Box. -Vol. I.] CAMPBELL— .VA/AS A.VD ZA.Vy/CHELLfA. 1 5 



The divisions of the spore-mother-cell are successive, and 

 the resulting pollen-spores of the bilateral type common to 

 most Monocotyledons. The successive stages in the division 

 of the spores is shown in figs. 17-22. 



It follows the usual form, and there were no remarkable 

 variations from the common type of nuclear division. The 

 number of nuclear segments could not be determined posi- 

 tively, but it is not very large, probably 8-12. During the 

 earlier phases of division the nuclear spindle was distinct 

 and relatively large, the nuclear segments not occupying the 

 whole of the equatorial diameter. The formation of the 

 connecting threads, the cell-plate, and finally the cellulose 

 division wall, agree with other Monocotyledons, and the 

 subsequent divisions of the cells, resulting in the tetrads of 

 spores (fig. 22), show nothing peculiar. Unlike most Sper- 

 maphytes, both spore-mother-cells and spores have thin walls ; 

 this is probably simply an accompaniment of the aquatic 

 habit, but it is interesting, as the spore-mother-cells in this 

 respect are more like those of the Pteridophytes than those 

 of the typical Spermaphytes. After the division is complete, 

 the young spores separate and round themselves off, but do 

 not develop the cuticularized exospore found in most plants. 

 This is no doubt simply the result of their submersed con- 

 dition, for, as they remain permanently under water, such a 

 condition would naturally be expected. The young spore 

 at this stage is a globular thin-walled cell, with nearly clear 

 contents, and contains a single, large nucleus. It soon, 

 however, divides into two cells of very unequal size, a large 

 vegetative cell and a small antheridial one. These are sep- 

 arated by a distinct membrane (fig. 23), and are also dis- 

 tinguished by the very unequal size and different appearance 

 of the nucleus. That of the vegetative cell is large, with a 

 distinct nucleolus, but comparatively poor in chromatin. 

 The nucleus of the generative or antheridial cell has a very 

 inconspicuous nucleolus, but stains deeply with ordinary 

 staining agents. After the division of the pollen-spore it 

 rapidly grows, becoming much elongated, and before it is 

 ripe there is a division of the generative cell into two. The 



