Shinkishi Hatai 11 



of one neurone enter into the nerve processes of other neurones, and thus two neurones 

 become continuous by means of these primitive fibrils. The observations were made 

 on Crustacea. Apathy's (1897) observation on the lower animals (annelids) contradicts 

 radically the neurone doctrine, for he was able to follow the primitive neuro-fibrils 

 which come from one ganglion cell and enter into the cell-body of another element, 

 where they become fused with the protoplasm. Anastomosis of the axones with den- 

 drites has been observed by several other investigators, for instance, Ballowitz (1893), 

 Heymansand Demoor (1894), and others; but in these cases always in the peripheral 

 system. Thus Gerlach and Golgi's hypothesis of a diffuse network of the nerve- 

 processes has been revived through a more careful investigation of modern 

 neurologists. 



It is impossible at the present moment to say which of these views is correct, since 

 we do not know absolutely which technique shows the tissue in most nearly normal 

 condition. But after examining the results obtained by several investigators, it seems 

 to be quite reasonable to say that Golgi's silver-nitrate technique is not effective 

 enough to bring out the minute structures of the neurones, and, further, it acts on the 

 tissue so irregularly that in some cases even the same tissue in the same condition 

 presents a widely different appearance. In addition, the ordinary silver-bichromate 

 method does not show the internal structure of the neurones. Consequently, for the 

 purpose of this discussion, results obtained by Golgi's technique can hardly be con- 

 sidered as at all conclusive. 



As has already been mentioned, the nerve-cells in the white rat present a fibrillar 

 structure owing to the parallel arrangement of the neurosomes. This structure, how- 

 ever, is merely a modified reticulum which has been very much elongated. In some 

 cases several of these parallel lines of the elongated meshes combine together and form 

 very thick strands. Further, these united filaments or strands are found throughout 

 the cell-body, forming a very complicated network. In the case of the dendrites these 

 united filaments are noticed most frequently. Now, comparing these figures with that 

 of Golgi's endocellular network previously mentioned, one might expect the two figures 

 to be identical, for this anastomosis of combined filaments in the cells in the white 

 rat occurs only around the nucleus and in the neighborhood of the axone hillock, 

 not in the hillock itself, and never occurs along the periphery of the cell-body, where 

 wide meshes of a polygonal shape are alone visible, Golgi's endocellular network has 

 a similar distribution within the cell-body. A similar arrangement has been observed 

 in the cells of cerebral and cerebellar cortex. The only difference between Golgi's 

 results and those of the present writer is that Golgi's network is much simpler than 

 the latter and does not show any minute meshes formed by delicate filaments. This 

 dift'erence is due very probably to an insensitiveness of Golgi's technique, so that it 

 does not bring out these minute structures. 



It has been suggested by some investigators that Golgi's endocellular network 

 might represent the system of the intracellular canaliculi of Holmgren. If Golgi's 



187 



