The Leaf-structure of Cordaites . 
95 
directly on the upper side of the centripetal xylem sometimes have 
quite thin radial walls as seen in transverse section, but as the 
nature of their markings is uncertain the outlines of these cells 
alone are given in the longitudinal section (fig. 6, si). 
In seeking to explain the sheath it is natural to turn to the 
recent Cycads for help, and we find in Worsdell’s 1 paper on trans¬ 
fusion-tissue that in support of his view of its origin from centri¬ 
petal xylem he mentions the cotyledons of Cycas as having more 
or less scattered transfusion-tracheides with bordered pits, which 
he considers are derived from the centripetal xylem. He follows 
Lignier 2 in tracing transitions between the normal transfusion-cells 
near the bundle and the xylem, and in considering the first to be 
derived from the xylem, but differs from him in considering the 
“ accessory transfusion-tissue ” of the leaf blade as a distinct and 
separate development on the part of the mesophyll of the leaf. 
It is possible that in Cycas (and very likely the same may hold 
good of other Gymnosperms), in addition to the transfusion-tissue 
which Worsdell has derived from centripetal xylem, and which, 
since it was presumably the first to appear in evolution, may be 
called “ primitive transfusion-tissue,’’ on the one hand, and to the 
accessory transfusion-tissue of the lamina, which may be called 
“ mesophyll-transfusion-tissue,” on the other, there is a third 
category of transfusion elements, developed from the parenchyma 
of the mid-rib in connexion with the primitive, which may be 
called “peridesmic transfusion-tissue,” on account of its origin from 
the parenchyma surrounding the bundle. For instance, in C. 
Seemanni transverse sections of the mid-rib shew a large develop¬ 
ment of transfusion-elements scattered widely in the ground-tissue 
and passing almost insensibly into it; first there are the true 
transfusion-tracheids with bordered pits flanking the centripetal 
xylem and even circling round the bundle and appearing outside the 
phloem ; secondly, mixed with these are other elements which 
resemble the first, except that their pits are faintly bordered ; and 
thirdly others again with simple pits shewing transitions into those 
with loose reticulations far from the bundle. The third of these 
categories distinctly suggests a phylogenetic origin from the paren¬ 
chyma of the mid-rib, but the presence of the transitional elements 
1 W. C. Worsdell, on “Transfusion-tissue.” Trans. Linn. Soc. 
1897, p. 307. 
2 Lignier, “ La Nervation teniopteridee des folioles de Cycas 
et la Tissu de transfusion.” Bulletin de la Soc. Linn, de 
Normandie, ser. IV., tome VI., fasc. 1, 1892. 
