100 THE BOTANICAL MAGAZINE. LVoi. xxxvi. No. 429. 



Pollen mother-cells : — When the extrusion occurs some nuclei 

 occupy excentric positions in the cell, while others remain in the 

 centre. This has been observed also by Kornicke in Crocus vernus, 

 All. (io). In Iris japonica the nuclear membranes are generally 

 indistinct (Fig. 1), though at times something like a nuclear membrane 

 or a blurred outline stained with methylene blue is observed. 



The most active time for extrusion seems to be the synizetic 

 stage as it has been pointed out by former investigators. But in 

 presynizetic, spireme, hollow-spireme and even near diakinesis stages 

 (Fig. 6) also the extrusion figures are found. The extrusion of the 

 chromosomes or spindle fibers which is described by Kornicke was 

 not observed in this plant. This may depend upon the time of 

 separation of the mother-cells from each other. 



It can not be said that all the extrusion from the mother-cells 

 in a loculus of Iris japonica. occurs in one direction only and simul- 

 taneously, though in certain mother-cells the extrusion seems to be 

 in one and the same direction. The directions are generally manifold 

 and the extrusions do not take place simultaneously. Cases are often 

 observed where bodies extruded from one cell enter into two neigh- 

 bouring cells, or bodies extruded from two cells enter into one, and 

 sometimes a cell, receiving the materials extruded from two other 

 cells, throws its nuclear substance into the next (Fig. 4). 



Regarding the path of the extruded materials in the cell membrane 

 Hottes in Vicia (9), Miehe in Allium etc. (11), Kornicke in Crocus 

 (10) and Schrammen in Vicia (15) lay stress upon the pores of plasma- 

 continuity (plasmodesm). Digby (l), Gates (7) and West and Lech- 

 mere (23) find respectively protoplasmic connections between the pollen 

 mother-cells in Galtonia, Oenothera, and Lilium. The plasmodesms, 

 or bundles of them, are also recognized in the favourable preparations 

 of Iris japonica (Fig. 17) and it is most likely that the nuclear 

 substance may pass through the pores of the plasmodesms or break 

 through the pits into the cytoplasm of the adjoining cell. The con- 

 nection between the extruded bodies and the nucleus from which they 

 originate is sometimes one thick cord, or often one or more fine 

 threads. It is not infrequently observed that each one of the small 

 extruded bodies in one side of the cell has a slender connecting thread, 

 while in other cases the, nucleus is connected with one large extruded 

 body by several fine threads. 



The phenomena of passing of nuclear bodies from pollen mother- 

 cells into tapetal cells or the reverse can not be found in Iris japonica 



