146 PROTOPLASM 



represent such structural relations in ganglion cells, I do 

 not for my part consider it necessary to give more. I will 

 only refer here to the very successful photograph (XIII.) of 

 a very thin section of a ganglion cell of Ltimbricus, which 

 permits the reticulate meshwork to be made out very plainly. 

 The preparation was stained with iron-hsematoxylin, which 

 also rendered distinct the numerous strongly stained granules 

 lodged in the nodal points of the network. These granules 

 can also be observed quite plainly in isolated and unstained 

 ganglion cells of this worm. In the latter I first observed 

 a perfectly definite radiate layer of alveoli at the surface of 

 the cell, as is depicted on Plate VIII. Fig. 3. Photograph 

 XIII. also shows this layer very well in places. That 

 the radiate layer of meshes occurs also round the conspicuous 

 nucleus of the ganglion cell, can be definitely seen in 

 isolated cells from the calf (Plate VIII. Fig. 2). 



As I have already pointed out, the fibrous or fibrillar 

 nature of the protoplasm only depends on the arrangement, 

 or rather on the extension and elongation, of the meshes. 

 There is indeed very often an appearance as if stronger and 

 thicker fibrils or " runners " (Frommann) passed through the 

 meshwork. In my opinion, however, this only depends, on 

 the one hand, upon the fact that the tracts of the frame- 

 work are more distinctly prominent according to the extent 

 to which they are ranked in a straight line one behind the 

 other, and on the other hand upon the close packing of the 

 granules, of which we have before seen clear examples. 



The processes arising from the ganglion cells always 

 have the structure of a meshwork forming distinct longi- 

 tudinal fibrils. In Fig. 1, Plate IX., I have represented the 

 structure of a broad protoplasmic process of a ganglion cell 

 of the calf, which was most especially clear and distinct. 

 I reckon the breadth of the meshes at about 0'8 /i. The 

 structure of such protoplasmic processes is always clearer 

 and sharper than that of the axis-cylinder to be described 

 later, which may well be the result of deposits of granules. 



For the view of Nansen (1886), namely, that nerve tubules 

 enter into the ganglion cell, spread themselves out as fibrous 

 tracts in its reticulate protoplasm, and finally emerge again, I 



