522 Boodle . — Comparative A natomy of the 
ance of a parenchymatous pith, ( b ) differentiation from part 
of this parenchyma of a central strand of sclerotic tissue 
surrounded by an endodermis and continuous with the cortex 
at the nodes 1 , (c) reduction in size of the central core of 
sclerotic tissue, and severing of its connexion with the cortex, 
leaving endodermal pockets as a vestige of the previous con- 
nexion between inner and outer endodermis, ( d ) disappearance 
of brown sclerotic tissue and of inner endodermis, leaving 
a parenchymatous pith, (e) transformation of the latter into 
sclerotic tissue (as found in most parts of the stem of vS\ dicho - 
toma and always (?) in 5. digit ata). 
If one now turns to a consideration of the internal tracheides, 
it is important to remember their relation to the inner endo- 
dermis. The two best-developed strands of tracheides were 
associated with the two best-developed examples of internal 
endodermis. One tracheide-strand, being quite unconnected 
with the xylem-ring, appears to be a vestigial structure ; the 
association with the inner endodermis also points to the same 
conclusion, for the tracheide-strand is an added structure in a 
part of the stem, which is presumably reverting to earlier 
characters . 
The exact significance of the tracheide-strands must be left 
an open question, but one may say that they probably indi- 
cate that the tissue within the xylem-ring (exclusive of the 
area taken up by the inner endodermis) is potentially xylem- 
tissue. This is suggested also by the fact that, apart from 
these definite strands of tracheides, one or two tracheides 
sometimes branch off from the inner face of the xylem-ring 
and pass a short way into the pith. This and other features 
often cause an indefiniteness of demarcation between the 
xylem-ring and the central parenchyma, such as impressed 
Prantl (’81, p. 24 ) in 5. Pennula , and gave one of his reasons 
for regarding the central parenchyma or sclerenchyma of 
Schizaea as belonging to the ‘ bundle,’ and not representing a 
‘ true pith.’ His other reason was the absence of an endo- 
1 The change described in (3), and possibly in (a) also, might begin at the nodes 
and spread to the intemodes. 
