254 
A. G. Tansley. 
of Zygopteris, particularly of Z. corrugata (Fig. 5), while quite recent 
discoveries of Kidston and Gwynne-Vaughan (’07) 1 make the 
connexion with Botryopterideze even plainer, by revealing the 
existence in Osmundites of types of structure clearly connecting 
the recent Osmundaceze with the simplest form of Botryo- 
pteridean stele, such as that of Renault’s Granniiatopteris Rigolottii. 
It would thus be most natural to deal with the Osmundaceze 
immediately after the Botryopterideze, and at the beginning 
of the series of living Ferns. This course was not adopted in the 
present lectures partly because the line of evolution in the Osmun¬ 
daceze stands, as has been said, apart from that of the great 
majority of living ferns and did not form a suitable introduction to 
the bulk of our subject matter, and partly because some of the 
phenomena displayed have led to the propounding of a view that 
the Osmundaceous series of types really represents a reduction- 
series from a dictyostelic form, instead of a progressive series 
starting from a simple protostelic form. This view could not be 
discussed adequately except in the light of our knowledge of the 
anatomy of the other Ferns. 
We will begin with a consideration of the two living genera 
Osinunda and Todea, which agree closely in the leading features of 
their structure. 
The stem is upright or obliquely ascending, in some cases 
(Todea barbara) rising some feet above the ground like a small tree- 
fern. Usually, however, it is quite buried in the soil. The diameter 
of the stem is very greatly increased by the thick, woody, overlapping 
leaf-bases, which completely surround it, and also by the numerous 
roots. The phyllotaxy is a complicated spiral such as ^ 8 T . 
A transverse section of the central cylinder in the majority of 
species (Fig. 81) shows a circle of xylem strands of different shapes 
(some U-shaped with the concavity directed inwards, and some more 
or less oval or circular in cross-section) separated from one another 
by rays of parenchyma, while the whole is surrounded by a 
complete ring of parenchyma (mesocycle) and phloem, bounded 
externally by pericycle and endodermis. 
The phloem characteristically forms wedge-shaped projections 
into the rays. At certain points, outside the ordinary vertically 
running sieve-tubes, and sometimes interrupting the circle of these, 
are tangentially elongated cells which have the structure of normal 
1 Published since these lectures were delivered, and partly still 
unpublished. 
