some Species of Gleichenia. 427 
stele, and these by division give rise to the full normal number. The two 
steles separate with open xylems, and in each of them the inner endodermis 
and phloem become continuous with the corresponding outer tissues. Thus 
a ramular gap is present (Plate XXIX, Fig. 9) but soon after their separa- 
tion the two steles becomes closed again. The stele of the branch is smaller 
than that of the main rhizome. Plate XXIX, Fig. 9, shows that the separa- 
tion of the two steles in monopodial branching bears a slight resemblance to 
the departure of a leaf-trace from the stele of the rhizome, but differs in the 
behaviour of the protoxylems and phloem ; this may be seen by comparing 
Plate XXIX, Fig. 9 with Plate XXIX, Fig. 6 . There is also the difference 
that in monopodial branching no nodal island is formed ; the central mass 
of sclerenchyma simply enlarges so as to follow the changes in size and 
shape undergone by the stele before the separation of the branch. 
Dichotomous branching is always in the horizontal plane, and the 
upper and lower protoxylems are equally concerned in supplying the two 
exactly similar branches. This necessitates an increase in number in both 
kinds of protoxylem. The stele becomes horizontally elongated, and 
divides first on the upper side, as shown in Text-fig. 3 1. In Text-fig. 3 2 
the separation is complete, and the two steles are still open. Within a short 
distance they close again and resume the normal form. This is the usual 
mode of dichotomy, but in one case the two steles were closed when 
they separated, owing to the elongated stele having become constricted, 
and then nipped off in the vertical plane. Thus in the case of dichotomy 
a ramular gap may or may not be present. 
One case was observed in which a dichotomy was closely associated 
with a node. A nodal island was formed in the xylem of a stele, which had 
become slightly elongated in the horizontal plane as a preliminary to 
dichotomy. The two processes of preparation for dichotomy and for 
separation of the leaf-trace then went on concurrently. The stele split on 
the upper side first, freeing one end of the leaf-trace ; then the dichotomy of 
the stele was completed, the two steles being open, and the leaf-trace hang- 
ing on to the end of one of them, but very soon afterwards becoming free. 
Theoretical Considerations. 
Having described the nodal structure and the branching of the rhizome 
in G . pectinata, we may now consider the origin of the solenostelic structure 
of this species. It is to be expected that important data tending to the 
solution of this problem may be obtained from the anatomical study of 
some fossil Gleicheniaceae in the possession of Prof. C. Bommer, of Brussels, 
and possibly also from an examination of ‘seedling’ plants of G. pectinata, 
when these can be procured. 
In the absence of evidence of this kind, an attempt may be made to 
