SummaryThe causal relationships responsible for the early organization and subsequent structural configuration of the vertebrate central nervous system are poorly understood. This is due in part to the complex nature of the subject but perhaps even more to the fact that causal neurogenesis has not yet attained full status as an independent biological discipline. Those general principles which are known derive largely from the separate fields of experimental embryology and descriptive neurology and are based, for the most part, on a random collection of non‐integrated facts. The source and nature of the data are probably adequate; what is needed is a conscious, co‐ordinated effort to envisage this field of study in its totality and, for immediate purposes, as an end in itself.1. The early formation of the medullary plate is dependent upon the inductive effect of the mesodermal substratum. Physical contact is probably necessary. Organization of neural ectoderm into medullary plate and tube in the absence of underlying tissue has been reported, but these results are probably due to the disintegrating action of the culture medium upon the inner, uncoated plate cells. The autolysis produced liberates neurogenetic substances responsible for the inductions obtained and the process is considered a pathological one. The morphogenetic forces responsible for the folding of a flat medullary plate into a hollow tube are probably intrinsic to the neural ectoderm itself. An increase in the contractile tension of the superficial gel layer of the external ends of the plate cells may be an important factor. Neither differential water absorption nor localized cellular proliferation seems to play a part in the process. The establishment of polarity and morphological gradients within the neural axis are relatively independent of substrate tissue. Throughout the early stages of neurogenesis the neural ectoderm plays a more active role than has previously been supposed.2. There exists no valid evidence that neuromeres represent an inherent metamerism of the early neural axis. The segmental arrangement of the spinal ganglia, and probably dorsal and ventral roots as well, is imposed from without by virtue of the proximity of the segmenting axial mesoderm. Mesodermal segmentation is primary; ectodermal segmentation is secondary. The organizing effect of the mesoderm upon the segmentation of the neural tube is more pronounced than certain aspects of the mesodermal‐ectodermal relationship obtaining in earlier stages.3. The theories of Kappers (neurobiotaxis) and of Bok (stimulogenous fibrillation) are inadequate for solving the complexities of neural differentiation. Neither theory is buttressed by critical experimental evidence and both are highly speculative in nature. Neurobiotaxis and stimulogenous fibrillation still remain thedues ex machinaof neurology.4. Nothing decisive is known concerning the important phenomena of cellular proliferation and differentiation within the central nervous system. Failure to delineate clearly between these two processes has caused some confusion in the past. Cell number is controlled to some extent by position of a part within the anterior‐posterior gradient; more anterior regions of the neural tube are less dependent upon such a position effect. The definitive number of differentiated cells constituting a specific region is regulated by the size of the peripheral field. This is true of intramedullary as well as extramedullary cells and is common to all major vertebrate groups studied. The mechanism whereby non‐nervous tissue effects this control of neuroblasts is not known. Experimental studies differ in their interpretation as to whether regulation occurs during the initial stage of differentiation or is implemented through the subsequent atrophy of already differentiated cells. Invasion of an undifferentiated field by nerve fibres from a neighbouring region has been thought to initiate the differentiation of the indifferent neuroblasts. Critical evidence for such a thesis is as yet lacking. There is good experimental reason to believe that the non‐nervous periphery affects differentiation alone, while proliferation of indifferent cells is influenced to a great extent by intracentral factors.5. The arrangements of nerve fibres into specific tracts and the polarity of their growth is generally supposed to be influenced by pre‐existing fibre pathways, particularly the fasciculus longitudinalis medialis. Regenerating intramedullary fibres have been observed to traverse preferential routes through the embryonic cord and to exhibit marked polarity in their growth. Substrate configuration and mechanical forces are probably of great importance in directing fibre growth. The reason for the formation of the standard commissures and fibre decussations is not known. The operation of purely mechanical forces within the central nervous system appears inadequate to explain all the intricacies of fibre pattern observed.6. The precise point of entrance of sensory components of cranial and spinal nerves into the central nervous system has been thought to be determined by localized attraction centres of high metabolic activity. Indirect evidence exists for such a supposition but critical proof is lacking. Ectopic sensory nerve roots usually enter the brain wall at definite loci and their dorso‐ventral (medio‐lateral) orientation is probably determined by the position of correlation tracts of corresponding function. The place of exit of motor roots can be experimentally altered, but nothing is known concerning the causal relationships involved in their normal egress.7. Structural regulability of the embryonic central nervous system reaches a high level in amphibians, less so in fishes and mammals. Specialized cell types are determined early in ontogeny but may be replaced from indifferent cells at relatively late stages. The cytogenesis of various cell types is probably influenced by factors extrinsic to the cells themselves. Anatomical and functional repair of intramedullary fibres is possible. Special inhibiting factors are probably responsible for failure of proper regeneration rather than any inherent difference between central and peripher