to resemble the motor paralyses which we have already passed under review. First, the cortical regions correlated with the various sense departments are, evidently, not well-marked and circumscribed; they always cover large areas of the brain surface, and even seem to overlap. Secondly: the disturbances, here as before, do not consist in any permanent abrogation of function. If the injury is restricted to a comparatively small area, they may be entirely compensated. If it affects a larger portion of the cortex, there will, it is true, be permanent sensory derangements, but they will express themselves rather in an incorrect apprehension of sense impressions than in absolute insensitivity to stimulus. Thus, dogs whose visual centre has been entirely removed will still avoid obstructions, and others, whose auditory centre has been extirpated, will react to sudden sound impressions, although they can no longer recognise familiar objects or the words of their master. They take a piece of white paper, laid in their path, for an obstacle which they must go round; or confuse bits of cork with pieces of meat, if the two have been mixed together.[34] All these phenomena indicate that the functions of perception have in such [p. 197] cases been abrogated or disturbed, but that the removal of the centrosensory areas is by no means and in no sense the equivalent of destruction of the peripheral sense organs. There is, further, one respect in which the terminations of the sensory conduction paths differ from those of the motor: while the derangements of movement point to a total decussation of the motor nerves, the disturbances of sensation, or at least of the special senses, are bilateral, and accordingly suggest that the fibres of the sensory paths undergo only a partial decussation in their course from periphery to centre. Figg. 82, 83 and 84 show roughly the extent of the visual, auditory and olfactory areas in the cortex of the dog, as determined by the method of abrogation. The frequency of the dots indicates, again, the relative intensity of the disturbances which follow upon extirpation of the area in question; the black dots correspond to crossed, the hatched dots to uncrossed abrogation symptoms. We notice that the visual centre is situated for the most part in the occipital lobe, though less marked disturbances may [p. 198] caused from a part of the parietal lobe and probably also from the hippocampus; the temporal lobe, on the other hand, is practically exempt. The auditory area has its centre in the temporal lobe, from which it appears to extend over a portion of the parietal lobe, as well as the callosal gyre and the hippocampus. The olfactory area has its principal centre in the olfactory gyre. Besides this, it seems to occupy the uncus and the hippocampus, while its share in the parietal region is but small. In the visual and auditory spheres, the crossed fibres have an undoubted preponderance; in the olfactory area, the uncrossed appear to be in the majority. The gustatory area cannot be made out with certainty: it probably lies on the two opposite surfaces of the intercerebral fissure in the anterior region of the parietal lobe.[35] On the other hand the area whose extirpation affects the sense of touch and the sensations of movement -- the two cannot be distinguished in this group of symptoms -- occupies a broad space on the convex surface of the brain. It has its centre, in the brain of the dog, in the anterior parietal region and extends from that over the whole frontal portion, and downwards and backwards to the margins of the temporal and occipital lobes. The centrosensory area for the sense of touch has, that is, precisely the same extent as the centromotor area for the general muscular system of the body; it can accordingly be illustrated from Fig. 81, which we have already employed in our previous discussion. This coincidence suggests the hypothesis that a distribution into smaller, overlapping centres for the various parts of the body will obtain in the case of sensations as we have found it to obtain in the case of movements. For the rest, the phenomena of abrogation which make their appearance after removal of the touch sphere run precisely parallel to the disturbances of the special senses described above: the permanent symptom is always the derangement of perception, and never the insensitivity sometimes observed as the direct consequence of operation.

The brain of the monkey so closely resembles the brain of man (cf. Fig. 64, p. 144), that the discovery of its cortical centres is a matter of peculiar interest. It was therefore natural that the attempt should be made, soon after the establishment of the centromotor points on the brain of the dog, to determine the corresponding points on that of the monkey. Experiments were carried out by HITZIG [36] and FERRIER,[37] who found, in agreement with the course of the pyramidal paths, that the stimulable centres lie for the most part in the region of the two central gyres, whence [p. 199] they extend as far as the superior portion of the subfrontal and medifrontal gyres. More exact determinations of the points were then made in further investigations by HORSLEY and SCHÄFER,[38] and by HORSLEY and BEEVOR.[39] Fig. 85 gives the results obtained by the two latter writers on a Bonnet Monkey (Macasus). They show in general that the cortical centres for the trunk and the hind limbs lie principally on the superior surface; those for the fore limbs somewhat lower down; and lastly, those for the muscles of face,

These results of stimulation are, on the whole, borne out by the results of extirpation of various regions of the cortex, if we allow for the greater margin of uncertainty which the abrogation method always leaves (p. 191). It is noteworthy, also, that the disturbances are apparently less quickly compensated in the more highly organised brain of the monkey than they are in the dog, so that the symptoms of abrogation and stimulation are here more nearly in accord. Nevertheless, according to the observations of HORSLEY and SCHÄFER, it is impossible to induce an approximately complete paralysis upon the opposite side of the body, even in the interval immediately following the operation, unless the whole centromotor zone is extirpated. If the area of injury is more limited, the muscles involved show only a weakening, not a total abrogation of movement.

The determination of the centrosensory centres is, again, far less certain; the interpretation of the symptoms presents very much greater difficulties. Hence for the brain of the monkey, as for that of the dog, the results may he regarded as reliable and assured only in so far as they refer to the general delimitation of the various sense departments. With this proviso, we may conclude from the experiments of HERMANN MUNK, with which those of other observers agree on these essential points, that the cortical surface of the occipital brain constitutes the visual centre, and that of the temporal lobe the auditory centre. The area for touch, taken as inclusive of all the organic sensations, coincides in position with the centromotor regions for the same parts of the body, i.e. is situated in the neighbourhood of the two central gyres and of the superfrontal gyre.[40]

We have, in the above discussion, left the question of the nature of the cortical functions untouched, save in so far as it is connected with the problem of the termination of the paths of conduction in the cerebral cortex. The [p. 201] question cannot come up for consideration in its own right until the following Chapter, when we review the central functions in their entirety. Even with this limitation, however, the experiments upon the terminations of the conduction paths still leave room for differences of interpretation. At the same time, physiologists are on the road to an agreement: it cannot be disputed that the ideas of the moderate party, ideas which compromise between the hypothesis of a strictly circumscribed localisation, on the one hand, and the denial of any local differences whatsoever, on the other, have gradually gained the upper hand. It is this middle course that we have followed on the whole, in the preceding paragraphs. It may be that the lines of the various motor areas will, in the future, be drawn somewhat more closely or somewhat more widely; but the fundamental assumption that the functional areas extend from definite and narrowly circumscribed centres, and that at the same time they frequently overlap one another, has established itself more and more firmly, as the most probable view, in the minds of impartial observers. GOLTZ has protested with great energy against the hypothesis of sharply defined localisations. His work has done a great deal, both by its positive contents and by the stimulus it has given to other investigators, to clear up our ideas upon the subject.[41] But the results which GOLTZ has obtained in his later papers do not differ in any essential respect from those of most other observers; and he himself has now come to accept a certain dissimilarity of central representation, which in its general features resembles the account given above. Cf. also Ch. VI , pp. 281 ff. below.

More serious are the differences of opinion regarding the functional significance of the various regions of what is called the sensory sphere. As regards the position of the centrosensory areas, Munk has concluded, on the basis of numerous experiments with dogs and monkeys, that we must distinguish between cortical areas in which the fibres of the sensory nerves directly terminate, and areas in which sensations are raised to the rank of perceptions. The phenomena which make their appearance after destruction of the former, he names, in the case of the two higher senses, cortical blindness and cortical deafness; the disturbances which result from extirpation of the centres of the second order, he terms mental blindness and mental deafness. According to Munk, the visual centre in the dog includes the portion of the brain lying posteriorly to the Sylvian fissure, and covered by the parietal bones; in the monkey, the whole surface of the occipital lobe (A Figg. 86, 87). This visual centre is then subdivided into a central area (A' Fig. 86), and a peripheral area, which surrounds the central on all sides (A). The centre is supposed, on the one hand, to correspond to the spot of dearest vision of the dye of the opposite side, and, on the other, to contain the elements in which memory images are deposited. Its destruction accordingly means loss of clear vision and, at the same time, of a correct apprehension of sensations. The peripheral portion, A, according to the same author, is on the contrary merely a retinal centre. Every point within it is correlated with corresponding points upon the two retinas, each half of the brain representing the same-sided halves of the retinas of the two eyes. Hence, if one occipital lobe is extirpated, the animal becomes hemianopic; it is blind to all the images which fall upon the same-sided halves of its retinas. Further, in dogs the correlation is symmetrical. The central visual area of [p. 202] each hemisphere corresponds to the smaller, lateral division of the retina of the same side, and to the larger, median division of that of the opposite side; so that, e.g., extirpation of the central visual area of the right hemisphere produces blindness over the extreme edge of the right retina, and

MUNK'S statements, and more especially those that relate to the distinction between direct sensory centres and what he calls mental centres, have, however, been challenged from many quarters. The physiological and psycho-[p. 204]logical assumptions which underlie this division of functions are hazardous in the extreme. That apart, there are two principal points in which MUNK'S conclusions are negatived by the facts as otherwise ascertained. In the first place, it is evidently incorrect to assert that the removal of any cortical area of the animal brain is followed by total blindness or absolute insensitivity to sound stimuli. There are many observations which show that rabbits, and even dogs, will react appropriately to impressions of light and sound after removal of the entire cerebral cortex. They avoid obstacles placed in their path, perform complex expressive movements, and so on.[45] In the second place, the symptoms consequent upon lesions of the cortex correspond in all cases to what MUNK terms mental blindness and deafness; they are, as GOLTZ puts it, symptoms of cerebral weakness. The removal of a cortical area is never the equivalent of destruction of the peripheral organ, or of a part of it.[46] LUCIANI conjectures, further, that the more profound sensory disturbances noticed by MUNK some time after the operation may perhaps be due to a propagation of descending degeneration to the lower centres of the thalami and quadrigemina. Only the relations of definite parts of the visual centre to definite regions of the binocular held of vision have found confirmation in the experiments of other observers:[47] a result which, as we shall see below, is also in agreement with the defects of the field of vision observed in man, after partial destruction of the visual cortex.

