30 Sargant . — Theory of the Origin of Monocotyledons 
The insertion of the plumular traces takes place as in 
Chlorogalum (PL IV, Fig. 6) : the three traces of the first 
leaf (A, B, and C, in Fig. 6) unite with each other and then 
divide into two branches ( Pl v PI 2 ) which join the two cotyle- 
donary traces (M l5 M 2 ). 
The root-stele when first formed suggests a diarch structure, 
and this appearance persists for some time. In Fig. 7 (PL 
IV) there are only two phloem groups, but the protoxylem 
elements internal to them are quite distinct (px 2 . ) px z ). They 
finally disappear a little further down, and the root appears 
truly diarch. 
Finally, in the seedling B : figured in Fig. 5, the group 
px 2 recovers itself and divides the phloem group lying outside 
it into two. The opposite phloem group remains single, and 
the corresponding group of protoxylem, px Zi is suppressed 
permanently. The root of course is triarch. 
In the two other seedlings examined both lateral proto- 
xylem groups reappear after a temporary eclipse, and the 
stele of the root is tetrarch. 
Arthropodium cirrhatum. The two bundles of the cotyledon, 
though slender and lying close together, are distinct from 
each other until they reach the base of the cotyledon. Here 
their protoxylem groups unite, and they assume the typical 
double structure. 
The plumular traces unite with those of the cotyledon just 
as in Chlorogalum and Anthericum (PL IV, Fig. 9). A diarch 
root-stele is formed (PL IV, Fig. 10). During the process 
two lateral protoxylem groups appear, and they have not 
quite vanished in Fig. 10 ( + , + ). The two massive phloem 
groups, however, formed by the union of a cotyledonary 
with a plumular mass of phloem on either side of the stele, 
are never broken up. The root remains diarch to the end. 
The vascular symmetry of Arthropodium is so far reduced 
from that characteristic of Chlorogalum , that it precisely 
resembles the Allium symmetry (Diagram VII). 
