248 MR A. C. SEWARD AND MISS S. O. FORD ON 
parenchymatous tissue traversed by obliquely running sclerenchymatous strands. The 
prominent wings of young petioles have been referred to in the account of the apex of 
the stem. The parenchymatous tissue of the stipules is prolonged upwards as three or 
more layers of green tissue covering the lower part of the leaf-stalk. We haye not 
noticed any of the slits between the epidermal cells of the petioles mentioned by 
Thomae *. ` 
A characteristic feature of the young fronds of the Osmundace:e is the presence of 
numerous filamentous hairs in place of the scaly ramenta of most Ferns. In acircinately 
coiled frond of Todea barbara, shown in Pl. 30, fig 51, a mass of these hairs is seenath; 
they consist of long and frequently branched filaments which arise from epidermal cells 
on the inner face of the young leaf. 
Tn a transyerse section of the petiole, shown in fig. 51, taken at the level a, the stele 
was found to contain 26 protoxylem groups alternating with as many islands of large 
mucilage-saes ; at this level lignification was confined to the small protoxylem tracheids, 
and, as shown in fig. 36, Pl. 29, a large space containing remnants of spiral bands (sp) 
had already been formed and appeared as a more prominent cavity than in the older 
part of the petiole (cf. fig. 45, pz). 
The most external elements of the phloem on the convex side of the leaf-stele exhibit 
the usual characteristics of protophloem, and in sections cut from the level « of the 
young frond (Pl. 30. fig. 51) they are already differentiated as small clear sieve-tubes. 
These protophloem elements ( pp, fig. 50) may be traced downwards into continuity with 
the tangentially-placed sieve-tubes in the phloem of the stem. We have already. 
expressed our belief that the thicker-walled and horizontal sieve-tubes in the stem 
(fig. 42, ¢, &e.) are not true protophloem. In other words, the first-formed phloem 
elements in the apical region of the stem are not continuous with the protophloem of 
the leaf. 
It is unnecessary to deseribe the structure of the roots; they are usually diarch or 
triarch, and in a few instances a tetrarch stele was met with. 
TODEA SUPERBA.—The stem of this species agrees in the general plan of the stele with 
T. barbara, but differs in detail. In the plant examined the stele had a diameter 
of 2:2 mm., being therefore much smaller than that of 7. barbara. The xylem is 
characterized by its greater continuity; it sometimes forms a completed sinuous ring 
(Pl. 30. fig. 41), or it may be interrupted by a single medullary ray occasionally one 
cell in breadth. In the section represented in Pl. 29. fig. 29 there are seven xylem 
strands. 
A few tracheids of exceptional breadth were found here and there in the xylem 
strands ; some of these, four or five times broader than the usual type, are represented at 
v, PL. 29. figs. 30 & 31; the large diameter and short length suggest a storage rather 
than a conducting function. The position of these large tracheids is shown at a in 
fig. 29; they occur sporadically in the stem and are not confined to any particular 
region. These large tracheal elements may be compared with the short and wide 
* Thomae (1386). 