The disturbances observed in man, as a result of lesion of the cerebral cortex, may take the form either of stimulation phenomena or of symptoms of abrogation. The former, which appear sometimes as epileptiform contractions, sometimes as hallucinatory excitations, hardly come into account for the question of localisation of functions, since they rarely accompany local and circumscribed injury of the cortex. We have, therefore, to rely upon the symptoms of abrogation; and these are the more valuable, the more limited the range of function which they involve. Nevertheless, it requires great care to separate them from the affections of surrounding parts, which are seldom absent at the beginning of the disturbance, and from the phenomena of restitution of function, which make their appearance after the lapse of time.[48] The observations which have been brought together, with due regard to these precautions, lead to results, more especially as regards the centromotor areas of the human cortex, which agree in their principal features with the experimental results obtained on the brain of the monkey. This will be seen at once from a comparison of Figg. 88 and 85 [p. 205] (p. 199), the former of which gives a schema of the localisation areas in man, based on pathological observations, while the latter shows the centromotor points of the brain of Macacus. It is evident that, in the cortex of man as in that of the monkey, the areas whose lesion produces motor paralysis are grouped in a

The phenomena of abrogation observed in pathological cases of partial destruction of the cortex inform us, further, of the position of the principal centrosensory areas in the human brain. First and foremost, the central terminations of the optic paths,

A similar complication obtains in the case of the auditory area of the human cortex. Pathological affection of the terminal areas of the acoustic nerve is shown, primarily, by abrogation or impairment of the power of hearing; but this is invariably accompanied by a profound modification of the faculty of speech. The connexion of the two is not surprising, since the motor and sensory areas of the sense of hearing lie side by side in the cortex (see Fig. 88, and Figg. 89, 90), and may therefore easily be involved in the same lesion. There is, however, a further factor that introduces a peculiar complication into the phenomena: in all affections of what we suppose to be the cortical terminations of the acoustic nerve, derangement of the direct motor and sensory terminals is, apparently, always accompanied by disturbances of connective paths or centres. Hence, most of the derangements of the faculty of speech -- those ordinarily distinguished by their symptoms as aphasia, word-deafness, agraphia, word blindness, etc. -- are of extremely complex character. And we must accordingly assume that, besides the direct auditory centre, there are always involved other central areas which, like the corresponding motor centre, lie in its near neighbourhood. We shall recur to the probable conditions of these derangements of speech in our consideration of the complex functions of the central organ (Ch. VII.). As regards the boundaries of the direct auditory centre, we cannot speak with complete assurance; its connexion with other central areas, which co-operate in the auditory functions, leave a margin of uncertainty. There can, however, be no doubt that the principal terminus of the acoustic path (Fig. 89) is the posterior section of the supertemporal gyre (T1 Fig. 88), the part that borders the end of the Sylvian fissure. That is, this region lies directly opposite to the motor areas of the subfrontal gyre, which are brought into activity in the movements of speech (Fig. 88).[55]

The demonstration of the centres for the fibres of the gustatory and olfactory nerves presents considerable difficulty, though for other reasons.[p. 209] The method of abrogation here fails us: the ambiguity of the symptoms, whether in man or in the animals, renders it practically impossible to determine the effects of cortical lesion. In this instance, therefore, we must still rely entirely upon the results of direct anatomical investigation of the course of the conduction paths. These results indicate that the olfactory area occupies the space marked in Fig. 90 by small open circles; i.e., that it extends on the one hand over a narrow strip on the posterior margin of the frontal lobe and over the callosal gyre, and on the other over the superior and inner margin of the temporal lobe, adjoining the posterior extremity of the callosal gyre. The parts in question, and especially the callosal gyre, are, as we know, much more strongly developed in certain animal brains, e.g. in the carnivores (Fig. 63, p. 143); so that the area of distribution accords with the relatively low development of the sense of smell in man. The taste area is supposed to lie somewhere in the neighbourhood of this olfactory centre. So far, however, it has not been definitely localised, either by anatomical or by functional methods.[56]

We return to safer ground when we seek to determine the central areas for the sensations of the general sense, i.e. more particularly for tactual and common sensations. Numerous observations go to show, in agreement with the results of operation on animals, that the centrosensory regions of the sense of touch coincide with the centromotor regions for the same parts of the body. Disturbances of tactual and muscular sensation are found to follow upon injury to the posterior portion of the three frontal gyres, the two central gyres, the paracentral gyre, and the parietal and subparietal gyres; i.e. to the whole region indicated in Figg. 89 and 90. The reference of special areas to the different parts of the body, and the separation of touch from common sensation, are matters of less certainty. On the former point, we can only say that, despite the general coincidence of the sensory and motor regions, it is still possible that the two kinds of centres are not wholly identical, but simply bound together by a close relationship of structure and function. On the latter, we have a few observations that point to a central differentiation of internal and external tactual sensations. Cases occur in which the sensation of movement is abrogated, while cutaneous sensation and motor innervation remain intact; and these isolated disturbances of articular and muscular sensations seem to be induced more especially by affections of the parietal and subparietal gyres.[57] But, after all has been said, the results so far obtained with regard to the localisations of the general sense leave us still in doubt upon many [p. 210] points of detail. In particular, all statements concerning the relation of these centrosensory cortical areas to the centromotor must be regarded, at present, as altogether hypothetical. They rest, not upon reliable observations, but for the most part upon some foregone psychological or physiological assumption.

If, in conclusion, we compare the whole group of results derived from pathological observation of the relations of the cerebral cortex to the several conduction systems with the outcome of the experiments made upon animals, we see at once that, where the facts are at all securely established, there is a large measure of agreement between the two methods. Thus, the position assigned to the centromotor areas in man and the animals is practically the same. In particular, the motor points of the central gyres are arranged in a similar order upon the human and monkey cortex. The same thing holds of the localisation of visual excitations in the occipital lobe. In the acoustic area, it is true, we find differences. The development in man of the cortical region connected with speech is offset, in most of the animals, by the greater bulk of the olfactory centres. There is thus a more pronounced dissimilarity in the structure of the anterior portion of the brain, and a consequent lack of correspondence of the cortical areas. According to the observations of FERRIER, MUNK and LUCIANI, the auditory centre of the dog, e.g., is forced, by the development of the olfactory gyre, relatively far back, into the posterior part of the temporal lobe. This apart, however, the auditory area appears, to all intents and purposes, to occupy an analogous position in the human and animal brain. And the same statement may be made, finally, with still greater confidence, of the cortical areas for tactual and common sensations, whose localisation refers us, in all cases, to regions which either coincide or interfere with the corresponding centromotor areas. So far, then, there is a general agreement between the results. The only difference of any considerable moment is that the derangements of function resulting from cortical lesions are as a rule more serious in man than they are in the animals. And this difference itself has only a relative significance, since it appears in the same way between various classes of animals, e.g. between dog and rabbit, or still more markedly between monkey and dog. It would seem, then, that the phenomena in point are simply illustrations of the general fact that the subcortical centres have a higher value, as centres of independent function, the lower the organisation of the brain to which they belong.[58] Lastly, having allowed this difference its due weight, we have again to say that the character of the disturbances produced by local lesions of the cortex is the same for man and for the animals, in so far as the derangement never amounts to an absolute abrogation of function, and is therefore by no [p. 211] means equivalent to the interruption of a peripheral conduction path. The nearest approach to such a result is given in the paralyses which follow upon destruction of the centromotor zones. Even these, however, are definitely distinguished by the possibility of comparatively rapid restoration of function.

We have made mention, in the preceding paragraphs, of all the areas of the human cortex that can lay claim, chiefly on the ground of pathological observations, to be considered as the termini of motor and sensory conduction paths. The motor area, shown in Fig. 88, and the visual centre of the occipital lobe were the earliest of these 'cortical centres' to be discovered, and are at the present day the two whose lines can be most sharply drawn. Much more precarious, for the reasons given in the text, is the status of the acoustic area: and this despite the fact that the derangements of speech, which stand in intimate connexion with it, have been under observation for a long period of time.[59] Finally, the correlation of the central gyres and the adjoining region (as mapped out in Figg. 89 and 90) with the sensations of the general sense (external and internal sensations of touch, pain, and organic sensations) may be regarded as sufficiently well established. The localisation was first suggested by TÜRCK, who noticed that lesions of these coronal fibres and of the crural fibres in the region of the capsula of the lenticula produced unilateral sensory disability.[60] In these cases, and still more in cases of destruction of the central gyres themselves, the symptoms are, however, invariably complicated by the simultaneous appearance of motor paralysis in the corresponding regions of the body. For the rest, the hemianæsthetic disturbances are usually distinguished from such hemiplegic accompaniments by their more irregular character; they may be confined to certain factors of the general sensitivity -- muscular sensations, pain, sensations of temperature, etc. -- or they may be combined with other sensory disturbances in the departments of special sense, and more especially with amblyopia.[61]

All these sensations of the general sense, sensations derived from the organ of external touch, as well as from joints, muscles, tendons and other bodily organs, have been grouped together by certain authors under the indefinite name of 'bodily feeling.' In accordance with this usage, the area ascribed to the general sense in Figg. 89 and 90 was termed by H. MUNK the 'area for bodily feeling.' The title has become current; and its employment is generally connected with various psychological hypotheses, which play an important part in the interpretation of the centromotor symptoms induced by lesions of this region. Thus, SCHIFF propounded the theory,[62] which has been accepted by MEYNERT,[63] H. MUNK,[64] and many other anatomists and pathologists, that the centromotor innervations are direct concomitants of the ideas of the respective movements. This means that the cortical region assigned to the 'area for bodily feeling' is to be regarded as a sensory centre, analogous to [p. 212] the centre of sight or hearing. The volitional process is then explained as a reflex transference, occurring, possibly, in the cortex itself, or, perhaps, in deeper-lying parts. This aspect of the theory is rendered especially plausible by the belief that 'will' is nothing else than an 'idea of movement,' and that consequently the 'cortical function' underlying voluntary action consists simply and solely in the excitation of a movement idea, i.e. in a sensory process. Now the assumption that 'will' is equivalent to idea of movement is, of course, a purely psychological hypothesis; it can be demonstrated or refuted, not by anatomical and physiological facts, but only by a psychological analysis of the voluntary processes themselves. Hence we cannot enter upon its examination in this place, though we shall take it up in due course. The physiological investigation of the conduction paths is properly concerned with the single question, whether the cortical areas under discussion are exclusively centrosensory in function, or whether they also evince centromotor symptoms. If the question is put in this way, and we repeat that this is the sole way in which, from the physiological standpoint, it can be put, then the only answer possible, in the light of the observations, is the answer given in the text. But it need hardly be said that that answer gives us no warrant for speaking of a 'localisation of the will' in the brain cortex. To do so would be as absurd as to say, e.g., that the subfrontal gyre and its surroundings parts are the seat of the 'faculty of speech.' The removal of a screw may stop a clock; but no one will be found to assert that the screw is what keeps the clockwork going. The will in abstracto is not a real process at all, hut a general concept, gained by abstraction from a large number of concrete facts. And the concrete individual volition, which alone has actual existence, is itself a complex process, made up in every case of numerous sensations and feelings. There can be no doubt, therefore, that it involves a number of different physiological processes. The hypothesis that a complex function, like speech or volition, is conditioned solely upon certain individual elements may accordingly be pronounced a priori as improbable in the extreme. Resides, all that follows from the observations is that those parts of the brain cortex which we claim as centromotor contain transmitting stations, which are indispensable for the transference of voluntary impulses to the motor nerve paths: the anatomical facts making it further probable that the regions in question contain the proximate stations of transmission from brain cortex to central conduction paths.[65]

There is one other fact that we may mention, in conclusion, as of importance for the psychogenetic interrelations of the different sense departments. According to the investigations of FLECHSIG, the fibre systems that radiate from the mesencephalon to the various cortical centres obtain their myelinic sheath at very different stages of embryonic (partly also of postembryonic) development, and therefore, we may suppose, assume the functions of conduction at these same intervals. In man, the fibres that ascend to the tactual centre from the sensory dorsal columns of the myel, together with a few others that enter the optic radiation, are the earliest of the coronal bundles to attain to full development. They are followed, at a somewhat later period, by fibres which in part supplement this pre-existing system, and in part trend towards the olfactory and visual centres. The myelinisation of the fibre system of the acoustic path is completed last of all, to some extent after birth. At the same [p. 213]
time, it does not appear that the animal series presents any thorough-going parallelism in this regard; the investigations of EDINGER prove that the olfactory radiation is developed very early in the lower vertebrates, while in man it belongs to the systems of later development.[66] And in the human brain, this general course of development divides into a large number of separate stages, each corresponding to the completion of some smaller fibre system. FLECHSIG himself has thus been led to distinguish no less than forty fibre tracts, running in developmental succession to definite regions of the cortex. He finds, in general, that the conduction paths of the 'association centres,' discussed in the following Section, are the latest to reach maturity.[67] It should be said, further, that FLECHSIG'S statements have been called in question from many quarters. Some authorities altogether reject the idea that the myelinic sheath developes system by system; others at least dispute the regularity of the development.[68] Moreover, it cannot be denied that the greater the number of the cortical centres that we are called upon to distinguish by the order of their completion in time, the smaller becomes the probability that each single centre possesses a peculiar functional significance. Nevertheless, the general result is noteworthy, that the conduction paths whose cortical centres receive special elaboration in the human brain are apparently also the latest to attain to individual development.[69]

