46 BOTANICAL GAZETTE [JANUARY 
The carpel arises from the same general region with the sta- 
mens. In figs. 8 and g is seen the nucellus, the projections on 
either side of which are to form the walls of the carpel. These 
projections grow rapidly, soon extending beyond the nucellus. 
A fully formed flower is shown in fig. 9, in which the carpel 
extends beyond the spathe and encloses the single half-anatro- 
pous ovule. Further discussion of the development of the carpel 
and ovule will be taken up in connection with the megaspore. 
The homologies of the floral arrangement are very perplex- 
ing. It may be true, as given in the usual accounts, that there 
are here two flowers, one carpellate, the other staminate, enclosed 
in acommon spathe. I can see no good reason for thus sepa- 
rating the carpel and the stamens. If the conditions are taken 
as they present themselves, it would seem as if this might be a 
single flower, with its parts probably spirally arranged, and 
enclosed within a spathe. If we are to take as evidence the 
supposed relatives of the Lemnacez, the aroids, we must recog- 
nize that the former are represented by forms so greatly reduced 
that we are safe in Saying but little as to the kind of aroids 
from which they have been derived. 
DEVELOPMENT OF THE MICROSPORES. 
Young stamens which are not yet represented by distinct 
branches show a group of archesporial cells3 immediately beneath 
the epidermis. One such stamen is shown in jig. 13. The two 
stamens at this time appear as obtuse outgrowths upon a very 
short stalk, the two together presenting the appearance of a 
single stamen in which two archesporial masses may be seen 
(jig. 14). Later stages show clearly that these are not two 
archesporial masses of one stamen, since as the anther grows 
each mass of archesporial tissue broadens and deepens and 
becomes divided by a plate of sterile cells which cease to take 
the characteristic stain of archesporial tissue. This is shown in 
3 While there may be some question as to just what cells should be called arche- 
sporial, for convenience I shall use the term in describing the cells of this region 
until they become clearly separated into sporogenous cells and those of the tapetal 
row. 
