40 BOTANICAL GAZETTE [JULY 
maturity might suggest that it is carpellary in origin (fig. 33). The 
stigmatic surface is prominent, and the cavity of the stylar canal is 
partially filled by a mass of similar cells, and long conducting cells 
extend downward on either side of the ovule, reaching the vicinity 
of the micropyle. At an early period the inner integument closes 
over the tip of the nucellus, but the outer integument does not close 
over the inner (jig. 33). 
The earliest satisfactory preparations of the embryo sac show 
the first division (fig. 36) and the second (fig. 37). The final stages 
show great irregularity. Fig. 38 shows a group of three cells (to 
the right) which suggests the egg apparatus; the two cells at the 
extreme left are certainly antipodals; while the two in the middle are 
evidently fusing polar nuclei. The solitary cell to the left suggests 
the third antipodal, but a male cell appears to be fusing with it. 
The ninth cell may be the other male nucleus. The embryo sac 
represented in fig. 39 contains only five cells, two of which are clearly 
fusing polar nuclei. 
The number of antipodals varies from two to eleven, and it is 
quite possible that they may sometimes be even more numerous 
(figs. 33, 34, 35, 40). me 
ENDOsPERM.—In Aglaonema the endosperm does not begin with 
free nuclear division, as in Dieffenbachia, Nephthytis, and Arisaema, 
but wall-formation begins at once. The growing endosperm first 
lines the side of the embryo sac which is toward the funiculus, but 
later fills the cavity and completely surrounds the embryo (figs. 33) 
40, 41, 42). 
Empryo.—The embryo much resembles that of Lysichiton kams- 
chatcense as figured by CaMpBELL.? Fig. 41 represents the position 
in which the proembryo was most frequently found, but in many 
cases it appears at the micropylar extremity of the sac. It appears 
quite as frequently adhering to the surface opposite the growing 
endosperm, and in at least one case was found resting on the antip- 
odals. There was no evidence of displacement by sectioning, S° 
that this variable position is probably due to the fact that the egg 
may lie in any part of the embryo sac. 
BLAIRSTOWN, N. J. 
? CAMPBELL, D. H., Notes on the structure 
; of the embryo sac in Sp ium and 
Lysichiton. Bor. Gazerre 27:153-166. 1899 ry’ pargan 
ie 
