DR. D. H. SCOTT ON BOTRYCHIOXYLON PARADOXUM, 971 
from the internal xylem, spreading out laterally in a somewhat Y-shaped form. All 
the outer part of the protrusion is composed of the secondary wood, here cut rather 
obliquely. The whole bears a striking resemblance to an early stage in the emission of 
a leaf-trace in Ankyropteris corrugata, as shown, for example, by Dr. P. Bertrand 
(09, pl. xii. fig. 87). The difference is simply that which distinguishes the steles as a 
whole—the outer wood of the appendage is primary in Ankyropteris corrugata, secondary 
in Botrychioxylon *. Internal xylem is, of course, present in the Ankyropteris just as 
in our plant, though unfortunately not mentioned in Dr. Bertrand's description ('09, 
p. 110). 
We can scarcely doubt that the protrusion of the xylem shown in PI. 87, fig, 2 is of 
the same nature as the similar structure in A. corrugata, namely, a leaf-trace preparing 
to detach itself from the stele. From the analogy of the plant referred to we should 
expect the leaf-trace, when free, to have two immersed groups of protoxylem near its 
two extremities, possibly with a third transitory group in the median position 
(Bertrand, ’09, pl. xii. fig. 88). In 4. corrugata all the three groups are connected with 
the internal xylem of the stele, just as is the case in A. Gray? (Scott, 12, p. 48). 
The section next above the one just mentioned shows a large strand in the cortex 
precisely corresponding in position to the stelar protrusion in the lower section (see 
Pl. 38, fig. 5). It is here separated from the stele by a tract of cortical tissue, and happens 
to be further pushed out by the intrusion of a Stigmarian rootlet; the true relations of 
position are, however, clear. There can be no doubt that the strand seen in this section 
is the same which was preparing to leave the stele in the section next below (compare 
Pl. 37. fig. 2 and Pl. 38. fig. 5). The structure of the strand is, however, in a high 
degree peculiar and different from what we should have expected (Pl. 39. fig. 9) The 
wood of the strand measures about 860 x 620 p, the longer axis being tangential; beyond 
the wood is a zone of phloem, then a layer or two of parenchyma, and a somewhat 
interrupted dark-coloured band of cells, which may possibly represent an endodermis. 
Nearly the whole of the wood is radially seriated and of a secondary character ; the 
preservation is good, and at some points the tabular cambium-cells are quite recognizable 
The secondary wood reaches a thickness of from 5 to 7 elements, the greater thickness 
being on the abaxial side. Between the two ares of secondary ood is à narrow band e 
somewhat compressed elements, no doubt representing the primary xy lem. This ban 
is concave towards the exterior of the stem. At each end of the primary band is a group 
of small elements (best shown on the right) which may probably be the distare : 
What view are we to take as to the nature of this strand i yi " UR possibilities . 
it might b > f a root, the stele of a branch, or a lea ciaoe unde. 
y ee is favourable to the first interpretation, for we know that the 
dary tissues are also consistent 
plant had diarch roots (Pl. 40. figs. 14 & 16). The secontaly” o g od (PL 40 
With this view, for some of the undoubted roots possess aes ay E : if a 
fig. 16). On the other hand, the dimensions are * good xam u ths lr st otai 
may judge from the other examples present in me spomi 
: , but this he 
* As a matter of fact, some radially seriated wood is shown (at rad.) in Dr. Bertrand's photograph, but this 
Connects with the departure of a root-strand, 
