CENTRAL NERVOUS SYSTEM IN VERTEBRATE ANIMALS. 
937 
speck; they are usually pear-shaped, and have a process at each end; they measure 
about 0*006 millim. long by 0*0045 millim. broad. There is probably also a network 
of fine fibres in the molecular layer of the cerebellum, such as is found in the correspond¬ 
ing layers of the wings ; but I have not been able to add anything to what I remarked 
on this subject in my paper on the brain of the Teleostei.* 
The intermediate layer consists of the Purkinje cells and a bundle of longitudinal 
medulla,ted fibres of small size, which appear to join eventually the crus cerebelli. The 
Purkinje cells resemble those of the Teleostei in general, in being arranged parallel 
to the long axis of the lobe : they are shorter and thicker than the corresponding cells 
in M. cephalus; their usual size is 0*018 millim. long by 0*016 millim. broad; some 
are as long as 0*022 millim.; the nucleus is about 0*009 millim. in diameter, and the 
nucleolus 0*002 millim.; there is a circular space round the nucleolus, which is clearer 
and smoother than the remainder of the contents of the nucleus : these cells then 
resemble those of the motor horn of the grey matter of the spinal cord, in the disk-like 
nucleolus and the clear space which surrounds it. Each Purkinje cell has two processes, 
one thicker, corresponding to the protoplasmic processes of the cells of the spinal cord; 
the other finer, corresponding to the axis-cylinder process of the same. I never could 
perceive any prolongation of this into the nucleus, but on the other hand, the axis- 
cylinder process was in one case distinctly traceable into one of the medullated fibres 
of the longitudinal bundles belonging to the intermediate layer. This fact confirms 
the statement of Koschenikoff,! and also accords well with the theory which I 
suggested in the description of the cerebellum in M. cephalus, and which only required 
this arrangement to complete the statement. 
It now remains to consider the granular layer of the cerebellum; Denissenko 
described a method by which he thought he had demonstrated the existence of two 
different species of cells in this layer. During the process of double staining with 
eosin and hsematoxylin, he maintained that one kind of cell takes up one colour and 
another kind absorbs the other colour; he thus had the so-called eosin cells and the 
hsematoxylin cells. Whatever may be the case in higher animals, in Fishes there 
exists only one species of cell in this layer; I tried the experiment several times, 
carefully following Denissenko’s directions; but I never could find that the cells of 
this layer exercised any selection as to the material they took up, but the result was 
that all the cells assumed a more or less light purple uniform tint according as the 
eosin or haematoxylin solutions were the stronger : from this one must infer that there 
is no difference of composition in the cells forming this layer in Fishes. In shape they 
resemble those of the third layer of the cerebellum in M. cephalus ; they belong to 
the smallest category, generally measuring about 0*004 millim. in diameter; they mostly 
possess a nucleus and nucleolus; some are even still smaller, being 0*0034 millim. long 
by 0*003 millim., but the nucleus is of the same size as that of the larger specimens. 
* Loo. cit. 
t Axencylinder Vortsatz der Nervenzellen im Kleingehirn, &c., Arch. f. Mile. Anatomie, Bd. v., 1869. 
6 D 2 
