374 Boodle . — Anatomy of the Schizaeaceae . 
surrounding a central pith. The ring of xylem is interrupted 
by the departure of the leaf-traces as seen in Fig. n, which 
represents the node of dichotoma , Sw. 
The rhizome of 5. digitata has a cortex of 6-8 layers 
of brown sclerotic cells. The endodermis contains mucilage, 
and the pericycle is usually 2-3 cells thick and its cells are 
rather thick-walled. The outer pericyclic cells generally 
have their radial walls coinciding with those of the endo- 
dermal cells, and the sieve-tubes sometimes correspond with 
the inner pericyclic cells. The phloem consists of sieve- 
tubes and parenchyma, and is 2-4 or 5 cells in thickness. 
The sieve-tubes are sometimes tabular and three or four of 
them may form a neat radial row, at other times they are 
smaller, more rounded and irregularly placed, with parenchy- 
matous cells among them. Some little doubt must be ex- 
pressed as to the nature of the elements here spoken of as 
sieve-tubes. In longitudinal section they are at first sight 
considerably unlike the sieve-tubes of Lygodium . Some 
of them are short elements, of the same length as the peri- 
cyclic cells or slightly longer, with straight or slightly inclined 
end-walls; but others are a good deal longer and may have 
long pointed ends. The points in which they resemble the 
sieve-tubes of Lygodium are best seen in the more elongated 
elements, in longitudinal sections unstained and mounted in 
glycerine-jelly, and are as follows : — the walls are rather thick 
and refractive, and oval or rounded pits are seen on some of 
the longitudinal walls, and bear refractive granules. The 
phloem-parenchyma-cells have rather scanty contents, except 
for the presence of a distinct nucleus, and their walls do not 
stain quite so strongly with haematoxylene as those of the 
sieve-tubes. One meets with doubtful elements, which might 
be either sieve-tubes or phloem-parenchyma, but those which 
most suggest sieve-tubes are usually destitute of a nucleus. 
The c sieve-tubes ’ occupy the usual position for sieve-tubes, 
are differentiated early as in other Ferns, and also when first 
formed are well differentiated by haematoxylene. All the 
characters point to the elements in question being sieve- 
