On Some Stages in the Life History of Gnetum. 235 
species described,* it is first recognised after germination. The marked 
difference in the sizes of the tube and generative nuclei by Karstenf is not 
observed in G. africanum ; here the tube nucleus, before it shows signs of 
deterioration, is usually somewhat smaller than the other (Plate XLVI, 
figs. 9, 10). 
In the pollen chamber in G. africanum the grain swells, splitting and 
casting off the exine (Plate XLYI, fig. 8, ex) ■ this is described also for other 
species of G-netum,| for Ephedra, § and for Welwitschia.|I The pollen-tubes 
usually follow a fairly direct line from the base of the pollen-chamber 
towards the summit of the sac. The diameter of the axial tissue of the 
nucellar cap is frequently so limited by the degeneration already referred 
to (p. 233), that the passage of three or four pollen-tubes almost completes 
its disorganisation, leaving a broad irregular cavity. 
The pollen-tube, so far as we have seen, always enters the embryo sac at a 
point very near its apex. In G. Gnemon, according to Lotsy,^ it sometimes 
penetrates the sac-wall at a lower point, but always in the upper half. When 
there are several pollen-tubes, the destruction of nucellar tissue is particularly 
great just above the summit of the sac (Plate XLYI, fig. 11, c). If the 
pollen- tube contains three nuclei at the time of its germination, one of these 
does not enter the tube (Plate XLVI, fig. 8). Of the other two, the second, 
soon after entering the tube, becomes incorporated in a cell— the generative 
cell (Plate XLYI, figs. 9, 10). The organisation of the protoplasm of this 
cell must be effected very rapidly ; no stages intermediate between the free 
nucleus and the apparently fully-organised cell have been observed. 
Soon after the organisation of the generative cell, two nuclei are seen in 
it (Plate XLYI, figs. 12, 13, 14 ; and Plate XLYII, fig. 15). The division 
which yields this 2-nucleate generative cell has not been seen. The corre- 
sponding division of the generative cell of Welwitschia has been described.** 
The two nuclei in Gnetum africanum are usually unequal in size or staining 
capacity, as they are in Welwitschia, and, as also in the latter, the advantage 
may lie either with the leading or the following nucleus. ft Their shape does 
not vary to the same extent as in the homologous nuclei of Welwitschia ; 
they are usually spherical or nearly so ; an exception is shown in Plate XLYI, 
fig. 14. Usually they are both retained in the same unit of protoplasm ; in 
* Karsten, 1893 b, p. 360, figs. 59-63 ; Lotsy, 1899, p. 94. 
t Loc. cit. 
X Karsten, loc. cit., p. 359. 
§ Land, 1907, p. 275; Berridge and Sanday, 1907, p. 132 ; Berridge, 1909, fig. 2, 
II Pearson, 1906, fig. 17 ; 1909, p. 358. 
^ Lotsy, loc. cit., p. 94, fig. 40. 
** Pearson, 1909, figs. 44, 48. 
tt In G. funicular e, etc., the two sperm nuclei are said to be quite alike. Karsten, 
1893b, p. 360. 
XX Pearson, 1909, p. 360. 
