Primitive Angiosperms . 171 
This similarity in structure between the seedling of the primitive 
Monocotyledon and that of the primitive Dicotyledon is precisely what was 
foreseen on theoretical grounds (ante, p. 160). It confirms the results 
derived from study of the evidence relating to both classes. It is com- 
pletely and simply explained by supposing the seedling of the Primitive 
Angiosperm to have possessed tetrarch vascular symmetry which it handed 
down with more or less modification to its immediate descendants, mono- 
cotylar or dicotylar. 
If this be accepted as a working hypothesis, does it follow that the 
embryo and seedling of the Primitive Angiosperms had two cotyledons ? 
The seedling of Anemarrhena has but one, yet its tetrarch symmetry 
is very clear. Might not a monocotylar Primitive Angiosperm give rise 
to tetrarch dicotylar forms, as well as a dicotylar Primitive Angiosperm 
to tetrarch monocotylar forms ? 
If the alternatives are followed out to their consequences, the second 
hypothesis — that of the dicotylar Primitive Angiosperm — is certainly far 
more probable. Any stele which receives traces from two similar and 
opposite cotyledons, each symmetrical about the same median plane, must 
itself be symmetrical about two vertical planes : that is, about the median 
plane common to both cotyledons, and about the intercotyledonary plane 
perpendicular to it. The reduced or diarch type of skeleton described 
in the seedlings of Delphinium and Nigella is as completely symmetrical 
about both planes as the tetrarch skeleton of Althea. Reduction within 
the seedlings of Dicotyledons does not tend to produce a skeleton sym- 
metrical about one plane only. On the other hand, a stele receiving traces 
from a single lateral member is naturally symmetrical about the median 
plane of that member, but not necessarily or even probably symmetrical 
about the vertical plane at right angles to it. 
Accordingly, among Monocotyledons, the tetrarch skeleton as it is 
found in Anemarrhena, Albuca , and Galtonia, is the only type symmetrical 
about two planes. Even in these cases symmetry about the plane bisecting 
the cotyledonary traces is not quite complete, for these traces are slightly 
displaced in one region of the axis by the insertion of the plumule (Sargant, 
72, Diagram VI (A), pp. 26 and 70, Fig. 3). 
That the tetrarch skeleton, symmetrical about two vertical planes, 
should be found in Monocotyledonous seedlings of an ancient type is 
exactly what would occur if the primitive Monocotyledon were derived from 
a dicotylar ancestor by the fusion of both cotyledons into a single member. 
The dual symmetry of the internal structure would naturally survive the 
external bisymmetry. On this hypothesis we may suppose that two 
cotyledons were characteristic of the main line of Angiospermic descent 
from the original Pteridosperm down to the evolution of the primitive 
