On Some Stages in the Life History of Gnetum. 255 
(2) the presence of archegonia in the former and their complete absence 
from the latter ; (3) the reduction of the oosphere of the former to a free 
nucleus in the latter. These differences are physiologically very considerable. 
Nevertheless, if the one sac is directly derived from the other, there is every 
reason to suppose that it is the consequence of a comparatively simple series 
of reductions, a series, perhaps, not less simple than those which have 
resulted in the reduced type of pollen-grain. Prima facie it is not 
improbable that parallel series of reductions affecting the development of 
the free microspore, and the macrospore retained permanently within the 
sporangium, should produce very divergent results. In the case of the 
pollen-grain the autophytic prothallus of the pteridophyte has lost its 
assimilating function and undergoes its shortened development under new 
physiological conditions. But the physiological changes which are bound 
up with the evolution of the prothallus of the seed from the free-living type 
are more considerable. While the development of the macrospore is much 
less shortened than that of the microspore, it has not only lost the power 
of assimilation, but, for a prolonged period, it leads the life of a parasite,, 
becoming in due course the host of a new parasitic sporophyte. It is not 
unlikely, therefore, that its stability is considerably impaired, and that 
a further series of reductions should produce results apparently out of 
proportion to the causes to which they are immediately due. 
If we may assume the existence of certain tendencies and their complete 
realisation in the final stages of the evolution of the Gymnosperm prothallus, 
the conditions now obtaining in the embryo-sac in certain species of Gnetum 
could be realised. These tendencies are: (1) The extension of the poten- 
tialities of an Archegonium-initial nucleus to every nucleus in the sac ; 
(2) the reduction of the Archegonium to one (or more) gametes ; (3) the 
organisation of these at a period antecedent to the formation of cell- walls. 
During recent years a number of Gymnosperms have been investigated 
in which the Archegonia (or archegonial initials) are very numerous and 
irregularly distributed upon or within the prothallus. These occur both in 
the Cycads (Microcycas) * and in the Conifers (Araucaria,t Agathis,^ 
Sequoia, § Widdringtonia,|| Callitris,^ Actinostrobus **). In Sequoia the 
pollen-tube appears to exercise at least a directive influence upon the 
development of the archegonial initial. ft In Widdringtonia, Callitris, and 
* Caldwell, 1907. 
t Seward and Ford, 1906, pp. 306, 367. 
+ Eames, 1913, p. 5. 
§ Lawson, 1904. pp. 14 and 599, fig-. 17. 
II Saxton, 1910 a, text-fig. 1. 
% Saxton, 1910 b, p. 560. 
** Saxton, 1913, p. 327. 
ft Lawson, 1904, p. 17 ; i. e. not confined to a group at the micropylar end of the 
prothallus. 
