8 CYCADE. 
There is no instance known in fact of an anther of 2-5 
cells, each quite distinct from the other, and opening along 
their mutual inner surfaces. 
But this aggregation, relation, degree and mode of dehis- 
cence, together with the situation on the under side of the 
leaf, from which they are exclusively derived, are all points 
common to the reproductive (spores) of certain Cryptogams. 
The first stage of development of the pollen observed, 
when the scales were—long—broad. 
The pollen cells at this time are perfectly formed : those 
.along the centralline being the largest: they present faint 
traces of the line of subsequent opening. "The cavity which 
is undivided, is one occupied by a loose mucilagino-cellular 
ETC which is resolvable into angular grumous bodies, with- 
- gut distinct membrane, each presenting a cytoblast, or as I 
think, a point of commencement of the membrane. The de- 
velopment is, I think centripetal, i. e. the cells of the peri- 
iphery are the most distinct. 
| There are also some remarkable points in the structure of 
the ovulum. The nucleus of the vegetable ovulum is al- 
ways solid, up to the period of its being penetrated by the 
. pollen-tube, it is homogeneous up to the period of the 
appearance of the embryo-sac. But in Cycus the conical 
apex is not solid, nor can it be said to be homogeneous, for 
it contains a conical cavity with a fundus presenting a central 
elevation, and its apex is produced into a cellular membrane 
which lines the canal of the foramen, and I believe adheres 
to it at its apex. 
Then again when there is more than one appearance of 
embryo-sac,* their development is synchronous, but in Cycas 
* They are not embryo-sacs. Embryo sacs do not exist for a long time — 
after the enlargement of the fruit if there be no fecundation. 
The appearances of the funicle of the ripe embryo is analogous to 
Loranthus. ; 
What are the above sacs, which at any rate are developments of n9 - 
