The brain of Acipenser. 79 
are similar to the PURKINJE cell in Phot. 38; two are imperfectly 
shown in the upper part of Phot. 36. These cells are thus provided 
with two sets of well developed dendrites which must have different 
functions, since one set is in relation with the fine fibres of the 
cerebellar crest while the other ramifies in the substance of the 
acusticum. The neurite arises from the side of the cell or from the 
base of one of the dendrites. 
There is a gradual transition from the cells last described to 
cells whose dendrites do not reach the cerebellar crest and which 
do not present any of the characters of PURKINJE cells. These are 
found in all parts of the acusticum and are smaller than the PURKINJE 
cells, measuring 6—22 by 16—112 uw. They may be considered 
as typical or primitive cells of this nucleus. They have ovoid, 
pyramidal, or irregular cell bodies with two or more dendrites. 
Their neurites become arcuate fibres to the opposite side of the 
medulla. 
In the median portion of the acusticum are a considerable 
number of fusiform cells with relatively small bodies, measuring 
12—16 by 32—80 u. From either end of the cell arises a single 
thick and very long dendrite, the long axis of the dendrites and 
cell being usually parallel with the internal surface. 
Finally, there are present at least in the cephalic portion of 
the acusticum a few cells with short neurites. The cells measure 
12—16 by 20—28 u, they are provided with two or three somewhat 
profusely and irregularly branched dendrites, and their neurites 
break up in end-branches within the acusticum. 
I have mentioned above that the neurites of certain cells in 
the acusticum form a small bundle accompanying the spinal V tract. 
The neurites of all other cells in the acusticum, except those of the 
II type, take one of two courses. The great majority of neurites 
of the three classes of cells described — PURKINJE, transitional, 
and primitive acusticum cells — run to the base of the medulla. 
These fibres are only occasionally impregnated. They are gathered 
in small bundles or run singly and can be traced with some cer- 
tainty by means of the darkened medullary sheaths. Some im- 
pregnated fibres I have traced across the ventral raphe, where they 
turn cephalad, but I believe that a considerable number turn for- 
ward without crossing. The fibres do not at once form a distinct 
bundle, but forward from the VIII nerve a bundle can be recognized 
which grows in size and compactness, gives off a bundle which 
