CENTRAL NERVOUS SYSTEM IN VERTEBRATE ANIMALS. 
759 
The latter author and Koschenikoff (75) demonstrated that they join double con- 
toured nerve-fibres, — a fact which would accord well with the above-mentioned views, 
and would imply that they join, not the small cells of the third, but the fibres of the 
fourth layer, and thus a junction would be formed in a roundabout manner between the 
minute elements of the third layer and the commissures of the other parts of the brain. 
According to Hadlich (77) the arrangement is similar to that above given, but he 
does not explain what becomes of the finer fibres of the Purkinje cells. Qber- 
steiner (78) describes small cells or corpuscles in the outer or first layer in the cere- 
bellum of the foetus, to which the fibres from the broad processes of the Purkinje 
cells are attached, — a fact which Sankey (83) has confirmed in the adult. I have not 
found any cells of this description in the external layer of the cerebellum in Teleostei ; 
but their presence would not interfere with the interpretation adopted here, for the 
fibres in question could yet pass through those corpuscles, and behave as above 
mentioned. 
The whole cerebellum in the Teleostei appears to correspond to a single lobulus in 
that division of the brain in the human subject. 
In Mugil cepl talus there are indications of the commencement of a second lobulus, 
which in Labrax Lupus are more decided. 
There is a vertical fissure in the central line of the cerebellum in Mugil, forming a 
ventricle, which extends from the posterior end as far forward as the posterior edge 
of the crura cerebelli ; it has an appearance as if the two halves had not properly 
united. 
The Valvula Cerebelli. 
The valvula cerebelli has precisely the same structure that the cerebellum itself 
has; the layer of straight fibres, the layer of Purkinje cells, and the layer of minute 
cells are all present : the only difference to be seen lies in the arrangement of these 
strata. 
The layer of straight fibres is a direct continuation of the external layer of the 
cerebellum, which, passing forward into the ventricle of the optic lobe nearly as far 
as its anterior extremity, turns back on itself and forms a fold ; so that the edges 
which would be external in the cerebellum face each other, leaving a narrow fissure 
between them which communicates with the external surface between the posterior 
end of the tectum, and the anterior surface of the cerebellum, by a foramen which is 
closed in by pia mater and gives access to vessels. The Purkinje cells bear the same 
relation to the layers of straight fibres that they do to the external layer of the cere- 
bellum. Then the mass of small cells resembling those found in the third layer come 
to be placed on the surface facing the ventricle of the optic lobe, and extend as wings 
on each side. In the Mugil these wings are simple rounded bodies, but in the Labrax 
a deep transverse fold of straight fibres divides them into two lobes, and in the Creni- 
labrus two such folds divide them into three lobes. This also appears to be the case 
MDCCCLXXVI1I. 5 e 
