The brain of Acipenser. 175 
are to be regarded as connective elements, establishing connections 
between successive segments of the cerebro-spinal axis. These are 
the granule cells in the cerebellum and the small cells found in the 
dorsal horn in all Vertebrates which send their neurites into the 
dorsal tracts, where they bifurcate to run up and down the cord to 
re-enter the grey substance at other levels. The absence of these 
elements in the acusticum is to be explained by the presence of the 
cerebellar crest. A large part of the root fibres of the nerves runs 
through the acusticum to the cerebellum, and the granule cells have 
become largely or wholly restricted to the cerebellum and send their 
neurites to the PURKINJE cells of the acusticum. 
4) Cells with short neurites. I have described these for the 
cerebellum and both parts of the acusticum in Acipenser, and they 
have been found in the cerebellum and dorsal horns in many 
Vertebrates. 
There is, then, a remarkable similarity — amounting, as it seems 
to me, to a unity — of structure throughout the entire region of 
the dorsal horn, acusticum and cerebellum. It has already been 
pointed out by GORONOWITSCH, KINGSBURY, HERRICK, and myself 
that the cerebellar crest is simply a part of the molecular layer of 
the cerebellum. I have also shown that the acusticum is a con- 
tinuation of the granular layer of the cerebellum. 
The course taken by the neurites which leave these nuclei to 
go to other parts of the brain or cord is not so well understood 
as the points above mentioned. However, from the nucleus funi- 
culi and acusticum the neurites form internal arcuate fibres. The 
course of the neurites of the corresponding cells in the cerebellum 
(PURKINJE cells) has become secondarily modified in all probability 
in all Vertebrates. 
I think it has now been shown satisfactorily that the dorsal horn 
of the cord is continued in the hind brain by the acusticum and 
cerebellum, that all these parts must have had the same structure 
in the primitive Vertebrate brain, and that in the brain of Acipenser 
there still exists a relatively near approach to this primitive condition. 
It is to be noted further that there is an almost perfect functional 
isolation of these centers from the lobus vagi. The centers under 
consideration receive the endings of all general and special cutaneous 
components (except those to end buds) and send their secondary 
fibres chiefly to the tectum. The only connection whatever between 
these centers and their tracts and the lobus vagi and its tracts in 
