founded on the Structure of their Seedlings . 19 
transition the two bundles resemble those in the cotyledon of 
Fritillaria imperialis (PL III, Fig. 2), but are rather nearer 
together, the protoxylem groups fused. As in Muscari 
armenaicum A 3 , and in Fritillaria also, the protoxylem is 
already on the way to become external, though the transition 
proper has not begun. 
During the transition in Eucomis nana the double bundle 
contributes an exact half to the tetrarch root-stele. Its two 
phloem groups are continued directly downwards, and the 
protoxylem, besides forming the ray separating these phloem 
groups, sends out branches to right and left, which unite with 
branches from the two lateral traces entering the stele. 
This is in effect the transition shown in Diagram II. In 
that case, however, the lateral traces were throughout smaller 
than the main traces, and the tetrarch root-stele was, when 
first formed, one-sided from this cause (Fig. 9, PI. II). In 
Eucomis nana the converse is the case. The lateral traces — 
on which are inserted not only the supernumerary traces of 
the sheath but also the plumular traces — are more massive than 
the reduced main bundles, and the tetrarch stele is at first less 
well-developed on the side from which the main traces 
enter it. 
Throughout this description of the transition in Eucomis 
nana I have assumed that the double bundle of the cotyledon 
represents the two main bundles of Muscari. This assump- 
tion is justified, I think, not only by the behaviour of the traces 
in the hypocotyl, but also by their orientation within the 
cotyledonary sheath. The double bundle is placed in the 
thickest part of the sheath, exactly opposite the first leaf, and 
this is also the position of the two main bundles in the sheath 
of Muscari. Compare the position of M x and M 2 in PI. I, 
Fig. 6, Hyacinthus romanus. It is by no means uncommon 
to find two separate bundles replaced by a double bundle in 
allied genera. For example, in Fritillaria (PI. Ill, Fig. 2) 
the bundles of the cotyledonary sheath are separate: in Lilium 
they are joined as in Eucomis. In Bloomeria they are separate : 
in Allium joined. 
C 2 
