382 Farmer and Hill.— Arrangement and Since lure 
periphery of the stem, the petioles of the leaf dying back 
to this region. Zeiller not only does not figure any leaf- 
traces as passing out through the root-zone, but in the 
text he expressly emphasizes their non-existence in this 
region. He explains their absence by supposing that, as in 
Caulopteris endorhiza , the original cortex was very thin, but 
that the parenchyma increased in amount behind the apex of 
the stem where the roots pass down through it. The external 
surface of the stem, together with the leaf-scars, is supposed 
to have become detached, leaving the root-traversed cortex 
exposed. But we find it difficult to accept this explanation 
without some reservation. It is not easy to see, even assuming 
the growth in thickness of the cortical parenchyma, how the 
vascular strands which originally passed through and across 
it, can have become so ruptured or have suffered such entire ob- 
literation as to leave no trace of their existence, if the tissue 
in question is thus correctly interpreted. Psaronius , however, 
resembles Angiopteris in possessing, besides those roots related 
to the leaf-traces, others that may arise from within the 
central plexus of vascular strands. We have observed several 
instances in which they could be traced running across the 
cauline sclerenchyma in their outward course to join the 
peripheral zone of sheathing roots. 
Turning to the parenchymatous tissue itself, in its relations 
with the roots, one is at once struck by a point of difference 
between Psaronius and Marattiaceae in an important par- 
ticular. In the latter plants, the peripheral parenchyma of 
the root is sharply defined from the stem cortex, owing to 
the discontinuity of the tissues in development. The root 
is obviously an intrusive organ in the Marattiacean cortex. 
But in Psaronius , in the best preserved sections that we have 
seen, the outer cortex surrounding the zone of small-celled 
sclerenchyma, and forming a peripheral ring in the roots, 
passes almost insensibly into the parenchymatous tissue 
between them. Furthermore the cells of the latter are very 
irregular in character, often producing the effect of elongated 
cells ‘ combed ’ in various directions. 
