376 BOTANICAL GAZETTE [NOVEMBER 
the expanse is increased.”” MATTE (6) assigns almost exactly the 
same cause. 
A definite cause for the girdling cannot be given at present, for it 
seems to be deeper seated than at first suspected. The even distri- 
bution of the four strands of each leaf in the vascular cylinder appears 
to be the dominating factor. When the primordium appears on the 
stem tip, its distance from the potential vascular cylinder is very short, 
and the same conditions that determine cell division in the develop- 
ment of the leaf, cause the diff tiation of cells along certain paths 
that run from definite points in the procambium to definite places into 
the developing leaf. After the traces have been started, they con- 
tinue to develop with the further growth of all the tissues; new organs 
appear and intercalary growth continues; thus the strands are 
lengthened more and more and their curves are widened to keep pace 
with the ever increasing growth of the plant. 
Summary 
1. The vascular cylinder of the embryo is a protostele, but becomes 
a siphonostele in the seedling. It is very short and squarish in cross: 
section, one of the diagonals of the section being at right angles to the 
inner faces of the cotyledons, and the other parallel with them. Near 
each of the four corners is a group of protoxylem cells, the Jon8 
diameter of whose section extends along the diagonal. 
2. The four protoxylem groups extend downward to form the 
protoxylem of the root. 
ds of each coty- 
arallel with 
strands 
One and the other branch (from the same group). runs into 
of the other cotyledon (opposite the first). jindet 
4. For each leaf or scale four strands leave the vascular 
