106 BOTANICAL GAZETTE [AUGUST 
appearance and reaction to various stains indicate that they are 
supplied with some substance which serves as food for the embryo 
sac. Hence the direction of growth of the embryo sac may be 
regarded in part as a response to chemotactic stimuli. In Loran- 
thus sphaerocarpus, according to TrEUB, the embryo sac is sur- 
rounded by a sheath consisting of a single layer of cells which are 
filled with starch. During the development of the long arm from 
the base of the sac, the portion within the floral axis has been slowly 
enlarging and.advancing upward. The sac as now seen is shaped 
like a hook, the short arm of which is within the floral axis, while 
the long club-shaped portion lies wholly within the tissue of the 
carpel. 
The history of the formation of the nuclei of the sac and their 
arrangement within it are no less interesting to follow than the 
form of the sac itself. The two nuclei in the chalazal end of the 
four-nucleate sac become the “antipodals”’ of the mature sac. 
They do not divide, but become somewhat enlarged. They are 
nearly always in close contact with each other, and by the time of 
the origin of the embryo they may be partially fused. Still later 
they may become wholly fused (fig. 30), while in other cases they 
never unite, but remain separate until a very late period in the forma- 
tion of the endosperm and embryo, when they finally disappear. 
Coincidently with the downward growth of the sac, the greater 
portion of the cytoplasm and the two micropylar nuclei, together 
with most of the food substances, move into the tubelike extension 
as it is being formed. This mass usually lies a short distance from 
the apex of the sac as it works its way up through the tissue of the 
carpel (figs. 32, 33). The cytoplasm which is in contact with the 
wall of the micropylar end of the sac does not contain starch. It 
1s very dense, finely granular, and stains more darkly than the cyto- 
plasm of the lower portion of the sac, in which the dividing nuclei 
and food substances are imbedded. With iodine it stains a yellow- 
ish dark brown. It is evident that this is a specialized portion of 
the protoplasm of the sac which probably secretes an enzyme for 
_— the tissue as it advances through the carpel. T his is 
oe i from the fact that a number of examples were found where 
Pex of the sac extended up between the cells of the tissue 
