302 JouRNAL o¥ CoMPARATIVE NEUROLOGY. 
medullary tube, up to the period represented by the 9 day 
chick, all mitotic figures in the cord are around the canal; all, 
with a very few exceptions, in the brain are around the ven- 
tricle. He counted the mitoses in 25 sections of both cerebral 
hemispheres of a mouse embryo, 2-3 mm. in length and found 
50 ventricular mitoses to 1 extra-ventricular. In the cerebellum 
and basal ganglia, however, he finds the mitoses always scattered 
and he considers that this fact is connected with the rapid 
growth of these parts of the central nervous system. Wherever 
growth is slow, asin the cord and hemispheres, cell division 
takes place around the ventricle,—wherever it is rapid, as in 
the cerebellum, thalamus and corpus striatum, cell division oc- 
curs in all layers of cells. This division of cells, he finds, ceases 
in birds and mammals at about the time when a posterior me- 
dian fissure is clearly to be seen, or when the canal has become 
circular and the ependymal cells are provided with cilia. 
VIGNAL (5) takes a stand different from either of the above. 
He finds that in embryos of chicks, rabbits and sheep, the 
mitotic figures are only in the ependymal layer nearest the ven- 
tricle, but does not therefore conclude that cell division is con- 
fined to this region. On the contrary, he considers that the 
greater part of it takes place in the gray matter. He explains 
this apparent contradiction by assuming that these cells divide, 
not by mitoses, but by the formation of ‘‘achromatic figures.” 
Naturally a description of such figures cannot be given, but as 
the nucleus, when dividing enlarges and loses some of its 
chromatin, the author considers that it is possible to distinguish 
the dividing cells from the resting ones. He is led to this belief 
by his failure to find mitotic figures in the developing cord aiy- 
where except directly around the canal (although he treated 
his material by FLemmrne’s method), and as there is every evi- 
dence of a more rapid increase in the number of cells than can 
be accounted for by the few ventricular mitoses, it follows that | 
the cells must divide in some other way than that ususally 
described. 
ScHAPER (6) confirms ALTMANN’s dictum as to the ventric- 
ular predilection of mitoses, but only for the earlier stages of 
