8 FILICALES 
nm 
ios) 
from a more complex condition, or that the system is itself in the upgrade, 
and an indication, as seen in the living examples, of the approximate 
limit which development had attained in the group. The former opinion 
has been elaborated by Jeffrey! and by Faull:? they hold the Osmundaceous 
stele to be a reduced form of “amphiphloic siphonostele,” and in support 
of their opinion they adduce the presence of an internal endodermis 
(O. cinnamomea and T. hymenophylloides), and the occasional presence in 
some specimens of O. céxnamomea of internal phloem also, locally in the 
Fic. 299. 
A =a representation of a portion of the xylem-ring of Osmunda regalis seen from 
without ; é¢=cut end of a departing leaf-trace; 2g=leaf-gap. (After Lachmann, from 
Kidston and Gwynne-Vaughan.) =a representation of a portion of the xylem-ring of 
Todca barbara, seen from without. Lettering as above. (After Seward and Ford, from 
Kidston and Gwynne-Vaughan.) 
neighbourhood of the branchings of the axis. There are good grounds 
for doubting whether the local and inconstant occurrence of internal phloem 
and endodermis will bear the weight of a far-reaching theory of reduction: 
the question has been argued sufficiently elsewhere,’ on grounds of anatomical 
comparison of living forms, and without acceptance of the reduction theory. 
Even on grounds of physiological probability it would appear less likely 
that a robust and large-leaved phylum of Ferns should show a reduced 
vascular system in its stock than that the stock should retain a primitive, 
though perhaps imperfectly efficient system. 
Apart, however, from such questions of probability, a good basis for 
' Phil. Trans. vol. excv., p. 119, etc. 
2“ Anatomy of the Osmundaceae,” Bot. Gaz., 1901, p. 381. 
*Scott, Mew Phytologist, vol. i., p. 209; Seward, /c., p. 255; Boodle, Ann. of 
Hot., 1903, p. 5183; Chandler, vv. of Bot., 1905, p. 406. 
