The Origin of the Seed-Leaf in Monocotyledons. 109 
These schemes are of two kinds: in Lilium and other genera of the 
Tulipeae, for example, the cotyledonary traces themselves change 
from an asymmetrical to a symmetrical orientation as they pass 
downwards through the hypocotyl. The Allies, on the contrary, 
depend on plumular traces to complete the symmetry of the 
root-stele. 
Had this asymmetrical ground-plan turned out to be essentially 
monocotyledonous, as I then thought it, two views of its origin were 
possible. Such a cotyledon as that of Lilium might be considered 
as originally terminal, but early pushed aside by the more vigorous 
plumule. This would suggest the superior antiquity of Monocoty. 
ledons, and the derivation of Dicotyledons from an ancestoi with 
one seed-leaf. The structure of Allium , on the other hand, admitted 
of a different explanation. Its cotyledon might conceivably be 
lateral in origin as well as in appearance. The second half of the 
root-stele might originally have been supplied by the traces of a 
second cotyledon opposite the first It was easy to conceive of a 
series of forms in which the traces of the plumule should step by 
step replace those of a gradually disappearing seed-leaf. 
When at length a form was reached from which others were 
clearly derived, I found to my surprise that it was truly symmetrical 
throughout. The ground-plan of Anemcirrhena is unmistakeable in 
the allied genera Aspliodelus and Asphodeliue. Through the inter¬ 
mediate forms of Chlorogalum and Anthericum , it is connected with 
Arthropodium , and the ground-plan of Arthropodium is very similar 
to that of Allium and its allies. 
The vascular systems of seedling Asphodeleas and Allieas are 
thus linked together by the probably ancient structure of Anema- 
rrliena. The vascular system of its cotyledon, hypocotyl, and primary 
root—already described elsewhere in detail 1 —is given in the series 
of diagrams 2-6. It is, as I have said, a symmetrical system. The 
cotyledon is a somewhat flattened green cylinder ending in a 
tapering green apex, which in young seedlings is still enclosed 
within the seed-coat. The plumule is developed rather late. In 
the seedling figured it is still quite embryonic, and is completely 
enclosed within the base of the cotyledon. The axis is slightly 
thickened just below the plumule, and it passes at once into the 
very well-developed primary root (fig. 1). 
Two massive bundles run the whole length of the cotyledon. 
In transverse section it is somewhat elliptical, and the bundles are 
1 K. Sargant, Aim. of Bot., vol. xiv., 1900, p. 633. 
