cxlviii KEKVOUS SYSTEM. 



Three modes of connection of cells with fibres are described. 1. From a cell, 

 which may have several branched outrunners, one stout unbranched process is con- 

 tinued into a nerve-fibre, at first naked, and probably representing only the axis- 

 cylinder, then acquiring a medullary sheath and dark borders, and finally a mem- 

 branous tube or primitive sheath. 2. From one or more finely divided branches of 

 a cell, or of more than one cell, equally fine fibrils are prolonged, which coalesce 

 into a pale fibre, having the characters of an axis-cylinder, which then, as in the 

 former case, may in its progress become a dark-bordered medullated fibre. 3. The 

 extreme ramifications of a cell or cells become connected, as in the last case, with 

 fibrils, which join into a nerve-fibre ; but the connection takes place by the inter- 

 vention of small bipolar cells, which are by one pole continuous with the branches 

 of the larger cell or cells, and by the other with fine fibrils which join into a pale fibre, 

 or into an axis-cylinder of a dark-bordered fibre. Gerlach, and after him Waldeyer 

 and others, have described this last mode of connection, as seen by them in the 

 cerebellum. In the cortical grey matter of the cerebellum there are well known 

 large cells generally with one undivided process directed centrally, and two or three 

 finely divided branches towards the surface (fig. LXXIX.). Scattered in the neighbour- 

 hood of these large cells, and also collected in a layer named the stratum ferrugineum, 

 or rust-coloured layer, are numerous small cells, often called granules (fig. LXXX. c) ; 

 and it is alleged by the above named authorities that fine ramifications of the large 

 cells join neighbouring small cells or pass inwards to join those of the stratum 

 ferrugineum, and that the small or intermediate cells are, on the other hand, con- 

 nected with filaments which coalesce into nerve-fibres as above described. This 

 statement derives support from the important observations of Mr. Lockhart Clarke, 

 on the structure of the olfactory bulb. Along with this indirect connection through 

 small intervening cells, Gerlach supposes that a process or processes of the large cells 

 pass directly into nerve-fibres ; and should such direct connection take place by the 

 prolongation of an unbranched cell-process into a nerve-fibre, the arrangement would 

 be analogous to that in the ganglia ; the simple origin, representing that of the 

 straight fibre from the ganglion-cell, whilst the ramified origin, with the intervening 

 small cells, might be compared to that of the superficial or spiral fibre, with its 

 interposed nuclei. 



The fibres of origin of a nerve, whether deeply implanted or not, on 

 quitting the surface of the brain or spinal cord to form the apparent origin 

 or free part of the root, are in most cases collected into funiculi, which are 

 each invested with a sheath of neurilemma. This investment is generally 

 regarded as a prolongation of the pia mater, and in fact its continuity with 

 that membrane may be seen very plainly at the roots of several of the 

 nerves, especially those of the cervical and dorsal nerves within the verte- 

 bral canal, for in that situation the neurilemma, like the pia mater itself, is 

 much stronger than in the cranium. The funiculi, approaching each other 

 if originally scattered, advance towards the foramen of the skull or spine 

 which gives issue to the nerve, and pass through the dura mater, either in 

 one bundle and by a single aperture, or in two or more fasciculi, for which 

 there are two or more openings in the membrane. The nerve roots in their 

 course run beneath the arachnoid membrane, and do not perforate it on 

 issuing from the cranio-vertebral cavity ; for the loose or visceral layer of 

 the arachnoid is prolonged on the nerve and loosely surrounds it as far as 

 the aperture of egress in the dura mater, where, quitting the nerve, it is 

 reflected upon the inner surface of the latter membrane, and becomes con- 

 tinuous with the parietal or adherent layer of the arachnoid. The nerve, 

 on escaping from the skull or spine, acquires its external, stout, fibrous 

 sheath, which connects all its funiculi into a firm cord, and then, too, the 

 nerve appears much thicker than before its exit. The dura mater accom- 

 panies the nerves through the bony foramina, and becomes continuous with 

 their external sheath and (at the cranial foramina) with the pericranium ; 



