CENTRAL NERVOUS SYSTEM IN VERTEBRATE ANIMALS. 
941 
caused by these foldings. These cells are of about the same size as the others, but 
differ in shape, being multipolar instead of bipolar. In other respects they belong to 
the same class, having a large oval or rounded nucleus and an easily distinguishable 
nucleolus. The specimen shown in the figure (fig. 17,/!) is a representative cell, having 
six processes, none of which can be singled out from the others as an axis cylinder 
process. This cell measures 0‘01 millim. long by 0*008 millim. wide; the nucleus 
measures 0*006 millim. by 0*004 millim., and the nucleolus about 0*001 millim. in 
diameter. 
These cells are situated among a network of fibrillse, which fills up the space between 
them; but I could not trace any of the processes of the cells into this network. 
In other respects the ridges at these curves and foldings have the same structure as 
elsewhere, the only difference being that at these places there is a larger space than 
ordinary between two contiguous ridges, which is consequently filled in with the cells 
resembling those just described. 
The basal portions of the wings, which form the greater part of the whole mass to 
which the ridges appear merely as appendages, are entirely composed of extremely 
small cells which correspond in every respect with those of the granular layer of the 
cerebellum in structure, together with very fine nerve fibres, which are accumulated in 
greater numbers towards the outer side. The fibres run in a direction from the roots 
of the wings towards the apex, and give a longitudinally striated appearance to the 
part especially near the outer surface. In places where a sufficiently thin section has 
been obtained to see their arrangement, these cells are found to be placed in long, 
narrow groups, one or two deep, following the direction of the fibres, which form a 
network of elongated meshes. The cells are attached to them by short stalks in the 
same way as is represented in the figure of the torus longitudinalis (fornix) in my 
paper on the brain of the Teleostei.* This arrangement becomes less apparent, and 
the cells are more irregularly scattered towards the inner side of the wing, where the 
fibrillse take a less decidedly longitudinal direction. The cells themselves are round, 
oval, or fusiform, sometimes with three, sometimes with two processes, one from each 
end, and occasionally with only one (fig. 1 7,g.); they measure about 0*003 millim. long 
by 0*0026 millim. broad, one larger than those figured was 0*0034 long, with a nucleus 
of the same size as that of the other. 
Marcusen was of opinion that the anterior lobes of the wings had a different 
structure from the remainder, but such is not the case. The anatomical arrangements 
are precisely the same throughout the whole extent of this very extensive organ, which 
might be considered extremely complicated were it not that it is entirely made up of 
repetitions of the same structures. 
The difference to be remarked between the anatomy of the ridges on the wings of 
the valvula cerebelli and that of the cerebellum is reduced to three points : first, the 
arrangement of the layers comes in a different order; secondly, the Purkinje cells are 
* Pig. 18, e. Phil. Trans., 1878. 
