THE SPINAL ANIMAL 33 
thetic efferent cells are not, however, present in all spinal segments, hut are, as ahnve noted, 
absent from the cervical and lower lumbar. They are distributed peripherally to the motor-cells 
of the pre-vertebral ganglia of the sympathetic chain, and constitute the system of fibres which 
Langley * calls pre-gangUonic. Their distribution to the ganglia is an overlapping one, so that 
each pre-vertebral ganglion contains endings from a short series of sympathetic efferent spinal 
roots. Hence it is that so small a muscle as the iris receives efferent fibres from three spinal 
roots ; in the monkey the first, second, and third thoracic. t 
It has by many been supposed that even in the cord of the higher vertebrata the collection 
of the perikarya of the efferent root-cells, corresponding as they do with that of the afferent root- 
cells, which form the spinal ganglia, must be massed into a group for each spinal segment, 
separated by a distinct interval from that of the next segment. { As a fact, not only is evidence 
of such an arrangement as this wanting, but the evidence as regards the mammalian cord refutes 
the supposition. § So complete a fusion of the individual segments has in the course of ages gone 
on that, as regards the position of the efferent cells inside the cord, although no demonstrable 
degree of segmental interlap exists, yet also there is now existent no demonstrable segmental 
interval. 
Just as the peripheral distribution of the afferent fibres of the spinal ganglia is an over- 
lapping one, so also is their central intra-spinal ramification. The central or stem-process of each 
afferent root-cell, after entering the cord, bifurcates, as was discovered by Frithjoff Nansen,|| 
the explorer. The headward branch of division ascends in the case of many fibres to the grey 
'nuclei' of the dorsal columns in the bulb; the aboral branch descends only a short distance, 
one or at most a few spinal segments. Each of these branches gives off collaterals into the grey 
matter, especially into the grey matter of the same segmental level as the root-ganglion, whence 
the root-fibre in question is itself derived. The intra-spinal extent of the afferent root-cell is 
therefore far more expanded than is that of the efferent root-cell. The dendrites of the latter 
seem to be practically confined to the segment in which the cell-body lies; the processes of the 
afferent fibres on the hand are traceable far into segments widely distant from that into which 
they plunge first on entering the cord. There is, therefore, a very extensive intra-spinal overlap 
in the central distribution of the afferent fibres derived from each afferent root-ganglion. 
Whether this applies equally to the afferent root-cells connected with the viscera, as to other 
afferent root-cells, is a question. Tlie focal and ciicumscript character of the areas of tenderness 
and referred pain studied by Head in cases of visceral disorder, and tlieir correspondence in 
situation with the skin areas of distribution of the individual spinal roots, suggests that the central 
end-distribution of these visceral afferent fibres is less diffuse and less extensive intra-spinally, and 
more concentrated within a single segment. On the limitations of intra-spinal root-overlap the 
study of the reference of visceral irritations may give most valuable light. Regarding the 
existence of segmental boundaries to the intra-spinal spread of impulses, observations such as the 
following illustrate the slightness of the evidence in support. If the central end of the afferent 
* ' Journ. of Physiol.,' xv, 1894. f Shun ington, 'Joiirn. of Physiol.,' xiii, 189Z. 
§ Kaiser, v. Argiitinski. || Bergen's Museum, 1887. 
I Schwalbe. 
E 
