428 BOTANICAL GAZETTE [DECEMBER 
The daughter nucleoli are reconstructed at the expense of the double 
segments. These undergo an active vacuolization and are then con- 
densed into a nucleolus in which the two substance are again mixed 
together. 
According to other views, there may be more than one kind of 
structure called a nucleolus. CARNoy (12a) makes four groups of 
these structures as follows: nucléolus nucléinieus, nucléolus noyaux, 
nucléolus plasmatiques, and nucléolus mixtes, the first one being con- 
sidered as portion of the chromatin network and the third concerned 
with the formation of the spindle. Other authors (ROSEN 62, DAvIs 
22, Wr1son 82) have recognized two kinds of nucleoli, true nucleoli 
and chromatin nucleoli, the latter being considered entirely of chro- 
matin. 
In Polysiphonia the nucleolus lies in various positions within the 
nuclear cavity, and is not connected with the chromatin network. 
The chromosomes are formed from the gradual transformation or 
rearrangement of the substance of the network, and the nucleolus takes 
no part in their development. In Corallina also Davis (18) clearly 
distinguishes the chromatin bodies from the nucleolus. 
Spindle formation.—The spindle fibers in Polysiphonia are 
meagerly developed and of short duration. During the prophase 
of mitosis the nucleolus remains unchanged and the two poles are 
marked by deeply staining bodies, but I regret that I have not been 
able to trace the process of spindle formation. In the sporelings, 
where the nucleus is comparatively large in size and the spindles are 
more conspicuous, it was noted that short slender fibrils are attached 
to the chromosomes when assembled irregularly in the middle region 
of the nuclear cavity (figs. 3,17). These fibrils are the remains of the 
nuclear network and are the only fibrillar structures ever seen within 
the nuclear membrane at this time. The centrosome-like bodies, which 
at metaphase seem to become the centers of the well-differentiated 
centrosphere-like structures, have no radiation into the-cytoplasm, 
though in the case of sperm mother cells a few fibrillae may be seen 
running from the centrosome-like bodies at prophase (fig. 64). I 
have not observed the entrance of spindle fibers from the kinoplasmic 
centers outside of the nucleus, as has been reported by Davis (18) 
for Corallina. 
