238 Transactions of the Royal Society of South Africa. 
protoplasmic layer. All the nuclei are in precisely the same stage of 
division. It may be assumed that all the divisions which occur in the sac 
before septation are simultaneous, as they appear to be in Welwitschia * and 
in Ephedra.t The chromosomes cannot be counted with certainty, but they 
are probably twelve— the number determined by Coulter in the reduction 
divisions of the microspore mother-cell of G. Gnemon.% 
Lotsy § suggested the possibility that all the free nuclei of the " fertile " 
(micropylar) region of the sac were descended from one of the two daughter- 
nuclei resulting from the first mitosis in the macrospore ; those of the 
chalazal end from the other. The absence of dividing nuclei from his 
material " made it impossible to follow this question up, yet it is a point of 
primary importance." In view of what is now known of the origin of the 
endosperm in Welwitschia and in Gnetum, the point no longer possesses 
that degree of importance which in 1899 was very naturally attributed to it. 
In the early stages figured by Strasburger,|| as in those seen by us, the 
arrangement of the nuclei does not support Lotsy's suggestion ; and the 
later stages already known, or now to be described, indicate that the primary 
nuclei of the sac differ from one another only in the accident of position. 
One of the many puzzling features of the life-history of Gnetum has 
been the occurrence of fertilisation before the formation of endosperm in 
some species^ {G. Bum/phianum, G. funicular e) and after it in others 
{G. Gnemon).** This difference of character has resulted in the informal 
separation of the species of Gruetum already investigated into " Karsten's 
Gi-neta " and Gnetum Gnemon,ff the latter being regarded as an intermediate 
form between Welwitschia and the Grneta described by Karsten.":]:^ 
Karsten's statements, howover, do not justify such a conclusion. While 
fertilisation does occur in Karsten's G-neta before septation begins, §§ this is 
not always the case, for, having described and figured the free nuclear sac of 
G. ovalifolium,\\\\ he proceeds, "In diesem Zustande nun erwartet der 
Embryosack den herannahenden Pollenschlauch. Eine weitere Differenzirung 
seines Inhaltes tritt nicht ein, hochstens spannen ein paar Plasmafaden im 
Scheitel von einer Seite des waudbelages zur andern, quer durch den 
Embryosack hindurch, oder es wird ganz in dem Chalazaende eine Endo- 
Pearson, 1909, p. 346, figs. 18, 19. 
t Jaccard, 1894, p. 15, figs. 24, 25 ; Land, 1904, p. 10, figs. 33, 35. 
X Coulter, 1908, p. 47. 
§ Lotsy, 1899, p. 92. 
II Strasburger, 1879, figs. 59, 61, 62, 63. 
^ Karsten, 1893b, p. 356. 
Lotsy, 1899, p. 93. 
ft Lotsy, loc. cit., p. 101 ; 1911, pp. 353, 359; Pearson, 1909, p. 357. 
Xt Lotsy, loc. cit., p. 103. 
§§ Karsten, 1892, p. 212, figs. 8-11 {G. funicular e). 
II II Karsten, 1893 b, p. 356, Taf. x, figs. 56, 57. 
