It would seem an appropriate time to re-examine the cellular structure of the mammalian nervous system for the following reasons. Firstly, there is considerable confusion in the literature about the appearance of the different kinds of neuroglia by light and by electron microscopy, and this is complemented by widespread disagreements among distinguished neuropathologists about the international classification of tumours of the central nervous system. Secondly, there is an increasing volume of experiments on the physiology and biochemistry of tissue cultures of neurons and different kinds of neuroglia, whose validity depends upon the accurate identification of both the parent tissue and also of the cells subsequently growing in culture. The biochemical classifica tion in recent years has often tended to become independent of the cellular identification, which makes the use of the neuroglial cell names doubtful and the significance of the biochemical properties of the cells difficult to relate to the physiological properties in vitro or in vivo (Table 1).