106 
GEOKCtI'] K. NlCHOLLy, 
the ependymal epithelial cells into intimate connection, and 
that they are possibly in some way connected with the control 
of the lymph supply in the ventricles. 
Holm makes no reference to the occurrence of Reissuer’s 
fibre, although lie seems to have seen the isthmic canal, of 
which in Myxine it may be said that it has persisted uierel}" 
because it gives free passage to that structure. In the case 
w'hich he describes (’01, p. 369) the isthmic canal (his “upper 
canal ”) was evidently much less reduced than is commonly 
the case. 
He has apparently followed Retzius (’94) in his erroneous 
identification of the ventricle of the mid-brain as the fourth 
ventricle. As I have pointed out above, the cavity of the 
mid-brain is greatly reduced, its anterior portion being repre- 
sented only by the sub-cominissural canal, while the whole of 
that large space (optocoel) , which in the Petromyzontidae (Text- 
fig. 5) lies above and behind the posterior commissure in the 
roof of the mid-brain, is reduced in Bdellostoma (Text-fig. 7) 
to a more or less complete annular space, and in Myxine 
(Text-fig. 8) to a short, dorsally directed and blindly ending 
canal. Behind, beneath the posterior portion of the corpora 
bigemina, the aqueductus Sylvii is, in the Myxinoids, 
reduced to two narrow passages, the isthmic and ventricular 
canals. 
The main cavity, then, in the mid-brain ventricle, which, 
with the ventricular canal, is identified by Sanders,^ Holm 
' In justice to Sanders, whose work appears to liave been largely- 
overlooked, it should be pointed out that his observations were really 
remarkably correct, and that in many particulars he has anticipated the 
results of moi-e recent workers. Not only did he identify correctly 
nearly all of the various brain-ventricles, but later work has also 
justified his identifications of various parts of the brain. He continued 
the observations of Retzius as to the absence of a cerebellum. Curiously 
enough. Holm (’01, p. 378) misrepresents Sanders as interpreting the 
corpora bigemina as the cerebellum, a mistake which, in view of 
Sanders actual statements, is altogether inexplicable. Thus (’94, p. 6), 
Sanders says, “ It is remarkable that the cerebellum ... is here 
entirely absent,” and again (op. cit., p. 22), “Myxine appears to present 
