528 TRANSACTIONS OF THE CANADIAN INSTITUTE [VoL VIII. 
to a similar conclusion with regard to the bulky piths of other species. 
Seward and Ford’s interesting discovery of a sporadically marked internal 
endodermis in Todea hymenophylloides provides the necessary bridge if 
one be required. 
(b) Internal phloem. 
The presence of internal phloem in Osmunda cinnamomea and Os- 
mundites Skidegatensis has been difficult of explanation on the upgrade 
theory. Tansley does not attempt it. Kidston and Gwynne-Vaughan 
suggest that there has been a “‘subsidence”’ of phloem through branch- 
gaps as in the Lycopsida. Support for this suggestion has been drawn 
from the fact of localization of internal phloem in the neighbourhood of 
forking in Osmunda cinnamomea. Investigations prove in opposition to 
this, that branching is not a seedling character. Further, the external 
phloem tends to dip into the bays and the leaf-gaps of the seedling, and 
internal phloem has been discovered in a portion of the stem of Osmunda 
cinnamomea far removed from any region of branching. Indeed, the facts 
connected with the occurrence of internal phloem in the Osmundacee 
most readily harmonize with the theory that it originated in this group as 
in the rest of the Pteropsida, and then suffered degeneration to a greater 
or a less extent. According to this view Osmunda cinnamomea stands 
as the bridge between Osmundites Skidegatensts ana such forms as Osmunda 
regalts. 
(c) Internal endodermis. 
On the upgrade theory the internal endodermis has been even more 
difficult of explanation than the internal phloem. The Gordian knot 
may be cut by denying that it is homologous with the external endodermis, 
but this would constitute it as an exception among the Pteropsida which 
is scarcely probable. Kidston and Gwynne-Vaughan recognize this 
difficulty, and suggest, though hesitatingly, that the external endodermis 
too, may have subsided through branch gaps. Butitis peculiar neither in 
form nor behaviour and is most readily understood by being considered 
as homologous with the external endodermis, and as having originally 
entered the stele in the manner characteristic of other ferns. 
(d) Nodal pockets. 
The occurrence throughout the Osmundaceze of pockets of paren- 
chyma extending down into the xylem from the inner ends of the medullary 
rays is a matter of considerable interest, because there is reason to believe 
that a knowledge of their origin would shed considerable light on the 
course of evolution of the stele in this family, and because of the possibility 
