vin CONDUCTIVITY AXD EXCITABILITY OF XERVE 91 



nerve are simultaneously investigated, they will be found only in 

 the immediate neighbourhood of the cut surface, and are but little 

 developed ; further on, the fibres show hardly any perceptible 

 changes, and the nerve also remains permanently excitable. 



"Waller's investigation of the roots of the spinal nerves threw 

 much light on the subject, showing plainly that the ganglion-cells 

 exert a "trophic" influence on the nerve-fibres connected with them, 

 as indeed does every nucleated cell upon its non-nucleated pro- 

 longations. Degeneration of the peripheral end invariably follows 

 the division of the anterior root of a spinal nerve, while the 

 central end, which is still connected with the cord, remains normal. 

 After dividing a posterior root the consequences vary according 

 as the section is between spinal cord and ganglion, or is peripheral 

 to the latter. In the first case the central end, which is connected 

 with the cord, degenerates, and the degenerated fibres may be 

 followed a long way into the cord itself (in the posterior columns) ; 

 the end connected with the ganglion, on the contrary, like the 

 ganglion itself, is unaltered. 



After dividing the nerve on the peripheral side of the ganglion, 

 the end connected with the latter is not affected, while the 

 peripheral end degenerates. From repeated experiments, there- 

 fore, we must conclude that the cells of the spinal ganglion are 

 the parent-cells (or trophic centres) of the posterior root-fibres, 

 while the cord, and its large multipolar cells in the anterior horn, 

 are the trophic centres of the anterior root-fibres. This agrees 

 with the development of the fibres, since the spinal ganglia 

 comprise the true parent -cells of the posterior roots. The 

 degeneration of divided nerve -trunks has been successfully 

 employed in vertebrates to discover the course of certain bundles 

 of fibres within the central organ (spinal cord), by means of the 

 easily -recognised alterations of the medullary sheath. The 

 arrangement of the posterior columns of the cord has been very 

 accurately determined by this method. 



We have thus far assumed that a nerve-fibre must be perfectly 

 homogeneous under normal conditions from its origin to its end, 

 at all points of its course, i.e. that it presents no essential 

 differences in regard to excitability and conductivity. This pre- 

 sumption is indeed not justified. We have long been in 

 possession of certain facts which show that different parts of the 

 nerve are anything but homogeneous. The sole gauge of 



