a 
1913] YAMANOUCHI—HYDRODICTYON AFRICANUM 77 
was found in fixed material. Consequently, the final product of 
the coenobium may be gametes, which escape from the cell and 
after fusion form resting spores. These directly or with the inter- 
polation of a polyeder stage may repeat the cycle. 
Not only does this form differ from H. reticulatum in size and 
habit, but there are points of fundamental difference in the structure 
of the cell, the most remarkable of which are the presence of numer- 
ous definite plastids and the formation of starch, not by the frag- 
mentation from the pyrenoid, but in the plastids. These characters 
which differ from H. reticulatum seem to us to be a sufficient basis 
for a new species, which may very properly 
be called Hydrodictyon africanum. 
The general cell structure, nuclear division, 
and starch formation deserve a brief account. 
The protoplast of the coenobium is surrounded 
by a thick wall and occurs as a thin parietal 
layer, a large vacuole occupying the center. 
The cell wall consists of three different layers. 
The protoplasm may be conveniently de- 
scribed as consisting of three parts: Haut- 
schicht, the outermost part, directly in contact fe 
with the cell wall; plasma membrane, the =. 39 
innermost part which lines the central ¥ my ee 
vacuole; and between these two the main is a Sank of o ec 
cytoplasmic body, which contains numerous tion of a single coeno- 
nuclei, plastids, pyrenoids, starch grains, and ium: fixed May 20, 
small vacuoles (fig. 2). partes 
The nucleus varies greatly in size at different stages of the 
division (fig. 3). In the resting stage it is very small, about as 
large as the shorter diameter of the plastid. It then gradually 
increases in size. There is a scanty chromatin network and a 
single nucleolus. When the nucleus has reached its full size, the 
nuclear network becomes impregnated with a large amount of 
chromatin. Knots appear in the network and eventually develop 
into 18 chromosomes. The nuclear membrane is always present 
until just after metaphase. A centrosome is recognizable from 
prophase to late anaphase. After telophase, new daughter nuclei 
