Bower.—On Medullation in the Pteridophyta . 561 
fossil there are air-spaces comparable with the endodermal space in Selagi- 
nella , though not in exactly the same place in the tissues. There is also 
a tendency to breaking up the xylem of the stele into separate strands, as 
in N. spinnlosci , and, lastly, in both of them there is a mass of softer tissue 
occupying the centre of the stele. This I regard in both cases as an intra- 
stelar pith, resulting from degradation of the central region of the xylem- 
core, a conclusion which is supported in both cases by the existence of 
intermediate conditions between the indurated tracheide and the thin- 
walled cell. Facts such as these seem to me to be incompatible with the 
statement of Professor Jeffrey that ‘the pith must in all cases be regarded 
as a derivative of the cortex’. But, further, there is already at hand 
evidence that the formation of an intrastelar pith can be artificially deter¬ 
mined by experiment. This has been shown by Flaskaemper in the roots 
of certain Leguminosae. 
The medullation of roots has not hitherto figured in these discussions. 
But as a central medulla is frequently present in them, it would be interesting 
to see how it originates, and then to draw the analogy with what is seen in 
the shoot. No doubt the pith of the root may frequently be in continuity 
with the pith of the axis. It certainly is so in many seedlings, and in that 
case it would be open to those who hold that its origin is always cortical to 
regard it as a further involution from the pith of the axis, and thus be 
cortical in origin, though, in a sense, at second hand. But recently experi¬ 
ments have been described by Flaskaemper, 1 which demonstrate in the root 
that pith may be formed de novo , within the stele, where no possibility of 
involution exists. He found 2 that, if after a seedling of Vicia or Phaseolns 
had germinated and the radicle had begun to elongate the cotyledons were 
removed, the plant was checked, and the root, previously pithed, lost its 
pith in the parts subsequently formed. But when the plant had recovered 
strength by continued exposure to favourable conditions, the pith was 
reconstituted in the later parts of the same root. In a typical case, where 
the seedling had formed its root 6 cm. long, the cotyledons were removed : 
the seedling was then kept under careful cultivation for six weeks, during 
which time the root grew on to 2 6 cm. An examination of its stele from 
the base onwards showed in the first 5 cm. a normal tetrarch structure 
with many pith-cells. At 6 cm. the pith appeared reduced, and at 7 cm. 
pith was absent. The pithless state continued till about 20 cm. from the 
base, but in the distal part of the root, formed after the plant had recovered 
the loss of its cotyledons, pith appeared again in the transverse section. 
Here, then, is a clear case of the origination of an intrastelar pith in a root, 
without the possibility of any involution of the cortex from without. This 
analogy, which has a very close parallel with what has been described by 
Faull in his seedlings of Osmunda , and with my own observations on 
1 Flora, 1910. 2 1 . c., p. 205. 
