250 HILL. [Vou. IX. 
of radially grouped cells. Each group is now made up of a 
larger number of cells than in the preceding stage, and is con- 
sequently larger. Many of the groups project for a consider- 
able distance into the cavity of the epiphysis. The cells 
composing the groups are pear-shaped, and are held together 
in a bundle by their nerve processes. Fig. 16 is a camera- 
drawing of such a group of nerve-cells as seen in a longitudinal 
section at the point marked X in Fig. 15a. In order not to 
confuse the figure only a few have been drawn of the large 
number of cells that could be seen in the section. 
If the group is a deep-seated one, its fibres may abut upon 
the walls of a blood-vessel between two of the folds; if the 
group is superficial, its fibres may reach the free surface of a 
fold and pass along the pia-mater. The fibres of each group 
may pass to the surface independently, or those of two or more 
adjacent groups may join into a single band, which passes to 
the surface. If the fibres reach the wall of a blood-vessel, 
they are seen to turn in a curve and follow the course of the 
vessel. Individual fibres soon become so intimately associated 
with the wall of the vessel, that it is not possible to distinguish 
them from it. These fibres probably follow the blood-vessels 
to the external surface to join the column of fibres in the stalk. 
The third region: In the distal part of the stalk there are 
no folds and no blood-vessels. The projection into the cavity 
of the stalk of the groups of nerve-cells gives its walls the 
appearance of being folded; but this appearance should be 
distinguished from the actual folds which occur in the distal 
part of the epiphysis. The fibres coming from the groups of 
cells in the dorsal wall of the stalk all turn toward the brain, 
and pass directly into the column of fibres which traverses the 
stalk. In this respect they appear to differ from the fibres in 
the distal part of the epiphysis; but it is likely that in all 
stages the fibres from each group of cells pass in one direction, 
z.e., toward the brain. The appearance which is sometimes 
presented of fibres passing in both directions does not probably 
correspond to the reality. It is more likely that it is to be 
explained by the fact that the fibres of each bundle separate 
from one another as the bundle approaches the pia-mater, in 
