82 SMITH : TETRADESMUS, A NEW COENOBIC ALGA 
completed in one of the daughter protoplasts and not in the other. 
The relation of the cleavage planes to one another shows that 
there are certain mechanical factors governing their position. It 
is well known [Wilson (20)] that in the cleavage of the animal egg 
changes often occur in the position of the cells with regard to one | 
another. Surface tension seems to be one of the factors governing 
this change.. In moss rhizoids, de Wildeman (18) found that the 
cross walls are always S-shaped in section, so that although the 
cross wall as a whole is placed diagonally, yet its junction with 
the outer walls of the rhizoid is at right angles. This curving of 
the cleavage planes where they meet the mother cell wall occurs in 
substantially the same manner in Tetradesmus (FIG. 13, 14). 
One view of the young four-celled colony, shortly after the 
second cleavage, when seen from the side, is shown by FIG. 13, but 
when the young colony is seen in a plane at right angles to that 
shown in this figure, we observe the very different appearance 
shown in FIG. 13A. FIG. 13 shows that even after the four proto- 
plasts of the young colony are formed, one and only one contains 
a pyrenoid. At this stage, or very shortly after, the old pyrenoid 
disappears (FIG. 14). This disappearance probably takes place 
quite rapidly, since no intermediate stages were found between 
those shown in FIG. 13 and 14. In no case was the pyrenoid 
recognizable at stages later than those of FIG. 13 and 14. 
The changes that follow in the young colony might be well said 
to constitute a period of maturation. The maturation consists in 
an elongation of each of the daughter protoplasts until it extends 
the entire length of the mother cell. The protoplasts do not elon- 
gate so as to lie in one plane, but the two in one (which we may call 
the lower) half of the cell slip up over the two in the upper half. 
This slipping occurs along the diagonal plane of the first cleavage- 
A stage in this slipping is seen in FIG. 15, although the drawing fails 
to bring out the fact that a portion of each of the two upper cells 
is hidden by part of the lower pair of cells. When this elongation 
is completed, the daughter cells are arranged in two planes, so that 
only two cells may be seen in any lateral view (FIG. 16). The 
growth of the four daughter cells seems not to be an equal elonga- 
tion of all parts, but rather an elongation of the end of each of the ; 
daughter cells farthest from the corresponding apex of the old 
