462 Smith . - Cytologic a l Studies in the Protococcales. /. 
same manner as Czapek ( 2 ) believes that the lipoids in the cell tend to 
migrate towards the surface film. An indication that surface tension 
governs the arrangement of these small plates is seen in the transverse 
cleavage of the cytoplasm, when after the new plasma membrane has been 
formed the starch plates are located along the middle of the cell, where there 
were none before. If surface tension is the factor governing the distribution 
of these starch plates it is easily seen that with the formation of the daughter- 
cells there is a change in the surface tension and a readjustment of the 
starch plates so that they are equally distributed over the surface of the 
new daughter-cells. On the other hand, if these plates were dragged in 
mechanically by the inward growth of the cleavage furrows from the plasma 
membrane, there would be an unequal distribution of the starch plates and 
a tendency for them to be more numerous on the edges of the newly formed 
plasma membrane near the mother-cell wall. Since, however, the starch 
plates are evenly distributed along the newly formed plasma membranes the 
assumption that surface tension is the controlling factor in their distribution 
does not appear unreasonable. 
The pyrenoids appear absolutely homogeneous in structure, with no 
light and dark areas as Timberlake ( 10 ) has shown for Hydrodictyon and 
Lutman (6) for Closterium . I have been able to demonstrate this irregularity 
in structure for Scenedesmus ( 7 ) and Pediastrum , so I believe that the failure 
to find the light and dark areas is not a matter of faulty technique. 
Mature cells of Characium Sieboldii are multinucleate. The number of 
nuclei varies, but is always (so far as I have observed it) a multiple of two, 
usually 32 or 64. The nuclei are distributed throughout the cytoplasm and 
are approximately equidistant from each other, varying in size from f to i-|/x 
in diameter. Usually the only structures noticeable in the nucleus are the 
single nucleolus and a sharply defined nuclear membrane (Fig. 1), but in 
some cases a few chromatin granules can be distinguished (Fig, 13). These 
granules are comparatively large, and occasionally are connected with one 
another by strands. The number of granules is approximately the same as 
the number of chromosomes, so that it is possible that the nucleus shown in 
Fig. 13 is not a resting nucleus, but is in the spireme stage. 
Owing to the small size of the nuclei it is impossible to make out clearly 
the details of karyokinesis. Nuclear division in the Chlorophyceae frequently 
takes place at night, but I have found no regular time for the occurrence of 
karyokinesis in Characium , and division figures are as apt to be found in 
material fixed during the daytime as in that fixed at night. I have not 
been able to determine the early stages in the formation of the spindle, but 
when the equatorial plate stage is reached the spindle is sharply defined. 
Karyokinesis takes place simultaneously in all the nuclei of a cell, this 
being contrary to the condition found by Timberlake ( 9 ) in the coenocytic 
Hydrodictyon^ where, although all the nuclei in a cell are dividing at the 
