On Some Stages^ in the Life History of Gnetiiw,. 
263 
type in which the spore -mother-cell formed a tetrad or row of four spores, 
one of which became the embryo-sac, as in the forms represented in Columns 
i, ii, and vii. This tendency prevails regardless of the final structure of the 
sac resulting from its action. A similar tendency is perhaps present in the 
G-ne tales, for, according to Lotsy, the macrospore- mother-cell of Gnetum 
undergoes only one complete division, the nuclei resulting from the homo- 
typic division being included in the embryo-sac* 
But there seems to be evidence in the facts presented in the table that 
this tendency may become operative in the later as well as, or instead of, 
in the earlier stages. 
In Modilewski's Onagraceae (Col. vii), in which the reduction divisions 
are normal, the oosphere appears in the 5th instead of in the 6th generation ; 
in Cypripedium only the homotypic division takes place in the sac, neverthe- 
less the oosphere belongs to the same generation as in Lilium — the 4th. If 
these cases are due to the prevalence of a tendency to shorten this part of 
the life-cycle, then Eujjfiorbia virgata (Col. i) must be regarded as a more 
primitive form or a reversion to a more primitive form than that which is 
established in the majority of Angiosperms. 
While the evidence is admittedly very meagre, there is still some reason 
to believe that the tendency to shorten the series of events between the 
mother-cell stage and the organisation of the gamete has been operative not 
only at the beginning of the series but also at later stages in the ontogeny. 
If so, the gamete which in Welwitschia appears in the 12th or 13th nuclear 
generation, in Gnetum probably somewhat earlier, would in later forms be 
organised in earlier stages, as in the 7th, 6th, 5th, or 4th, as in existing 
Angiosperms. It is at least to the credit of the hypothesis that the tendency 
which, it is suggested, has yielded the Angiosperm-sac, is also believed to 
exist among the lower Gymnosperms,t and may have led to the evolution of 
the free gametes of Gnetum from the intrasporic vegetative prothallus of the 
lower forms. 
In brief, the hypothesis now outlined derives the Angiosperm-sac from a 
primitive form whose essential characters are preserved in Gnetum. The 
evolutionary tendencies believed to have led to this result are (1) a 
tendency to shorten the series of simultaneous divisions of the sac- 
nuclei, with the results that (a) the nuclear generation in which the 
gamete or gametes appear is brought nearer to the macrospore-niother-cell 
and (b) the number of nuclei belonging to this generation is reduced in 
geometrical progression — involving perhaps the grouping of the reduced 
number of sac-nuclei at the poles of the sac ; (2) a tendency to the functional 
differentiation of the nuclei so grouped, which involved the constitution of a 
number of cells within the sac and consequently a still further reduction in 
* Lotsy, 1899, p. 102. 
t See p. 254. 
19 
