102 BOTANICAL GAZETTE [AUGUST 
their lower part. Lateral placental outgrowths divide the ovarian 
cavity into two loculi, but from this central placenta two ovules 
usually arise, both of which, except in rare cases, form in the 
same loculus. The funiculi are curved downward so that the 
ovules lie with the chalaza in the base of the loculus and the 
micropyle toward the style. Several hypodermal cells (‘‘ primary 
mother-cells’’), possibly archesporial in nature, become deeply 
situated on account of their own transverse divisions and form a 
multicellular sporogenous tissue, sharply differentiated laterally, 
as in pteridophytes. Later it is also well differentiated toward 
the micropyle, but not in the region of the chalaza, where a strong 
“intercalary growth” takes place, suggesting the origin of spo- 
rogenous tissue from cells not arising from the primary sporog- 
enous cells. Of the sporogenous cells,’ some function as spore 
mother-cells, some remain small and are soon resorbed, while in 
some species others become tracheids, recalling the formation of 
elaters among the liverworts. Those that function as spore 
mother-cells divide, each forming a row of four. Thus with 
pteridophytic sporogenous tissue, there is spermatophytic arrange- 
ment of megaspores. Some of these enlarge, forming embryo 
sacs, aS many as a score of sacs being counted in a single 
ovule. They elongate more or less antipodally, some even 
penetrating the chalaza to the funiculus. Differing from almost 
all other spermatophytes, the functionless megaspores of Casu- 
arina are not resorbed, but remain among the fully developed 
embryo sacs. Whatever the number of ovules in an ovary, only 
one is fertilized. The sacs in the unfertilized ovules enlarge, 
become binucleate, but form no antipodal prolongations. In the 
fertilized ovules not all the megaspores that enlarge form pro- 
longations; those not centrally placed are often arrested early. 
In the fertile ovule only one sac is fertilized. 
The formation of cells within the sac was not definitely 
traced,” but developing sacs with one nucleus, and sometimes 
with two nuclei, were found. A certain nucleus, of unstated 
*Treub applies the term acre ’ cells to the tissue both before and 
after the formation of megaspore 
? It will be remembered that this work was done twelve years ago. 
