Such fibres are constituent elements of the facial, glosso-pharyngeal, and in some animals also the vagus cerebral nerves (Fig. 443), in connexion with the ganglia of which these epibranchial placodes are formed (Froriep and Streeter). The observations of Professor J. P. Hill upon embryos of Echidna seem to suggest that in mammals these gustatory neuroblasts are derived from the entoderm. When first formed, the neural tube is compressed from side to side and presents an elliptical outline in transverse section (Fig. 444). The two side walls are very thick, whilst the narrow dorsal and ventral portions of the wall are thin, and are termed the roof-plate and floor-plate respectively (Fig. 444). The cavity of the tube in transverse section appears as a narrow slit. The wall of the neural tube consists at first of low columnar epithelium arranged in a fairly regular series, but with a certain number of large spherical so-called germinal cells scattered between the columns. But this regular disposition as a single layer FIG. 444.-DIAGRAM OF TRANSVERSE SECTION OF EARLY NEURAL TUBE. of cells does not last long. For even by the second week the rapid proliferation of the cells has led to a marked increase in the thickness of the side wall and a scattering of the more numerous nuclei, apparently irregularly, throughout its substance (Fig. 444). The latter consists of a network of protoplasm in which definite outlines of cells cannot be detected. As growth proceeds the innermost part of this nucleated protoplasmic syncytium becomes condensed to form a delicate membrane termed the internal limiting membrane, which lines the lumen of the tube, whilst its outermost part presents a similar relation to an external limiting membrane, which invests the outer surface of the tube. Toward the end of the first month the side walls of the tube show signs of a differentiation into three layers. Next to the central canal there is an epitheliallike arrangement of the innermost cells of the syncytium, forming the ependyma Then there is an intermediate layer crowded with nuclei, hence known as the nuclear or mantle layer. On the surface is a layer singularly free from nuclei, which is called the non-nuclear or marginal layer. The germinal cells are placed in the ependymal layer between its radially arranged cells as they pass in towards the internal limiting membrane; and the protoplasm of the germinal cells forms part of the syncytium. At one time it was imagined that the germinal cells were embryonic nerve-cells, the parent-cells of the real neuroblasts, and that the whole of the rest of the syncytium represented the supporting tissues, which in the adult form the neuroglia. But it is now known that from the proliferation of the germinal cells, in which mitotic figures can usually be seen, some cells are formed which become ependymal epithelium, and others which migrate peripherally into the mantle layer. There, while forming part of the mantle syncytium, they undergo further proliferation and some of the resulting cells develop into spongioblasts, which constitute the supporting framework, the embryonic neuroglia; others become rudimentary nervecells or neuroblasts, and others again are known as indifferent cells. The latter are destined to undergo further subdivision and become the parents of more spongioblasts and neuroblasts. From this it is clear that the greater part-all except the germinal cells— of the syncytium, which is known as the myelospongium, is not merely supporting neuroglial tissue, as was once supposed, but is the rudiment of both neuroglia and true nervous tissues. The details of the process by which the neuroblasts become dissociated from the neuroglial network are quite unknown. It is commonly supposed that a spherical cell in the mantle layer that is to be transformed into a neuroblast frees itself from the syncytium, and remains for a time independent and wholly unattached amidst the meshes of the neuroglial network: it is supposed further that its true nature as a neuroblast becomes revealed when it takes on a pearshape, and a protoplasmic process, the stalk of the pear, pushes its way into some other part of the nervous system, or out of it into the mesoderm to reach some muscular or glandular tissue, and becomes the axis cylinder process or axon of the nerve-cell. Such an interpretation of the appearances exhibited in the walls of the neural tube at the end of the first month is adduced in support of a view concerning the constitution of the nervous system known as the neurone theory. "Neurone" is the term applied to a nerve-cell and all its processes; and the neurone doctrine assumes that there is no continuity whatever between the substance of one neurone and that of another, such as occurs in Hydra (Fig. 439), and that the functional connexions between them are brought about merely by the contact of the processes of one element with the processes, or the cell-body itself, of another element. In accordance with this conception the facts of embryology are supposed (by His) to demonstrate that when the axon grows out from a previously spherical and unattached cell it is able to push into the surrounding tissues, and, as it were guided by some instinct, eventually finds its way to that particular area of skin, muscle, gland, or other part of the body where nature intends it to go. This is the current teaching in regard to the neurone-theory; and it is supposed to have been conclusively demonstrated by the facts revealed not only by embryology and the study of the minute structure of the nervous