Primitive A ngio sperms. 1 5 1 
leaves appear lateral. Are we therefore to conclude that the first two or 
three leaves in Sparganium ( 36 , pp. 648-52) and the first seven or eight in 
Pistia ( 56 , pp. 692-3) are really cauline structures which have become 
undistinguishable from true leaves in response to the demands of the 
environment, while the succeeding leaves alone have a morphological right 
to the name ? 
Those who shrink from this conclusion are faced by two alternatives. 
They must suppose either that the leaves are only apparently terminal 
through suppression of the axis, or — with Celakovsky ( 16 ) — that leaves 
were originally terminal members, that the stem began as a symposium of 
leaf-bases, and that this primitive construction survives only in the embryo 
and young seedling of a few Monocotyledons. The gradual transition from 
the terminal to the lateral position, which can be traced in the successive 
leaves of such forms, would then reproduce a similar transition in the 
history of the race. 
The species which on this view are most primitive among living 
Angiosperms are, of course, those in which most terminal leaves are formed 
before the appearance of an axis. In Hegelmaier’s researches Spcirganium 
was found to have two or three terminal leaves, Pistia seven or eight. Both 
are aquatic forms. Pistia is very highly specialized to that habit. The 
seed before germination remains under water, but soon after germination 
begins it floats up to the surface, and further development proceeds there. 
The embryo of Pistia is much reduced : neither a suspensor nor a primary 
root is formed. 
Fleischer’s best examples are J uncus glaucus, a semi-aquatic form, and 
Luzula multiflora , an Alpine species. 
My own experience is that the reduction in structure so characteristic 
of aquatic species is very strongly marked in their young seedlings. For 
this reason I have rarely found such seedlings of much use in my own 
work. Ancestral features have commonly disappeared or become obscure 
in the general loss of differentiation. Thus the fact that terminal leaves 
seem characteristic of aquatic embryos suggests very strongly that their 
peculiar position is due rather to suppression of the axis than to any 
reminiscence of a stemless period in the history of the race. 
If we adopt this view, the complete suppression of the axis in the 
embryo and seedling of Pistia , for example, must be due to adaptation of 
the embryo itself to its surroundings. If the absence of an axis were 
a mature character which had become embryonic in the course of evolution, 
we should expect traces of a stem in the embryo which would disappear in 
the mature plant. But the converse is the case. The stem exists in the 
mature plant, and is absent in the embryo and seedling. 
In the case of Pistia the conditions of germination are peculiar, and 
will account for very great adaptations in seedling structure. But less is 
