founded on the Structure of their Seedlings . 39 
always odd. and the lateral bundles are always arranged 
symmetrically on either side of the midrib. 
The cotyledon of the same species, however, commonly 
possesses two equivalent bundles, and if others are present 
they are arranged symmetrically with regard to both. These 
main bundles are often distinct and widely separate, as in 
Albuca , Galtonia , Anemarrhena , Bloomer ia, Aloe, and other 
genera. When they approach each other more closely, as 
in Muse aid armenaicum (PI. II, Fig 5) and Fritillaria 
imperialis (PL III, Fig. 2), they occupy the position of 
a midrib. This is still clearer in Allium , Lilium and many 
other genera, in which the two main bundles are much reduced 
in size, and are united by a common protoxylem group either 
in the sheath of the cotyledon only or throughout its length 
(PL V, Figs. 1 and 2). 
The cotyledon of Zygadenus with its solitary bundle seems 
at first sight to be an exception to this rule. During the 
transition, however, this bundle opens out into a double 
structure perfectly comparable with that of Allium. In the 
allied genus Veratrum the central bundle of the cotyledon 
does not open out in this way during the transition, and but 
for the relationship with Zygadenus we might take it for 
a true midrib. 
Another very striking example of the same kind is found 
in the two species Yucca aloifolia and V. gloriosa. The 
cotyledon of Y. aloifolia contains three bundles, and the 
central one occupies the position of a midrib. During the 
transition this bundle opens out into the characteristic double 
structure (ante, p. 35). In the cotyledon of Y. gloriosa there 
are four bundles, but just before the transition begins the two 
central bundles approach each other so closely that their 
protoxylem groups unite. There can be no doubt as to the 
double origin of the ‘ midrib * thus formed, and it strengthens 
the presumption that the central bundle of F. aloifolia is also 
double. 
From these examples it is quite clear that the whole case 
for the absence of a true midrib in the cotyledon of Liliaceous 