The whole group of fibres that pass upwards in the myel and, reinforced by additions from the posterior brain ganglia and the cerebellum, finally radiate into the corona of the cerebral cortex, is ordinarily termed the projection system of the central organs. The name was first employed by MEYNERT, and is intended to suggest the idea that the system in question represents the various peripheral organs in determinate regions of the cerebral cortex. In metaphorical language, the periphery is 'projected' upon the brain surface. The fibre masses of this projection system, some of which enter the coronal radiations as direct continuations of the crura, while others are derived from the mesencephalic ganglia, quadrigemina and thalami, and yet others issue from the cerebellum, are crossed at every point of their path to the cerebral cortex by foreign fibre masses, which connect various regions of the cortex with one another. This second group is known (the term was again coined by MEYNERT) as the association system of the cerebral cortex.[69] Both names, as employed here, have, of course, a purely anatomical significance. The projection system has nothing at all to do with what, e.g., is called in physiological optics the outward 'projection' of the retinal image, and the association system, similarly,[p. 214] has nothing to do with the psychological 'association of ideas.' The point must be sharply emphasised, because, as a matter of fact, confusions of this sort, due to obscurity in psychological thinking, have often played -- nay, continue to play -- a part in discussions in which the terms are employed. Now, as regards the projection system, the anatomical facts are sufficient to prove that, if it represents a sort of projection of the peripheral sensory surfaces upon the brain cortex, it can at best be accredited with but a partial performance of this duty. For, on the one hand, the various sensitive areas of the bodily periphery appear, in most cases, to be connected at the same time with several points upon the cortex; and, on the other, the different fibre systems which terminate in a given area of the cortex may correspond to distinct external organs. All this means that the projection system is at least as deeply concerned with the central connexion of the bodily organs as it is with that central representation of them from which it takes its name. As regards the association system, there is not the slightest reason for bringing it into any kind of connexion with the associative processes of psychology. The only hypothesis that we have the right to make about it, on the score of function, is that its fibres serve in some manner to effect the functional unity of separate cortical areas.

The association system like the projection system, may be divided into various component systems, distinguished in this case partly by the direction of connexion, partly by the distance separating; the connected regions of the cortex. We thus obtain the following subordinate systems of association fibres:

(1) The system of transverse commissures. This is principally composed of the callosum or great commissure, but is supplemented as regards the temporal lobes by a portion of the precommissure, which also contains the decussation of the olfactorius fibres (Fig. 91; cf. above, p. 132, and Fig. 53, p. 127). The callosum represents a strongly developed cross-connexion; its fibre masses connect not only symmetrical, but also, to some extent, asymmetrically situated cortical regions of the two hemispheres. The callosal fibres cut across the coronal radiations at all points except in the occipital region, where the two sets of fibres separate into [p. 215] distinct bundles (m', Fig. 58, p. 135; cf. also Fig. 57, p. 134). The connexion effected by the callosum between symmetrical parts of the cortex is fullest, as might be conjectured from its marked increase of size in transverse section as we proceed from before backwards, for the cortex of the occipital region. This is why a defective development of the callosum, as observed in cases of microcephaly, is accompanied by a marked atrophy of the occipital lobes.

(2) The system of longitudinal connective fibres (Fig. 92). This system takes an opposite, direction to the foregoing; its fibres connect remote cortical areas of the same hemisphere. Dissection of the brain reveals several compact bundles of this kind,

(3) The system of intergyral fibres (fibræ propriæ, Fig. 92). This system serves to connect adjoining cortical areas. The fibres are for the most part deflected round the depressions of alba formed by the cerebral fissures (cf. also fa, Fig 58, p. 135).

The association systems that thus connect the various regions of the [p. 216] cerebral cortex may be again divided into three classes, according to their mode of origin and termination. They may (1) connect different areas of the projection system, i.e. centromotor or centrosensory regions, with one another. They may further (2) connect determinate areas of the projection system with other areas, in which no projection fibres directly terminate. Finally, (3) it is probable that in certain parts of the cortex associative fibres of different origin run their course together; so that these areas are connected with the projection system only indirectly, by way of the association fibres that issue from them and terminate in other cortical regions. Areas of this sort, which must be regarded exclusively as terminal stations of association fibres, have been termed by FLECHSIG 'association centres.'[70] They occupy, according to this investigator, practically the whole region of the cerebral cortex which is not taken up by the sensory centres: i.e., in the human brain, the parts which are left unmarked in Figg. 89 and 90. If, then, we consider every continuous surface of this sort as a separate central area, we shall have to distinguish three association centres: an anterior or frontal centre, which covers the larger part of the frontal brain; a middle or insular centre, which extends over the cortex of the insula and its immediate neighbourhood; and a posterior, parietotemporal centre, of wide dimensions, taking in a considerable portion of the parietal and temporal lobes. Between these association centres and the projection centres there lie, still according to Flechsig, intermediary areas and marginal zones, in which projection fibres, either intermixed with the others or grouped in distinct bundles, terminate along with the association fibres. The validity of this distinction of special centres, wholly deprived of direct connexion with the projection system, is disputed by many authorities; and the statements made with regard to the boundary lines and dimensions of the fields in question, and more especially with regard to the extent of the marginal zones and mixed areas, are open to doubt on many points. Nevertheless, it appears to be an established fact that certain areas of the human cortex are supplied for the most part by association fibres, and that these are, in general, the areas whose destruction shows itself not so much in direct centromotor or centrosensory symptoms as in more complicated anomalies of function. On the other hand, it must not, of course, be forgotten that the 'direct' motor and sensory centres themselves cannot possibly be regarded as simply projections of the peripheral organs upon the brain surface. The symptoms of abrogation are decisive: the disturbances set up are of a complicated nature, and their compensation may later be effected, within wide limits, by vicarious functioning of other parts. This result is in harmony with the further fact that there is no region of the brain surface [p. 217] that does not receive association fibres as well as projection fibres: indeed, it is probable that the former constitute the large majority of the coronal fibres in all parts of the human brain. On this count, therefore, it would seem that the search for specific differences is altogether in vain. But if any and every derangement of function due to central interference, in whatever part of the brain it may occur, is of a more or less complicated character, it follows that there can be no question of contrast or antithesis as between the functions of the various areas. The difference is always a difference of degree; we have to do, in the particular case, with a closer and more direct or with a remoter and more indirect relation between a given cortical area and certain peripheral functions. This fact should be kept in mind, further, in all attempts to put a functional value upon the different distribution and relative dimensions of the projection and association centres. It has been found, e.g., that the association centres, together with the main bundles of association fibres that run between the different parts of the brain (Figg. 91, 92), attain very much larger dimensions in the human than in the animal brain: in many cases, indeed, their presence in the animal cortex cannot be demonstrated at all. This statement holds more particularly of the frontal association centre, whose high degree of development determines in large measure the peculiar conformation of the primate, and more particularly of the human brain. Finally, it is worth remark that the cortical area which contains the most extensive representation of peripheral organs, the region in the neighbourhood of the central gyres correlated with bodily movements, the sense of touch and organic sensations, also evinces the most extensive connexions with the association centres.

The existence of 'association centres,' as defined by FLECHSIG, has in recent years been the subject of animated discussion among the students of brain anatomy and brain pathology. RAMON Y CAJAL, EDINGER and HITZIG (the latter with certain reservations) declared themselves in favour of the hypothesis; while DÉJERINE, VON MONAKOW, SIEMERLING, O. VOGT and others pronounced the distinction altogether impracticable.[71] There are no cortical areas, say these authorities, to which projection fibres cannot be traced; just as there are, by general admission, none which are not supplied with association fibres. The question as such is, of course, a question in anatomy pure and simple. We can here do no more than point out that its settlement can hardly be of such importance, from the physiological point of view, as it might, perhaps, appear to us from the anatomical. The occurrence of cortical areas which, in all probability, are connected with peripheral organs only indirectly, by way of the conduction paths that lead to other centres, may, no doubt, be con-[p. 218]sidered as evidence on the one hand of the extremely complicated structure of the particular brain, and, on the other, of the peculiarly complex function of these areas themselves. But the conditions do not, surely, warrant us in ascribing to them a specific function, and setting them off, as 'psychical centres,' from the 'projection' or 'sensory centres.' As a matter of fact, there is nothing to indicate that the structure of the cerebral cortex conforms to any such simple design as that certain parts shall be, so to say, reflections of the peripheral organs, while others shall be reserved as centres of a higher order, serving to bring the direct centres into mutual connexion. On the contrary, it is an essential characteristic of every part of the central organ that it brings together elements which, while spatially separate at the periphery, nevertheless co-operate for the unitary discharge of function. In the case before us, it is first of all the lower central parts, and then, in the last resort, the whole body of the cerebral cortex, that mediate connexions of this kind. What is called the 'visual centre,' e.g., is by no means a repetition of the retinal surface within the cortex. The retina itself is, as we know, nothing else than a part of the cortex that has been displaced far forwards; so that, in this instance, the central duplication would really be a piece of quite needless self-indulgence on the part of Nature. It is because the visual centre contains, along with the conduction paths that connect it with the retina, other paths, whereby it is able to connect the retinal excitations with further functional areas, e.g. the motor, concerned in the act of vision, -- it is for this reason that the visual centre is a true 'centre,' and not a mere duplicate of the peripheral organ. If, now, there are portions of the cerebral cortex whose elements stand in no sort of direct connexion with the conduction paths that run to the periphery, then we must simply say that these areas are possessed, in an unusually high degree, of an attribute which determines the character of the central organ at large, and which must therefore be predicated in some measure of all the other areas, whose connexions with the periphery are more or less direct. To speak of a number of 'psychical centres,' one must have made assumptions that are equally impossible whether from the standpoint of physiology or of psychology. There is, in reality, but one psychical centre; and that is the brain as a whole, with all its organs. For in any at all complicated psychical process, these organs are brought into action, if not all together, at any rate over so wide a range and in such various quarters as to forbid the delimitation of special psychical centres within the functional whole.

The investigation of the fibre systems by physiological experiment, by pathological observation, and by anatomical dissection comes to a natural end at the point where these systems pass over into the cerebral cortex itself. If we wish to inquire further regarding their mode of termination within the cortex, and more particularly regarding the mutual relations of the different parts of the projection and association systems that run to one and the same cortical area, we must gather up the results of the histological examination of the structure of the cerebral cortex. Now our knowledge of this extraordinarily complicated formation has, it is true, not advanced so far that we can establish, beyond the reach of doubt, the [p. 219] terminations of all

We notice, first of all, that in certain structural outlines the cerebral cortex shows uniformity of design over its whole extent. It consists throughout of several strata of nerve cells. In the human cortex we can distinguish, according to the size, direction and position of the cells, eight or nine distinct layers. The order in which these strata are arranged is, on the whole, the same for all regions, though their relative thickness and the number of the elements characteristic of each layer differ very considerably from part to part. Figg. 93 and 94 show the structural relations obtaining in two typical cases. Fig. 93 gives a microscopical transsection through the postcentral gyre, i.e. through a part of the centromotor region of the cerebral cortex; Fig. 94 gives a similar section through the occipital cortex of the human brain. In these, as in all other sections, from whatever part of the brain they may be taken, the outermost and innermost layers are practically identical in constitution: they are characterised by spindle-shaped cells, in the former set crosswise in a fibrillar reticulum whose general trend is in the horizontal direction, and in the latter placed lengthwise among longitudinally directed fibre masses. There are, on the other hand, well-marked differences in the depth of the layers of large and small pyramidal cells, and of those composed of large and small stellate cells. In the centromotor regions, the pyramidal cells form the great majority of all the cell elements. This may be clearly seen in the section from the postcentral gyre shown in Fig. 93, whose formation lies midway between that of the typical 'motor' and the typical 'association' cortex (superfrontal, subtemporal, etc. gyres); but it is still more apparent in the precentral gyre, where the pyramidal cells ('giant' cells) are especially large and extend far down into the seventh layer of spindle and triangular cells. In the visual cortex, on the contrary, the pyramidal cells, and particularly those of the larger class, are greatly reduced both in bulk and in range of distribution, while there are large accumulations of stellate cells, that send out dendrites in all directions (4 and 5, Fig. 94).

