248 The Botanical Gazette. (July, 
The antipodal cells occupy the entire posterior end of the 
embryo-sac, which is narrow and always remains so (figs. 5, 
7, 13). As in Senecio vulgaris, they lie in a longitudinal row 
(fig. 5), or one below and two above (figs. 1, 3); but I find 
here a much greater variation in the number of cells than 
Strasburger figures for Senecio vulgaris. In several instances 
only two cells were present (fig. 7), separated by a much 
swollen wall. The lower cell is longer than the upper and 
the nucleus is greatly lengthened in the direction of the axis 
of the embryo-sac; it contains several nucleoli and may have 
arisen from the fusion of two nuclei. Fig. 3 shows an em- 
bryo-sac in which the lower antipodal cell has divided, and 
in the upper one of these the nucleus has doubled. The 
doubling of the nuclei in the antipodal cells is, however, a 
frequent occurrence. In fig. 4 it will be seen that each anti- 
podal cell has divided, making six in all. A curious cond 
tion of things is shown in fig. 6. The wall, +, was well 
developed, but the others, «’,2’, were extremely delicate. 
n many cases the walls separating the antipodal cells were 
slightly swollen (fig. 5). These facts seem to indicate that 
the antipodal cells represent a structure that shows a tef- J 
dency toward further development, suggesting a muc ei E 
duced: prothallium or rudimentary structure of some ae 
and a comparison of it with a prothallium of the vasc 
cryptogams seems reasonable. 
The fusion of the two nuclei to form the endosperm — 
takes place after the formation of the egg apparatus 4” . 
tipodal cells. In no case were the antipodal cells peccgihe 
lying free without cell walls at the time when the a 
arent to unite to form the endosperm nucleus; the w: 4 
always well developed at this stage. : 4 
Of the two oA given off into the cavity - ape q 
bryo-sac, the one which comes from the antipod# mgs . 
erses the greater distance and meets its fellow clos’ i 
egg-cell where they unite (fig. 1). The nT ie tin 
invariably lies very near the egg, apparent: gs ‘ilar tO 
many cases (figs. 4, 8, 9, 10). Its structur =~ fe- 
that of the a8 mele except it is much larger ee 
spect. It is surrounded by rather dense granular P 
from which radiate delicate threads connecting mi sac. 
delicate layer of protoplasm lining the cavity “Hi we 
this particular fig. 9 is the most interesting: 
