64 MR. R. H. COMPTON: AN INVESTIGATION OF THE 
contiguous ones at each end of the major axis and one separated group at 
each end of the minor axis. 
At the base of the hypocotyl a ninth xylem appears, being budded off 
from one of the polar groups; the other polar group appears to be about 
to bud off a tenth xylem, but this does not actually happen. Each of the 
nine groups now broadens inwardly, becoming somewhat V-shaped. As we 
pass up the hypocotyl the xylem groups become more and more spread out 
tangentially ; so that at the top of the hypocotyl, just beneath the cotyledons, 
the xylem tissue consists of two long bands arching round the ends of the 
ellipse, and mesarch to endarch in development, and a lateral triad of xylem 
groups on each side of the ellipse. The polar groups become almost com- 
pletely endarch and pass out bodily into the cotyledons. The lateral groups 
go straight up into the epicotyl and become the median strands of the 
primordial leaves. Other small intercalary bundles from the epicotyl also 
appear at the cotyledonary node, being inserted on the cotyledon traces. 
The cotyledons remain lateral to the main axis of the seedling, being 
twisted to one side at their short stalks, much as in the Viciew. The 
vascular tissue on entering the cotyledon immediately gives off a downward 
branch, the main mass of the trace forming the midrib. 
(ii) In the second seedling a quite different condition of the vaseular 
strands was found. The main root was found to be pentarch throughout, 
there being five long radial 2-3-seriate plates of primary xylem surrounding 
a fair-sized pith. As we pass up the main root into the hypocotyl each of 
the five xylem groups spreads out tangentially as in seedling No. 1. ; the 
size of the individual groups varies much, however, and there is no sign 
of the grouping into polar cotyledonary traces of the kind in the previous 
seedling. Each xylem divides into two halves, and rearrangement takes 
place, the result being that at the node each cotyledon takes one whole root- 
pole and half of each of the adjoining ones ; two half-bundles are left over 
in this process and pass up into the epicotyl. The diagrams will help to. 
explain this (P1. 8. figs. 110, 111). 1-5 are the five root-poles, each of which 
dividesinto two. Cotyledon i. takes both halves of xylem 1, half of 2, and half 
of 5; the half of 5 becomes the downward branch into the cotyledon after 
a brief fusion with the midrib, which is composed of the rest of the vascular 
tissue. Cotyledon ii. similarly takes the whole of 4 and half of 3 as midrib, 
and half of 5 as downward branch. The remaining portions of 2 pass up 
into the plumule together with other new plumular bundles which appear 
at the cotyledonary node (not shown). 
Van Tieghem (1870, p. 219) mentions that in the hypogeal Dolichos 
lignosus the root has a solid tetrarch xylem star, and that the cotyledons take 
two of the root protoxylems, the primordial leaves the other two. 
Dodel (1872, p. 149) has publisheda very fullaccount of theseedling structure 