These differences in the representation of characteristic cell forms are paralleled by differences in the arrangement of the cortical fibre systems. Where the pyramidal cells are the prevailing type, the fibres in general take a longitudinal course, ascending vertically from the alba to the cerebral surface. A large number of these longitudinal fibres issue directly from the pyramidal cells: the neurites of the larger cells (A Fig. 95), in particular, are continued without break to the myelic columns. We have, then, in these longitudinal fibres of the centromotor region, the direct point of departure of the pyramidal path. All the other processes of the pyramidal cells are dendritic in character. The stoutest of them leaves the [p. 221] cell body on the side opposite to the neurite and runs, still longitudinally, to the periphery of the cortex, where it is broken up in the nervous reticulum of the outermost layer. We thus have, in all probability, a direct centrifugal conduction, as indicated by the arrows in the Fig., beginnings in the peripheral layer of the cortex and continuing through the pyramidal cells into the motor paths. This system is, however, cut across by other longitudinal fibres, some of which issue from smaller cells whose neurite runs a brief course and then splits up into terminal fibrils, while others ascend in more connected fashion from the alba, and again break up in the fibrillar reticulum of the outermost cortical layer. The former (B) constitute, it is supposed, links in the association system, which, as we know, sends fibres to every cortical area; the latter (D) are probably centripetal neurites of deeper lying cells, situated in the mesencephalon. Putting these facts together, we may describe the structure of the motor cortex as follows. (1) Its most characteristic constituent is the centrifugal path (A), mediated by the pyramidal cells. It contains further (2) a centripetal fibre system (D), probably connected with the path (A) in the nervous reticulum of the outermost cortical layer, and representing the terminal sphere of a neurone territory which belongs to deeper lying portions of the brain; and (3) an association path (B) mediated by intercalary cells, which, like the other two, takes in general a longitudinal course and discharges into the fibrillar reticulum of the outermost layer. The direction of conduction in this path is indeterminate; it may possibly vary with the direction of the incoming excitations. Finally, we must mention (4) the plexus of stellate cells, which, while it plays but a small part in the centromotor regions, is never entirely wanting. This, if we may judge by the character that attaches to it in the occipital cortex, is to be regarded as the [p. 222] terminal station of paths of sensory conduction. Its position in the motor cortex is, as we have said, comparatively insignificant. It may be distinguished from the other constituents by its plexiform structure, the fibres running in all possible directions. The presence in the centromotor region of a formation which is characteristic of the sensory centres, may, perhaps, be taken to mean that this region is sensory too, as well as motor. Such an interpretation would be in accordance with the fact that physiological experiment and pathological observation place the centre for the general sense in this part of the cortex.

We now turn, by way of contrast, to the 'visual' cortex, as a typical illustration of a pre-eminently sensory region. We are at once struck by the marked difference in the course of the fibres. Plexiform formations, with fibres running in all directions -- horizontally, therefore, as well as vertically and obliquely -- are very strongly preponderant, while the longitudinal fibre masses characteristic of the terminal areas of the pyramidal paths find but scanty representation (Fig. 96). These plexuses, C, are constituted of the large and small stellate cells, sending out processes in all directions, which appear conspicuously in the section of Fig. 94; in all probability, they consist simply of interlocking neurones, of relatively limited range. The characteristic systems of the motor area also appear, only in lesser numbers, in the visual cortex; just as the formations which we connect with the sensory functions are present, in some degree, throughout the centromotor region. We notice, in particular, the longitudinal centrifugal fibres, connected with the pyramidal cells (A). There are, [p. 223] further, the centripetal fibres, ascending to the cortex from deeper lying cell groups; and, lastly, the supposed association fibres with their intercalary cells (B). We must accordingly infer that the visual cortex discharges centromotor, as well as centrosensory functions. The muscles that can be innervated from it are, we may suppose, more especially the muscles of the eye, though it is possible that other motor organs, correlated with the ocular muscles, are also under its control. H. MUNK has observed movements of the eyes, in animals, as a result of stimulation of the visual cortex.[72]

Such are the differences that obtain between the two main types of cortical structure, the sensory and the motor. Minor differences are found in the various parts of the motor cortex, and again between the visual area and the other predominantly sensory regions. The former have already been discussed. As regards the latter, we notice that in the olfactory cortex of man the pyramidal cells are even rarer than in the visual; the smaller pyramidal cells are altogether wanting. The auditory cortex is characterised, on the other hand, by its great wealth of stellate cells and by the extent of its sensory fibrillar reticula. In the 'association cortex,' finally, these plexuses become less conspicuous, and the granular layers, containing for the most part intercalary cells of varying form, play the leading part.

Putting all this together, we may sum up as follows the general outcome of investigation into the structural peculiarities of the cerebral cortex. Not only are the essential morphological elements the same, for all divisions of the cortex, but their general arrangement also presents no really significant differences. At the same time, there are several layers which, with their characteristic elements, attain to very different degrees of development according to the special functions of the various parts of the cortex. Two kinds of cellular elements, in particular, with the arrangement of fibrillar processes that goes with them, appear to be of symptomatic importance in this regard: the pyramidal cells, with their longitudinally directed fibres, and the stellate cells, with their fibrillar reticula, -- the former characteristic of the centromotor regions, and therefore, we may suppose, serving in the main as points of departure of the great centrifugal conduction paths; the latter characteristic of the sensory regions, and therefore, in all probability, serving in the main as terminal stations of paths of centripetal conduction. To these we must apparently add, as a third characteristic constituent, varying greatly in extent of development, certain [p. 224] cells with limited neurone territory, set longitudinally and connected with longitudinal fibre systems, which may perhaps be looked upon as the substrate of the 'association' paths. Lastly, in all regions of the cortex, the outermost layer with its reticular fibre arborisations seems to form a meeting-place, in which conduction paths of the most various kinds come into contact with one another. The question whether the different regions of the cerebral cortex possess a specifically different structure, which may at the same time serve as the basis for the discrimination of functions, or whether their constitution is, in essential features, the same throughout, has often been under discussion in recent years. MEYNERT, in his epoch-making studies of the human cortex, declared himself for uniformity of structure;[73] and he has been followed by GOLGI [74] and KÖLLIKER.[75] Other investigators, and more especially RAMON Y CAJAL,[76] to whom we are at the present time most deeply indebted for our knowledge of the structural conditions here prevailing, uphold the hypothesis of specific differences. This divergence of opinion seems, however, when closely examined, to be much less radical than its phrasing indicates; it hinges, apparently, upon the different interpretation put by the parties to the controversy upon the term 'specific,' as applied to structural peculiarities. RAMON Y CAJAL'S enquiries have themselves furnished conclusive proof of the extraordinary degree of similarity obtaining in the structure of the various regions, and have shown that the differences are in every case merely relative differences in the number of particular elements and in the development of the layers. Nay more: since they have made it probable that the different centromotor, sensory and 'associative' functions are bound up with definite cell and fibre systems which, as a general rule, are found in all parts of the cerebral cortex, and that these functions are in the main conditioned simply upon differences in direction of conduction, which in their turn depend upon differences in mode of connexion with peripheral organs and with other cortical areas, they have really done away with any possibility of the correlation of specific elementary substrates with the 'specific' functions of the various departments of the cortex. They rather force us to the conclusion that the different modes of cortical activity are founded not upon the specific character of the structural elements, but solely upon their different modes of connexion. Nevertheless, as we shall see in what follows, modern brain anatomy presents us with the very curious spectacle of a science that holds with extreme tenacity to the hypothesis of specific functions, while its own results are constantly rendering this hypothesis less and less practicable, and indeed, if the evidence of structural relations is to count at all, bear striking testimony against the specific nature of the elementary nervous functions.[p. 225]

It is almost inevitable that the student who is tracing the course of the conduction paths, and their concurrence and interrelation in the various divisions of the central organs, should be led into hypotheses concerning the functions of these different parts. Hence it is not surprising to find that, as a matter of fact, physiological conclusions have often been based upon anatomical data. The value of such inferences must, of course, always remain problematical, seeing that they always need to be supplemented by direct physiological analysis of the functions themselves. At the same time, it is evident that we may look to the conditions of conduction to indicate points of departure for functional analysis. They will, at any rate, warrant us in ruling out certain ideas, from the outset, as inadmissible, and in accepting others as more or less probable; and they will do this altogether apart from any functional reference or physiological knowledge. Thus, in view of the complicated character of the acoustic conduction, as represented in Fig. 77 (p. 183), it will be granted without hesitation that an hypothesis which should explain the process of tonal perception as due merely to sympathetic vibrations of some sort of graded nervous structures within the brain must be pronounced wholly improbable. In the same way, the idea that the act of spatial vision is effected by a direct projection of the retinal image upon elements of the visual centre, arranged in mosaic on the analogy of the rods and cones of the retina, could hardly be reconciled either with our knowledge of the relations of the optic conduction to other, and more especially to motor paths, or with the observations made upon the structure of the visual cortex. In this sense, then, the argument from anatomy to physiology forms a useful preparation for the considerations of the following Chapter. Now that we have concluded our discussion of the conduction paths which are of particular importance for the psychological functions of the nerve centres, we may, accordingly, pause to point out the main lines of interpretation suggested by the general results of the preceding enquiry, and to attempt their formulation in certain laws or principles.

We head our list with the principle of manifold representation. This first and most general principle was formulated long since by MEYNERT, himself the first to make a systematic study of the microscopical structure of the brain. It declares that, as a general rule, every region of the bodily periphery which is controlled by the central organ has not one but several means of representation at the centre. In other words, if we have recourse to the analogy of a mirror and its images -- an analogy, however, which, as we shall soon see, cannot really be carried through -- it says that every [p. 226] sense organ and every organ of movement, together with every least part of such an organ, every sensory or motor element, is reflected not once only, but several times, in the central organ. Every muscle, e.g., has its proximate representation in the myel, from which (under the right conditions) it may be stimulated, or its excitation inhibited, without the interference of higher central parts. It is then represented a second time in the regions of the mesencephalon, the quadrigemina or thalami; and again, a third time, in the centromotor regions of the cerebral cortex. Finally, we must assume that it is indirectly represented in the parts of the cerebellum, and in the association centres, with which these regions stand in connexion. Now it is by no means necessary that the whole group of central representatives shall co-operate in every discharge of function by the peripheral elements. On the contrary, there can be no doubt that the lower centres are often able to exercise an influence upon the peripheral organs, in which the higher representations are not involved at all. When, however, the peripheral effect is the result of activity in the higher neurones, the excitation, under ordinary circumstances, is either mediated directly by lower centres, or at least arouses concomitant excitations in them. In this sense, therefore, the principle of manifold representation appears as an immediate consequence of the complex nature of all central functions. At the same time, it shows that there is an ascending progression in the co-operation of the various central representatives of one and the same peripheral region, and thus leads at once to the following second principle.

