McAllister. — Nuclear Division in Tetraspora lubrica. 689 
Fig. 20 shows a definite central spindle in a very early telophase. Later, 
as in Fig. 23, the rudiments of a cell- plate may be seen in the central part 
of the central spindle as a collection of granules. Whether these granules 
arise as thickenings of the spindle fibres could not be determined with 
certainty. Stages such as are shown in Fig. 22, in which the central spindle 
and the cell-plate do not extend out to the cell-wall show clearly that cell- 
plate formation is initiated between the daughter nuclei, and the wall 
extends out to the wall as is the case in the higher plants. This, as above 
mentioned, is the case in Oedogonium , and may be of common occurrence 
in the green Algae, though in Spirogyra and Cladophora the cross wall 
formation is initiated at the wall and extends inward. 
The splitting of the cell-plate seems also to be from the centre outward 
(Figs. 25 and 26). Frequent cases were observed in which there was a 
widely separated cleft between the two daughter-cells which did not extend 
to the surface of the cell. The wide separation of the cleft is of course due 
to plasmolysis. If the cell-plate had been split completely to the wall one 
could expect that the shrinkage would manifest itself on the outside of the 
cells, rounding up the newly formed corners and edges. This was not the 
case in any of the cells observed. It is clear that the cleavage does not 
begin at the surface and extend inward. 
The single disc-shaped pyrenoid remains unchanged through the three 
nuclear and cell-divisions of the vegetative cell (Figs. 18, 27, and 28), so that 
at first but one of the eight cells which are to form the gametes has a 
pyrenoid (Figs. 28 and 29). The gametes at the time that they become 
motile are all equipped with a pyrenoid, the origin of which was not deter- 
mined. It is clearly evident, however, that here the pyrenoid does not 
arise by the division of a pre-existent pyrenoid, but that it is formed anew 
from the cytoplasm. In the vegetative multiplication of the cells of the 
Alga the pyrenoid divides to form the pyrenoids of the daughter-cells. 
It seems also clear that the pyrenoids do not disappear upon the 
initiation of the steps leading to gamete formation as is the case in 
Hydrodictyon and in Chlamydomonas according to Klebs (26, 27), in 
Cladophora according to Strasburger (45), and in Volvox on the authority 
of Overton (40). Fig. 29 shows the persistent pyrenoid in the eight-celled 
stage before the differentiation of the gametes. 
The pyrenoids commonly appear as flattened disc-like bodies, one 
flattened surface being toward the nucleus. Usually there is a plane of 
cleavage extending through the centre of the mass parallel with the 
flattened surface (Figs. 8 and 9), separating it more or less completely into 
two parts. This apparent doubleness may be similar to that described by 
Timberlake (48) for Cladophora , in which The differentiation of the pyrenoid 
into two parts takes place in such a way as to divide it by a plane passing 
through its longer axis. In many cases the pyrenoid is actually split into 
