146 JOURNAL OF COMPARATIVE NEUROLOGY. 
in the same size, form, appearance, and distribution. I also have 
given some evidence in an earlier paper, that the neurosomes can 
hardly be regarded as an artificial product, and so far as modern 
histological technique is concerned, there is no reason to favor 
BETHE’S view on this point. 
However, I cannot agree with HELD who thinks that there 
are two kinds of network, one nervous and the other non- 
nervous, and that the former may coincide with BETHE’s Golgi- 
netz. HELp’s non-nervous network has been demonstrated by 
using the molybdenum technique of BrErHr. According to 
HELD, this technique brings out so-called ‘‘Gliaschnirringe ” 
which send filaments to form the non-nervous network. This 
point seems to me quite doubtful, for if the Gliaschnuirringe as 
well as the non-nervous or neurokeratin network were stained, 
one would expect to find neurokeratin structures along the 
medullary sheath. The latter, however, were not observed by 
him. My own technique, which is able to stain both neuro- 
keratin and glia, fails to show the existence of the non-nervous 
or neurokeratin network about the cell-body. One can see fre- 
quently the neuroglia fibers which surround the cell-body ; 
these, however, merely pass over the cell-surface and never stop 
on it. I noticed very often the neurokeratin rings or ‘‘Gliasch- 
nirringe’ of the medullary sheath in my preparations but was 
unable to observe any process from these rings such as has 
been described by Hetp. Therefore I am unable to confirm 
HELD on this last point. 
From the above I believe that there is only one kind of 
pericellular network which is formed by the terminals of the 
axones, and further, I believe that GoLGi's or BETHE’S network 
is identical with the network described by HELD as the nervous 
network from the (1) coincidence or indentical distribution, and 
(2) from the reticular structure observed in all cases. A differ- 
ent appearance shown by different technique, may be due to 
the fact that the Gore! and BerHe methods produce a precipt- 
tation of silver chromate and molybdenum salts respectively, 
which covers the finer meshwork, thus obscuring the internal 
structure and giving it a coarse appearance. 
