AXIS-CYLINDERS IN SECTION 151 



however, T must state unhesitatingly that, after having 

 studied many such constrictions, I have observed, as a rule, 

 no narrowing of the axis-cylinder, but have been, on tin- 

 other hand, able to trace it as a strand of even and equal 

 thickness rm-ht through the constriction. 



o o 



I obtained preparations just as beautiful, in fact simply 

 splendid, showing the structure described in the axis- 

 cylinder, by making macerations of the spinal cord of the 

 calf. The macerations were done partly in 10-1 5 per cent 

 alcohol, stained yellow with iodine, partly in Miiller's fluid. 

 In these axis-cylinders, which, as is well known, can be 

 particularly easily isolated on account of their lacking the 

 sheath of Schwann, I have seen quite plainly that the 

 meshes are not rectangular, but, as was to be expected, 

 somewhat polyhedric, so that the apparent fibrilhe do not 

 run absolutely straight, but are lines with slight zigzag- 

 bends (Plate IX. Fig. 3). 



If now one takes into comparison the appearances shown 

 in cross sections of the axis-cylinder of the frog, the structure 

 already found is still further explained. The sections were 

 made from material treated either with picro-sulphuric 

 or picro-sulphuric-osmic acid, and had the thickness of about 

 I yu. They were partly stained in Delafield's haematoxylin, 

 partly with iron-haBmatoxylin, or in gold chloride as well, 

 and were examined in water or methyl alcohol. In such 

 preparations it may be made out in the plainest manner that 

 the axis-cylinder does not at all show the dotted appearance 

 of sections through isolated fibrils, but has a very beautiful 

 net-like meshwork in cross section (see Plate IX. Fig. 4, a-c, 

 and the Photographs XL and XII.). The nodal points of 

 the network stand out very prominently as a rule, and it 

 also seems as if some fine granules were lodged in them. 

 It is further of especial interest that the most external layer 

 of meshes in the network are here again distinctly directed 

 vertically to the surface of the axis-cylinder, and that the 

 internal meshwork is by no means always quite irregular, but 

 frequently exhibits a structure as of fibres running either 

 perfectly straight, or slightly sinuously. The meshes are 

 arranged just as we have frequently observed in other 