The central organs of the higher vertebrates are, very evidently, subject to a law of ascending complication. The number of branch paths, and therefore of the relations mediated by them between centres which, while functionally distinct, are still somehow interrelated by the needs of the organism, increases rapidly as we pass from below upwards. In the myel, the main lines of conduction are brought together compactly in the peripheral nerves; and the connection between principal path and secondary paths is of a comparatively simple and limited character. In the oblongata and the mesencephalon, these connexions begin to show a considerable increase, both in number and complexity. In the mesencephalic portion of the acoustic and optic paths, for instance, we find that the connexions with motor and with other sensory centres, which in the myel were arranged on a relatively simple pattern, are repeated in a very much more elaborate and complicated form. This ascending progression of conductive connexions reaches its final term in the cerebral cortex. Every part of the cortex, however diverse its proximate connexions and however distinct [p. 227] its proper function, is the meeting-place of conductive systems of the most varied kinds; so that what we term a 'visual' centre, e.g., always possesses something of the 'motor,' and something even of the 'associative,' along with its sensory character. It is, therefore, a corollary from this law of increasing complexity of representation in the ascending direction, as bearing more particularly upon the functions of the cortex, that every cortical area in the brain of man and of the higher animals is, in all probability, itself the seat of a manifold representation. Every portion of the visual cortex, that is, will contain, besides the representation of a part of the peripheral retina, further representations of motor areas connected with the function of sight and, possibly, of other, functionally related sensory areas; and finally, in all likelihood, indirect representations, mediated by the association fibres, of more remote functional centres that are again in some way concerned in the act of vision. Hence the idea that a sensory centre is, in essentials, nothing more than a central projection of the peripheral sensory surface, -- the visual centre, e.g., a projection of the retina, the auditory centre a projection of the 'resonance apparatus' of the labyrinth, -- even if it were admissible on physiological and psychological grounds, could hardly be held in face of the grave objections arising from the anatomical facts.

This principle of increasing differentiation in the ascending direction enables us to explain a further fact, suggested by the results of the gross anatomy of the brain, but brought out with especial clearness by histological examination of the conduction paths and by the phenomena of function that we describe in the following Chapter. This is the fact that many, perhaps most of the functions which in man and the higher mammals are finally integrated and co-ordinated in the cerebral cortex, appear in the lower vertebrates to be completely centralised in the mesencephalic ganglia: so, more especially, certain sensory functions, such as sight and hearing. Even in the lower mammalian orders, e.g. the rodents, the cortical representations of these organs do not attain anything like the extent and the functional importance that they possess in man. That is to say, the central organ provides itself with new representations, only in proportion as a more complicated co-ordination of functional units becomes necessary. When this happens, there is a relative reduction of the existing central stations in the same degree. This accounts for the comparative insignificance of the mesencephalic region in the brain of the higher animals and of man.

At this point the question naturally arises, whether or not the investigation of nervous conduction has furnished any evidence of specific differences [p. 228] in the functions of the central elements and of their conductive processes. We answer it by saying that one, and only one such difference may probably be inferred from the anatomical and physiological relations of the conduction paths. This is the difference in direction of conduction, connected with the twofold mode of origin of the nervous processes, which was first suggested from the anatomical side by RAMON Y CAJAL, and which receives confirmation from certain elementary facts of nerve mechanics (p. 99). In the older physiology, the establishment of determinate directions of conduction was ascribed to the nerves themselves, though it could hardly be brought into intelligible connexion with the properties of the nerve fibre. We are now able to refer it to a peculiar process of differentiation in the nerve cell. The explanation has been given above, in Ch. III. Every cell, as we there set forth, is the seat of excitatory and inhibitory processes, which under the influence of this differentiation are distributed in different proportions to definite cell regions. Originating in this way, however, the principle of different directions of conduction can hardly be looked upon as a law of universal validity. It is rather a principle of development, entirely compatible with the persistence of an undifferentiated condition in certain of the central elements. We have, as a matter of fact, found this condition to obtain in various types of cells, in which the twofold mode of origin of the conduction paths is neither anatomically proved nor physiologically probable. And it is noteworthy that these cells always occur in situations where the functional requirements do not include a differentiation of the directions of conduction, or rather -- for this is really the more correct expression -- where there is no demand for inhibitions of an excitation that has come in from a given direction. Thus, the differentiation is more or less doubtful in many cells of the sensory system, from those of the spinal ganglia onwards; and there are many intercalary cells, some lying within the central organs, some displaced far outwards in peripheral sense organs, which physiologically give no ground whatever for the assumption of a definite direction of conduction, and morphologically offer no sign of a twofold mode of origin of their processes. Cf. above, pp. 158 f.

The differentiation of the directions of conduction is, then, the result of a process of differentiation peculiar to the nerve cell, and apparently connected with an especial modification of the cell structure. It is, at the same time, the sole form of functional difference that the investigation of the conduction paths has brought to light; and, we must add, the sole form that a simple determination of these paths can ever reveal to the investigator. For the enquiry has, of course, its definite limits. It can tell us where, between what terminal stations, and (under favourable conditions) in which directions the processes are conducted; but it cannot [p. 229] tell us anything of the nature of the processes themselves. Nevertheless, it is a point of importance for the physiology of the central functions that, apart from this differentiation of the directions of conduction, no qualitative differences in the central elements can be demonstrated by morphological methods or inferred from the mechanics of innervation.

We have yet to mention a circumstance from which, in very many cases, the principle of the manifold representation of peripheral areas undoubtedly derives its peculiar significance: the fact that bodily organs which lie more or less widely apart, but yet function in common, are oftentimes brought into spatial as well as into functional connexion in their central representations. This means, of course, that the integration of functions may be mediated with the least possible circuity by conduction paths running directly between the central stations. Thus, the nerve paths that are brought into action in the locomotor movements of man and the animals issue from the myel at very different levels. But there are several places in the central organ (mesencephalon and cerebral cortex) where the centres for these movements lie close together; so that a suitable co-ordination can be effected, for voluntary as for involuntary, purely reflex movements, along well-trodden paths of cross connexion between neighbouring centres. In the same way, the intimate relation that obtains between the rhythm of auditory ideas and the rhythm of bodily movements becomes intelligible when we remember the number and variety of the connexions, at comparatively short distance, between the acoustic and the motor centres. It would seem, then, to be one of the most important features of conduction within the central organ at large that it serves, literally, to centralise: it unites the various functional areas, and thus renders possible an unitary regulation of functions that are separated in space but belong together in the service of the organism. And this means, further, that all the separate functions distinguished by us, since they are known solely in this their centralised form, must in reality themselves consist of an union of many functions, distributed over different, in many cases over widely remote peripheral organs. We may, therefore, reject without discussion any such view of the conductive connexions of the central organ as maintains, e.g., that there is a central act of vision, independent of motor innervations and of the mutual relations of different retinal elements: for the central organ of vision is not a mere projection of the retina upon the cerebral surface, but an extremely complicated structure, in which all the partial functions concerned in the visual function find representation. And we may reject, similarly, any theory that pro-[p. 230]poses to isolate the rhythmical form of successive auditory impressions, as a form of excitation peculiar to the auditory centre, and thus to separate it from the associated motor impulses. Every psychophysical function that falls under our observation is already, in point of fact, a centralised function, i.e. a synergic co-operation of a number of peripheral functions. What the retina or the peripheral auditory organ could contribute of itself to the formation of our perceptions, we do not know, and we can never find out: for the functions of eye and ear and of all other organs come under our observation always and only in this centralised form, i.e. as related to the activities of other functional areas.

Among these combinations of remote peripheral organs for unitarily centralised, synergic functions, a place of special importance is taken by the connexions dependent upon decussation of the paths of conduction. In their case, the separation and rearrangement of the paths are carried, by their passage to the opposite half of the brain, to the highest conceivable point; and, as a result of this, the functional significance of such central rearrangement is shown with the greatest possible clearness. Among the decussations themselves, that of the optic nerve, which appears in well-marked developmental sequence through the whole animal kingdom, shows the most obvious relations to the visual function. Where the compound eye occurs in its earliest stages of development, as in the facetted eyes of the insects, the retinal image forms a rough mosaic whose spatial arrangement -- since every facet represents a relatively independent dioptrical structure -- corresponds to that of the external object; what is above and below, right and left, in the object has precisely the same position in its image. Such an eye, therefore, if it has a muscular apparatus at its disposal, does not move about a point of rotation situated within itself, but is seated upon a movable stalk, i.e. turns (like a tactual organ) about a point lying behind it in the body of the animal. Under these conditions, it does not appear that the optic paths undergo any appreciable decussation; indeed, it is characteristic of the invertebrates at large that the great majority of the nerve paths remain upon the same side of the body. When, on the other hand, we turn to the lowest vertebrates, we are met at once by a complete reversal of the picture; the optic paths are now entirely crossed, so that the retina of the right eye is represented exclusively in the left, that of the left eye in the right half of the brain. RAMON Y CAJAL conjectures, with great acumen, that this arrangement may serve to compensate the inversion of the retinal image effected by the dioptrical apparatus of the vertebrate eye: he reminds us that the eyes of the lower vertebrates are, as a general rule, set laterally in the head, and that they accordingly cannot furnish a common image of the objects seen, although their two images may supplement each other in the sense that the one eye sees parts [p. 231] of an extended object which remain invisible to the other. The hypothesis also furnishes an explanation of the fact that the decussation changes from total to partial in proportion as the eyes, placed in front of the head, acquire a common field of vision, -- as they do in many of the higher mammals, and more especially in man. RAMON Y CAJAL, however, bases his interpretation of these facts upon the assumption that the retinal image is directly projected upon the visual cortex. Suppose, e.g., that the eyes are so situated, laterally, that the right eye images precisely the half ab, the left the half bc of the object abc (Fig. 97) It follows at once, from the fact of inversion, that the two retinal half-images aß and ßg  are out of their right positions: for if we regard ßg  as the direct continuation of aß, ß should be joined to ß, and not g  to a. If, now, the optic fibres undergo a total decussation, this incongruity will disappear in the projection on the central visual surface: the two halves of the image can be put together as half-images in just the same relation that they sustain in the external object (a'ß', ß'g').

RAMON Y CAJAL believes that the decussation of the optic nerves, thus necessitated by the optical construction of the eye, has formed the point of departure for all the further decussations of conduction paths; that it was followed, first of all, by the crossing of the motor paths of the ocular muscles, and then by that of the other, sensory and motor paths, correlated with these.[77] The hypothesis is ingenious in the extreme; and it is entirely probable that optic decussation and binocular synergy are closely interrelated. Nevertheless, the theory cannot be carried through in its present form. It is based upon assumptions that conflict not only with everything else that we know of the nature of the act of vision, but also, in the last resort, with everything that we know of the character and course [p. 232] of the conduction paths and of their terminations in the brain cortex. The retina itself is, as we have remarked above, a part of the central organ that has been pushed outwards to the periphery. It is, then, first and foremost, somewhat surprising that the disorientation of the image on the retina should be of no consequence, and its derangement on the cortex seriously disturbing. Those who hold such a view evidently rest it upon the belief that consciousness resides directly in the cortex, and there takes cognisance of an image of the external world, which must therefore, at this point, exactly repeat the real position of the external objects. That certain difficulties arise from the presence of the cortical gyres is admitted by RAMON Y CAJAL himself. To meet them, the further hypothesis must be made, that the disorientation of the images produced in each individual brain by the convolution of its outer surface is compensated by a remarkably accurate adaptation in the distribution of the crossed fibres. But then there is still another difficulty. If the image on the central visual surface corresponds exactly to the spatial properties of its object, then we must expect not only that the asymmetry arising from inversion of the image is binocularly compensated by the decussations in the chiasma, but also that, in each separate eye, there is an analogous compensation with regard to the vertical dimension. What in the retinal image is above, must in the visual centre be below, and conversely. The right and left decussation of the optic fibres would then be accompanied, in every optic nerve, by a second, vertical decussation. But this has not been demonstrated. Even in cases of what is called cortical hemianopsia in man, its existence has never been suspected. On the other hand, it is noteworthy that this cortical hemianopsia is much more obscure in its symptoms than the hemianopsia observed in cases of interruption of the fibres in the optic tract and thalami; minor defects, in particular, may pass without symptoms of any kind, or may be connected simply with a diminution, not with complete abrogation of sensitivity to light.[78] This, however, is just the opposite of what we should expect, if an undisturbed reproduction in the visual cortex of the space relations of the object were the one thing necessary for the act of vision. Moreover, we must remember that there is a very much simpler and more plausible explanation of the fact that eve see objects upright, despite the optical inversion of their images, than would be given with this hypothesis of a vertical decussation. Wherever the visual organ has become a dioptrical apparatus, involving inversion of the image, the point of ocular rotation lies, not behind the eye in the body of the animal, as it does in the stalked eyes of the invertebrates, but in a point d within the eye itself (Fig. 98). Hence, when the central point of vision in the yellow spot moves in the retinal image from below upwards [p. 233] in the direction aß, the external point of fixation in the object moves from above downwards, in the direction a b. That is to say, the displacement of the point of rotation to the interior of the eye brings with it a direct compensation of the inversion of the image. For we estimate the space relations of objects in terms of the position and movement of the line of fixation before the point of rotation, not behind it, and not either in terms of the retinal image -- whose position is really just as little known to us as are the space relations of the hypothetical image in the visual centre: and of that we do not even know whether or not it exists at all. Intrinsically, it is, without any question, far more probable that we should assume in place of such an image a system of excitations, corresponding to the various sensory, motor and associative functions simultaneously concerned in the act of vision.

