244 Transactions of the Royal Society of South A frica. 
Soon after the completion of the primary endosperm there commences at 
a point not far from the centre of its axis the differentiation of a small-celled 
tissue, which later becomes a very marked feature.* Longitudinal and 
transverse divisions result in the formation of a tissue of cells (Plate L, fig. 
26). These divisions extend upwards and downwards from the point at 
which they commence, and ultimately produce a columella-like cylinder, 
which extends without interruption from growing point to growing point, 
and is broad enough to be conspicuous to the naked eye in a bisected 
endosperm. Towards and into this cylinder the early growth of the pro- 
embryos is directed. In its early stages the cells contain no starch ; later 
this region as well as the larger-celled peripheral portion of the endosperm 
are well provided with starch. The position, form and cell-characters of 
this central cylinder " and its relation to the proembryos suggest that it is 
a simple form of absorbing, storage and conducting tissue. Whatever its 
precise functions, its presence and that of the two growing points indicate a 
considerable degree of differentiation.f There is no doubt that the endo- 
sperm of G-netum is not only a *' special formation," but " an organism." % 
4. The Oospore and Proembryo. 
Cells believed to be oospores are shown in Plate XLIX, figs. 23 and 24. 
But as fertilisation has not been seen, nor the relations between these cells 
and the proembryos determined, their identification requires confirmation. 
It is not clear from Lotsy's account that he saw the fusion of the gametes. 
He states that the oospores are distinguishable by the large size of their 
nuclei § and that they occur in pairs, each pair resulting from the action of 
the two sperm-nuclei from one pollen-tube. These oospores were later seen || 
to produce tubular proembryos. In G. ovalifolium and G. Bymphianum 
also each of the two sperm-nuclei of the pollen-tube fuses with a female 
gamete,^ but in G. africamim the number of oospores is certainly not regu- 
larly twice the number of pollen-tubes, since in some cases the sperm-cell 
contains only one nucleus when it reaches the sac (p. 236). 
Information regarding the details of fertilisation is greatly to be desired. 
In the case of Welwitschia, the protoplasm of the male cell plays an impor- 
tant part in the constitution of the oospore.** Since in G-netum the female 
gamete is a free nucleus it might be anticipated that the same would hold 
* Coulter, loc. cit. 
t Cf. Pearson, 1909, p. 371. 
: Lotsy, 1899, p. 93. 
§ Lotsy, 1899, p. 96. 
|] Lotsy, loc. cit., p. 97 ; fig. 55. 
^ Karsten, 1 893, p. 372. 
Pearson, 1909, pp. 361, 362. 
