FARR: QUARDIPARTITION IN SISYRINCHIUM 57 
of Monocotyledons, in which she finds that contrary to the 
general opinion the nuclei of most Monocotyledons are small. 
It is only in the Iridaceae and parts of the Liliaceae, Amary]- 
lidaceae and Convallariaceae that large nuclei are found. How- 
ever, her study is not concerned with pollen-mother-cells 
especially. It would be interesting to find how great the 
differences in size of pollen-mother-cells may be among Mono- 
cotyledons, and their relation to the size of the nuclei. The 
tetranucleate pollen-mother-cells of Sisyrinchium just before 
division are 36-40 microns in diameter, which is just about the 
size of those of Nicotiana. 
The mother-cell wall is likewise thickened to about the same 
extent as is that of Nicotiana, that is, on the average about one- 
tenth of the diameter of the cell. The process of thickening 
of the cell-wall begins about the time of synapsis, but the greatest 
development is during interkinesis. The cells are more compact 
within the anther than they are in either Nicotiana or Magnolia, 
but nevertheless are quite loosely disposed during diakinesis 
and later stages of reduction. They are somewhat closer 
together toward the end of the process than in the earlier stages 
doubtless due to the enlargement of the cells and the thickening 
of the walls. There seems to be no evidence that the gelatinous 
substance enveloping the protoplast during the later stages of 
reduction-divisions is a secretion from the protoplast, as Valleau 
(26) suggests, rather than a product of the cell-wall. The 
middle lamellae remain quite distinct throughout these stages 
and careful observation of the material both of Sisyrinchium 
and of the other forms previously studied leads the writer to 
the conclusion that it is the secondary lamellae of these pollen- 
mother-cells which during reduction-divisions take on the 
colloidal property of imbibition and swell to many times their 
original thickness. 
The heterotypic division results in two discoid nuclei being. 
organized at opposite ends of a rather long central spindle. No 
indication of a cell-plate or other equatorial differentiation is 
to be discerned at any time during interkinesis. Nor is there 
any indication of furrowing. The nuclei gradually become more 
nearly spherical (Fic. 1) and finally, before the second nuclear 
division begins, they are almost perfect spheres. They appar- 
ently do not come into contact with the plasma membrane. The 
fibers of the central spindle disappear almost entirely during in- 