This compensation of the inversion of the retinal image by the motor mechanism of the single eye evidently furnishes the nearest analogy for the reciprocal orientation of the right and left retinal images as it occurs in binocular vision. Here, too, we must suppose, the motor mechanism has not adapted itself, after the event, to the relation obtaining between the two hypothetical images in the visual centre, but has from the first exercised a determining influence as regards orientation within the common field of vision. Now for the visual organ with laterally placed eyes, the fundamental characteristic of the field of vision is that it consists of two entirely different halves; in the ideal form of such an organ, the separate fields will just meet in the middle line. Vision of this sort may fitly be termed, with RAMON Y CAJAL, 'panoramic' (as opposed to 'stereoscopic') vision. It covers a wide range; but it mediates only a superficial image, and gives no direct idea of the third dimension. Further, a correct orientation of the two halves of the panoramic image is possible only if an object that travels continuously from the one half of the field of vision to the other shows no discontinuity in its movement; and this condition, again, is fulfilled only if similarly situated eye muscles are symmetrically innervated as the movement continues. If the object has been followed, e.g. by the line of regard of the right eye in Fig. 97 from a to b,, then the movement must be carried on, without interruption, from b to c, by innervation of the line of regard of the left eye; that is to say, the innervation of the left rectus internus, whose direction of pull is indicated by [p. 234] the dotted line i2, must be so co-ordinated with the innervation of the right rectus internus, i1, that it promptly relieves its predecessor, to give way in its own turn to the innervation of the left externus, e2. Now there is, as we have said, no ground for supposing that the visual centre is the scene of any kind of pictorial projection, even remotely resembling that which we have on the retina. It is, however, not improbable that the mechanisms of release, by means of which sensory are transformed into motor impulses, are here arranged in a certain symmetry: in such a way, that is, that if e.g. the road to the rectus internus is thrown open in the visual cortex of the right hemisphere at a definite point e', it must in the left hemisphere be thrown open at a point e'', functionally co-ordinated with e' and lying symmetrically with it to the median plane. This arrangement will, naturally, come about by process of development. Let us take, as the first term of the series, a state of affairs where the eyes of the two sides stand in no sort of functional relation to each other: the condition seems, as a general rule, to be actually realised, e. g. in the visual organs of the invertebrates. Here, we must imagine, these mechanisms of release, like all the other central elements, will in each optic ganglion be arranged symmetrically to the median plane: whatever lies farthest to the right on the right-hand side of the body will do the same on the left side, and conversely. Suppose, again, that the development is carried a step further, and that the eyes are to co-operate for a panoramic vision of the kind described above. The symmetrical arrangement would now be insufficient; it would prevent the regular sequence of the movements of the two eyes, required for an adequate apprehension of objects. The arrangement in the visual centre of the hypothetical mechanisms of release is, of course, unknown to us, apart from the probability of their symmetry with reference to the median plane; fortunately, it is also, for our present question, a matter of indifference. We will assume, for simplicity's sake, that the points of release lie inwards for the interni and outwards for the externi. Then, if the object is followed in fixation from a to c, the externus e1, will first be innervated from a central point e'. As the internus i1, comes into action, the central innervation will move from e' to i'. Next, upon the entrance of the object into the visual field of the left eye, it will pass without interruption to i2, i'', and thence, finally, to e2, e''. If there were no decussation, the central point of release corresponding to a point upon the nasal half of the retina would, on the contrary, lie inwards, and the point of release corresponding to a point upon the temporal half lie outwards, on both sides of the median line. The arrangement of the points, from right to left, would then be i' e' e'' i'', and the innervation would first of all travel on the right, from within outwards, and then shoot across to the left visual centre, to execute a similar movement there. We must add that the [p. 235] arrangement which controls movement naturally determines as well the localisation of the resting eyes; so that bc is seen as the direct continuation of ab. The reason, once more, is not that this is the order in which images of the separate points are projected upon the brain, but that it is the order in which there is congruity of the sensory and motor functions that work together in every instance of spatial perception. Presumably, therefore, the total decussation of the visual organ with panoramic function does not represent an arrangement which was first found good on the sensory side and later extended to the motor. It must rather be regarded as an arrangement which, from the outset, applies to both sensory and motor areas, and which mediates their co-operation.

We are, then, justified in positing an organisation that extends over the entire sensorimotor system and is determined by the necessary co-ordination of particular sensory and motor points. The hypothesis has the further advantage that it, and it alone, can adequately explain the obvious connexion between the change from total to partial decussation of the optic fibres, on the one hand, and the passage from panoramic to stereoscopic vision on the other. Suppose that we have reached the stage at which the eyes are set so far forward that their fields partially overlap, to form a common field of vision, embracing the same objects. The conditions are now very different from those obtaining in panoramic vision. The synergy of the eye movements, in so far as it is controlled by the nearer objects in the common field of vision, is no longer laterally symmetrical, no longer such, i.e., that right corresponds to right and left to left upon the two sides, but has become medianly symmetrical, so that those points are homologous that lie on either side at the same distance from the median plane. Stated directly in terms of eye movement, this means that the synergy in panoramic vision is that of parallel, in stereoscopic vision that [p. 236] of convergent movement of the lines of sight. The difference will be clear at once from a comparison of Fig. 99 with Fig. 97. In Fig. 97 where we have a representation of the limiting case in which the two fields of the laterally placed eyes pass into each other without break, the point of contact b is the sole point seen in common by both halves of the visual organ. If the object is given the direction in the third dimension indicated in Fig. 99, only this one point upon it remains visible to the two eyes. For the visual organ with a common field of vision (Fig. 99) the case is different. Here the object, despite its position in the third dimension, remains entirely visible to both eyes; the image ag  upon each retina corresponds to that side of it which, as it recedes in space, is turned towards the eye in question. Along with this change in scope of vision goes, further, a change in the conditions under which the eyes must move in order to bring all parts of the object upon the yellow spot. The movements are not touched off successively, as in the previous instance, -- the left eye taking up the movement at the point where the excursion of the right eye reaches its limit, -- but have become simultaneous: right and left eye range over the object, from a to c and back again from c to a, both at once. And the muscles co-ordinated for the purposes of these movements are muscles placed symmetrically not as regards external space, but as regards the median plane of the body and, consequently, as regards each separate eye: internus and internus, externus and externus. This means, once more, that there is a radical change in the conditions under which the movements are touched off centrally by light stimuli. The farther inward, towards the median plane, a retinal point lies in either eye, the farther out, in the third dimension of space, lies the corresponding point of the object situated in the common field of vision. If, therefore, a point is stimulated upon the nasal halves of the two retinas, the interni are brought into action at i' and i'', and a movement is produced by symmetrical increase of convergence in the direction ca. If, on the other hand, a retinal point is stimulated on both sides that approximates to the near limit of the field of vision, the binocular organ will traverse the object with symmetrically decreasing convergence; the two externi are brought into synergic action at e' and e''. Recurring to our hypothesis that the elements in the central organs are arranged upon a principle of median symmetry, we may say, therefore, that the requirements of symmetrical convergence will be fulfilled if the mechanisms for the release of motor impulses by light stimuli are distributed symmetrically to the median plane in the visual centre of the same side. Since distant points in space correspond to retinal points that lie nasalward, and near points in space to retinal points that lie temporalward, the arrangement of the mechanisms of release will be in conformity with the medianly symmetrical disposition of the parts of the brain if the musculi [p. 237] interni are represented in the brain, too, on the inside, nearer the median plane, and the musculi externi on the outside. Let us assume that this arrangement actually holds. Then the inversion of the image, due to the dioptrical structure of the eye, is precisely the position that will satisfy the needs of stereoscopic vision and of the mechanism of convergence. It is, accordingly, a condition of the adaptation of the visual organs to these needs that the optic paths remain uncrossed for the whole extent of the common field subserving stereoscopic vision. But the common field is only a part of the total field of vision. In man himself and in the animals with a like visual endowment each eye has, besides, its own particular field, represented by the inner portions of the retinas, and governed, of course, by the conditions of panoramic vision and the special laws of eye movement that they bring in their train. Here, then, the two visual organs are laterally symmetrical, and the necessary decussation of the optic paths -- and, with them, of their centromotor releases -- is effected on the pattern of Fig. 97. In this sense, therefore, the partial decussation of the optic paths gives an accurate picture of the state of affairs resulting on the one hand from the co-operation of the two eyes in binocular vision and on the other from the coordination of their independent functions. We may add that, in all probability, the decussation in the chiasma possesses this functional significance not only for the terminations of the optic paths in the visual cortex, which we have had primarily in mind in the foregoing discussion, but also for the terminations in the mesencephalic region. According to the plan of the opticus conduction laid down in Fig. 78 (p. 186), the two centres are similarly disposed in the essential matter of sensorimotor connexions.

There are many other instances of the decussation of conduction paths, recurring at all levels from the myel upward. In no case, however, is the functional interpretation of the phenomenon as obvious as it is in that of the opticus crossing. Nevertheless, we have no right to conclude that all the other decussations are simple consequences of this transposition of the optic fibres. On the contrary, the same synergy that is apparent here obtains also for the other organs of sense and of movement, and more particularly for the relations between sensory excitation and motor reaction, and may therefore lead independently at any point to analogous results -- though the results, once produced, may very well reinforce and support one another. This relative independence of the different decussations is further suggested by the fact that in the lower vertebrates, where the optic decussation is total, other decussations, as e.g. that of the motor paths of the skeletal muscles, are much less complete than they are in man. Moreover, in all the vertebrates up to man, the myel evidently retains the character of a central area in which the conduction paths remain for [p. 238] the most part upon the same side of the body, whereas the oblongata shows at once a large number of partial decussations, due to the correspondingly large number of motor functions of bilaterally symmetrical nature -- movements of respiration, of mastication, of swallowing; mimetic movements -- that have their centres in this region. Similar decussations occur also in the olfactory and acoustic areas, where they are again, in all probability, connected with motor synergies.

There is, finally, a further circumstance, affecting those cerebral areas which, as seats of the more complex functions, receive association paths from various sensory and motor centres, that presumably stands in intimate relation to the phenomenon of decussation: the circumstance that certain centres, which exist potentially in both hemispheres, are preponderantly developed upon one side. This applies especially to the 'speech centre,' which we discuss in the following Chapter. In the great majority of cases, the speech centre has its principal seat in certain frontal and temporal areas of the left hemisphere. Since this left side, in consequence of the decussations of the motor paths, contains the centres for the motor innervation of the right half of the body, we may suppose that the arrangement is connected with the disproportionate development of the muscular system on the right and, more particularly, with the right-handedness of the ordinary man. The latter phenomena may themselves be referred partly to the character of human movements, which involve a preference for the one side, and partly to the asymmetrical position of other bodily organs, especially the heart.

