or plates, just inside the sheath. The only methods which affords 
satisfactory staining of these processes is the chromo-silver method, 
but by its use I have betn able to trace them to their extreme 
terminations (vide fig. g4, ppr; 104 — 105). I have, however, also in 
preparations stained with borax-carmine (fig. 93) or stained according 
to Heidenhains hæmatoxylin method (fig. 95) been able to trace 
the protoplasmic processes with their branches through the white 
substance and very close towards its periphery if not entirely to 
the external sheath. 
The same ganglion cell may have protoplasmic processes, the 
various branches of which penetrate to the periphery, both on the 
ventral side and on the dorsal side of the nerve-cord (vide fig. 93, 
9S, 105, 106). 
The slender protoplasmic processes and their branches resemble 
the processes of the neuroglia-cells very much in their aspect and 
it is indeed extremely difficiilt to draw any distinction between them, 
Both have the same appearance, both come from the grey substance, 
and both traverse the white substance towards the periphery where both 
as a rule terminate under the sheath enveloping the spinal cord. In fig. 
109, a protoplasmic processes issues from the small ganglion cell gc, but 
this process can not be distinguished from the many neuroglia-fibres, /, 
in any respect but that it issues from a ganglion cell.') A difference is 
sometimes that the neuroglia-fibres have a straighter course more 
directly towards the periphery of the white substance than the pro- 
toplasmic processes have. I have found no constant and reliable 
difference in the chromo-silver staining of these processes similar to 
what GOLGI has stated as regards the neuroglia cells and proto- 
plasmic processes of the Mammalia, and what I also have had 
opportunity to observe myself in higher Vertebrata. Sometimes 
there was, however, also in Myxine a tendency towards a more red- 
dish staining of the processes of the neuroglia cells. 
The branches of the protoplasmic processes of one cell do not 
anastomose, neither do the processes from various ceUs anastomose. 
Though I have examined a considerable number of preparations, 
and although they have been most perfectly stained, in this respect, 
I have not yet observed a single case of indubitable anastomosis 
between protoplasmic processes. 2) I believe, thus, that I am entitled 
■) Vide also fig. 93,gCo,. 
^) In my previous paper on Myxine (1. c. 1886) I have mentioned one 
case from the brain vvhere I was in doubt whether there was an anastomosis or 
not. The more [ have examined this preparation, the more do I doubt that an 
