782 DR ALEXANDER BRUCE AND DR JAMES W. DAWSON ON 
(p. 742), and the other, that these are new-formed fibres of regeneration from the anterior 
roots, in the sense of the collateral regeneration of NaGEorTe (p. 726) in tabes. 
For the first we must assume that in early foetal life a meningeal lymphangitis had 
caused a diversion of the growing axons, so that instead of passing into the spinal 
nerves a certain number of them had become side-tracked into the meninges, and thence 
made their way along the vessels of the cord and in the pial spaces ventrally and 
laterally. ‘The fact that in the immediate neighbourhood of the anterior roots the 
fibres composing the strands were few in number compared to the great mass of the 
fibres composing the nodules, in no way tells against this view, for the importance does 
not depend upon the number of the axis-cylinders. It is well known that one axis- 
cylinder can furnish a large number of fibres by breaking up into its constituent fibrils : 
these, according to accessory circumstances, can twist up into whorls and form nodules 
when they come even from an extremely restricted number of fibres. Such neuromata 
would be analogous to stump-neuromata. But while it is possible in Weigert- and 
silver-stained preparations to explain the fibres as outgrowths from the axis-cylinders 
and their divisions, it is impossible to account for the nucleated tubes and fusiform cells 
with which the nerve fibres were continuous as outgrowths from axis-cylinders. 
For the second possibility it is necessary to pre-suppose a degeneration, in order 
that a regeneration might occur from the preserved end. Such neuromata would again 
be analogous to stump-neuromata. Nacxorre’s position has been already stated, and it 
has also been pointed out that such an explanation cannot be accepted for the fibres 
forming the neuroma, for no degeneration of extra-medullary anterior root could be 
traced. Further, if evidence of such regeneration had been present, the explanation 
is still awanting of the presence of the nucleated tubes and fusiform cells with which 
the nerve fibres are continuous. 
If, now, we turn to the medulla and pons, it is clear that neither of the two 
possibilities that at first presented themselves as accounting for the cord nodules need 
be considered. There is no question of tracing the fibres composing even the most 
fully formed nodules to aberrant nerve fibres nor to a regeneration of fibres. Further, 
if such connections existed it would be the connection between fully differentiated 
fibres and immature fibres—a connection inconsistent with the explanation of the 
axis-cylinder as an outgrowth. 
We have still to consider one further possible explanation for the cord neuromata 
before we have cleared the way for attempting a unification of the processes at work. 
This explanation is related to the multicellular structure of the peripheral nerves. In 
stump-neuroma, KmEnNepy has found that the new-formed fibres terminate in proto- 
plasmic tubes and fusiform cells within which axis-cylinder and myelin-sheath 
differentiation is taking place. The new-formed fibres are in direct continuity with 
a nerve which preserves its connection with the centre, and, according to the cell-chain 
theory, the new fibres have arisen within the proliferated cells of the sheath of — 
Schwann, and only later become continuous with the fibres of the preserved end. 
