CENTRAL NERVOUS SYSTEM IN VERTEBRATE ANIMALS. 
945 
surround a central vacant space. In addition to these fibres there are also a few cells 
scattered throughout the layer; they resemble the largest sized cells of the fourth 
layer, but are on an average of slightly larger dimensions ; sometimes they have three 
processes, one of which can be traced for some distance into the complex of fibres 
forming the exit of the nerve. The sixth layer is formed of finely granular material 
resembling that of the third layer; it occupies the anterior end of the tuberculum, 
and forms a sort of kernel round which the other layers are placed. 
We have now described those parts of the brain of Mormyrus which differ from., and 
ore in fact additions to, the nervous centres of an ordinary Teleostean ; the remaining 
portions show very slight variations from the general plan. 
The Central Cavities. 
In tracing the central canal of the spinal cord and its prolongations forward, it is to 
be remarked that tbe cavities show a tendency to contraction and almost obliteration. 
At a very short distance in front of the point where the central canal of the spinal 
cord merges into the fourth ventricle, this latter cavity (fig. 1 ©) becomes plugged with 
a mass of finely granular material, which in the central line shows indications of its 
original separation into two halves ; this contains part of the ganglion of the vagus. 
Farther forward this ventricle exhibits an extremely deep narrow fissure (fig. 9), which 
owes its depth and narrowness to the great development of the vagal tuberosities. At a 
point beneath the posterior end of the tuberculum impar the ventricle (fig. 8) becomes 
again plugged by a development of what appears to be connective tissue, corresponding 
to that material w r hich surrounds the central canal of the spinal cord. There are, 
however, two spaces left one above the other, the lower one lined by an endothelial 
layer of cells. They both lie beneath the space which occupies the centre of this part of 
the tuberculum impar. Towards the anterior end of the tuberculum impar the fourth 
ventricle (fig. 7) becomes reduced to an extremely small canal, triangular in outline, 
with the apex directed downward; in some specimens the lumen is scarcely perceptible. 
This corresponds to the narrower portion of the fourth ventricle in M. cephalus; after 
this it again enlarges (fig. 6) beneath the cerebellum, and shows a triangular lumen 
which seems to correspond to the anterior end of the fourth ventricle in other Teleostei; 
beyond this it again becomes contracted as the aqueduct of Sylvius (fig. 4) into a very 
small perpendicular slit, which soon extends through the hypoarium nearly to the 
inferior surface of the brain (fig. 3). This slit is occupied in its centre by a small coil 
of an epithelial tubular structure, which appears to correspond to the saccus vasculosus 
of the ordinary Teleostei. After this the aqueduct opens into what would be the 
ventricle of the optic lobe, were it not that it is quite open to the external surface, 
and by the peculiar position of the tecta, has come to be above instead of below 
them. The growth of the wings in fact has quite forced open this ventricle and 
separated the valvula cerebelli and themselves from the tori semicirculares and the 
6 e 2 
