140 
Agnes Robertson. 
ignorance general criteria are not available, so each case that arises 
must be dealt with on its own merits. In certain instances, as for 
example in those Lycopodium prothalli which have taken to a sapro¬ 
phytic mode of life, ancestral characters are obviously swamped by 
adaptive ones. Even in less extreme cases the paramount necessity 
of securing sufficient nourishment for the egg, and a satisfactory 
start in life for the young embryo, must tend to overrule vestigial 
characters in the female organ. Thus it is in the antheridia, or in 
the male gametophyte if the prothallia are dioecious, that we should 
be disposed to expect a more marked survival of ancestral characters, 
as natural selection would have operated less actively in eliminating 
them. This is borne out by the fact that Heim finds structural 
differences between the antheridia of various groups of ferns, which 
are not paralleled by corresponding differences in the archegonia. 
It is now generally believed that the Gymnosperms have deve¬ 
loped from fern-like ancestors, and hence we must suppose that the 
very ancient primaeval ancestors of our modern Conifers had a 
well-developed prothallus which has since undergone a process of 
reduction. In all the Gymnosperms of the present day the female 
prothallus survives in the form of the endosperm—retained, pre¬ 
sumably, because it could be turned to account as a food-tissue. 
But the male prothallus is extremely reduced, consisting at the 
most of two cells, or is even absent altogether. The most archaic 
types now surviving are those in which we should expect to find the 
clearest traces of such a prothallus. The Cycads and Ginkgo , 
which are recognised on other grounds as comparatively primitive 
types, possess respectively one and two prothallial cells. That traces 
of the prothallus should survive in these plants is just what we 
should expect, but we further find that the possession of two pro¬ 
thallial cells is a character shared by the Abietinete, perhaps the 
most highly modified of the Coniferales, and also by Podocarpus, a 
genus referred to the Taxacese and thus very remote both from the 
Abietineae and from Ginkgo ! Two families among the Pinaceae 
(Taxodieae and Cupresseae) and one among the Taxaceae (Taxeae) 
are alike in shewing no trace of a male prothallus. So at first sight 
the presence of prothallial cells seems to be a capricious character, 
appearing unaccountably in groups which we are forced on other 
grounds to regard as standing at different morphological levels. 
I think, however, that we are here dealing with the case of an ances¬ 
tral character obscured by an adaptive one. Somewhere in the 
course of the evolution of the Gymnosperms a stage must have been 
