192 MISS WINIFRED SMITH ON THE ANATOMY 
groups separated by parenchyma (Pl. 26. fig. 9). Two of these primary groups or 
* double bundles” pass into each cotyledon, and from the outer extremity of each group 
a lateral is given off (a’) (fig. 10). From this branches another lateral (@"), and from a" 
another (a). Gradually the remaining inner parts of the primary groups approach one 
another and fuse to form the midrib at 2:7 mm. above the node. 
In Bumelia tenax we have a type outwardly very unlike that of Payena Leerii. 
The seed is much smaller than that of any of the other species examined. It is egg. 
shaped; the long axis measures 8 mm., the short axis 6 mm. Inside the hard testa the 
cotyledons are enclosed in a thin covering layer containing little or no nutritive material ; 
its origin has yet to be investigated. "The cotyledons are very fleshy and packed with 
food-material. The radicle of the embryo shows no differentiation of xylem, but in the 
procambial strands of the cotyledons there are strands of xylem one or two cells thick. 
The seedlings are much more slender than those of any other species examined (fig. 11). 
They grew very quickly and, a month after planting the seed, most of them measured 
10cm. The cotyledons are hypogeal and petiolate. The petioles are fused at their sides, 
forming a sheath to the base of the first internode. The shoot ruptures this sheath 
laterally, the cotyledons are pushed to one side, and at the stage examined the shoot 
often measured 4 or 5 em. Above the rupture the petioles are seen again fused and the 
base of the cotyledons also forms a tube. "They are only separated at a third of their 
length. 
The anatomy of Bumelia tenax seedlings strikes one, at first sight, as so different from 
that of other types that it is difficult to believe them all to belong to the same order. 
On further investigation, however, it becomes evident that the essentials are similar, but 
that the anatomy is profoundly modified by the habit of the plant. 
The root has, as a rule, a hexarch stele with a small pith (fig. 12). All the six proto- 
xylems behave similarly at the root-apex, each arising as a single cell, which is gradually 
added to centripetally as higher levels are reached. In the upper part of the root and 
in the hypocotyl four of the vascular bundles are distinguished from the other two 
in the following ways :— 
(a) It is from these four bundles alone that rootlets arise. 
(5) They alone always retain their protoxylems throughout their course. 
(c) They alone are continuous with the bundles going to the cotyledons. 
These four I regard as primary bundles and equivalent to the four vascular groups of 
the tetrareh Payena type. The two other groups lose their protoxylems while the 
primary bundles are still exarch and they may fork into two strands. For convenience 
I have called them “ accessory bundles." Either they may both lie in the plane perpen- 
dicular to that of the cotyledons (Pl. 26. fig. 13), or one may be so situated and the other 
may be in the cotyledonary plane (text-fig. 3). 
A little higher the accessory bundles are barely recognizable; but at the node 
several xylem groups appear on the long sides of the rectangular stele, with or without 
spiral elements (fig. 14). They pass into the first internode. The stem in the case in 
question (fig. 11 a) is not a vertical continuation of the hypocotyl, but is bent at an angle 
to it, and this fact allows one to observe the gradual break up of the hypocotyledonary 
