196 J. B. JOHNSTON, 
the surface. These so-called neuroglia cells are probably my type 
A cells, the neurites of which Fusari has missed. 
3) Layer: containing numerous large stellate cells and bundles 
of nerve fibres. The neurites of the cells join the bundles of nerve 
fibres in the same layer. These cells correspond to at least a part 
of my cells D. A large part of the fibres of this layer belong, ac- 
cording to Fusari, to the Stabkranz. These fibres, if they exist, 
must be the Mantelbiindel of EDINGER (92) in Selachians, whose 
fibres cross close behind the chiasma and join the lateral part of 
the pallium with the tectum (?). I have described above (page 128) 
a small bundle of coarse fibres to which I have given the name of 
Mantelbiindel, but these would represent only a very small fraction 
of the fibres of the middle zone in Acipenser. It therefore seems 
to me that Fusarı was dealing not with fibres to the fore brain 
(Stabkranz) but with the tractus tecto-lobaris. 
4) A thick layer of spindle and oval cells, some with short and 
others with long neurites. The long neurites join the fibre bundles 
of the third layer. The cells with short neurites doubtless correspond 
to those of my cells A which are removed from the central cavity. 
The cells with long neurites probably correspond to part of my 
cells D, the disposition of both dendrites and neurites being similar. 
5) Layer of optic fibres, mostly longitudinal. 
6) A thin layer of round cells, few in number. 
7) A superficial vascular layer beneath the pia. 
BURCKHARDTS results are too meager to require attention. 
VAN GEHUCHTEN (94) has described the tectum of the trout 
as composed of three layers: ependymal, granular, and molecular. 
The granular layer is much thicker in proportion to the molecular 
than in Acipenser, due probably to the embryological character of the 
trout material. VAN GEHUCHTEN describes four types of nerve cells 
which he says were not described by Fusarı. These are: 
1) Ovoid cells in the granular layer with single vertical dendrites 
from which the neurites go off to join a layer of fibres separating 
the granular and molecular layers. These correspond to part of 
Fusarr’s spindle-shaped glia cells and to my cells B. 
2) Similar cells with single vertical processes which VAN GE- 
HUCHTEN considered to be neurites, without dendrites. These are 
my cells A, the single process being the dendrite and the neurite 
has not been impregnated in VAN GEHUCHTEN’s preparations. 
3) Pyramidal or stellate cells with ascending dendrites and neu- 
