2 2 Sargant. — T heory of the Origin of Monocotyledons 
Very great individual differences occur within the limits 
of this species. The transition follows type 5 — that is, the 
phloem of each cotyledonary trace branches in three directions, 
the xylem in two. This would normally lead to a tetrarch 
root, and does so in one seedling (A 2 ). In another (A 3 ) a very 
interesting variation is found (Diagram IV). 
The phloem groups at first behave in the usual way (Diagram 
IV, B), but before the branching is accomplished and four 
definite groups are formed, the original groups re-assert them- 
selves, and the phloem elements collect once more in two 
large masses. The four protoxylem groups are still distinct, 
A 
C 
B 
Diagram IV. 
but the right-hand branch of one group is not separated from 
the left-hand branch of the other by phloem elements, and 
the four groups naturally fuse in pairs (Diagram IV, C). The 
diarch root-stele is then complete. 
If we neglect the very small plumular trace which inserts 
itself on the stele during the transition without affecting its 
symmetry, this is precisely the structure characteristic of the 
Tulipeae, in which, however, the two cotyledonary traces are 
commonly reduced to a slender double bundle. But in 
Ornithogalam exscapum the size of the cotyledonary traces, 
their independence, and their behaviour during the transition 
in seedling A 2 , leaves no doubt as to their identity with the 
main bundles of Galtonia and Albnca. 
Before leaving the Scilleae, I must mention an anatomical 
peculiarity of the root-cortex which is found in a number of 
species belonging to it. 
