424 The Botanical Gazette. [September, 
nucellus (figs. 8,9, 10). Their disorganized remains now occupy a 
narrow Cavity above the enlarging embryo-sac. 
The further development of the embryo-sac is perfectly normal. 
The primary nucleus soon divides (fig. g), one of the resulting nuclei 
passing to the upper and the other to the lower end of the sac. Fur- 
ther details need not be given here. 
In this plant the mature embryo-sac does not reach the integuments, 
but is covered at its summit by a cap of about six cells in thickness, 
formed from the tissue of the nucellus. The cavity of the embryo-sac 
when nearly matured is very large, and is often readily seen by the 
unaided eye in section. Both the antipodal cells and the egg-appa- 
ratus are also very large. Of the former especially is this true, for 
they occupy about half of the entire length of the sac containing them 
(fig. 11). Shortly before they are absorbed, however, the embryo-sac 
has enlarged greatly, so that this relation is not maintained. The 
nuclei of the antipodal cells are also very conspicuous, while even the 
protoplasmic bridges or threads are plainly visible (fig. 11). 
In the lower end of the synergide and below the nucleus of each 
there is a vacuole (fig. 11). One is also to be seen in the upper end 
of the oosphere by focusing, but as this cell lies below and is partially 
covered by the synergidz, it is obscured from view. 
The ovules of this plant are anatropous, but in one instance ob- 
served, an ovule was borne on a very long stalk which arose at right 
angles to the surface of the placenta, then by a long curve extended 
upward in the cavity of the ovary to a height of the second ovule 
inserted above it (fig. 12). 
The foregoing may be summarized as follows: 
1. The embryo-sac arises as an hypodermal cell at the apex of 
the nucellus. 
2. This cell divides first into two cells. Each of these cells agai? 
divides, thus forming four. The lower of these alone under- 
goes further development and becomes the embryo-sac. 
3- The upper daughter-cell is occasionally divided into two by an 
almost vertical wall. 
4. The antipodal cells are unusually large. 
This work was carried on under the direction of Prof. D. M. Mot- 
tier, who has very kindly supplied me with an abundance of properly 
preserved material—Frank M. Anprews, Bloomington, Ind. 
EXPLANATION OF PLATE XXVIII.—Fig. 1. Longitudinal section of very youDs 
ovule showing hypodermal cell at apex of nucellus.—Fig. 2. Same, undergoing 
divisio —Fig. 3. Division completed.—Fig. 5. Both cells again undergoing 
nuclear division.—Fig. 6. This is completed and walls formed.—Fig. 7. TBE 
upper of the four cells has divided into two by an almost vertical wall.—Figs- 
