Influence of the Adult Plant upon the Seedling. 329 
throughout the whole length of the seed-leaf. In a very few cases 
the phloem may divide into two masses in the base of the cotyledon ; 
it is, however, in the upper region of the hypocotyl that each seed- 
leaf-trace divides completely into two, each half rotating in such a 
way that the protoxylem becomes isolated and bounded on each side 
by metaxylem and phloem elements. In this process a certain 
amount of variation may obtain : generally the bifurcation of the 
bundle and the rotation of its two halves is well-marked and takes 
place in a short vertical distance; in such cases the protoxylem 
takes up its exarch position, and, as it were, moves outwards very 
quickly. On the other 
hand, it sometimes hap¬ 
pens that the protoxylem 
remains relatively sta¬ 
tionary so that the as¬ 
sumption of the exarch 
position is much slower 
and is arrived at rather 
by the rearrangements of 
the other vascular tissues. 
Finally the corres¬ 
ponding metaxylem and 
phloem groups fuse to¬ 
gether so that the root- 
structure is organized in 
the upper part of the 
hypocotyl. 
In a normal plant the 
number of poles in the 
root will obviously cor- 
FiG. 12. P.lanceolata. Longitudinal section of a cotyledon shewing paren¬ 
chyma (par.), phloem (/>/;.), protoxylem (px.), and transfusion tracheides 
(t.t.). X 280. 
respond to the number of cotyledons, but when polycotyledony 
obtains, this number may be less than that of the seed-leaves. This 
is especially the case in the Coniferte, and the study of the behaviour 
of their cotyledonary traces has led to the classification of their 
seed-leaves into whole-cotyledons, half-cotyledons, and subsidiary 
cotyledons. 1 Also it is very general to find that the number of 
leaves in the first foliage-whorl and the number of cotyledonony 
buds, when present, correspond to the equivalent of the number of 
whole-cotyledons. 
1 Hill and de Fraine : Gymnosperms II, loc. cit. 
