COMPARISON OF EXTERNAL CHARACTERS 625 



dominant in our comparisons; and the burden of the argument must still 

 rest upon the facts derived from the sporophyte generation. We shall then, 

 excepting for an occasional reference, leave the gametophyte aside in the 

 present discussion, and review the characters of the Fern-plant in its relation 

 to the general theory of the sporophyte. 



EXTERNAL CHARACTERS. 



The Ferns are the characteristic megaphyllous members of the Pterido- 

 phyta, and thus differ markedly in habit from the smaller-leaved strobiloid 

 types. It is necessary first to inquire what are their probable relations to 

 these series. In point of time the distinction of habit dates back as far as 

 the earliest known fossils, and accordingly it is only by comparison that any 

 opinion can be formed as to their origin by descent, and then only as a 

 probability, not as a demonstration. The similarity of life-history shows, 

 however, that the sporophyte of the Fern as a whole corresponds to that 

 of the strobiloid types : the further question will then be as to the 

 correspondence of the parts, especially the axis and leaf. 



The chief difference lies in the proportion of leaf to axis, and in the 

 branching of the leaf, not in the fundamental relations of those parts as 

 regards origin or position : this is specially obvious in upright growing 

 species, with radial symmetry of the shoot. In the Ferns, as in other 

 Pteridophytes, there is reason to regard the radial type of the shoot as 

 primitive, notwithstanding the fact that a very large proportion of living 

 Ferns are dorsiventral. Among the Ferns of the Primary Rocks no dorsi- 

 ventral type of shoot has been described, unless it be the Permo-Carboniferous 

 genus Glossopteris, the relation of which to the true Ferns is still a matter 

 for discussion. It is possible that a creeping rhizome may have existed as 

 the base of insertion of some of the unattached fronds, but still in the 

 absence of demonstration of this the evidence points to the radial type as 

 having been prevalent. This is the case with the various stems designated 

 Caulopteris, in many of which the leaf-arrangement is on a spiral plan : 

 even those designated Megaphytum t where the leaves are distichous, were 

 of radial character, and all evidence indicates that their position was upright. 

 Among the best known of the early forms are the Botryopterideae, which 

 had relatively thin axes with leaves in some cases closely aggregated, in 

 others more laxly disposed : both types are of radial construction. Thus 

 the evidence, so far as it goes, indicates that the radial type of shoot was 

 prevalent, if not indeed exclusive, for the early Ferns. It is exemplified 

 by the Botryopterideae, the Marattiaceae, and the Osinundaceae, all early 

 types. 



That large-leaved forms would be mechanically unstable structures is 

 obvious, especially where the stem is thin and the internodes of appreciable 

 length. There is an inherent probability that such axes should become 



oblique or prone, with a dorsiventral development as a natural consequence. 



2 R 



