170 JOURNAL OF COMPARATIVE NEUROLOGY. 
are larger than in the leech, and beads of perifibrillar substance 
are scattered along the course of the fibrillae, often at their 
points of union. These networks correspond more nearly to 
those which ApATHy demonstrated in the leech with methylen 
blue, but they are not so extensive. 
My observations thus confirm the statements of APATHY 
that fibrillar networks occur in both ganglion cells and neuropil. 
The evidence, however, does not support his view that the 
networks within the cells are formed by neurofibrillae which 
differ from each other in both structure and function. This 
conclusion is, moreover, in entire agreement with BETHE’s ex- 
periment which proved that the cells are not the centers of ner- 
vous activity as APATHY supposed. As to APATHY’S assump- 
tion of a diffuse fibrillar network in the neuropil, there is no 
evidence to show that such a condition exists. There are rather 
numerous small networks, each limited to a definite region in 
the neuropil, and putting comparatively few fibrillae into com- 
munication with one another. 
Fibrillar structures in my preparations of both H/zsudo and 
Astacus, confirm BETHE’s statement that many neurofibrillae are 
found in the nerve elements which are entirely independent of 
the ganglion cells. In A/zvudo all the neurofibrillae of a nerve 
fiber may be put into communication with one another by net- 
works before they enter the cell. 
These facts all point to the existence of fibrillar continuity 
between the nerve elements; in both H/zvudo and Astacus appar- 
ent cases of continuity between two neurones have been ob- 
served. Great weight cannot be laid on only two observations 
of such connection, for there is always the possible danger of 
optical error in tracing such exceedingly small structures. But 
the very existence of independent neurofibrillae in the nerve 
elements and the presence of fibrillar networks in the neuropil 
are incompatible with the idea that the nervous system is com- 
posed of anatomically independent, cellular units. On the con- 
trary such conditions can be explained only by assuming that 
between the nerve elements fibrillar continuity exists. 
The proof of such continuity does not, however, annihi- 
