248 The Botanical Gazette. [June, 
has increased in size apparently by keeping pace with the 
elongation of the nucellus, as the cross wall separating it from 
the cell above is horizontal. Its nucleus is a little larger 
than those of the cells above, and a large vacuole occupies 
the upper third of the cell. 
In fig. 25 it is evident that five cells have been formed from 
the initial cell. It is quite impossible to determine the exact 
order of cell-division in this case. It may have been in the 
manner indicated in figs. 20, 21, and 22, or the uppermost 
cell of the four resulting from the two successive cell-divisions 
(fig. 23 or 24) may have undergone a further division. Judg- 
ing from my experience in this work and the appearance of 
the preparation from which fig. 25 was drawn, I am inclined 
to regard the latter process as the correct one. The condi- 
tion presented by fig. 26 seems to support this view, for there 
can be little doubt that the uppermost cell of the series di- 
vided again at almost right angles to the usual plane of 
division. 
gether have about the same volume as the egg-apparatus (fig. 
29). With the gradual increase in size of the embryo-sac 
. 
prior to fertilization and some time afterwards, the antipodal 
upper ends. The nuclei multiply by fragmentation. 
idea of the phenomenon just mentioned. In this embryo-sac 
the endosperm consisted of many free nuclei regularly dis- 
posed in the protoplasmic lining of its cavity, z. e., the endo- 
ll 
(To be concluded.) 
