96 
M. C. Stopes on 
makes it clear that no sharp histological line can at present be 
drawn between it and the first or “ primitive ” type. 
In the fossil Cordaites the transfusion tissue is more compact 
and forms two definitely organised sheaths. It is possible that 
the inner sheath abutting so intimately upon the centripetal xylem, 
may represent but little modified elements of it, and thus be 
primitive transfusion-tissue, while the outer cells may be really 
phylogenetically a parenchyma sheath which has acquired bordered 
pits and is thus peridesmic in origin. 
The completeness of the sheath, its direct attachment to the 
centripetal xylem, and the tracheal nature of its walls and pittings, 
it may be conjectured, present a primitive stage in the development 
of transfusion-tissue from centripetal xylem, and is an interesting 
example of the way in which this wood seems to have more than 
“ held its own,” 1 in the leaf when ousted from the stem by the 
centrifugal development. 
A careful comparison of our preparations with Renault’s 
figures 2 of C. principalis leads to the conclusion that the “ partie 
centrifuge du faisceau ” there figured may perhaps be identical 
with the sheath here described. Renault’s sections correspond 
perhaps to a somewhat different type of preservation in which the 
softer elements of the bundle were lost; he does not shew any 
phloem or xylem-parenchyma, and it is more than likely that in his 
slides the inner sheath was separated from the outer, and, the 
phloem being absent, was interpreted as centrifugal xylem such as 
occurs in other species of Cordaites, e.g. C. lingulatus (phot. 1). 
It appears then, as M. Renault has stated, that these sheath-cells 
are elements of the wood, though it seems probable that they are 
to be considered as derived from the flanks of the centripetal xylem 
rather than constituting the centrifugal wood in this particular 
case. 
That the two leaves are identical and ours therefore C. princi¬ 
palis is, as we have said, strongly suggested by the comparison of 
the figures (phot. 2, with Renault’s fig. 6, pi. xvi.), when it 
will be seen that our leaf resembles C. principalis entirely in its 
arrangement of vascular and sclerotic tissues, upon which Renault 3 
1 D. H. Scott, “The Old Wood and the New.” Nitw Phyto- 
i.ogist, Februar}', 1902, p. 29. 
2 Renault, Sur quelques tiges, pi. xvi., fig.. 6 
3 “ Sur quelques tiges.” p. 301. “ J,a feuille se distingue par trois 
on quatre petites baudes de tissu liypodermique placees entre les 
bandes priucipales qui accompaguent les faisceaux vasculaires la 
face inferieure il 11’y a qu’uue seule bande d’hj'poderme entre deux 
nervures, mais elle est plus considerable.” 