The study of the conduction paths has been dominated, up to the present time, by the view, natural to the adherents of a strict localisation theory, that every bodily organ must find its representation at some point of the brain cortex. The other and at least equally tenable idea, that all parts of the brain and, very especially, all parts of the brain cortex are intended to mediate the interconnexion of different conduction paths, has, in consequence, been forced unduly into the background. The strict localisation theory, as held by H. MUNK and by other physiologists and pathologists, assumes that the surface of the brain is made up of a number of sensory centres, which are in reality simple reflections or copies of the peripheral sensory surfaces; so that e.g. every point upon the retina has a corresponding point in the visual cortex. Secondarily, it is true, recourse is had to the subsidiary hypothesis (mentioned above, pp. 201 ff.) that the immediate neighbourhood of these direct sensory centres is taken up with special ideational centres, to which the direct sensory impressions are in some way transferred: this hypothesis is to explain the origination of memory ideas. The point of view persists, practically unchanged, in FLECHSIG'S and RAMON Y CAJAL'S theory of specific 'association' centres. For it appears to be the opinion of these authorities that, if the more complicated functions are thus handed over to independent centres, the sensory centres proper are for that very reason all the more securely established in their character as direct reflections of the bodily periphery. As a matter of fact, however, this view is [p. 239] sufficiently refuted, even apart from the complex nature of the disturbances consequent upon cortical lesion, by MEYNERT'S principle of manifold representation. It is unjust to what is by far the most important aspect of central organisation, the combination into an unitary resultant of component functions that are oftentimes separate at the periphery. GOLTZ has been, from the first, its earnest opponent; and in so far as it leaves altogether out of account the true character of the central functions, his opposition is justified. On the other hand, he and his school have plainly gone too far in denying localisation of function outright. Localisation, in a certain sense, is a direct corollary from the fact that centralisation is never universal, but is always confined primarily to the integration of a limited number of components. Within these limits, the course and distribution of the conduction paths point, without any question, towards a certain localisation, though they and, in the last resort, the structure of the cerebral cortex itself, point not less clearly to an interconnexion of functions at all parts of the organ. If we look at the facts as a whole, we may safely say -- while abstracting entirely from functional derangements, whose interpretation is oftentimes doubtful -- that modern brain anatomy has furnished overwhelming proof that the older idea of the brain cortex, as a reflection or copy of the totality of the peripheral organs, supplemented at most by a few special areas reserved for higher psychical needs, is altogether untenable.

The problem of the decussation of conduction paths, in the strict sense of the term, is less difficult of solution than that of the unilateral representation of function, such as has been demonstrated, in particular, for the functions involved in speech. Originally, it would seem, the function in these cases was symmetrically disposed upon both halves of the brain, but practice has for some reason been predominantly unilateral, and the one half has consequently gained the ascendancy. It is natural to connect this difference of development with the right-handedness of the majority of mankind. And the connexion is sustained by the fact that, in various instances, derangements of speech have been observed in left-handed persons as a result of apoplectic effusions in the right half of the brain.[79] Now it is not difficult to explain this relation, if we once accept the possibility of unilateral development. In civilised man, who is right-handed, writing is a function of the right hand; and the associations that colligate the various speech functions are so many and so varied that the unilateral development of writing alone might bring with it a corresponding unilateral localisation of the related factors. But since men write, as a rule, with the right hand, and are in general more practised in the mechanical control of the right hand than of the left, the left half of the brain must, by reason of the decussation of the motor conduction paths, receive a larger measure of practice than the right. This practice will, of course, be shared by the speech centre. The numerous instances of restoration of the speech functions, with persistence of the central lesion, may then be referred, along with all the other possible forms of vicarious function, to the substitution under special conditions of the right for the left half of the brain. We have a similar substitution in the case of the external organs: a patient who is paralysed on the right side is able, by practice, to use his left hand for actions previously performed by the right, e.g. for writing; and the change of habit at the periphery naturally carries with it a new course of practice at the centre. But when all this is granted, the initial question still calls for answer; the question why the right-hand side of [p. 240] the body should ever have been preferred at all. In seeking to answer it, we must bear in mind that most mechanical functions, to a certain extent even that of walking, make a heavier demand upon the one side of the body than upon the other; and that, under this condition, the right side is naturally marked out for special favours by the general asymmetry in the position of the organs of nutrition in the higher animals. Here again, the plan of arrangement is, as we all know, governed by the close interdependence of the individual organs. The placing of the liver on the right means that the great reservoirs of venous blood also lie on the right, so that the arterial system is necessarily relegated to the left. In the rare cases in which this disposition is reversed (cases of what is called situs transversus viscerum), it is the rule that all the asymmetrical organs are involved in the rearrangement of parts. Now the central organs that stand in greatest need of protection are the circulatory organs. Consequently, most mammals in combat with their enemies are apt to put their right side in the forefront of the battle; and this habit must react favourably, by way of stronger development, upon the muscles of the right side of the body. In man, the upright position brings with it a special need for the protection of the central organs of circulation, and at the same time helps to render this protection easy and efficient. On the other hand, it is probable that the left-handed situation of the circulatory organs has furthered the development of the left-hand side of the brain.[80] Since, then, the more developed half of the brain must correspond to the more developed half of the body, it is on the whole intelligible that the peripheral paths of the right side should in the main be represented in the left half of the central organ, and those of the left side in the right. On this assumption, it is possible that the decussation of the pyramidal paths in man and the mammals may itself be a simple consequence of the asymmetrical practice imposed by outside conditions upon the bodily organs and their central representatives.

[1] Cf. the general discussion of reflex excitations in Chap. iii., pp. 85 ff. above.

[2] The most fruitful of these methods may find mention here. They are: GOLGI'S method of impregnation by a metal, and more particularly by silver; and the methods of staining by haematoxylin and methylene blue, introduced respectively by WEIGERT and EHRLICH. For details regarding these and other methods, see OBERSTEINER, Anleitung beim Studium des Baues der nervösen Centralorgane, 3te Aufl., 1896, 7 ff., and EDINGER, Vorlesungen über den Bau der nervösen Centralorgane, 6te Aufl., 1900, 3 ff.

[3] FLECHSIG, Die Leitungsbahnen im Gehirn und Rückemark des Menschen, 1876. C. VOGT, Etude sur la myélisation des hémisphéres cérébraux, 1900. Cf. §§ 5, 6, below.

[4] L. TÜRCK, Sitzungsber. d. Wiener Akad., nth. math.-naturw. Cl., vi., 1851, 288; xi. 1853, 93. CHARCOT, Ueber die Localisationen der Gehirnkrankheiten, trs. FETZER, i., 1878, 159 (Lecons sur les localisations dans les maladies du cerveau, 1875); FLECHSIG, Ueber Systemerkrankungen im Rückenmark, 1878.

[5] GUDDEN, Arch. f. Psychiatrie, ii., 1870, 693. FOREL, ibid., xviii., 1887, 162. VON MONAKOW, ibid., xxvii., 1895, I, 386.

[6] MEYNERT, art. Gehirn in STRICKER'S Gewebelehre, 1871, 694 ff. (Buck's trans., 650 ff.); Psychiatrie, Pt. I. 1884. LUYS, Recherches sur le systéme nerveux cérébro-spinal, 1865. The Brain and its Functions, 1877 and later (Internat. Sci. Series).

[7] The olfactory, optic and acoustic nerves are purely sensory: the oculomotor, trochlear and abducent ocular, the facial and the hypoglossal are purely motor; and the trigeminal, glossopharyngeal and pneumogastric and accessory resemble the myelic nerves, i.e. become mixed at a short distance from their place of origin. The sensory root of these latter has a ganglion, which is wanting in the sensory nerves proper.

[8] C. BELL, An Idea of a New Anatomy of the Brain, 1811; An Exposition of the Natural System of Nerves, 1824; The Nervous System of the Human Body, 1830.

[9] This analogy appears quite clearly as early as JOHANNES MÜLLER. Cf. his Lehrbuch d. Physiol., 4te Aufl., i., 1844, 623. MÜLLER, however, leaves the question undecided. The first physiologist who expressed the decided opinion that the function of the nerves is determined solely by the organs with which they are connected was, as MÜLLER tells us, J. W. ARNOLD.

[10] LUDWIG and WOROSCHILOFF, Ber. d. sächs. Ges. d. Wiss., math.-phys. Cl., 1874, 296. MOTT, Philos. Transact., clxxxiii. 1892, I.

[11] VAN DEEN, in MOLESCHOTT'S Untersuchungen zur Naturlehre des Menschen, vi., 1859, 279. SCHIFF, Lehrbuch der Physiol., 1858, i. 238. PFLÜGER'S Arch. f. d. ges. Physiol., xxviii.; xxix. 537 ff.; xxx. 199 ff.

[12] LUDWIG and WOROSCHILOFF, op. cit., 297.

[13] H. E. HERING and C. S. SHERRINGTON, in PFLÜGER'S Arch. f. d. ges. Physiol., lxiii. 1897, 222. Cf. p. 93, above.

[14] Concomitant sensations with pain stimuli are described by KOWALEWSKY (trans. from the Russian in HOFFMANN and SCHWALBE, Jahresber. f. Physiol., 1884. 26); 'conjugate' sensations, with arousal of tickling, by E. STRANSKY (Wiener klin. Rundshau, 1901, os. 24-26). In my own experience, the larynx and the external auditory meatus are especially liable to concomitant sensations.

[15] PFLÜGER, Die sensorischen Functionen des Rückenmarks, 1853, 67 ff.

[16] FLECHSIG, Ueber Systemerkrankungen im Rückenmark, 30 ff. BECTHEREW, Die Leitungsbahnen im Gehirn und Rückenmark, 1899, 17 ff.

[17] CHARCOT, Leçons sur les localisations, etc., 145 ff. FLECHSIG, Ueber Systemerkrankungen, 42 ff. EDINGER, Vorlesungen, 6te Aufl., 86, 358

[18] EDINGER, op. cit., 317. RAMON Y CAJAL, Beitrag zum Studium der Medulla oblongata, 1896, 52 ff.

[19] Cf. RAMON Y CAJAL, Medulla oblongata, 43, 122 ff. EDINGER, op. cit., 86 ff.

[20] EDINGER, op. cit., 272. BECHTEREW, Leitungsbahnen im Gehirn und Rückenmark, 439.

[21] This exposition follows HELD, Arch. f. Anatomie, 1893, 201 ff.

[22] BECHTEREV, Die Leitungsbahnen, 169 ff.

[23] For a detailed account of the terminal nervous apparatus of the retina, cf. below, Ch. VIII § 4.

[24] RAMON Y CAJAL, Die Structur des chiasma opticum. Trans. J. BRESLER. 1899.

[25] GUDDEN, Arch. f. Opthalmologie, xxv., I. VON MONAKOW, Arch. f. Psychiatrie, xxiii., 1892, 619; xxiv., 229. GILLET and VIALET, Les centres cévébraux de la vision, 1893.

[26] MONAKOW, Arch. f. Psychiatrie, xx., 1889, 714 ff. BECHTEREW, Die Leitungsbahnen, 199 ff. RAMON Y CAJAL, Les nouvellas idées sur la structure du systèmenerveux, 1894; Studien über die Hirnrinde des Menschen, i. Die Sehrinde, I 00. Besides the reflex path rr to the nidi of the oculomotor nerves, there is another, which mediates the reaction of the pupil to light stimulation. Its course has not yet been fully made out; but it seems to be altogether divergent from the other, since the pupillary reaction is not abrogated by destruction either of the pregemina or of the genicula. Cf. BECHTEREW, op. cit., 215.

[27] I avoid the use of the simpler terms 'motor' and 'sensory,' in order to indicate -- from the outset, the essential difference that obtains between the conditions of conduction here and in the peripheral nerves. The other terms in current use, 'psychomotor' and 'psychosensory,' seem to me to be objectionable, for the reason that they suggest a participation of consciousness or of mental functions which, to say the least, is hypothetical. It must also be remembered that there are many central parts besides the cerebral cortex, e.g. the brain ganglia, that are also endowed in certain measure with the properties under discussion.

