1904 | SARGANT: . EVOLUTION OF MONOCOTYLEDONS 339 
what differentiated from the primitive stock of angiosperms by 
adaptation to a geophilous habit, and was thereby tending to a 
monocotyledonous structure, 
The development of the embryo within the embryo sac of 
monocotyledons has already been described at some length, and 
the value of the phylogenetic argument based 
on it has been criticised. Nevertheless, since 
vind ad great importance has been attached to the appa- 
rently terminal position of the single cotyledon, it is 
worth while to consider how the facts appear from a 
new standpoint. If the cotyledonary 
member be derived from the two coty- 
ledons of an ancestor, its rudiment 
cannot be really terminal, but must 
‘represent the congenital fusion of two 
lateral members. The terminal posi- 
tion is readily understood by compari- 
son with such a 
seedling as that of 
Delphinium nudt- 
caule( figs.rand2). gots: 
In this species 
the leafy stem bud 
develops in the same season as 
the cotyledons, and so soon as 
| the first leaf attains any size it 
i 1.—Del- breaks through the cotyledon- 
ealize at ary tube near its base. 
eight weeks after FO! @ Short time this leaf 
seed wassown, appears to be laterally 
inserted on the coty- 
ledonary axis (fig. r), but as the leafy 
stem develops the cotyledons are pushed to 
ecnber (ig. 2) Tien coon aaa a pe 
nts y COT- nudicaule. Seedling 
rected; the stem bud is seen to be terminal, fourteen weeks after 
and the tube of the united cotyledons a fusion ee 
of two lateral members. 
