236 Transactions of the Royal Society of South Africa. 
some cases the cell becomes divided into two uninucleate portions, probably 
as a result of a simple mechanical strain as is suggested for Welwitschia.* 
Both these conditions are recorded by Karsten for G. Bumjphianum and G. 
ovalifolium.f In G. funiculare the protoplasm of the binucleate cell 
remains undivided until it enters the embryo sac. 
The nucleolus of the sperm nucleus is large and the surrounding nuclear 
substance contains numerous coarse chromatin granules. The male nuclei 
of G. ovalifolium and G. Gnemon, figured by Karsten]; and LotsyJ respec- 
tively, possess no nucleoli but " more or less regular balls of chromatin accu- 
mulated in a mass which resembles a raspberry somewhat." § These 
characters have not been seen in our material. Lotsy's figure represents 
these nuclei just before their entrance into the sac. In earlier stages figured 
by Karsten, 1 1 they appear to be very similar to those seen by us in (r. afri- 
canum. We have identified the sperm nuclei just before they enter the sac 
they still retain the characters of the Inucleolus and chromatin described 
above. 
We have not succeeded in obtaining certain proof that the smaller of the 
two sperm nuclei is in process of disintegration, though its condition (e. g.^ 
Plate XLVI, figs. 10, 13) frequently suggests this conclusion. In the lower 
part of the nucellar cap it is frequently not possible to find the second sperm 
nucleus (Plate XLYII, fig. 16). It is therefore not improbable that one of 
them may disappear as is sometimes the case in Welwitschia.^ If this is so,, 
there is little doubt that the disappearing nucleus is a sperm nucleus and 
not the homologue of the so-called " stalk-nucleus " of Ephedra and the 
lower gymnosperms. The history of the male gametophyte of G-netum and 
Welwitschia is shortened, not only by the partial elimination of the pro- 
thallus, but also by the complete suppression of one of the later divisions, 
almost certainly the first division of the " body cell " of the lower forms. 
The tendency to reduce the series of events intervening between the spore- 
mother-cell and the functional gamete is seen in the history of the micro- 
spore as in that of the macrospore. 
The tube-nucleus usually maintains its spherical form to the end (Plate 
XL VII, fig. 16). Sometimes it becomes crescent-shaped, as occasionally in 
Welwitschia.** Karsten doubtfully recognises the partially disorganised 
tube-nucleus in the embryo-sacff in G. Bumjphianum and G. ovalifolium; in 
* Loc. cit. 
t Karsten, loc. cit., p. 361. 
X Karsten, 1893b, figs. 68, 69 ; Lotsy, 1899, fig. 41. 
§ Lotsy, loc. cit., p. 95, 
II Karsten, loc. cit., p. 360, fig. 64. 
1[ Pearson, 1909, p. 360. 
** Loc. cit., fig. 43. 
ft Karsten, 1893 b, p. 368, fig. 71. 
