272 SOME CHAPTERS ON THE 



growth of new fibres. The cells of the neurilemma multiply 

 actively. They first act as phagocytes in removing the debris of 

 the degenerated nerve, and then elongating, form themselves into 

 long chains of cells ; but if no nerve fibres enter these chains, they 

 never, as far as can be seen, form either medullary sheath or axis 

 cylinder. So that although these neurilemmal cells doubtless have 

 important nutritive and other functions (we have already referred 

 to the suggestion that the failure of regeneration in the spinal 

 cord is due to the absence of the neurilemma), yet by themselves 

 these neurilemmal cells do not form new nerve fibres. 



III. The third possibility, that regeneration may occur from 

 new fibres which wander into the cut peripheral end from an alien 

 central end, is one that gives rise to many interesting questions. 



It has been known for a long time that intentional suture of 

 nerves originally distinct leads to regeneration, but even when 

 actual suture is not performed, growth may take place if there 

 are any cut nerves in the vicinity, and it is occurrences of this 

 description that invalidate much of the evidence for autogenous 

 growth, as has been explained in the previous section. 



There would appear to be degrees in the chemiotactic attrac- 

 tion which determines the direction of the fibres ; efferent somatic 

 fibres unite most readily with their own kind, less readily with 

 pre-ganglionic fibres, and apparently not at all with post-ganglionic 

 fibres ( 35 ). Of the various possibilities of union thus summarised 

 there is one which has attracted special attention, as it has an im- 

 mediate bearing on surgical procedure. This is the particular case 

 in which different efferent somatic fibres are made to unite with 

 each other. Ballance ( 50 ) was the first to make a practical applica- 

 tion of the fact, although he did so under the impression that 

 regeneration would occur from the peripheral end. He sutured 

 part of the spinal accessory nerve to the peripheral end of the 

 facial, and in his most successful case the patient recovered the 

 use of the paralysed facial muscles, but with the drawback that 

 there was a synchronous lifting of the shoulder. 



Kilvington ( 28 ) has studied the experimental conditions under 

 which union of this class takes place, and two of the many 

 interesting facts he observed may be noted. 



The optimum arrangement, as judged by the functional recovery 

 of movement, is shown in Fig. 5. Here the internal and external 

 popliteal nerves are cut across, the central end of the external 



