On Some Stages in the Life History of Gnetum. 
233 
1. The Nucellus and Pollen-chambek. 
The pollen-chamber of G. Gnemon has been described and figured by 
Xotsy.* In 6r. africanumf the chamber is formed in the same manner, but 
it becomes deeper and narrower than in G. Giiemon (Plate XLV, figs. 2-4). 
In G. scandens it is broad and shallow as in the latter species and in 
Welwitschia^ (Plate XLY, fig. 1). 
After pollination the micropyle becomes blocked by tissue in the manner 
already described § ; we have frequently seen large papilla-like cells || pro- 
jecting from the lower end of the blocking tissue into the cavity of the 
pollen-chamber (Plate XLY, fig. 2). 
Soon after pollination, sometimes before this occurs, the tissue of the 
nucellar-cap below the level of the floor of the pollen-chamber, may undergo 
disintegration similar in character to that described by Coulter for G. 
Gnemon. % Occasionally, as in the latter species, the epidermis becomes 
infolded and a slit-like constriction appears in this region of the nucellus. 
More commonly, however, in G. africanum (Plate XLV, fig. 3) the epidermis 
remains unaffected and the disintegration begins in the underlying tissues. 
It gradually extends inwards and downwards producing a broad, irregularly 
funnel-shaped cavity which may extend to the top of the embryo-sac 
(Plate XLV, fig. 4). During a stage in the later growth of [the endosperm 
its micropylar apex may turn aside into part of the cavity thus formed 
(Plate XLV, fig. 5), and it may thus acquire a somewhat irregular form. 
At a still later stage, with the destruction of all but a thin layer of crushed 
nucellar-cap cells, the endosperm assumes the more regular form already 
described.** 
There is no indication of the meaning of this curious deformation of the 
nucellar cap. It is clearly not incited by pollination nor by the growth of 
the pollen-tube, for it may occur in unpollinated ovules, and often there is 
no sign of it in ovules which have been fertilised. Its occurrence does not 
appear to be detrimental to the normal development of the seed, and it is 
therefore unlikely that it is a stage in the nucellar disease described by 
Lotsy. It imposes limitations on the possible directions of growth of the 
pollen-tubes, though in its absence their course is not materially different. 
The cells of the cap not affected by this degeneration are packed with 
starch grains. 
* Lotsy, 1899, p. 94, figs. 18, 35. 
t Thoiay, 1911, %. 14. 
X Bertrand, 1878, fig. 10 ; Pearson, 1906, figs. 28, 29. 
§ Thoday, loc. ext., p. 1120. 
II Cf. Rerridge, 1911, fig. 4. 
H Coulter, 1908, fig. 5 a. 
Thoday, 1911, Text-fig. 9. 
