CELL DIVISION BY FURROWING IN MAGNOLIA 
393 
until the cell is completely partitioned, and the, cell plate gives rise to 
plasma membranes between which the cell wall is formed. No indi- 
cations of furrows or similar structures are discernible in these cells. 
It is observed that this cell plate formation occurs simultaneously 
with the organization of nuclear membranes and the union of the 
chromosomes to form the spireme. It is practically finished before 
the nuclei begin to enlarge and separate. The furrowing in both the 
heterotypic and homoeotypic divisions, on the other hand, is subsequent 
to and not simultaneous with these nuclear changes. It thus appears 
that the furrow and the cell plate are not homologous structures 
though they accomplish the same end. The formation of a furrow 
in the pollen mother cells must thus be regarded as taking place only 
when no cell plate has been previously formed. In other words, if a 
cell plate is for any reason not formed there is a furrow developed at a 
later time. The conditions incident to the two processes may not be 
at all the same. It is entirely likely that the conditions for furrowing 
exist in all cells of both plants and animals, but in plants they can not 
usually express themselves on account of the previous cell plate for- 
mation. Thus animal and plant cells may be potentially identical as 
to furrowing, but most plant cells in addition have the power of cell 
plate formation. The writer has previously suggested (6) that the 
latter process is related to the presence of a cellulose cell wall and the 
inability of the cell to enlarge in response to osmotic pressure, and 
hence is characteristic of plant cells alone. 
The cause of the heterotypic furrow's remaining arrested until 
after the second nuclear division is a matter of considerable interest. 
The suggestion has been made by several writers that the lack of 
cytokinesis between the first and second karyokineses of the pollen 
mother cells of many plants is due to too short a period of interkinesis. 
The observations of the writer on Magnolia indicate that the period 
of interkinesis in this form is by no means short as compared with 
the time required for mitosis. The greater number of cells in my 
preparations were in interkinesis rather than in either the heterotypic 
or homoeotypic mitoses. If, as the writer has previously suggested 
(6), the furrowing is the consequence of a mutual attraction between 
the nuclear membranes and the plasma membrane, then an arrest of 
the furrow would be due either to a change in this force of attraction, 
or to the fact that the nuclei have such a size and position as to cause 
the resultant of the attracting forces to become zero after a certain 
