June, 1922] FARR MEIOTIC CYTOKINESIS OF NELUMBO 
division. But in each case I was able to satisfy myself that this was simply 
due to the crossing of the fibers in this region. It is evident that if the 
halves of two homologous chromosomes lie side by side at the poles as they 
do in telophase, and if the halves of each pair are connected by spindle 
fibers, these fibers will cross and may even touch in the equatorial plane. 
In this case the appearance will be that of a much attenuated letter X with the 
upper and lower angles of the figure very small and the lateral angles very 
large. This is exactly the appearance which the equatorial plane of the 
central spindle in some of these cells presents. It is not followed by the 
obvious thickening of the fibers such as one finds in the heterotypic division, 
and consequently could not be taken as evidence of the development of a 
cell plate. Figure 2 is of a cell which is precisely at the stage when the 
early stages of cell-plate formation should be taking place, if they are to 
occur at all. It is at this stage of the first division that the transitory cell 
plate puts in its appearance, and in all other forms that I have studied the 
cell plate if it is formed at all is associated with this stage of the karyokinesis. 
As the nuclei become organized they do not take the form of flat discs 
such as are found in Magnolia (7), but round up directly (fig. 3). As they 
round up and enlarge, the spindle fibers become fewer and fewer. No cases 
were found in which spindle fibers appear to be entirely lacking as in the 
stages of interkinesis, but they do become very scarce indeed. It is obvious 
that if a cell plate were being formed there would in all probability be a 
great increase in, the number of spindle fibers especially in the peripheral 
region, but such is not found to be the case. The nuclei gradually migrate 
toward the plasma membrane as they enlarge, in a very similar manner 
to that found in Nicotiana (6). It must be that a relatively long period of 
time is involved in this stage of the formation of pollen. This is indicated 
by the frequency of these stages in the sections studied, and also by the 
enormous enlargement of the nuclei and their migration to the plasma 
membrane. An examination of figures 2, 3, and 4 suggests the degree of 
this enlargement. The volume of the nuclei shown in figure 4 is approxi- 
mately four times that of the nuclei shown in figure 3. Furthermore, the 
cell itself has undergone an enlargement, that shown in figure 4 being about 
twice as large as that shown in figure 3. Even more marked than either 
of these changes is the enlargement of the nucleoli, which in the latter stage 
are many times their former size. The prochromosomes so distinct in the 
earlier stage are now quite indefinite in appearance. All the evidence seems 
to indicate that a long interval elapses between the completion of karyo- 
kinesis and the initiation of cytokinesis. 
Following this, which is called the tetranucleate stage, there occurs the 
quadripartition of the cell by furrowing. This begins by the appearance 
of a structure which may at first seem to be a centripetally forming cell 
plate. A chip from the superficial portion of the protoplast may make it 
appear that there is a continuous plate across the equatorial plane. Such a 
