The Hymenophyllacecz. 
1T 7 
If such a central protoxylem is lost however, the advantage of 
precocious formation of peripheral elements for the attachment of 
root-steles may well come into play. The local differentiation of 
such elements in various Hymenophyllaceae and other ferns is 
common. But actual continuous spiral protoxylem in an extreme 
peripheral position (cxarchy) is quite rare in ferns, and though this 
condition is sometimes thought to he primitive in all Pteridophytes, 
the evidence, from the ferns at least, is too scanty to be convincing. 
Again starting from a central protoxylem, if we imagine the 
protoxylem strands to increase in number by the decurrency into 
the stem of strands from successive leaves, or of several strands 
from one leaf, as we apparently see happening in species of 
TricJioinanes, and at the same time the stele increases in diameter 
relatively to the leaf-trace owing to crowding of traces or for 
other reasons, the trace protoxylems will naturally come to run 
in a more peripheral position, and thus we may have the origin 
of the type of niesarchy that we get for instance in Gleichenia. 
Then if the stele enlarges sufficiently in diameter owing to other 
causes (a topic to which we shall have to return in later lectures) 
the central xylem will eventually be replaced by pith. In the 
middle and higher ferns, where this is the case, the xylem is very 
frequently endarch again, i.e., the protoxylem are placed on the 
inner limit of the xylem ring, and sometimes mesarch, either of 
Fig. 30. Hymetiophyllum dilatation var. Forstcrianum. Strand of base of 
petiole, a short distance above insertion of axillary branch. X 260. From 
Boodle. 
Fig. 31. Ilyinenophyllutn dilatation var. Forstcrianum. Strand of petiole. 
X 260. From Boodle. 
