farr: cytokinesis of pollen-mother-cells 269 



quadripartition and bipartition (73), with a predominance of the 

 latter. Harper (28^) asserts that in the former cases the materials 

 of the spindles of the heterotypic division may be used in the 

 homoeotypic nuclear divisions. Hofmeister (3 2b) reported centrip- 

 etal division of the cytoplasm in the spore formation of certain 

 Abietineae, but this has never been confirmed. He also notes 

 the absence of a cell-plate and the swelling of the mother-cell-wall 

 in the divisions in Pinus. 



i. Pollen- formation in monocotyledons 

 In the monocotyledons a considerable number of instances of 

 quadripartition have been reported. Mottier (44c) shows it 

 in his figure of the embryo-sac of Helleborus, where the nuclei 

 are arranged in a rhomb, with five spindles and upon each a cell- 

 plate. The same investigator gives an interesting figure from the 

 embryo-sac of Lilium. The four nuclei at the micropylar end of 

 the sac may be arranged either in one plane or tetrahedrally, the 

 latter condition being shown in the drawing. Cell-plates are 

 formed between each pair of nuclei, and also one on the fibers 

 which extend from the egg into the cytoplasm of the embryo-sac. 

 This is one of the few instances of cell-plate formation on fibers 

 which do not run between nuclei. Others are the embryo-sac of 

 Ephedra studied by Strasburger, and the embryo development of 

 Picea as shown by Hutchinson (33) in his figure 41. 



The strong tendency to believe that all cell-division in the higher 

 plants proceeds according to the classic type of Strasburger's 

 central spindle and cell-plate formation has been noted. None- 

 the-Iess the literature contains some indications that in many 

 cases of quadripartition the process may be a furrowing from the 

 surface inward, quite as in the animal and many lower plant cells. 

 Berthold (6) ini886 wrote (p. 217) : " Beidersimultanen Viertheilung 

 der Pollenmutterzellen scheinen bemerkenswerthe Besonderheiten 

 nicht aufzutreten. Die neue Zellplatten bilden sich hier entweder 

 wie gewohnlich bei den hoheren Pflanzen und auch in den Spore- 

 mutterzellen von Anthoceros und Isoetes zuerst frei im Zellraum, 

 um erst nachtraglich an die alte Membran sich anzusetzen, oder 

 aber sie dringen von dieser aus, Cladophora und Spirogyra ent- 

 sprechend, gegen das Zellinnere vor, wie oben schon erwahnt, in 

 Form dicker, plumper Leisten," 



