116 A. G. Tanstey. 
concavity of the leaf-trace into the stele of the stem is a phe¬ 
nomenon found in several relatively primitive ferns, and is probably 
a mechanism to strengthen the mechanical connexion of pctiolar 
strand and stem-stele. 
T. spicatum (also an upright species) differs from the forms 
hitherto described in having protoxylem elements ( which are 
scalariform, not spiral) scattered through the mass of the metaxylem 
of the rhizome. The protoxylem of the leaf-trace is again central. 
In T. scandens, a creeping form, there is a large stele (Fig. 29) with a 
Fig. 29. Tricliomanes scandens. Stele of rhizome; xylem of same type 
as in T. radicans, but exarch, x about 100. From Boodle. 
mass of large tracheids separated by a network of parenchyma, and a 
distinctly peripheral spiral protoxylem, as in the Schizaeaceous 
genus Lygodium. 
Boodle considers that these types may have been derived from 
a form like T. reniforme 1 by increase of xylem, due to the crowding 
of the leaf-traces consequent on the erect habit, and by loss of 
spiral protoxylem in T. spicatum through diminished rapidity of 
growth in length. T. scandens, he thinks, may have arisen from 
the erect forms and re-developed spiral protoxylem in a new 
position. For reasons given in the second lecture (p. 57) and in 
view of its apparently general occurrence in Botryopterideae, 
together with its occurrence, more or less modified by dorsiven- 
trality, in both rhizome and petiole in the central types of 
Hymenophyllaceae, it seems to me very likely that a radially 
symmetrical endarch protostele is really primitive in the Ferns. 
It is likely, I think, that the primitive type had more xylem and 
was less markedly dorsiventral than T. reniforme. 
i 
