110 J. B. JOHNSTON, 
from side to side and are traced as far as the region of the nucleus 
anterior, forming a part of the common longitudinal bundle of this 
region (page 112). Whether these fibres are related to the nucleus 
anterior or to some more distant nucleus, I can not decide. I have 
been unable to find either cells of origin or end-branching. In ad- 
dition to these three sets of fibres, there are a few fibres which 
leave this decussation to join the dorsal decussation of the tectum. 
I consider that this is due merely to an incomplete separation of 
the two decussations. This may account also for the fact that oc- 
casional fibres go from the epiphysial decussation to join the tractus 
tecto-cerebellaris I, since this tract is made up in part by fibres 
from the cephalic part of the dorsal decussation. 
In some of my later GOLGI preparations I have been fortunate 
enough to have nerve fibres in the pineal stalk impregnated. Although _ 
these preparations are not altogether satisfactory, I find a number 
of slightly varicose fibres which have a very direct course in the 
proximal portion of the pineal stalk. On reaching the base of the 
stalk these fibres turn laterally over the dorsal wall of the epi- 
physial sac at the cephalic border of the epiphysial decussation. 
Some of the fibres, at least, then turn cephalad and enter the 
ganglia habenulae. I have traced about an equal number into each 
ganglion. Whether all these fibres have the same destination I am 
unable to say. It is possible that some of them are in relation 
with the epithelial cells of the epiphysial sac. 
2. The Ganglia habenulae. 
In frontal sections through the epiphysial sac the bundles of 
MEYNERT appear at either side of the third ventricle a little ventral 
to the epiphysial sac, and surrounded by cells of the central grey 
matter or of the nucleus praetectalis (Fig. K). Passing forward, 
the cephalic wall of the epiphysial sac is seen to be formed by new 
masses of cells projecting in from either side, but much larger on 
the right. These are the cells of the ganglia habenulae. They are 
covered laterally (ectally) by a fibre layer which here consists chiefly 
of the bundles of MEYNERT. In the next two or three sections ce- 
phalad the pineal stalk rises dorsally, and immediately in front of 
and below it the fibre layer of the two sides unites to form the 
commissura habenularis (or commissura superior). The cell layer 
is here ventral and mesial (internal) to the fibre layer, although 
there are a few cells scattered among the fibres. Immediately in 
