1908-9. | THE STELE OF OSMUNDA CINNAMOMEA. 521 
foretells the twenty-first trace, but it does not connect with the pith, and 
above the node is a very short-lived gap. Once more the pith dies out. 
But a pocket comes into existence at the succeeding node, and by the time 
the twenty-third node is reached there is once more a long narrow pith. 
All this time the xylem has been slowly increasing in amount and has 
become quite bulky. 
G.—The base of this seedling was also lacking, and presumably the 
lowest node remaining was about the twenty-fifth or thirtieth. By this 
time there was a small permanent “‘pith”’ and there were two overlapping 
leaf-gaps. ‘The roots occurred with regularity about 90° around from the 
corresponding leaf-traces, and not near the latter as in the adult stems. 
Ten leaf-traces were passed before the first endodermal cell appeared at the 
inner entrance to the last leaf-gap. The course of the development of the 
endodermis calls for no further description than that already given, other 
than to re-iterate the fact that as it expands it shifts its position some- 
what, and shows a tendency to send diverticula out into the newly formed 
medullary rays. 
Tue ADULT. 
(a) Branching. 
The stem of Osmunda cinnamomea may branch occasionally, some- 
times once, sometimes not at all. There is no reason to suppose that it 
may not branch more than once as do its two native sister species, but it 
has never been my fortune to find such a case. The branching has been 
described as ‘‘dichotomous,”’ but this implies a phenomenon of the exis- 
tence of which in this family we have absolutely no proof. Indeed, 
Mettenius (9) and Sadebeck (11) hold that dichotomy does not occur in 
the Filicinez:, and that here, as in similar cases throughout the group the 
branching is of the lateral type. 
One of the apparently peculiar features in connection with the branch- 
ing of Osmundacinnamomea is the frequent accompaniment of ramular gaps. 
These are of varying degrees of openness; in some instances they gape so 
widely that there is a broad connection between the pith and the cortex, 
in other instances, there is no indication of a gap, even in the xylem, 
and between these extremes there are numberless intermediate grades. 
That these gradations or even the phenomenon of branching itself have 
any phylogenetic meaning as far as the nature of the stele is concerned, 
remains to be proved. 
(b) Leaf-gaps and cladosiphony. 
