246 MR. A. C. SEWARD AND MISS S. O. FORD ON 
D. Leaf-traces and the Course of the Protorylem Strands. 
The phyllotaxis of Todea barbara appears to be represented by the fraction j3. 
In any transverse section of the stem, one or more xylem groups situated at the 
rounded angles of the polygonal stele are seen to be bending outwards (Pl. 29. 
fig. 26, a; text-figs. 1 & 2, æ) preparatory to cutting off their outer portions as 
the single curved band of vascular tissue which passes into each leaf. A single 
zroup of protoxylem occurs on the concave side of these tracheal strands, and in 
one or two eases two groups of spiral elements were noticed. After the arch of 
a horseshoe-shaped band has been detached, the two ends of the arms, or compensating 
strands (Pl. 29. fig. 38, c), remain in the stele of the stem. It is difficult to follow 
the course of the compensating strands in Todea owing to the small degree of 
separation between the xylem groups of the stem. The course of the xylem strands is 
illustrated by the diagram shown in fg. 27, Pl. 29. Beginning with the lowest leaf- 
trace n, it is seen that the left-hand compensating strand soon unites with its left-hand 
neighbour and the joint strand gives off a leaf-trace, +3, at a higher level. Similarly 
the left-hand strand of 2+1 unites with the adjacent xylem group—the right-hand 
strand of leaf-trace »—and the common strand gives rise to leaf-trace 44-4. The right- 
hand compensating strand left in the stem after the exit of a leaf-trace continues as a 
single strand for a short distance, and then unites with its neighbour on the right. The 
joint bundle gives off a leaf-trace 4-5 internodes higher up the stem. 
Zenetti has given an interesting account of the origin of the leaf-traces in Osmunda 
regalis, which in essentials might be extended to Todea. As a xylem strand arches 
outwards and becomes elongated in a radial direction (text-figs. 1 & 2, 4), a few tracheids 
begin to bend out from the shoulders of the depressed arch to form the steles of the two 
roots that are in-close connection with each leaf-trace. The roots do not usually arise at 
precisely the same level; they may appear as at r, r in fig. 38, Pl. 29, or, as more 
frequently happens, one may appear before the other (fig. 39, 7). In some instances the 
steles of the rcots are not given off until the leaf-trace has passed into the extra-stelar 
tissue, but as a rule one at least may be detected before the foliar bundle has left the 
stele. The tangentially elongated sieve-tubes are replaced by narrower and more 
vertically placed sieve-tubes in the region immediately external to an outgoing leaf- 
trace. The crest of the arched xylem band is gradually detached from the lower ends of 
the two limbs, which remain behind as the compensating strands (text-fig. 2,¢, c; also 
fig. 38, Pl. 29): these may be small oval strands, one of which only contains a mesarch 
protoxylem group (fig. 38, c, c), or one or both of the compensating strands may be in 
direct continuity with a sinuous xylem band (fig. 39, Pl. 29; text-fig. 2). At about the 
stage when separation of the trace occurs, a conspicuous group of sieve-tubes appears on 
the concave side of the foliar strand (text-fig. 2, s; figs. 38, 39, s ) Zenetti* has 
deseribed how this apparent island of phloem is in reality the lower end of a strand of 
the stelar phloem from a higher level, which bends upwards in the form of a U-shaped 
loop to constitute the inner phloem of the concentric leaf-trace. This author gives 
* Zenetti (1895), pl. 2. fig 11. - 
