1906] YAMANOUCHI—POLYSIPHONIA VIOLACEA 421 
Centrosphere-like structures also appear unmistakably at the meta- 
phase of this mitosis (fig. 142), as well as in the previous mitosis 
(fig. 137). : 
The tetraspore mother cell (fig. 147) increases rapidly in size, 
soon becoming very much larger than the stalk cell. With the 
growth of the cell the nucleus also increases in size, and in the resting 
condition shows a conspicuous linin network. This network con- 
sists at first of lightly staining anastomosing threads, having knots 
here and there which stain a little darker (figs. 147, 147a). The 
cytoplasm presents a fine granular structure, with small alveoli. 
The nuclear network undergoes gradual change in such a manner 
that the lightly staining threads become somewhat thicker, and the 
knots grow into large irregular masses (figs. 148, 148a). Some por- 
tions of the threads connecting the knots become thinner and more 
slender, at last fading away; while other portions of them become 
thicker, and then the knots gradually diminish in bulk; so that, by 
and by, the anastomosing chromatin network becomes transformed 
into long continuous threads of irregular thickness, finally broadening 
into ribbons (figs. 49, 149a). 
hese chromatin threads or ribbons derived from the network 
Now spread and become distributed throughout the nuclear cavity in 
4 continuous and tangled fashion, presenting no free ends. It is diffi- 
cult to decide whether there is a single continuous thread or a double 
structure, but probably the latter condition is present, for the threads 
Senerally run side by side in pairs (figs. 150, 150a). Most of the 
chromatin threads become more tangled, twisted, and massed at one 
- Side of the nuclear cavity, only a few traversing the cavity to the 
Opposite side of the nuclear membrane. Synapsis is now generally 
believed not to be an artifact, and a careful study of this stage in 
Polysiphonia convinces me that the uniform chromatin threads which 
Tun parallel in pairs actually fuse into a single thread in certain por 
tions, although at the same time they may be separated in other parts. 
All of this probably means that the two continuous threads resulting 
from the transformation of the chromatin network are of distinct 
origin, paternal and maternal, and that they come in contact where 
they run closely parallel, and finally fuse together in the tangled and 
Contracted condition of synapsis, according to the recent interpretation 
