CH. XIV.] REGENERATION OF NERVE 153 



union and consequent regeneration in the peripheral end occur. He 

 also draws attention to the olive-shaped swelling at the free end of 

 each growing axis-cylinder. These are also figured by Marinesco, 

 who calls attention to the fact that these terminal swellings, although 

 they may roughly be described as olive-shaped, vary a good deal in 

 external form; this is shown in the accompanying drawings (figs. 

 155 and 156), and is quite intelligible now that we have Boss 

 Harrison's description of the constant changes of form they exhibit 

 in embryonic history. 



The two next figures (157 and 158) are drawn from preparations 

 stained by Cajal's new method, and they require but little comment. 



They show the new fibres penetrating the cicatricial tissue of the 

 junction from the central end in a peripheral direction ; they show 

 the absence of any new axons developed autogenetically in the 

 peripheral end. Such preparations ought to carry conviction to 

 those who have any lingering belief in auto -regeneration, that the 

 Wallerian view is the only possible one to adopt. 



It must not, however, be supposed that the peripheral end is 

 entirely inactive ; for while degeneration is progressing in the axons 

 and their fatty sheath, an active multiplication of the cells of the 

 primitive sheath or neurilemma is taking place. These neuri- 

 lemmal cells probably play a nutritive action towards the more 

 important structures within them, and Graham Kerr, in a recent 

 study of nerve growth in the fish Lepidosiren, has supported in a 

 very conclusive and entirely independent way the view that Mott 

 and I advanced some years ago of the value of the neurilemma in 

 maintaining the nutrition of the axis cylinder. There is but little 

 doubt also that these cells act as phagocytes in the removal of the 

 degenerated products of the other portions of the nerve-fibre. But 

 after this is accomplished they elongate and unite into long chains. 

 It is this appearance that has led some observers into regarding them 

 as true nerve-fibres ; they have jumped to the conclusion that the 

 neurilemmal cells are also able to form a conducting core, and so 

 have regarded auto-regeneration as a histological possibility. But 

 all recent observations by the best methods, as I have already stated, 

 have failed to discover either an axial core or a fatty sheath in these 

 " embryonic fibres," as they have been termed. Howell and Huber 

 put it very well twenty years ago, when they said the peripheral 

 structures are able to prepare the scaffolding, but the axon, the 

 essential conducting core of the fibre, has an exclusively central origin. 



The change in the neurilemmal cells which occurs in the peri- 

 pheral segment is even more vigorous at the central termination of 

 the cut nerve ; here its nutritive function (or apotrophic function, as 

 Marinesco calls it) is effective, and provides for the nourishment of 

 the actively lengthening axis cylinders. At the peripheral end, 



