AND AFFINITIES OF STEPHANOSPERMUM. 385 
improbable. The pollen-grains of Stephanospermum, large from the first, have already, 
in the stage usually preserved, sojourned in the pollen-chamber for a considerable time, 
and have there undergone a further development. Had they formed rhizoid-like 
attachments to or actual tubes in the tissue of the wall of the chamber, indications could 
hardly be wanting. It may, I think, be safely assumed that structures of that kind 
were not produced. That mobile antherozoids were liberated in Slephanospermum may 
be inferred with some degree of confidence from the whole organisation of the pollen- 
chamber. Whether all or but a part of the internal cells of the pollen-grains were 
concerned in the production of antherozoids, there is no evidence to show. That papille 
(as in Pl. 42. fig. 15) arose from the periphery of these cells, liberating antherozoids, is not 
inherently improbable. To the occurrence of perforations, whether on the surface 
of the grains or, as is so frequently the case, on their partition-walls, just this 
significance may be attached, viz., that these walls were readily destructible, and that 
consequently a very limited number of papillae might have served for the discharge of 
all the antherozoids produced in one of these multiple pollen-grains. 
The view which I adopt in regard to this seed is that no tubular, rhizoid-like 
attachments were formed by the pollen-grains on the walls of the pollen-chamber, but 
that the large grains, after a process of maturing in the chamber, liberated antherozoids 
by some simple means, such, perhaps, as the extrusion of short papille, the tips of which 
became perforated. Finally, that the antherozoids were formed in considerable numbers 
and that the distance they had to traverse by swimming was relatively great—as compared 
with recent Cycads, for instance. 
In Stephanospermum caryoides the number of primary internal cells is much reduced. 
They are partly enclosed in an extensive peripheral cell recalling the tube-cell of the 
pollen of recent Gymnosperms. The two internal cells produced secondary complexes . 
of antherozoid mother-cells, whilst the outer cell seems to have facilitated the liberation 
of the antherozoids. Whether the blister-like outgrowth (c) of Pl. 44. fig. 40 indicates the 
natural means of exit of the antherozoids must remain an open question till additional 
specimens are available. In this seed, as in S. akeniotdes, the freedom of the pollen from 
special nutritive relations with the tissues of the nucellus seems to mark an archaic 
condition not preserved in any recent Gymnosperm. The discovery of this seed ina 
stage of development apparently more advanced than any of the numerous specimens of 
S. akenioides that have passed through my hands, encourages the hope that the 
phenomena attendant upon fertilisation in the seeds of this period may ultimately 
become known to botanical science. 
The Difference in Pollen-structure. 
‘From the description which has been given of the pollen-grains in the two Stephano- 
sperma, it is evident that considerable divergence exists in their complexity in these 
two seeds. Whilst in the former the pollen-grains are filled with a tissue of some twenty 
cells or more, in the latter is contained a group of two. If, therefore, the close 
association of the two seeds be justified, as I incline to think on the near agreement of 
SECOND SERIES.—BOTANY, VOL. VI. 3K 
