54i 
Origin of Meclullotion in the Ophioglossaceae . 
prepares for the departure of the first leaf-trace (Fig. 7), one margin of 
which is defined by the gap in the xylem-ring; but for some time the 
leaf-trace strand remains connected by the other margin with the xylem 
of the stele, as it is seen to be in Fig. 7. Meanwhile, the endodermis 
remains continuous all round, but it shows some irregularities in position 
of its cells in the neighbourhood of the leaf-trace. These lead to an 
involution, first on the side on which the opening of the ring began (this 
is indicated in Fig. 7), and subsequently on the other side. These involu¬ 
tions of the endodermis approach one another as the foliar strand moves 
outwards, and meet, as seen in Fig. 8. The endodermal limit of the stele 
is thus never interrupted, and there is no direct communication between 
the cortex and the stele. After a time the xylem-ring closes again, com¬ 
pletely surrounding the central pith. But very soon it opens on the 
opposite side, for the issue of the second leaf-trace, which is also effected 
without any interruption of the continuity of the endodermis. The stele 
meanwhile assumes again a solid xylem-core, except for a few parenchyma 
cells irregularly scattered as at first among the tracheides. But soon 
the pith is reconstituted as before, and the process may be repeated at 
the exit of later leaf-traces. One point worthy of special notice is that 
as the leaves of the sporeling are successively larger, and their traces 
stronger, the incurving of the endodermis into the pith (which also becomes 
more massive as the plant strengthens) becomes more marked, indicating 
gradual steps towards that degree of encroachment upon the pith, and 
formation of that internal endodermis, which was first suggested by Van 
Tieghem. This is shown in slight degree in Fig. 9, and higher leaves of 
the same plant showed still deeper intrusion. Since, however, the degree 
of intrusion of the foliar pocket thus formed appears to depend upon the 
size of the leaf, it would not be expected to be large in so weak a plant as 
this sporeling from the Breadalbane Hills. 
Another of the same batch of weak plants from the hills, which was 
cut into sections by Dr. Lang, showed the following interesting features. 
Passing upwards from the base, the vascular structure begins with a solid 
stele, in which, however, a distinct pith appears at the departure of the 
first leaf-trace. But there is no communication with the cortex, the 
endodermis being continuous, as in the previous case. The trace with its 
surrounding parenchyma is nipped off from the stele by involution of the 
endodermis, as in the specimen already described (compare Fig. 8). But the 
next two leaf-traces differ slightly in their behaviour, for in their departure 
they show the formation of a band of endodermis, which passes internally 
to the leaf-trace, before the actual separation takes place. The stelar 
sheath is thus completed before the outer band surrounding the stele and 
the departing trace is interrupted. This band of endodermis appears to 
be a new formation, not directly referable to involution, as in the case of 
