The brain of Acipenser. 187 
paired Anlage in the cephalic end of the medulla. I have given 
evidence that the cerebellum is continuous with the acusticum and that 
it must be derived from the acusticum. Beyond this and the for- 
mation of the valvula by a transverse fold, the similarity between 
Acipenser and the Teleosts is probably slight. SCHAPER (94b) 
states that the valvula is not to be reckoned as a cerebellar structure, 
as it has not the characteristic histological structure of the cerebellum, 
but it is to be referred to the velum medullare anterius of higher 
forms. This interpretation will certainly not apply to Acipenser, 
since the structure of the valvula is identical with that of the body 
of the cerebellum, it receives fibres from the same sources as do 
other parts of the cerebellum, and its granule cells send their neurites 
into the molecular layer. The body of the cerebellum is formed in 
Teleosts by the great thickening of the lateral parts which eventu- 
ally fuse together in the median plane. In Acipenser, although 
' there is considerable thickening of the cephalic end of the 
acusticum (acusticum proper and ascending V nucleus of my de- 
scription) which is continuous with the body of the cerebellum, 
the structure of the body shows that it has been formed chiefly 
by an infolding along the median line, extending caudally from 
the mid-brain-cerebellar furrow. It appears probable that at an 
early stage of development the cerebellum of Acipenser consists o 
paired lateral lobes extending up from the cephalic end of the acust- 
icum and connected across the dorsal surface by a commissure 
(composed of neurites of granule cells). Such a stage would cor- 
respond closely with the condition in the 57 day trout embryo figured 
by SCHAPER (94a, figs. 5—7; ’94b, tab. 18, figs. 7—9). Further 
growth of these lateral portions results in Teleosts in thickenings 
which fuse in the median plane. In Acipenser further growth leads 
to an infolding along the mid dorsal line by which the outer sur- 
faces of the two lateral portions are brought into apposition, and the 
molecular layers thus brought together fuse into the common median ~ 
molecular mass and keel of the adult. The lateral portions also 
bulge out to form the lateral lobes. 
In Selachians ScHAPER (98) finds that the granular layer is 
not everywhere present. The neurites of the PURKINJE cells enter 
the granular layer either directly or after a longer or shorter course 
over the membrana limitans interna. He has found no collaterals 
although he notes that SAUERBECK (97) figures what may be col- 
lateral branching of PuRKINJE neurites. The cells of the molecular 
