farr: cytokinesis of pollen-mother-cells 281 



effected by means of cell-plates. That some dicotyledons form 

 their microspores by two successive divisions with cell-plate forma- 

 tion, can scarcely be doubted, as is shown in Asclepias, Cerato- 

 phyllum, Apocynum, and Rafflesia. But it is generally recognized 

 that the majority of pollen-mother-cells of dicotyledons do not 

 divide in this way. The best evidence of quadripartition by cell- 

 plate is that of Strasburger in his work on the alcoholic material 

 of Tropaeolum, which has been shown above to be not beyond 

 question. On the other hand there is some evidence that cleavage 

 furrows may be involved in the process. The work of Naegeli, 

 Wimmel, von Mohl, Sachs, and Hofmeister suggest it, though 

 such early work should not be taken as by any means final. Guig- 

 nard and Samuelsson directly afhrm it in Magnolia and Anona 

 respectively. 



There is also a considerable list of additional papers which in 

 one way or another indicate that furrowing may be involved in 

 the division of the pollen-mother-cells. As early as 1880 Bar- 

 anetsky (4) in addition to confirming Strasburger's observations 

 on successive bipartition by cell-plates in the pollen-mother-cells 

 of Tradescantia, studied this stage in four different dicotyledons. 

 He shows two figures of Hesperis matronalis with incipient furrows 

 along the equator. He does not describe them, but says, "Stras- 

 burger gibt an, dass 'in alien Fallen' der Theilung der secundaren 

 Kerne die Bildung einer Zellplatte zwischen ihnen vorausgeht. 

 Bei den von mir iiberhaupt beobachteten Dicotylen : Pisum sativum, 

 Lathyrus odoratus, Hesperis matronalis, Ipomaea tricolor, konnte 

 ich diese Angabe nicht bestatigt finden." 



Ishikawa (3 4) in his figure, number 1 1 , of Dahlia shows a pollen- 

 mother-cell in which four nuclei are tetrahedrally arranged. They 

 do not lie near the plasma membrane, but spindles are shown ex- 

 tending between them without the slightest indication of a cell- 

 plate. At the equator on one side there is a concavity, and the 

 other two sides are flattened. No mother-wall is shown, and the 

 paper contains no discussion of the process. 



Osawa (506) shows several figures of division of the cytoplasm 

 in the pollen-mother-cells of Taraxacum. Figure 22 is of a pollen- 

 mother-cell with four nuclei tetrahedrally arranged ; no mother- 

 wall is shown. The spindle fibers are thickened at the middle 

 throughout about one half their length. One side of the cell is 



