CENTRAL NERVOUS SYSTEM IN VERTEBRATE ANIMALS. 
939 
belonging to this layer are of extreme fineness and are unprovided with a medullary 
sheath, being at these parts merely axis cylinders. 
Each ridge is inserted into the main body of the wing, like a nail into a plank, with 
a conical' insertion consisting only of the molecular layer. The granular-basal part of 
the wings forms a conical projection between each pair of molecular layers, and the pia 
mater passes down between these, forming a double membrane which at the summit of 
this conical process of the basal part divides into two, one accompanying each molecular 
layer to the extremity of its insertion where it disappears. 
We have thus a series of ridges in alternately reversed positions, occurring in fact 
in pairs back to back. Where the wings are comparatively straight, the only contact 
that the ridges have with their basal portion is by the insertion at one edge; but 
where curves and foldings occur the ridges become more or less distorted, and may be 
in contact with those parts by the sides as well as by the edges. With regard 
to the minute structure of these ridges and the remainder of the wings we will begin 
with a consideration of the molecular layers. These are generally separate from each 
other at the summit of the ridge, but in some parts this layer of one ridge is continuous 
with that of the next; but in those parts where the wings end in a free edge, and 
the ridges are on the external surface, the end of each ridge where it abuts on this 
free edge, goes round to join the next, forming a loop. The striated arrangement 
of the molecular layer is very apparent and the palisade cells (as one may translate 
Sited a’s term “ stiftzelle ”) are very well developed. The pia mater generally penetrates 
between each pair of molecular layers, and where this is so the striations are directed 
at right angles to the width of the ridge as seen in section, or in other words parallel 
to its thickness; but in the ridges along some parts of the dorsal surface of the wing, 
where they are external, the pia mater does not penetrate between them but simply 
covers their free edge ; in this case the striations are not directed at right angles but 
parallel to the width of the ridge; this seems to show that the striation is due to the 
palisade cells and not to the ramifications of the protoplasmic processes of the Purkinje 
cells, to which I was formerly inclined to attribute it. Besides this, the prolongations 
of these cells may very often be seen actually sending processes into the molecular layer. 
I can also confirm Boll’s statement that the palisade cells forms a membrane distinct 
from the pia mater; for this latter membrane can be torn off leaving the layer of 
palisade cells intact. 
As previously mentioned, the arrangement of the layers in these ridges is not quite the 
same as in the cerebellum ; the granular layer becomes interposed between the molecular 
and the intermediate layers, instead of the reverse being the case; this granular layer 
is continuous with the basal portion of the wings, but differs somewhat in this, that 
in the basal portions of the wings the cells are more uniform in size, whereas in the 
ridges the cells vary more in dimension; the smaller ones (fig. 17, b.) are oval or 
rounded, with nucleus and nucleolus ; generally they are provided with two or three 
processes; they measure usually about 0-004 millim. or less ; one of those figured is 
