Sinnott.—Foliar Gaps in the Osmundaceae . 113 
with the question of leaf-gaps, as the plants he examined had not progressed 
beyond the protostelic stage ; and of Faull on O. cinnamomea and O. Clay - 
toniana, which distinctly show that, although the young condition is a tubular 
stele, the continuity of the xylem ring is always broken at the departure of 
a leaf-trace. 
Another region of the plant which is rapidly coming to be recognized 
as very retentive of ancestral characters is the vascular supply of the leaf. 
The occurrence of centripetal wood in the leaves of Cycas and Preplans, and 
in the cotyledon of Ginkgo , when it has almost, if not quite, disappeared 
from the rest of the plant, are good illustrations of this. A still more 
striking case was recently investigated in this laboratory by Mr. A. J. Eames 
(11), who has found centripetal xylem in the vegetative and reproductive 
leaves of species of Equisetnm , though it disappeared ages ago from the 
stem of the ancestors of living Equiseta. 
In the Osmundaceae the leaf-bundle is somewhat like a flattened arch 
in shape, with incurved ends, and consists of a band of xylem elements more 
or less completely surrounded by phloem. It thus presents a rough resem¬ 
blance to a portion of a siphonostelic central cylinder. It was thought 
that the relations of the vascular supply of the pinnae to this leaf- 
bundle would be of interest as showing the probable primitive method of 
departure from the stele of a foliar trace. The structure of the bundle in 
the rachis was consequently investigated in the three species of Osmunda , 
and in the four species of Todea already mentioned in this paper. 
In all the species, traces to the pinnae go off from the leaf-bundle near 
the incurved ends of the arch. In O. regalis the first indication of this 
separation is the bending out of the arch at either end. This progresses so 
far that the xylem is broken at the adaxial end of the young trace. Such a 
condition is shown in Fig. 11. The trace to the pinna remains connected 
at its other end with the foliar strand for a long time, however, even until 
the gap in the leaf-bundle becomes closed up again behind it. Fig. 12 
shows this state of affairs, which would seem to indicate that the trace 
has gone off without leaving a break in the xylem. The gap is thus a 
very short one. Through it, however, the fundamental tissue inside the 
bundle, or ‘ pith becomes continuous with that outside the bundle, or 
‘ cortex \ This cannot be seen at any one height, however, for the ground- 
tissue between the arms of the pinna-trace is cut off by the closing stele 
from its connexion with the ‘ pith ’ before it becomes continuous with the 
‘ cortex ’ by the complete separation from the leaf-bundle of the vascular 
tissue of the free end of the trace. At a certain level, therefore, there is 
apparently an island of fundamental tissue completely surrounded by 
vascular elements. 
The rachis of a leaf of a very young plant of O. regalis was also 
examined. Though as yet very small, and composed of but few cells, the 
I 
