farr: cytokinesis of pollen-mother-cells 297 



individual spindles seem to be pressed together along the line 

 where they are in contact, each, one being thus quite definitely 

 delimited. This results in many instances in the appearance of a 

 triradiate figure composed of bundles of fibers (fig. 28). 



As the furrow deepens to the level of the outer margins of the 

 nearest two nuclei it again becomes somewhat sharper. Mean- 

 while the nuclei slowly draw away from the plasma membrane 

 (fig. 29). The two edges of the furrow approximate an angle of 

 90° with each other, but the edge of the furrow is always rounded 

 (fig. 30). By this time the spindle has almost lost its inflated 

 appearance. The peripheral fibers are no longer arched, as in the 

 earlier stages. 



The question as to the relation of the furrowing to the growth 

 of the plasma membrane is a difficult one. It seems certain that 

 growth of this membrane must take place at some time during the 

 process of cell-division. It is not easy to determine whether the 

 fibers are pushed before the invading membrane in the equatorial 

 plane, as in animal cells, or whether the middle portions of^ the 

 fibers as they successively come in contact with the invading mem- 

 brane become absorbed in it, the remainder of the fiber between 

 either nucleus and the new membrane persisting as radiating fibers. 

 Such a transformation of fibers into plasma membrane would 

 resemble that described by Harper (286) for free cell formation 

 in the ascus. The existence of such radiating fibers from the 

 nucleus to the plasma membrane of the furrows may be found 

 in all later stages of the division process (figs. 30-36). It seems 

 not improbable that in Nicotiana also the spindle fibers are in 

 part transformed into the plasma membrane. It is not, however, 

 by any means proven that this is the only source of new plasma 

 membrane material; but it is not surprising to find that here as 

 well as in the cases where the cell-plate is the precursor of the 

 plasma membrane, we may have the origin of that membrane from 

 the fibers (kinoplasm) of the cell. As evidence accumulates it 

 seems more and more likely that this kinoplasmic material arises 

 from the region of the nucleus; this is then only further evidence 

 that the material of the plasma membrane has its primary origin 

 in the nucleus of the cell. The presence of radiating fibers from 

 all sides of the nucleus to the plasma membrane, not only during 

 division but as well in the young microspores both those of dico- 



