260 
A. G. Tansley. 
the leaf-traces departing from its outer surface without making any 
gap in the cylinder (Fig. 87). Apparently this is sometimes also 
the case in T. superba (Fig. 86). 
In Osmunda cinnamomea, on the other hand, we meet with a 
more complicated type of structure than that characteristic of the 
central forms of the family. In this plant Faull (’01) discovered an 
internal endodermis throughout the stem, separating the pith from 
the xylem, and, in addition to this, locally developed internal phloem, 
consisting of quite normal sieve-tubes betweerf the internal endo¬ 
dermis and the xylem. This internal phloem sometimes forms a 
continuous ring, but often occurs in more or less detached groups. 
It occurs only in the neighbourhood of the branching of the stem, 
both above and below the actual fork (Fig. 88). At the branching 
there is often, though not always, a complete break or gap in the 
A B 
Fig. 88. Osmunda cinnamomea. A.—Transverse section of stele showing 
almost continuous internal phloem and endodermis. B.—Steles of the two 
branches of a stem showing the “ subsidence ” of the external phloems through 
the xylem-branch-gaps, continuous with the internal phloem of the parent 
axis below. After Faull. 
vascular cylinder, placing the pith in continuity with the cortex, the 
internal endodermis in continuity with the external, and the internal 
with the external phloem. In other cases there is a less extensive 
interruption of the tissues, the central portion of the pith below the 
fork being continuous with the cortical ground-tissue above, but the 
inner and outer endodermis not coming into connexion. In rare 
cases there is a union of internal and external phloem also through 
the xylem leaf-gaps. 
In the adult stem of Tudca hynienophylloides, Seward and Ford 
also found an internal endodermis surrounding a sclerenchymatous 
pith. This internal endodermis does not, however, occur in the 
