The granules of the cerebellum have a curious developmental history. Every neurone in the body has a lifelong existence. Except for the rare accident of its destruction by disease it occupies its station to the hour of death. But at the time of birth many neurones are still immature. Not all the granules of the cerebellum have yet assumed their permanent form or situation. Beneath the pia mater there is still a layer of minute undifferentiated cells. These, as they grow into granules, elongate, in the first instance, into long spindles. Subsequently they sink down through the molecular layer and between the cells of Purkinje, leaving the poles of the spindle as the right and left divisions of the axon (Fig. 21). It is interesting to learn that such a migration is possible. It is also of interest to find that a tiny granule of the cerebellum goes through the same stages in attaining its adult form as one of the large cells of a spinal ganglion. [Illustration: FIG. 21.—THE GROWTH AND MIGRATION OF GRANULES OF THE CEREBELLUM. Half a dozen nuclei of as yet undeveloped granules are seen lying beneath the pia mater. From this level to the bottom of the drawing granules are shown in successive stages of growth. These developing granules, selected from various preparations of the cortex of the cerebellum, were drawn from nature.] There are many different types of neurone. Any attempt to describe them, or to give an account of the various details of structure which recent improvements in technique have enabled anatomists to observe, would fill a lengthy treatise; and would, moreover, be beside our aim, which is limited to obtaining such an idea of the unit of the nervous system as will enable us to form a conception, however crude, of the way in which it works. From the brief account that has been given, it will be evident that anatomists are approaching to an understanding of the mechanism. It will also be evident that they have already more information than they can apply. They are cognizant of many details of structure which they cannot interpret in terms of function; and at the same time are aware of wide gaps in their knowledge regarding facts which are essential to the construction of any scheme. This much is clear: A sense-cell on the surface or beneath it is touched (probably entered) by the ultimate twig of the outer limb of a neurone whose cell-body lies in a spinal ganglion, while its inner limb, as a fibre of a posterior root, enters the spinal cord. In the spinal cord the root-fibre splits into an ascending and a descending division which rain branches into the grey matter over a considerable area above its point of entrance, and a smaller area below it. The finest twigs of these branches are to be seen in the vicinity of the cell-bodies and dendrites of certain other neurones. The axons of these second links arborize in a similar way in the vicinity of large motor cells,