190 J. B. JOHNSTON, 
there are a few cells of the II type which serve for further spreading 
of the impulses, and some probably for making connections directly 
with the bodies of PURKINJE cells (see below). The PuRKINJE cells 
send their neurites beyond the limits of the cerebellum to make 
connections with the motor apparatus. In Teleosts, according to 
SCHAPER’S account, the ascending fibres must make connections 
directly with the PURKINIE cell dendrites or at least with the II 
type cells of the molecular layer. In this case the granule cells, 
and indeed the whole granular layer, seems to have no function to 
perform. It is of course possible to suppose that the impulses are 
first received by the large oval cells which lie among the PURKINJE 
cells and send their dendrites into the molecular layer, and are 
transmitted to the granules by the short neurites of the oval cells. 
If so, the further course of the impulses would be as traced above. 
But SCHAPER is by no means sure that the cells in question are 
cells of the II type, and he has not described them in Selachians. 
If they are cells of the II type and if their neurites end in the 
granular layer, it does not appear that they are numerous enough 
or evenly enough distributed to receive all in-coming impulses. Aside 
from physiological difficulties it does not seem to me probable that such 
great differences can exist in the minute structure of the cerebellum 
in Teleosts and Ganoids, and I am theorfore inclined to the opinion that 
ScHAPER has been mistaken in his identification of ascending fibres 
and that at least the greater number of such fibres must end in 
the granular layer as in Acipenser. 
SCHAPER mentions one kind of cell whose neurite is lost among 
the PURKINJE cells, which he thinks may correspond to the Mam- 
malian Korbzellen. In seeking an explanation for the peculiar va- 
riety of cells of the II type described on page 95 I have been struck 
by the close similarity between the neurites of these cells and those 
of the Korbzellen. A comparison of Phot. 39 with KÖLLIKER’S 
fig. 535 (96) will serve to bring out the resemblance. The thickenings 
of the neurite and the manner of branching and the formation of 
indefinite baskets by the terminal branches all suggest strongly that 
these are true Korbzellen. 
The course and destination of the PURKINJE neurites has been 
a most perplexing question. It has been impossible to trace them 
far enough from their cells, owing to their sinuous course and the 
irregularity of the cell bodies, to determine in which layer theyrun. So 
far aS my preparations go they lead me to think that the fibres in 
