250 GENERA INCERTAE SEDIS [ch. 



most part results of relatively recent homoplasy in stocks originally distinct. 

 The new classificatory problem will be, how far can the method of examining 

 facts historically rather than objectively be carried into effect? The nature of 

 the difificulties which beset a study so conducted will be better understood by 

 considering some examples in detail than by pointing out the many genera 

 and species, the phyletic position of which is still held in the balance. 



Cystopteris Bernhardi 



The genus Cystopteris wdiS established by Bernhardi in 1806, with the type 

 Polypodiuni fragile Linnaeus, 1753. This Fern was subsequently known as 

 Cystopteris fragilis , though Schkuhr {Farnkrmiter, Taf 54, p. 53) described it 

 as Aspidiuin fragile. The genus Cystopteris now comprises 13 species: 

 Sir W. Hooker notes it as a very natural genus, and may assuredly be 

 considered a connecting link between the Davalliaceae and Aspidiaceae. 

 He placed it between Davallia and Lindsay a (Sp. Fil. I, p. 196; Syn. Fil. 

 p. 103). Presl ranks it in his Tribe HI Aspleniaceae {Tent. p. 92): Mettenius 

 in his Tribe IV Aspidiaceae {Fil. Hort. Lips. p. 96), close to Onoclea and 

 Woodsia. Luerssen gives a whole page to its synonymy {Rab. Krypt. Fl. Ill, 

 p. 451). Christ follows Mettenius {Farnkrduter, pp. 8, 280). Diels includes it 

 in his Woodsieae-Wood^iinae (E. and P. I, 4, p. 159). But Christensen (Mono- 

 graph <:>{ Dryopteris, I, p. 64), writing of his sub-genus I. Endryopteris, remarks 

 that '' Eudryopteris in most characters agrees with Cystopteris. — To me it is 

 probable that Endjyopteris and Cystopteris are closely allied to each other, 

 and that it is unnatural to place them in two different tribes." In other words, 

 he holds that the natural relation of Cystopteris is with Ferns where the sorus 

 is typically superficial. The alternative is the relation pointed out by Sir W. 

 Hooker, with Davallia and Lindsaya, where the sorus is typically marginal 

 In origin. The question would then appear /r/;;/rty^?aV' to be one of develop- 

 mental fact. 



The fully developed sori of Cystopteris fragilis are clearly superficial, 

 being inserted upon a vein which extends beyond the receptacle (Fig. 747). 

 The indusium is basal, and forms an inflated investment. But in sections of 

 the young leaf-margin traversing the sorus in a median plane, the actual 

 margin and the receptacle appear in close relation to one another (Fig. 748, 

 A, B): still these sections suggest the origin as superficial. But Professor Von 

 Goebel describes, with drawings reproduced as Fig. 748, C, D, an earlier stage 

 of the sorus in Cystopteris inontana, as follows: "The placenta is initiated 

 either from the leaf-margin, or in the nearest proximity to it, on the under- 

 side of the leaf, on which also the indusium grows out. But the development 

 (or further development) of the leaf-margin is at first delayed, while the 



