174 J. B. JOHNSTON, 
at once evident on comparison of the accompanying photographs of 
these two structures (Phots. 8—13) which are taken from haemato- 
xylin sections. GORONOWITSCH also figures and describes (’89, fig. 46 ; 
96, figs. 14 and 9) fibres passing from the vagus lobe into the 
acusticum and others into the “Lobus trigemini”. I have shown 
that these fibres are only neurites from cells in these nuclei passing 
by the lobus vagi on their way to the base of the medulla. The 
only other similarity between the acusticum and the lobus vagi at 
present known to me consists in the fact that in the Mammalian 
brain the cells surrounding the fasciculus solitarius are said to send 
arcuate fibres to the raphe in the same way as do the cells of the 
VIII nuclei. But as we do not know the destination of these fibres, 
it is unnecessary to discuss this distant similarity between the two 
centers. Finally, the descriptions given in the earlier part of this 
paper (cf. page 83 ff.) have shown that the lobus vagi differs as 
widely as possible in its internal structure from the other centers 
under consideration. HERRICK (99) states as clearly as I have here 
and in my previous paper (98b) that the lobus vagi is entirely se- 
parated from the other sensory centers in the medulla. 
A comparison of the nerve elements described in the several 
nuclei will show that they fall into the following natural groups, 
each of which is represented in all the nuclei, with certain ex- 
ceptions noted below. 
1) Cells of medium or large size whose dendrites ramify freely 
among the afferent sensory fibres and whose neurites pass to the 
base of the cord or medulla to make connections with other centers. 
These are the larger cells found in the dorsal horn and nucleus 
funiculi, the medium and large cells in the acusticum with little or 
no connection with the cerebellar crest, and similar cells in the 
eranular layer of the cerebellum. 
2) Large cells whose dendrites are marked by characteristic small 
spines and are situated in the molecular layer. These are the PurR- 
KINJE cells of the acusticum, lobus lineae lateralis, and cerebellum. 
They are a modification of the cells of group 1) (cf page 191 ff. 
below). That these cells seem to be absent from the dorsal horn 
is due simply to the absence of the fine fibres in sufficient numbers 
to form a molecular layer and cause the development of the peculiar 
dendrites. | 
3) Small cells whose neurites form longitudinal tracts to end 
at higher or lower levels within the same set of nuclei. These 
