428 Boodle and Hiley. On the Vascular Structure of 
determine which is the most primitive type of stele among living species 
of Gleichenia , and provisionally to derive the other types from it. 1 
The genus Gleichenia , if Platyzoma be excluded from it, falls into two 
subgenera or sections, 2 Eugleichenia and Mertensia. Of these, Eugleichenia , 
which includes G. circinnata , G. dicarpa , &c., presents a series of apparently 
reduced forms, while most of the species of Mertensia , to which G.flabellata 
and G. pectinata belong, are not manifestly reduced. 
The conclusion that reduction has taken placepn ^Eugleichenia is based 
primarily on the comparative anatomy. In this subgenus the group of 
sclerenchymatous elements, which, as in most species of Gleichenia , is present 
in the nodal island, and is continued into the petiole, 3 has but slight con- 
nexion with the similar elements of the cortex, and in G. circinnata (Boodle, 
’01, p. 72 6, Tansley, ’ 07 , p. 138) it has no connexion whatever with the 
cortex. That this group of fibres was originally connected^with the cortex 
of the petiole may be deduced from its resemblance to the cortical scleren- 
chyma, and from comparison with Mertensia , in which the sclerenchyma of 
the nodal island is continuous with the sclerenchymatous cortex filling the 
concavity of the arched petiolar bundle. The isolation of the group of 
fibres in G. circinnata is strong evidence that reduction has taken place, and 
the nearly circular or subcordiform petiolar bundle of Eugleichenia may be 
held to have been derived from the horseshoe-shaped bundle of the Mer- 
tensia-ty^e by contraction, and by fusion of the free ends of the horseshoe, 
the sclerenchymatous cortex in the concavity of the latter thus becoming 
enclosed (see Boodle, ’01, Fig. 23), or at a higher level suppressed. One 
species of Mertensia , viz. G. linearis , has undergone a similar reduction as 
regards its petiolar bundle, and is thus exceptional. 
In two species of Mertensia , viz. G. pectinata and (one rhizome of) 
G . flabellata, a small amount of phloem was found on the abaxial side of the 
nodal island (Plate XXIX, Fig. 4, ab. phi). As described above, this was 
a short strip, which had no connexion with any other phloem. Now from 
Plate XXIX, Figs. 2-7, it will be seen that, if this group of phloem-elements 
were slightly extended and continued up into the petiole, it would fill the 
gap in the phloem at present existing on the adaxial side of the petiolar 
bundle. This suggests, firstly, that in the ancestral type of Gleichenia the 
phloem probably formed a continuous band round the xylem-arch in the 
1 Cf. Boodle (’ 01 ), pp. 73 7, et seq., Tansley (’ 07 ), p. 142, &c., Bower (’ 08 ), p. 562, where 
this subject is dealt with. A point of view suggested by the branching of the rhizome makes 
a re-statement of the theoretical position advisable. 
2 ‘Subgenera’ according to Diels (Engler and Prantl, Natiirl. Pflanzenfam.), who includes 
Platyzo?na as a third subgenus ; ‘ sections ’ according to Hooker and Baker (Synopsis Filicum), who 
exclude Platyzoma . 
3 In some species it is only continued for a short distance, i. e. into the basal part of the petiole, 
and in G. Boryi this group is entirely absent (Poirault, ’ 93 , p. 171). This would appear to be a case 
of extreme reduction. 
