194 
their ,,abnormal‘ position, for it will be seen that in Raja there are 
ample grounds for concluding their normal presence in the mesoderm. 
Some of the macro-ganglion cells!) proceed to wander out from 
the cord and into the mesoderm. The direction they take is towards the 
dorsal or neural portions of the muscle plates (epimeres or myotomes). 
Figure 1 is given as an example, but the variety of pictures one 
gets of this wandering is endless, and a number of plates could readily 
be filled with drawings of it, 
no two of which would be 
alike. 
In the series under re- 
view many of the ganglion 
cells are seen to be in the 
act of spinning processes (axis 
cylinders) towards the tips of 
the myotomes. Instances may 
be noted when there is a 
longish stretch of axis cylin- 
der, extending from a nucleus 
N and looking as if spun out by 
. the latter, just as the thread 
ane Gor te eat denen of a spider is formed by that 
in this and the following figures see the foot- animal. Some of these axis 
note at the end of the paper. cylinders proceed from gang- 
lion cells lying in the meso- 
derm, others extend from similar cells situated in the neural region 
of the spinal cord. In some cases nerve-forming nuclei are present on 
the nerve-fibre, in others they are entirely absent, and in these latter 
the „spinning“ of the axis cylinder is very obvious. 
Of this series of sections I have a note to the effect that some 
of the fibres may arise before any of the cells have taken on ganglionic 
characters and in these cases they would appear to be due solely to 
the activities of nerve-forming nuclei. 
Again, some of the macro-ganglion cells may be closely applied 
to the dorsal portion of a spinal ganglion, where they look as if 
really originally parts of the latter. Their relationships to the spinal 
ganglia cannot be discussed here: suffice it to say that no genetic 
connection seems probable. 
1) These cells will be spoken of as macro-ganglion cells and 
these nerves as macro-nerves. 