system, but also by the phenomena of degeneration and regeneration. Harrison has shown that the outgrowth of processes can be witnessed in the living nerve-cells of the frog. There are certain facts, however, which have always led some anatomists to refuse to believe in the validity of the neurone doctrine as a true expression of the real constitution of the nervous system. It has been clearly demonstrated by Graham Kerr that at a very early stage of development the neural syncytium of the spinal medulla (of the mud-fish Lepidosiren) is in free and uninterrupted continuity with the protoplasm of the muscle-plate, which lies in contact with the neural tube; and no stage is known in which these connexions do not exist. When, in the course of the subsequent growth of the embryo, the muscle-plate becomes removed further and further away from the central nervous system the protoplasmic strand, which links them the one to the other, gradually becomes stretched and elongated. As the neuroblast matures its chemical constitution becomes modified; it becomes specialised in structure to fit it for the peculiar functions it has to perform. These changes manifest themselves first in the body of the neurone itself and thence spread along its processes. With the knowledge that protoplasmic bridges exist long before the time His supposed the axon of his neuroblast to push its way outward, it seems not unreasonable to suppose that it is the chemical modification of these existing bridges which has been revealed in stained specimens, as it spreads from the cell body outwards into its processes. It is now a well-recognised fact that soon after the neural tube becomes closed the outlines of its constituent cells become blurred and then disappear, and a continuous protoplasmic network or syncytium is formed. No one has ever been able to detect the process of detachment of embryonic nerve-cells (neuroblasts) from this syncytium; and it is at least a possibility that the free anastomosis of the protoplasmic processes of many of the cells is not destroyed in the way demanded by the neurone doctrine. The known facts might be interpreted, at least as reasonably, by supposing that when nerve currents begin to traverse the syncytium (Fig. 444) structural modifications occur around the nuclei of the cells affected, and gradually spread along their processes, so as to give the appearance (in sections stained by special methods) of processes growing out from each neurone. Impulses brought from the skin by the sensory nerves, the nutrition of which is controlled by the cells in the sensory ganglion (Fig. 443), are carried into the wall of the neural tube, where they are received by processes of intercalated cells, which in turn transmit their effects directly or indirectly to (a) motor nerve-cells (or other kind of efferent nerve-cells), which stimulate a muscle, a viscus, or other active tissue to perform some work, or (b) to intercalated cells, the axons of which proceed to some other part of the nervous system, perhaps above or below the place where the sensory nerve enters (Fig. 444, funicular cells). As the walls of the neural tube increase in size the various neurones gradually become drawn apart, and the protoplasmic links uniting them become stretched and extended to form processes of varying length. It is right to explain that most writers give an explanation of the process of development which is at variance with that just sketched. The neuroblast is supposed to originate as a free-lying spherical cell, which is stimulated by some unknown force, sometimes assumed to be of the nature of a chemical attraction (chemotaxis), to protrude a process, which gradually elongates and pushes its way through the tissues, perhaps to some particular patch of skin, muscle, gland, or some other nerve-cell. The difficulty involved in such a conception is not only that it is opposed to all that is known of the early stages in the evolution of the nervous system, but also that it is difficult to conceive that every one of the millions of nerve-cells, muscle-cells, visceral and cutaneous elements can each have some specific attractive power which leads every individual nerve fibril to its appropriate and predestined place in the body. The Efferent Nerves.-The efferent cells of the neural tube are distinguished by the fact that their axons leave the central nervous system and traverse the mesoderm for a longer or shorter distance to end in relation to some muscle, gland, or other tissue outside the nervous axis. At an early stage of development (Fig. 445) such efferent fibres pass not only to muscles but also to viscera and other kinds of tissues. In the course of the growth of the body these various structures supplied by efferent fibres become removed progressively further and further from the central nervous system; and in this process a distinction can be detected in the behaviour of the efferent fibres proceeding (a) to the striped or voluntary muscles, (c) and the viscera and unstriped muscle, respectively. The efferent cells (a) which innervate voluntary muscles retain their positions in the central nervous system, their axiscylinder processes (motor nerves) becoming elongated in proportion to the migration of the muscle from its original situation. But the cells (c) innervating non-striped muscles and viscera behave in a different manner. As the viscus or muscle migrates (Fig. 445, B), the nerve-cell (c) follows it more or less closely, being as it were dragged out of the wall of the neural tube by its axon into a peripheral position, where