Seedling Structure of Gymnosperms. IV. 325 
Below the sucker the transition to root-structure rapidly proceeds; 
the four bundles undergo a centripetal displacement; at the same time, the 
individuals of each pair of strands come more closely together, and, at 
a slightly lower level, the corresponding groups of protoxylem elements 
effect a junction. Thus there are four masses of phloem, four of metaxylem, 
and two collections of protoxylem elements (Diagram 3, Figs. 7 and 8). 
While these changes are being effected two broad bands of thick- 
walled fibrous tissue are differentiated on the outer sides of the phloem 
elements (/. Diagram 3, Fig. 9). 
At first the two isolated groups of protoxylem elements are tangen¬ 
tially elongated, but a rearrangement soon takes place, resulting in the 
formation of two rounded masses (Diagram 3, Fig. 8). Then the corre¬ 
sponding metaxylems fuse, and, at a lower level, the metaxylems and pro- 
toxylems effect a junction (Diagram 3, Figs. 8 and 9). The four groups 
of phloem elements retain their identity for some considerable distance 
downwards, as in the case of Ephedra described above. Fusion does take 
place ultimately, we believe, although it is difficult to say with absolute 
certainty because the soft-bast in the root is very poorly differentiated ; 
indeed, in transverse section it is hardly distinguishable. 
The endodermis is first visible at a level just below the region of the 
foot, appearing at the same time with the fibrous tissue already alluded to. 
At lower levels the endodermis may be two or three cells in thickness. 
The occurrence of a band of a form of transfusion tracheides in the 
hypocotyl of Ephedra has already been described above. Welwitschia also 
shows them, and in this plant they are very prominent and have a disposi¬ 
tion exactly the same as in Ephedra, although in the former plant they also 
show an extension into the central ground-tissue (Diagram 3, Fig. 3). 
Abutting on to the margin of these two bridges are a number of cambiform 
cells, and possibly also some phloem elements, but no sieve-tubes have been 
observed. The presence of these tracheides is not due in the first place to 
the presence of a cambium, but rather to the differentiation of the ground- 
tissue elements. A cambium, however, may soon put in an appearance, 
and thus lead to their increase. The material at our disposal was too 
young to show differentiated epicotyledonary strands, but it is clear from 
Bower’s account and illustrations, taken in conjunction with our own obser¬ 
vations, that there is a wonderfully close agreement between Ephedra 
and Welwitschia in the inter-relationships between the plumular and 
cotyledonary bundles and these transfusion tracheides. 
It may be remarked in this particular connexion that Bower’s account 
and ours appear to differ: this is due to the fact that Bower’s seedlings 
were older, with the plumular traces developed ; and as these mask the 
presence of the transfusion tracheides, it would be impossible to realize 
their nature and significance without younger stages. 
