CTJTAlTEOTrS NEEVE-TEEMIlirATIONS IN MAMMALS. 583 



different cells being connected with each other and with the 

 central nerve-centres by one or several fibres. Once there, what 

 seems to occur is as follows. The epidermis is, as is well knoAvn, 

 continually being renewed by the addition of new or younger 

 cells, probably by the addition of wandering or embryonic cells 

 to its lower surface. If these living wandering cells apply them- 

 selves at a point where lies a nerve-fibril connecting two cells in 

 a ganglionic group, that nerve-fibril probably becomes entangled 

 in the epidermis. Once fairly entangled, the fibril cannot free 

 itself; for the continuous development of young cells on the 

 lower surface of the epidermis keeps going on, and the fibril 

 becomes only more deeply involved among the epidermic cells. 

 The younger cells keep pushing the fibril before them towards 

 the free surface ; and as the ganglionic basal cells are com- 

 paratively fixed, the fibril connecting them is forced to elongate 

 itself, and becomes bent, with its convexity towards the free sur- 

 face. As tlie fibril becomes pushed further on, we find it appearing 

 like a narrow arch (as in if, fig. 19, Plate XV.) resting on two 

 high piers or columns, each of which has its base upon or in a 

 nerve-cell in the ganglionic group. By-and-by the arch gives 

 way through extreme tension, as at y, fig. 19, or through being 

 carried to the free surface of the epidermis, leaving in either case 

 the two columns standing as two fibrils perpendicular to the 

 surface. Every stage of the process as we have described it can 

 be seen in one or other of our preparations. Even when the 

 arch is broken, there is still no chance for the fibrils to retract 

 upon their respective base-cells ; for they are as firmly held 

 between the epidermic cells growing outwards as a rope would 

 be if grasped and drawn along between a series of toothed wheels 

 moving in one direction ; and so the fibrils continue to be drawn 

 out in spite of themselves, to moulder and crumble away at their 

 free ends in the inert portions of the epidermis, and to be thrown 

 off with the dead epidermic cells. 



But the interesting process which we have described is not 

 merely confined to the fibrils connecting the cells in a ganglionic 

 group, which do not form the hundredth part of the iutraepidermic 

 fibrils seen in any locality. The advantage of watching the pro- 

 cess going on in such a group arises from the fact that the cells 

 form a comparatively fixed point of observation, whose relation 

 to the intraepidermic offshoots from them can always be calculated 

 Once, however, the process has been verified in such a locality, it 



