The Anatomy of Nephrolepis volubilis J.Sin. 265 
In preparation for the origin of a branch, the gap on the side 
on which the branch is to arise is closed by the bundles approaching 
each other and finally fusing ( B) by their margins. The portion of 
the xylem where this fusion occurs, at the same time bulges out 
sideways as an arc (C) seen in section, to the formation of which the 
two strands have contributed equally. This arc is the branch-trace. 
It soon frees itself first from one ( D ) and then from the other 
cauline strand, and at this stage passes out as a C-shaped strand 
into the branch, with its gap facing the main rhizome. Very soon, 
however, its opposite side opens, and the internodal structure of the 
main axis is thus established in the branch (E). 
The mode of origin of the leaf-trace is almost identical with that 
of the branch, and need not be described fully. 1 On entering the 
petiole the C, which is formed exactly like the branch-trace, but is 
smaller, becomes very much attenuated on its abaxialside, and may 
even split into two strands corresponding to the two internodal 
strands. This division, however, is only momentary, and the leaf- 
tracg, very soon assumes the form shown in PI. IV, fig. 8, 2 its 
adaxial margins turning outwards and becoming engaged in 
supplying the pinnae. In addition to peltate scales (5) the sides of 
the rachis bear curious stiff, mostly unicellular hairs ( h ). 
(6) N. altescandens (Colla) Baker. This species from Juan 
Fernandez, is closely related to the above, 3 and the agreement 
extends to anatomical characters. The internode has exactly the 
same structure as in that species. The origin of the leaf-trace is 
different; each internodal strand constricts off a small strand which 
forms half of the leaf-trace (Fig. 7, PI. IV) which thus at its origin 
consists of two separate small strands instead of a single large one. 
The two leaf-strands on entering the petiole join each other and the 
resulting strand assumes the form shown in Fig. 8, PI. IV. 
The origin of the branch is exactly the same as in N. ramosa. 
Fig. 7, PI. IV shows a section which happens to have passed 
through a point where a leaf and a branch were both being given 
off from the rhizome, diametrically opposite to each other. (Such 
a case was met with only once, the usual course being for the 
leaves and the branches to come off from the rhizome at consider- 
1 In one specimen from the Kew Herbarium (PI. IV, fig. 6) the tracheids 
near the middle of one of the internodal strands were either very narrow or 
absent, forming a break in the xylem, though the phloem passed round the 
bundle in the normal way. This interruption was not seen in a specimen from 
the Cambridge Herbarium. 
2 This figure will do equally well for N. ramosa and N. altescandens. 
2 Christ, l.c., p. 288. 
