1908-9. ] THE STELE OF OsSMUNDA CINNAMOMEA. 525 
sieve-tubes appeared in the thickening band of parenchyma between the 
internal endodermis and the xylem, to be succeeded by a continuous layer 
of phloem two or three cells thick, separated from the internal endodermis 
by parenchyma and likewise from the wood. Preceding this strengthening 
of the internal phloem there was a shrinking of the now parenchymatous 
pith and a thinning of the wall of the cylinder of xylem. The leaf-traces 
during the later stages were reduced and their separation made little 
impression on the xylem. A short distance above the establishment of 
the internal phloem, a considerably larger leaf-trace took its origin from 
one end of an elliptical (in transverse section) stele, accompanied by a 
thickening of the xylem and internal pafenchyma at that end, the for- 
mation of a few isolated tracheids in this internal lining of parenchyma 
and a very narrow gap. ‘The internal phloem extended out somewhat 
in the direction of this gap, but did not reach into it (Fig. 10). The 
stele rapidly closed up, subsequent leaf-traces originated cladosiphonically 
and the pith and internal endodermis disappeared, the xylem now sur- 
rounding a layer of parenchyma and a considerable mass of phloem in 
the centre, which also soon died out. 
(c) Accessory strands. 
Kidston and Gwynne-Vaughan call attention to bundles lying 
just within the cylinder of xylem on one side of the pith in the section of 
Osmundites Skidegatensis examined by them. Strands of an apparently 
similar nature sometimes occur in Osmunda cinnamomea varying all the 
way from single tracheids to groups of considerable size. The ones 
represented in Fig. 12 were taken from a region of branching and are 
exceptionally prominent. They are nothing more than portions of the 
cylinder of xylem shut off from the main part for a longer or a shorter 
distance by parenchyma. 
It is interesting to note in this connection that isolated tracheids 
laterally situated in the pith (centrad of the internal endodermis) have 
been observed several times. Such tracheids are surrounded by a ring of 
endodermal cells, and have obviously been ‘“‘pinched off’’ from the stele. 
CONCLUSIONS. 
Regarding the phylogeny of the Osmundacez, there is a wide diversity 
of opinions ,—a diversity due to fundamentally different conceptions of 
the evolutionary history of the stele. Van Tieghem (14) and subsequently 
Zenetti (15), basing their theory on the structure of Osmunda regalis, 
held that the Osmundaceous siphonostele was derived from a protostele 