it becomes a constituent element of one of the so-called sympathetic autonomic ganglia. As these sympathetic cells migrate from the central nervous system, each of them appears to draw out with it the axon of an inter or 1 calated cell (d); and it is customary to distinguish these latter elements (within the central nervous system) as splanchnic efferent cells. It is, however, a matter of fundamental importance to recognise clearly that the real splanchnic efferent cells, the homo logues of the somatic efferent cells, are found in the sympathetic ganglia, and that the elements to which this term is usually applied are in reality intercalated cells. This account is at variance with the customary description of the development of the sympathetic system, according to which the cells of the sympathetic ganglia are said to be wholly derived from the sensory ganglia; but it offers a reasonable explanation of the A Roof Plate B Alar Lamina facts (i.) that the cells FIG. 445.-DIAGRAM OF A TRANSVERSE SECTION THROUGH THE LEFT HALF OF THE in the sympathetic ganglia are of the efferent, and not of the sensory, type, NEURAL TUBE REPRESENTING TWO STAGES IN THE DEVELOPMENT OF THE and (ii) that the fibres from the central nervous system establishing relations with them emerge along the motor nerves. Moreover, the information brought to light by recent research in embryology (Froriep, Kuntz, and others) affords positive evidence in support of this view. Elliott, however, opposes this interpretation (Journal of Physiology, 1907, p. 438). Many, if not all, of the sympathetic cells are derived from the walls of the neural tube, and they migrate along the pathways formed by the motor, rather than the sensory, nerves. In the case of the spinal medulla they pass out chiefly along the anterior roots, and from the brain along the motor nerves-the oculomotor, and the motor divisions of the facial and vagus nerves. Nerve Components. From the statements in the preceding paragraphs it must be evident that there are several varieties of afferent and efferent nerves respectively entering and leaving the central nervous system. The cells of origin of the efferent nerves are all placed in the ventral part of the side wall of the neural tube; and for this reason this part of the wall becomes swollen at an early stage of development (Figs. 445 and 446). It is called the basal lamina. Most of the cells that emit afferent fibres are situated in the sensory ganglia outside the central nervous system, so that their growth can have no direct influence upon the form of the neural tube; but their central processes become inserted into the dorsal part of the side wall of the tube, which is called the alar lamina; and groups of intercalated cells collect around the entering fibres to form receptive or terminal nuclei. growth of these terminal nuclei leads to an expansion of the alar lamina which is analogous to, but much less extensive than, that seen in the basal lamina. This unequal swelling of the dorsal and ventral parts of each side wall of the neural tube leads to the development of a longitudinal groove, sulcus limitans, as a demarcation between the alar and basal laminæ. The The nuclei of origin of the efferent fibres, which are found in the basal lamina, may be divided into two (and, in some regions of the nervous axis, three) main groups. There is first the group of large multipolar nerve-cells which emit fibres to innervate the ordinary striped voluntary muscles. This is commonly called the somatic efferent nucleus. Then there is a group of small multipolar cells, the axons of which pass out into sympathetic ganglia, and indirectly control the involuntary unstriped muscles and other active parts of viscera. These cells form the splanchnic efferent nucleus. In the upper cervical and lower cranial region a portion of the somatic efferent nucleus is set apart to innervate the striped muscles developed in the branchial arches. This is the lateral somatic or intermediate efferent nucleus. Many recent writers are of the opinion that this nucleus is splanchnic; but its fibres directly innervate striped voluntary muscles, which are developed from the same material FIG. 446.-DIAGRAM OF A TRANSVERSE SECTION THROUGH THE RIGHT HALF OF THE FOETAL RHOMBENCEPHALON AND EPITHELIAL AREAS ASSOCIATED WITH IT TO ILLUSTRATE THE DIFFERENT CATEGORIES OF NERVE COMPONENTS AND THEIR CENTRAL NUCLEI. (myotomes) from which the other striped muscles are formed (Agar and Graham Kerr). The alar lamina also can be subdivided into a series of functional areas (Fig. 446). At the dorsal edge is the somatic afferent terminal nucleus, which receives impulses coming from the skin. In one region a part of this nucleus is specialised for the reception of impulses coming from the internal ear (acoustico-lateral terminal nucleus). Then there is a group of cells collected around the incoming visceral sensory nerves-the splanchnic afferent terminal nucleus. A part of this is specialised to receive taste impressions-the gustatory nucleus-but this has not yet been clearly demarcated from the rest of the nucleus. This analysis of the various functional elements that may enter into the constitution of the various cerebral and spinal nerves is made use of in elaborating the theory of nerve components, which will help us to understand many features of the structure of the nervous system that otherwise would be unintelligible. Nerve-cells. We have already noticed that there is a broad distinction between the nerve-cells which are found in the ganglia of sensory nerves and those |