— 143 — 
the same two types previously described. Some processes retain 
their individuality, and pass directly to form nerve-tubes, whilst giving 
of slender side-branches (fig. 55 — 57; 87, c; Sg, gc), other processes 
lose their individuality and are broken up into slender branches losing 
themselves in the plaiting of the dotted substance (fig. 87, a, h, d, e). 
I have been able to trace both these types of processes, only for 
comparatively short distances through the dotted substance, but 
still the difiference between them is generally possible to distinguish.') 
As before said, a great many of the nerve-tubes of the dotted 
substance go to constitute the peripheral nerves. Man}^ of these 
nerve-tubes are, consequently, direct continuations of the nervous 
processes from ganglion cells of the one type ; these nerve-tubes 
probably form peripheral nerve-tubes directly. A great many 
of the other nerve-tubes and fibrillæ, constituting the nerves, 
are either continuations of the slender branches arising from the 
subdivisions of the nervous processes of the other type (these 
fibrillæ are extremely slender, and in the roots of the nerves they 
unite to form thicker tubes), or they spring from a union of similar 
fibrillæ or branches; perhaps the side-branches given off from the 
nervous processes of the first type also contribute to the forming of 
these nerve-tubes. 
As constituents of the peripheral nerves, present in their roots, 
we have thus elements of two kinds; viz. i) nerve-tubes which are 
direct continuations of nervous processes, but which, however, are 
partly strengthened by the junctions of slender side-branches, and 
2) nerve-tubes arising from a union of slender tubes or fibrillæ.-) 
^) As is mentioned in my previous paper on the nervous system of the 
Ascidia (1. c. 1886), small ganglion cells occur in the dotted substance (vide 
fig. 87,/',^). Those cells have generally a tripolar or multipolar form (bipolar cells 
also occur). Fig. 57 represents such a cell, isolated, in a macerated preparation. 
The protoplasmic processes of these cells are, generally, directed peripherically 
towards the external layers of the brain, whilst the nervous process has very often 
a longitudinal course through the dotted substance. I have often, both in macerated 
preparations and in sections, been able to trace such nervous processes for some 
distance of their course through the brain, but have rarely observed any sub- 
division of them. I think it therefore probable that they, to a certain extent, are 
directly continued into peripheral nerve tubes; some of them are, however, broken 
up into branches losing themselves in the dotted substance. 
^) It may here be mentioned that, I have vbserved ganglion cells situated in 
the roots of the nerves and sending their nervous processes directly into the nerve. 
In fig. 88, which represents a part of a longitudinal section through the anterior 
end of the brain of a Corella, such a ganglion cell (gc) is seen situated at the 
root of one of the anterior nerves. I am not in a position to state whether the 
nervous processes of these cells send off side-branches. 
