114 A. G. Tansley. 
family, if we accept the theory which derives the filmy from the 
normal leaf. The large spore-output is another fact pointing in the 
same direction. 
Trichomanes radicans has a much stouter rhizome and a much 
larger stele than T. renifonne, but this is really of almost the same 
type, though there are many more tracheal elements (Fig. 26). 
Two spiral protoxylems are present through most of the internode, 
hut just above the node, tracing the stele downwards, they fuse 
into one, and the place of the second is taken below the node by 
that belonging to the incoming trace. The central parenchyma is 
confined to a small aggregation round each protoxylem, though the 
whole of the xylem is inter-penetrated by a network of parenchy¬ 
matous cells. 
At the node the stele (traced upwards) becomes elongated in 
the horizontal plane and divides into two somewhat unequal 
portions, each containing a protoxylem strand. The smaller portion, 
whose protoxylem has split into two (Fig. 27, B). again divides in 
Fig. 27. Trichomanes radicans. Diagrams^ showing separation of 
strands of the axillary shoot and petiole from the'stele of the rhizome; see 
text. Xylem with diagonal lines, phloem with interrupted lines. From 
Boodle, x about 13. 
the same manner, the adaxial, slightly larger, member becoming the 
stele of the axillary branch, while the abaxial member becomes the 
petiolar strand. This at first has the same character as the stem 
stele (Fig. 28 B), but soon the phloem thins out on the adaxial 
(upper) side, the xylem ring breaks at that point, and the protoxylem 
