230 Lang. — Studies in the Morphology and 
An indication of internal endodermis is found before this trace separates, 
and the gap in the external endodermis is repaired almost as soon as 
made. 
Above the origin of these two first leaf-traces the diameter of the stele 
of the branch diminishes until it is about that of the original main axis. 
The complete stele below the origin of the third leaf-trace has no indication 
of internal endodermis (Text-fig. io, k ). As the xylem of the trace separates, 
but while the external endodermis is still continuous, two strips of internal 
endodermis develop (/), and, on the separation of the trace, these join the 
edges of the gap in the external endodermis, and the internal endodermis 
becomes a complete lining to the stele (m } n). The branch thus on the 
whole repeats the development often found in plants developed from an 
embryo, and it is probable that the departure of the third leaf-trace would 
be a gradual one with a long leaf-gap. 
The structure of the other branching specimens may be more briefly 
described, and the vascular relations in each case will be clear from the 
diagrams of transverse sections without the need of reconstructions. 
The main axis, which bore the second branch, had had its apex cut 
off ; the single branch had originated a considerable distance behind the 
wounded end on what must have been a mature region of the rhizome. 
The portion of the rhizome which bore the branch had the general structure 
characteristic of a marked intermediate region. The leaf-trace subtending 
the branch departed very gradually, and the long leaf-gap had an internal 
endodermis. Text-fig. 1 2 , a, shows the previous leaf-gap still open and 
the subtending leaf-trace ready to depart ; the gradual separation of the 
trace can be followed in Text-fig. 12 , a-g. In this plant pericyclic 
xylem was developed in the stele of the main axis (Text-fig. 1 2 ,g; Phots. 
23, 34). At a lower level a band of pericyclic xylem had developed on the 
abaxial side of the departing leaf-trace (Text-fig. 12 , b, c\ Phots. 32 , 24 ). 
This soon dies out and plays no part in the vascular supply to the branch. 
The xylem of the subtending leaf-trace then shows an adaxial extension 
leading to the formation of a complete ring of xylem (Text-fig. 12 , d , e\ 
Phot. 33 ). This is clearly comparable to the adaxial xylem in the subtending 
leaf-trace of the first branching specimen, where there was no development 
of pericyclic xylem. As in that case, the adaxial xylem becomes separated 
from the departing leaf-trace, and ultimately gives rise to the vascular 
supply for the branch. The organization of the stele of the branch as 
followed upwards is, however, more irregular in this specimen. The adaxial 
xylem, after becoming detached from the leaf-trace (Text-fig. 12 ,/), forms 
an irregular group of tracheidal strands instead of becoming at once 
arranged to form the stele of the branch (Text-fig. 1 2 ,g; Phot. 34 ). On 
passing a little higher up these become rearranged to form the small 
medullated stele of the branch, which at once departs from the main 