[28] FRITSCH and HITZIG, Arch. f. Anat. u. Physiol., 1870, 300 ff. HITZIG, Untersuchungen über das Gehirn, 1874, 42 ff. FERRIER, The Functions of the Brain, 2nd ed., 1886.

[29] LUCIAN, Arch. ital. de biologie, ix., 268.

[30] FRANCK, Les centres motrices du cerveau, 1887.

[31] FRANCK and PITRES, Arch. de physiol., 1885, 7, 149. BUBNOFF and HEIDENHAIN, Pflüger's Arch. f. d. ges. Physiol., xxvi., 137.

[32] LUCIANI and SEPPILLI, Die Functionslocalisation auf der Grosshirnrinde, 289 ff, (Le localizzazioni funzionali del cervello, 1885). HITZIG, Berliner klin. Wochenschrift, 1886, 663.

[33] On this point, cf. in particular GOLTZ, Ueber die Verrichtungen des Grosshirns, 1881, 36. 119 ff.

[34] GOLTZ, PFLÜGER'S Arch. f. d. ges. Physiol., xxvi., 170 ff.; xxxiv., 487 ff. LUCIANI and SEPPILLI, op. cit. (German), 50 ff.

[35] SCHITSCHERBACK, Physiol. Centralblatt, v., 1891, 289 ff.

[36] HITZIG, Untersuchungen über das Gehirn, 126 ff.

[37] FERRIER, The Functions of the Brain, 1886, 235 ff.

[38] SCHÄFER, Beiträge zur Physiologie. C. LUDWIG gewidmet, 1887, 269 ff.

[39] HORSLEY and BEEVOR, Philos. Transactions. 1890, clxxix., 205; clxxxi., 129.

[40] H. MUNK, Ueber die Functionen der Grosshirnrinde, 2te Aufl., 1890. Berichte der Berliner Akademie. 1892, 679; 1893, 759; 1895, 564; 1896, 1131; 1899, 936.

[41] Ueber die Verrichtungen des Grosshirns, Abth. i.-vii. In PFLÜGER'S Arch. f. d ges. Physiol., 1876-1892.

[42] Cf. p. 188, above.

[43] On the relation of eye movement and retinal sensation, see below, Ch. XIV § 2.

[44] E. A. SCHÄFER. Proc. of the Royal Soc., 1887, 408. BAGINSKY, Arch. f. Physiol., 1891, 227 ff.

[45] CHRISTIANI. Zur Physiologie des Gehirns. 1885, 31 ff. GOLTZ, in PFLÜGER'S Arch. f. d. ges. Physiol., li., 570 ff.

[46] GOLTZ, in PFLÜGER'S Arch. f. d. ges. Physiol., xxxiv., 459, 487 ff. CHRISTIANI, op. cit., 138 ff.

[47] FERRIER, in Brain, 1881, 456; 1884, 139. LOEB, in PFLÜGER'S Arch. f. d. ges. Physiol., xxxiv., 88ff. LUCIANI and SEPPILLI, op. cit. (German), 145.

[48] For the criteria to be applied, cf. NOTHNAGEL, Topische Diagnostik der Gehirnkrankheiten, Einleitung.

[49] NOTHNAGEL, Topische Diagnostik, 438 ff. H. DE BOYER, Études cliniques sur les lésions corticales, 1879. EXNER, Untersuchungen über die Localisation der Functionen in der Grosshirnrinde des Menschen, 1881. VON MONAKOW, Gehirnpathologie, 1897, 282 ff.

[50] FERRIER, Localisation der Hirnkrankungen, 12 ff. (Localisation of Cerebral Disease, 1878). NOTHNAGEL, Topische Diagnostik, 549. VON MONAKOW, Gehirn. pathologie, 376 ff.

[51] NOTHNAGEL, Topische Diagnostik, 389. LUCIANI and SEPPILLI, op. cit. (German), 167 ff. VON MONAKOW, Gehirnpathologie, 445 ff.

[52] VON MONAKOW, Arch. f. Psychiatrie, xxiv., 229 ff. Cf. also the conditions found to obtain in the brain of LAURA BRIDGMAN, a deaf-mute, who became blind in early childhood. It may be remarked, further, that the whole development of this brain suggests an elaborate system of vicarious functioning, especially in the sensory region. DONALDSON, Amer. Journal of Psych., iii., 1890, 293; iv., 1892. 503 ff.

[53] Cf. the cases described by FÜRSTNER (Arch. f. Psychiatrie, viii., 162; ix., 90) and REINHARD (ibid., 147), and by VON MONAKOW, Gehirnpathologie, 468 ff.

[54] Cf. the following Chapter, and the account of visual ideas given in Ch. XIV.

[55] WERNICKE, Der aphasische Symptomencomplex, 1874. KAHLER and PICK, Beiträge zur Pathologie und pathologischen Anatomie des Centralnervensystems, 1879, 24, 182. LUCIANI and SEPPILLI, op. cit. (German). 217 ff. VON MONAKOW, Gehirnpathologie, 506 ff.

[56] FLECHSIG, Gehirn und Seele, 2te Aufl., 1896, 61; Die Localisation der geistigen Vorgänge, 1896, 34.

[57] EXNER, op. cit., 63 ff. LUCIANI and SEPPILLI, op. cit. (German), 321 ff. On disturbances of the sensations of movement, consult further NOTHNAGEL, Topische Diagnostik, 465 ff. VON MONAKOW, Gehirnpathologie, 362 ff.

[58] On this point see below, Ch. VI., §§ 5 and 6.

[59] See below, Ch. VI., § 7.

[60] CHARCOT, Leçons sur les localisations, etc. (Vorlesungen über die Localisation der Gehirnkrankheiten, 120 ff.). NOTHNAGEL, Topische Diagnostik, 581 f.

[61] VON MONAKOW, Gehirnpathologie, 364 ff.

[62] Arch. f. experimentelle Pathologie, iii., 1874, 171.

[63] MEYNERT, Psychiatrie, 1884, 145.

[64] Arch. f. Physiol., 1878, 171; Ueber die Functionen der Grosshirnrinde, 44.

[65] Cf. with this WUNDT, Zur Frage der Localisation der Grosshirnrinde, in Philos. Studien, vi., 1891 1 ff.

[66] FLECHSIG, Die Localisation der geistigeVorgänge, 13 ff. EDINGER, Vorlesungen, 6te Aufl., 161 ff.

[67] FLECHSIG, Neurol. Centralblatt, 1898, no. 21.

[68] Cf. DÉJERINE, Zeitsch. f. Hypnotismus, v., 1897, 343; O. VOGT, ibid., 347 ; SIEMERLING, Berliner klin. Wochenschrift, xxxv., 1900, 1033. See also p. 217, below.

[69] MEYNERT, in STRICKNER'S Gewebelehre, 117 (POWER'S translation, ii., 481 f.); Psychiatrie, 40.

[70] FLECHSIG, Gehirn und Seele, 2te Aufl., 1896; Neurol. Centralblatt, 1898, no. 21.

[71] RAMON Y CAJAL, Die Structur des chiasma opticum, 56. EDINGER, Vorlesungen, 6te Aufl., 228. HITZIG, Les centres de projection et d'association, Rapport lu au xiii. Congrès internat. de Med. à Paris, Le Nevraxe, i., 1900 (criticism by FLECHSIG, ibid., ii., and reply by HITZIG, 1900). VON MONAKOW, Monatsschrift f. Psychiatrie. viii., 1900, 405. O. VOGT, Journ. de physiol. at pathol. gén., 1900, 525. See above, p. 213.

[72] RAMON Y CAJAL, Studien über die Hirnrinde des Menschen. German trans, by BRESLER. Heft 1: Die Sehrinde; Heft 2: Die Bewegungsrinde, 1900. Comparative Study of the Sensory Areas of the Human Cortex, in Decennial Volume of the Clark University, 1899, 311 ff.

[73] MEYNERT, Vierteljahrsschrift f. Psychiatrie, i., 91, 198; ii., 88. Also in STRICKER'S Gewebelehre, ii., 704 ff. (POWER'S trans., ii. 381.)

[74] GOLGI, Sulla fina anatomia degli organi centrali, 1886.

[75] KÖLLIKER, Gewebelehre, 6te Aufl., ii., 809 ff.

[76] RAMON Y CAJAL, Studien über die Hirnrinde des Menschen, Heft i., 5 ff. Cf. also FLECHSIG, Die Localisation der geistigen Vorgänge, 82 ff.

[77] RAMON Y CAJAL, Die Structur des chiasma opticum, 22 ff.

[78] VON MONAKOW, Gehirnpathologie, 450.

[79] OGLE, Medico-chirurgical Transactions, liv., 1871, 279.

[80] According to GRATIOLET, the frontal gyres develope more rapidly on the left than on the right; in the occipital brain, the reverse order appears to obtain (Anatomie comparée dusystèmenerveux, ii., 242). GRATIOLET'S results are, however, questioned by ECKER (Arch. f. Anthroplogie, iii., 215). W. BRAUNE (Arch. f. Anatomie, 1891, 253) has also failed to find confirmation of OGLE'S statement that the left hemisphere is, almost without exception, heavier than the right. On the other hand, it is a fact easily verified that, in all primates, the fissures are more asymmetrically arranged in the anterior than they are in the posterior part of the brain. Moreover, the left frontal gyres, according to BROCA, are usually more complicated than the right. These observations accord with those made by BROCA and P. BERT upon the differences in temperature found in man at different parts of the head; the left half of the frontal region is on the average warmer than the right, and the frontal region as a whole warmer than the occipital (P. BERT, Soc. de biologie, 19 Janv., 1879).

[1] It should be said, with regard to Fig. 79, that the schema there given simply shows the relative positions of the optic fibres in the nerve and optic tract, with reference to the various parts of the retina, as they are to be inferred from the investigations of HENSCHEN (Brain, 1893). VIALET (Les centres de la vision, 1893) and others. The central arrangement of the fibres corresponds to this peripheral schema only in so far as the bundles proceeding from the macula lutea, find representation in the occipital cortex of both hemispheres. Cf. the account of the whole matter in BECHTEREW, Die Leitungsbahnen im Gehirn und Rückenmark, 209 ff. On the other hand, the definitive position of the lateral and median bundles in the cortical centres is, as we show below (pp. 206, 235) for the most part the precise opposite of their position in these proximate conduction paths; experiments on animals and pathological observations on man prove that the lateral portions of the retina are represented on the same, the median on the opposite side of the brain. In seeking to explain this fact, we must bear in mind that the proximate termination of the optic paths lies in the mesencephalic centres of the genicula, and that it is accordingly from this point that the final assignment of fibres to the cortex is made. In view of the complexity of the relations involved, and of the somewhat ambiguous symptoms which follow from injury to the cortex, many investigators, like HENSCHEN, and, to some extent, VON MONAKOW, have recently taken the position that the connexion of the various parts of the retina with the cortical centres at large has not yet been finally settled; and HENSCHEN is further inclined to restrict the visual centre to a limited area within the calcarine fissure (O' Fig. 65, p. 145), instead of allowing it the fairly extended region in the occipital cortex (Fig. 89, p. 206) to which it is usually referred. VON MONAKOW, however, while admitting that he is in doubt as regards the definitive correlation of retina and cortex, does not hesitate to express his conviction that the central representation of the various parts of the retina is, in any event, based upon their relation to the centres of ocular movement (Ergebnisse der Physiol., 1 Jahrg., 1902, 2 Abth., 600). This statement is, as the reader will see, in full agreement with what is said below (pp. 229 ff.) of the theory of decussations in general, and of the decussations of the optic paths in particular, as against the simple copy-theory of RAMON Y CAJAL. -- Later note by AUTHOR.