On Some Stages in the Life History of Gnetum. 
237 
G. Gnemon it usually, if not always, remains inside of the pollen- tube."* 
We have never been able to distinguish the tube-nucleus in G. africanum^ 
either in the sac or in a pollen-tube which has reached the end of its growth. 
3. The Embeyo-sac and Endosperm. 
Several stages in the development of the embryo-sac and endosperm 
have been described. 
It seems probable that more than one macrospore mother-cell is 
organised,! though perhaps only one usually gives rise to macrospores ; but 
according to Karsten,;]: "aus den Embryosack-Mutterzellen geben nun durch 
weitere Quer-und Langstheilungen eine ganze Anzahl von Embryosack- 
anlagen hervor." He further states § that he never saw more than three 
cells formed by the transverse division of a single mother-cell. According to 
Lotsy,|| the embryo-sac of G. Gnemon arises by the division of the mother- 
cell into two, each of which may become an embryo- sac. Our material gives 
no information regarding these early stages. Usually only one embryo-sac 
is organised in G. africanum ; in G. Gnemon, as stated for this and other 
species by other observers, two or three are not infrequently present.^ 
The earliest stage of the sac shown in our preparations contains four free 
nuclei. Later stages up to and including the septation of the lower part of 
the sac and the fusion of the free nuclei in each compartment have been 
described.** In G. africanum, as in G. Gnemon, the embryo-sac before 
septation is hardly more than four times as long as it is broad. In G. 
scandens it is much longer (Plate XLV, fig. 1) in proportion to its breadth, 
a character which appears to be constant in this species. The sac here 
outlined (Plate XLV, fig. 1) shows the aggregation of free nuclei at the 
chalazal end. There is yet no indication of septation. Later stages have not 
been seen in this species. 
Neither Karsten ff nor Lotsy |+ saw any stage of mitosis in the free 
nucleate sac. We have seen one only (Plate XL VII, fig. 17). The ovule in 
which it occurs is pollinated, and pollen-tubes have already penetrated about 
one-third of the length of the nucellar cap. In this sac the central vacuole 
is still present, and the nuclei are uniformly distributed throughout the thin 
* Lotsy, 1899, p. 95. 
f Strasburger, 1879, p. 116, Taf. xiii, figs. 51-55. 
X Karsten, 1893 b, p. 354. 
§ Loc. ext. 
II Lotsy, 1899, p. 102. 
T[ Strasburger, 1879, Taf. xiv, figs. 59, 60 ; Karsten, 1892 ; p. 210 ; Lotsy, 1899, 
p. 90; Pearson, 1915 a, fig. 2. 
Pearson, loc. cit., p. 318 (on p. 319 of this paper, "multinucleate," the last word 
in line 8 from the top, should read "uninucleate." 
ft Karsten, loc. cit., p. 356 ; tt Lotsy, 1999, p. 92. 
