234 
Transactions of the Royal Society of South Africa. 
2. The Pollen-tube and Sperm-cell. 
There is no conclusive evidence as to the pollinating agent in any species 
of Grnetum * ; the results of its action are very variable in G. africanum. 
Very frequently, as already stated, the ovule is unpoUinated ; otherwise one, 
two, or three pollen grains are commonly found in the pollen-chambers ;, 
frequently the number is larger. In one case, in G. Gnemon (Ceylon 
material), the large number present in the micropyle (Plate XLYI, fig. 7) 
recalls the condition often seen in Welwitschia.f There is every indication 
that pollination in the latter genus is much more efiicient than in Gnetum 
africanum. 
The adult pollen-grain in the microsporangium of G. af ricanum contains, 
three free nuclei. :[: The same condition for G. Gnemon is described by 
Lotsy § and for G. funiculare, G. Bumphianum, G. latifoUum, and G. Gnemon 
byKarsten.jl In no species has a generative cell been seen in a pollen grain. 
In G. funiculare the pollen-grains on the nucellus contain only two free 
nuclei,^ one having disappeared. In G. Gnemon, either two or three free 
nuclei are present in grains found in a micropyle (Plate XL VI, fig. 7). 
Presumably the same is true for G. africanum. ; in our material of this 
species pollinated ovules all show the pollen in the course of germination ; 
in many cases a third nucleus is seen (Plate XL VI, fig. 8) in the proximal 
end of the tube, in others there is no trace of it. In all these species then, 
as in Welwitschia,** three free nuclei are characteristic of the mature pollen- 
grain ; one of the three plays no part in its germination ; it disappears either 
before or soon after pollination or, if it persists, never, as far as is known, 
enters the pollen-tube. Such a nucleus must be regarded as constituting 
the last vestige of the prothallial tissue. It is noteworthy that this very 
considerable reduction of the prothalliai tissue, to a condition occasionaMy^ 
found in the Angiosperms,tt is characteristic both of Gnetum and Welwit- 
schia ; in Ephedra the prothallus is represented by one or two fully con- 
stituted cells with at least a partial formation of cell- walls. J + 
The pollen-grains of Welwitschia and Grnetum do, however, differ in the 
organisation of the generative cell. In Welwitschia this cell is organised 
before pollination. §§ In G. africanum (Plate XL VI, figs, 9, 10) as in the other 
* Cf. Karsten, 1892, p. 213 ; Lotsy, 1899, p. 94. 
t Cf. Pearson, 1909, fig. 11. 
X Pearson, 1912, p. 618, fig. 20. 
§ Lotsy, 18r;9, p. 94. 
II Karsten, 1893 b, p. 359. 
^ Loc. cit. 
** Pearson, 1906, p. 281, figs. 13-16. 
ft Coulter and Chamberlain, 1903, p. 135. 
XX Strasburger, 1892, pp. 10, 11 ; Land, 1904, p. 6 ; Berridge, 1909, fig. 2. 
§§ Pearson, 1906, p. 281, fig. 16. 
s 
