346 BOTANICAL GAZETTE [NOVEMBER 
When the young pollen grains are free in the anther, therefore, 
only a disorganized mass of protoplasm is in the position for- 
merly occupied by the tapetum. This substance is very soon dis- 
tributed among the pollen grains, where it possibly serves as 
nutriment. The tapetal cells of Potamogeton never contain two 
nuclei. In this respect, therefore, they differ decidedly from 
Convallaria. 
THE ARCHESPORIUM AND MOTHER-CELLS. 
Division in the primitive archesporium ceases at an early 
period, after which the development is confined to growth and 
constitutional changes in the cells already formed. The defini- 
tive archesporial cells are at first quite small, but during the 
long period of growth that now commences they double or even 
triple their original size. The mature pollen-mother-cell con- 
tains a very large nucleus surrounded by abundant cytoplasm. 
Unlike most monocotyledons, the cell wall here remains very 
thin, and does not become irregularly thickened, as in Conval- 
laria and Lilium (fig. 35). A similar condition has = 
observed also in Naias and Zannichellia. A very short me 
therefore, is required for disintegration, which undoubtedly 
accounts for the almost immediate separation of the yous 
pollen grains. 
THE ARCHESPORIAL NUCLEUS. 
Potamogeton belongs to an entirely different class ht 
Convallaria so far as the nuclei are concerned. The Lilium 
type, to which the latter plant belongs, possesses the wel a 
dense spirem and the large oblong chromosomes. The aon 
of Potamogeton are apparently very poor in chromatin. 
few chromosomes are small and spherical and the spirem ted 
meager. A detailed comparison with Convallaria, therefore, 
will be especially interesting. at 
The very young archesporial nuclei in Potente 
scarcely different from the surrounding vegetative nuclei. 
are surrounded by a definite membrane, have a large — 
like body, and a very poor linin network, which lies close t° 
|-known 
