1910] YAMANOUCHI—CHROMOSOMES IN OSMUNDA 7 
and thickening proceeds until metaphase, when the structure, instead 
of being of irregular density, becomes evenly compact and homo- 
geneous. The two chromosomes of the pair separate in metaphase 
(fig. 13) and proceed during anaphase to the poles. But before they 
reach the poles, there occurs a genuine longitudinal splitting of the 
univalent chromosomes in preparation for the second division. The 
splitting generally does not proceed throughout the whole length 
of the chromosomes, one end remaining unsplit and the parts already 
divided diverging so that there naturally results a V-shaped chromo- 
some. ‘Therefore in the heterotypic division there is no longitudinal 
splitting of chromosomes. The two chromosomes lying side by side 
simply separate at metaphase; there is, of course, a single longi- 
tudinal splitting in metaphase of the heterotypic division, but this 
is a provision for the second division. 
The V-shaped chromosomes in late anaphase of the first division 
(jig. 14) gather into a group at the pole. Vacuolation occurs and a 
nuclear membrane is formed. 
FORMATION OF CHROMOSOMES IN HOMOTYPIC MITOSIS 
The group of vacuolate chromosomes is distinctly recognizable 
after the formation of the nuclear membrane. As the vacuolation 
proceeds farther, the chromosomes become very alveolate, but as 
the process is more active in the lateral parts of each chromosome, the 
central part remains as a rather thick strand, so that for a considerable 
period after the organization of the daughter nucleus the V-shaped. 
chromosomes could be traced with perfect distinctness. 
Progressive vacuolation with the accompanying nuclear growth 
tends to change the general aspect of the ragged chromatin network 
of the newly formed nucleus. The process, however, before proceed- 
ing so far as to result in a resting stage, begins to reorganize the 
chromosome primordia out of the ragged chromatin reticulum. 
The chromosome primordium appears first in V-shape (fig. 15), 
but the location does not seem to agree exactly with the location of 
the daughter chromosomes as last seen in the previous telophase. 
In the nuclear cavity, during the period intermediate between the 
last telophase and the present prophase, there might have occurred 
some movement of parts of the chromatin network. But the uni- 
