On Some Stages in the Life History of Gnetum. 
249 
ment that " the nuclei which, in G-netum and Welwitschia, fuse to form the 
nuclei of the primary endosperm -cells are potential gametes which lose this 
potentiality by fusing," would appear to be fully justified. 
A more detailed examination of the facts established for G-netum and 
Welwitschia shows that of the three courses which a sac-nucleus may follow, 
(1) its exclusion from further participation in the life-history is due entirely 
to the accident of position in both genera ; (2) in Gnetum its fertilisation or 
inclusion in the endosperm-fusions is determined {a) by the time of arrival 
of the male gamete, {h) by its position with reference to the point in the sac 
to which the male gamete is carried by the pollen-tube. In Welwitschia 
this alternative is decided mainly by its position in the sac — the greater 
elongation of the micropylar region making for the retention of sexual 
characters, the lesser growth of the chalazal and middle regions for endo- 
sperm-fusion.* 
Lotsy,t in criticising this conclusion (that the free nuclei of the sac are 
all potential gamete-nuclei), considers that all nuclei of an x generation are 
of equal value, and one can, if one wishes, consider all to be potential 
gamete-nuclei. This seems to depend upon the exact meaning of the term 
potential gamete." This term was formerly used, J and is now used, of 
nuclei which, given the opportunity, will fuse with a male gamete to produce 
a new sporophyte. There is, as we have tried to show, good reason to believe 
that all the sac-nuclei after the last free-nuclear mitosis in Gnetum and 
Welwitschia are potential gamete-nuclei in this sense. This is not the case, 
for example, in Pinus : the gamete is not present among the free sac-nuclei ; 
it appears at the end of a series of divisions which are undergone by a few 
of these nuclei only. The rest form prothallial cells, which, like all their 
descendants, are entirely vegetative in function ; they are not potential 
gametes, nor do they produce potential gametes. It is true that there seems 
to be a tendency among the Coniferae towards the condition in which all the 
sac-nuclei are potential gametes, as they are in Gnetum and Welwitschia 
(see p. 248) . But no Conifer so far described has yet reached this condition. 
Whether the physiological difference between a gamete-nucleus and a 
vegetative nucleus, as these appear in the Conifer sac, is great or small, the 
difference exists. Our position is that this difference no longer exists in 
G-netum and Welwitschia ; all the sac-nuclei present in these genera before 
septation begins are believed to be potential gametes in the sense defined 
above. Lotsy's comparison of the nuclear fusion which occurs in the 
prothallial cells of Taxus, Ephedra, and other of the lower gymnosperms with 
the nuclear fusion in the endosperm compartments of Welwitschia is not 
valid because (1) the fusing nuclei in the former cases are not gametes, and 
* Pearson, 1909. 
t 1911, p. 340 
X Pearson, 1909. 
