152 BOTANICAL GAZETTE [MARCH 
tion, the division of the generative cell, which presumably takes 
place, although I was not so fortunate as to observe it, must 
occur after the pollen tube begins to form. 
ORIGIN OF THE MACROSPORE. 
The macrospore invariably has its origin in a hypodermal 
cell at or near the apex of the nucellus (fig. ro). Sometimes 
there are two or three hypodermal cells which by their size and 
intense staining indicate their sporogenous nature (fig. 77). A 
few cases were found in which two macrospores had developed 
to the fertilization stage, so it is evident that more than one of 
the sporogenous cells may continue its development. The 
usual appearance of an ovule before the differentiation of the 
archesporium is shown in fig. 72. In the nucellus six hypoder- 
mal cells, three of which are represented in the drawing, might 
be called archesporial cells, but I have not applied this term to 
a cell until it shows by its denser contents and reaction to stains 
that it has the characteristics of an archesporial cell. 
The archesporial cell divides into a primary tapetal cell and 
a sporogenous cell which is the mother cell? of the macrospore 
(fig. 13). The primary tapetal cell sometimes gives rise to a tier 
of five or six cells, resulting in a deep placing of the macrospore. 
Usually, there is a tier of two or three cells; but occasionally 
the primary tapetal cell does not divide (figs. rg-7g). All of 
these variations were found in Salix glaucophylla, and an exami-— 
nation of several ether = indicated a similar lack of uni- 
de ieee pment. of the macrospore mother ca | 
ti nt variations. Almost always it divides 
