158 Sargant . — The Reconstruction of a Race of 
But if embryological evidence fails to answer a comparatively simple 
question such as the method by which two cotyledons in the ancestral 
Corydalis , Carum , or Ranunculus became reduced to one in certain living 
species, how can we expect to settle the number of cotyledons possessed by 
a form so remote as the Primitive Angiosperm by evidence of the same 
kind drawn from the structure of its living descendants ? 
So far the comparison with Pseudomonocotyledons has served merely 
to confirm previous doubts as to the phylogenetic value of characters 
belonging to the young embryo. It is pretty clear that such characters are 
mainly adaptive. Ancestral features seem to be obliterated quickly and 
very completely at that early period. Two causes probably co-operate 
to this end: the severity of the competition which follows germination, 
and the plastic nature of the tissues in the young embryo. A very slight 
start in the race may mean survival to the seedling, and some minute 
variation in the structure of the mature embryo may ensure such a start. 
At the same time slight changes in the form of a mass of meristem are 
easily effected. 
What conclusions can be drawn from the history of the embryo within 
the embryo-sac ? We have discussed three possibilities with regard to the 
number of cotyledons possessed by the Primitive Angiosperms. First, that 
they had one cotyledon, which gave rise to the pair found in living Dico- 
tyledons by fission. This possibility is usually thought to involve the 
existence of a truly terminal cotyledon both in the Primitive Angiosperms 
and in living Monocotyledons. If this were so, we might well hesitate 
to accept an explanation which involved such a consequence. The 
difficulties arising out of this hypothesis have already been discussed. But 
the development of the embryo in the Pseudomonocotyledons examined 
by Hegelmaier, Schmid, and Sterckx, and in genera from the Dioscoreaceae 
and Commelinaceae described by Solms-Laubach, shows how a cotyledon 
of undoubtedly lateral origin may assume a terminal appearance, pushing 
the stem-apex to one side. The apparently terminal cotyledon of Alisma 
and other Monocotyledons may be derived from a single lateral cotyledon 
through intermediate forms of this kind. 
The possibility of a monocotylar race of Primitive Angiosperms need 
not then be abandoned on the ground that their single cotyledon must 
needs be terminal. Nor is it impossible to derive a dicotylar embryo from 
such an ancestor. The single cotyledon might divide to form a pair of 
members which by lateral shifting would come to occupy the same relative 
position as the cotyledons of living Dicotyledons. The view so amended 
has, however, lost its chief recommendation — that of simplicity. The 
cotyledons of Capsella can no longer be derived by fission only frorcqthe 
single member found in Alisma . 
