535 
Anatomy of the Osmundaceae. 
structures Is often to some extent a matter of personal opinion, and all that 
can be said is that, after making every allowance for lack of preservation. 
Dr. Kidston and I still adhere to our original statement, that c if medullary 
rays actually were present in the living plant they must have been extremely 
narrow and very rare ’. l 
As regards the lack of transitional forms, the supporters of the extra- 
stelar pith theory in the Osmundaceae are very badly off indeed. As 
I understand it, they regard all the known forms, except those with solid 
protosteles, as stages in a series of reductions. In consequence they have 
not a single form to show of all the advancing stages that must have 
occurred from the first pocketing into the solid protostele up to the produc¬ 
tion of the perfectly dictyostelic form that is supposed to have antedated 
Osmundites skidegatensis. 
As regards evidence of pocketing in the sporeling, I maintain that 
a distinction must be drawn between the pocketing of the peripheral tissues 
of the stele into the xylem and the pocketing of the external tissues of 
the stem into the stele. It is clear that the former must precede the latter. 
That the outer tissues of the stem have, in many cases, a tendency to invade 
the central tissues in the axils of the leaf-traces is well known, and these 
xylem-sheath pockets represent the initial and the simplest possible expression 
of this tendency . The next step is represented by the Lindsay a type of stele 
in which the phloem has followed the xylem-sheath, but the pocketing is 
still intrastelar. The simpler stages of extra-stelar pocketing are met with 
in certain Gleichenias and in Davallia pinnata , leading on to the formation 
of solenosteles and dictyosteles. Finally, even the epidermis and the sur¬ 
rounding atmosphere may join in the invasion and penetrate into the 
central tissues of the stem, as is shown in Onoclea , Cystopteris , and Aneimia . 
In the sporeling of Osmunda, however, there is no evidence whatever of the 
pocketing of the external ground tissue or of the endodermis into the stele. 
The xylem-sheath pockets are intrastelar and nothing else. It is only by first 
of all assuming that the reduction theory is true that they can be imagined 
even as indicating the position of previously existing extra-stelar leaf-gaps. 
The arguments brought forward by Sinnott 2 in favour of the primitive 
existence of leaf-gaps in the stele of the Osmundaceae based on the existence 
of gaps in the petiolar meristele made by the departure of the pinna-traces 
necessitate the acceptance of the preliminary assumption that the leaf is 
equivalent to an axial branch system. Even if this be granted, there is no 
reason to expect that the caulome and the phyllome should undergo the 
1 Professor Fanil ( 1 . c., p. 530), referring to the presence of scale-leaves in Osmumlites Dunlopi, 
represents us as admitting that this plant is reduced. This is obviously a misunderstanding. The 
scale-leaves occur in regular succeeding zones as they do in Osmunda regalis , and no doubt they 
served the same purpose of protecting the apex during adverse seasons. We do not regard either of 
these plants as reduced. 
2 1. c., p. 113. 
