306 JouRNAL oF ComparATIVE NEUROLOGY. 
only, and at first sight one does not notice any striking differ- 
ence, but a more careful study reveals almost as many types of 
dividing cells here as in the later specimens. It will be more 
convenient to describe these cells as they are found in the differ- 
ent cell layers, rather than at the different stages of develop- 
ment, for the character of the dividing cells is very much the 
same in the corresponding localities in both early and late 
specimens. 
Taking first the ependyma of the cord, we find that the 
great majority of the dividing cells are of the kind described 
by His as germinal cells,—round or oval cells, measuring, in 
the white rat, from 7.5 X 6 to 12 X 10.5 p (Fig. 5). These 
cells usually lie directly bordering on the canal, but in the cord 
of the foetus they may be the distance of one or two cell 
layers from the edge. In the cord after birth the germinal cells 
are usually smaller than in the foetus. These germinal cells 
form the majority of all dividing cells in the foetus and are 
found in appreciable numbers around the ventricles in the later 
specimens. They are found in all the stages of mitosis, but, 
except for the different arrangement of the chromosomes, they 
show no great variations. There are, however, occasionally 
dividing cells in the ependyma of the cord which cannot be 
classed with these typical germinal cells, but which differ greatly 
from them. These are long cells, much larger than the typical 
germinal cells of His, pointed at one or at both ends and in 
several instances sending a process to the internal limiting mem- 
brane of the canal (Figs. 6, 7), the cell shown in Fig. 6 
measuring 24 X 7.5. These long pointed cells are situated 
always at the ventral and dorsal extremities of the canal, in the 
region of the ventral and dorsal plates of His, where, accord- 
ing to His, no neuroblasts are formed, the ventral plates be- 
coming converted into neuroglia, the dorsal filled with spon- 
gioblasts. Certainly these cells are not the germinal cells of 
His, but resemble in every way, except their great size, the 
spongioblasts which are so abundant in these regions... It is 
interesting to note in this connection that, as the development 
of the cord proceeds, the dividing cells around the canal tend 
