Sinnott.—Foliar Gaps in the Osmundaceae. 115 
the internal and external endoderms almost meet (Fig. 17). The state of 
affairs here is somewhat intermediate between that found in O . Claytoniana 
and that in T. hymenophylloides. 
In the last species, T. superb a, the trace to the pinna is very small. 
Opposite to its point of departure, a gap is formed in the leaf-stele around 
the ends of which the internal and external endoderms become continuous. 
This gap is not wide enough, however, to permit the connexion through it 
of the e pith ’ with the ‘ cortex ’, the two rows of endodermal cells lying 
side by side, as shown in Fig. 18. 
In all the species examined, therefore, gaps in the leaf-stele at the point 
of departure of the vascular supply of a pinna were observed, though in 
O. regalis and O. cinnamomea they are very short, and at certain heights 
apparently absent. 
There seems to be a general relation between the size of the trace to 
the pinna and the character of the gap which it subtends in the leaf-bundle. 
The pinnae of O. regalis are fewer, and consequently larger than those of 
the other two species in the genus, and the gaps subtended by their 
traces are consequently much wider. T. barbara and T. hymenophyl- 
loides , also, whose pinnae are rather large in proportion to the leaf, possess 
more prominent c pinna-gaps ’ than does T. superba , where the primary 
divisions of the frond are much smaller. A leaf of Dicksonia antarctica was 
looked at in this connexion. Near the middle of the leaf the pinnae are 
large, but diminish in size towards the base till they become very small. It 
was found on examination that the traces to the large pinnae left wide gaps, 
through which the inner and the outer fundamental tissue became freely 
continuous, while in the case of the small pinnae the gap was minute and 
affected the xylem only, not breaking through the whole stele. 
The evidence from the leaf-trace, therefore, distinctly confirms that 
derived from fossil forms and from the structure of the young plant, in 
pointing towards the primitive existence of the foliar gap in the Osmundaceae. 
Conclusions. 
The ancestry of the Osmundaceae, as we have above remarked, is 
much a matter of doubt. From such fossil evidence as we possess, however, 
that bears on the matter, and from what we have observed in the structure 
of the young plant and of the leaf, both very tenacious of primitive 
characters, it seems reasonably certain that in all those ancient members of 
the family which possessed a true parenchymatous pith, the leaf-trace, as it 
departed from the stele, always formed a gap in the xylem, and very 
probably, as indicated by the structure of Osmundites Skidegatensis , in the 
whole vascular ring. The evidence also seems to point towards the 
correctness of Jeffrey’s and Faulks view of the Osmundaceae as a reduction 
series, for the occurrence, in so many cases where it can hardly be explained 
I 3 
