EXTERNAL CHARACTERS 631 
leaves of young Ferns, and lies at the basis of the architecture of the mature 
leaves, finds its counterpart in the dichotomy seen in certain strobiloid 
Pteridophytes. 
It is true that the Leptosporangiate Ferns show a very distinctive mode 
of segmentation both of apex and margin of the leaf. But it has been 
shown that in this character the Osmundaceae form an intermediate step 
from them to the Marattiaceae, while the latter diverge clearly from the 
Leptosporangiate type. It is thus seen that the definite segmentation of 
the Leptosporangiate type is no essential character of the Fern-leaf at large. 
Such considerations point to' the justness of the view that the Fern-leaf, 
however different in size, in continued apical growth, and in its segmentation, 
is essentially comparable with the’ smaller and simpler leaves of the strobiloid 
‘forms. We shall therefore accept the conclusion that in the evolution of 
Ferns some such leaf-enlargement as is faintly indicated in certain strobiloid 
Pteridophytes, and notably in the Ophioglossaceae, was carried out to a 
higher degree than in any other Archegoniate Plants. It would appear 
probable that the Ferns, developing early towards megaphylly, worked 
out to the fullest such methods of leaf-enlargement as are outlined in 
some other early types; in fact, that they were ultimately derived from a 
smaller-leaved ancestry, with a strobiloid shoot not unlike that which 
remained in the rest persistently small-leaved. 
As regards the differentiation of their leaves, Ferns show a comparatively 
low position. In a very large proportion, in which are included most of 
the types which are held as primitive, the leaves are general-purposes leaves : 
each serves at first for protection of the apical bud, and on unfolding is 
at once an organ of assimilation and of propagation. The differentiation 
of trophophylls and sporophylls is usually marked by a reduction of the 
assimilating surface where the sporangia are borne: examples are seen in 
Struthiopteris and Blechnum, in Acrostichum and Flatycerium, and the 
distinction is to be held as a morphological advance which had, however, 
already made its appearance in the Ferns of the Carboniferous Period. A 
good example of this is seen in the Hymenophyllaceae, where the leaves 
are undifferentiated in Hymenophyllum ; but in certain species of Z7ichomanes 
(§ Feea.), the genus which on other-grounds is held to be more specialised 
than Hymenophyllum, a distinction of sporophylls from trophophylls is seen. 
Sometimes the differentiation may be between parts of the same leaf, as 
in Osmunda, and the fact that within this genus the relative position of the 
sterile and fertile parts may vary indicates that the distinction is not very 
deep-seated. Innumerable middle-forms between the sterile and fertile 
conditions further indicate how imperfect the differentiation actually 
is. A further specialisation of certain leaves as protective scales is seen 
in Osmunda and in some elongated rhizomes: in such cases the rudimentary 
leaf-apex shows that these are potentially normal leaves diverted to the 
protective duty. It thus appears that the differentiation of the leaves in 
Ferns is not on a high scale: that they are all essentially of one type, and 
