The brain of Acipenser. 95 
wise. That such is not the case is shown by the great difference 
in the size and character of dendrites of these cells as compared 
with the granule cells, and by the fact that the neurites run in the 
granular and not in the molecular layer. If these are not cells of 
the II type, they must be considered as cells which send their neu- 
rites downward to the medulla through the granular layer and the 
acusticum. 
The second variety of cells of the II type have very peculiar 
neurites (Phot. 39). The cells measure 10—22 by 20—80 u, are 
ovoid or stellate in form, and possess several dendrites which present 
in greater or less degree the roughness characteristic 
of PURKINJE cell dendrites. The neurite is medium sized and 
presents at intervals club-like thickenings. The neurite divides at 
the thickenings but is not richly branched. The smaller branches 
have numerous moniliform varicosities instead of the club-like 
thickenings, and often are very profusely branched at their termin- 
ations or present in a very vague way the appearance of baskets. 
The neurites vary in thickness in proportion to the size of the cell. 
Although these cells vary greatly in size and in the degree of de- 
velopment of the spines peculiar to the dendrites of these and the 
PURKINJE cells, the character of the neurites seems sufficient basis 
for classing them together as a special variety of II type cells. They 
are most numerous in the part of the valvula and body near their 
junction with the tectum and lateral lobes (cf. page 190). 
8. The lateral Lobes. 
In the lateral lobes the granular layer is much greater in vol- 
ume than the molecular layer, and the former is only in part co- 
vered externally by the latter. Upon the latero-cephalic free surface, 
as noted above, no molecular layer is present. The molecular layer, 
composed as it is of fine fibres which reach the dorsal surface from 
the median keystone portion of the body, is narrower in cephalo- 
caudal direction than the surface of the lateral lobe. It thence 
continues laterad along the dorsal surface of the lobe and, growing 
narrower as it proceeds, it turns caudad to become the cerebellar crest 
on the dorsal surface of the acusticum. The number of fibres in the 
molecular layer when it leaves the body is many times greater than the 
number in the cerebellar crest. What becomes of the fibres between 
these two points is not easy to determine by direct observation. 
One large bundle leaves the cerebellar crest just in front of the 
