MULTIPLE NEUROMATA OF THE CENTRAL NERVOUS SYSTEM. 155 
anterior horn, and from this nodules have branched off, with the vessel branches, to the 
different areas of the grey matter. When the secondary fibrotic changes, following 
the nodule formation, have not occurred to any marked extent, a very beautiful inter- 
lacing of the fibres or intermingling of fibres from different nodules takes place at the 
borders of white and grey matter. Such appearances are best understood from Van 
Gieson sections, where the fibres of each nodule, as it were, become unwound and the 
disentangled threads of each nodule intermingle to form an inextricable maze of fibres 
that break up at their extremities into whorls of individualised nucleated elements. 
Many of these isolated elements in the loose meshwork are distinctly the cut portions 
of these terminal strands, but many could be conclusively proved to be isolated fusi- 
form nucleated cells. Sometimes these fusiform elements unite end to end to form 
chains of cells with imbricating processes (figs. 45, 46). 
In the terminal ramifications of the fibres of each nodule all the transitions between 
nucleated fibres, chains of nucleated cells, and nucleated fusiform cells could be followed. 
The calibre of the nucleated fibres is greater than that of the fibres with which they are 
in continuity (fig. 45). In each nucleated fibre there can be recognised a distinct axis- 
cylinder, a homogeneous yellowish-green surrounding layer, with a fine pink outer 
border. In the fusiform elements forming the chains are darkly stained elongated 
nuclei, and a homogeneous—with polychrome-methylene blue—slightly granular proto- 
plasm, and in some a faintly stained filament can be definitely demonstrated lying along 
one side of the nucleus or extending from its opposite poles. This appearance was more 
evident in preparations from the medulla and pons, and its significance will be dealt 
with in discussing the nodules in these regions. 
In the structure of the compact nodules we noted that there was little interstitial 
tissue, but here, in the breaking up of the nodules, the fusiform elements were inter- 
mingled with an increasing fibrosis. As these elements scatter in this fibrous tissue 
they become more and more unrecognisable, and progressively may be confounded with 
connective tissue elements. It is only when this fibrosis is absent or very slight that 
the specific nervous nature of the nucleated chains and fusiform cells can be recognised. 
(c) Origin of the Fibres fornung the Nodules. 
The majority of the nodules, it has just been stated, can be traced to fibres passing 
ventrally or laterally from the immediate vicinity of the emerging anterior nerve roots. 
It will now be necessary to discuss one or two points which are of importance in relation 
to the possible precise point of origin of the pial fibres so traced. 
OBERSTEINER thus describes the histological structure of the posterior root where it 
enters the cord: “ At the point where the posterior root pierces the pia mater it is 
constricted, and sometimes to a marked extent.” ‘This reduction in size takes place 
entirely at the expense of the myelin sheath. Consequently, when stained by Weigert’s 
method, the root at this point remains colourless, and a bright band, usually convex on 
