96 GYMNOGRAMMOID FERNS [CH. 



structure fluctuates between solenostely and dictyostely, while the leaf-trace 

 is undivided and saddle-shaped in transverse section. Even the median 

 attenuation of the meristele of Anemia finds its counterpart in Cheilanthes 

 and Pcllaea. Some of the larger Gymnogrammoids have more highly 

 disintegrated vascular tracts, as in Ceratopteris or Gymnogi'minne japonica:, 

 but those smaller types which compare sorally with Anemia and Mohria 

 approach these genera also anatomically. In Mohria dermal scales are 

 present, as in all the Gymnogrammoids excepting Janiesonia. Thus struc- 

 turally the comparison is upheld. 



The details of the collenchymatous thickenings of the cell-walls in the 

 prothalli of AdiantiLm and of CJieilantJies and Notholaena (Horvat) have 

 been shown to find their parallel in those of Lygodiiim (Heim), and oi Mohria 

 and Anemia (Bauke). This in itself may appear a slender ground for wider 

 comparisons: but when taken in conjunction with the similarities of these 

 Ferns in other respects, such facts acquire increased importance as indica- 

 tions of affinity. It should be frankly realised that the argument for a phyletic 

 relation between the living Schizaeaceae and the Cheilanthoid Ferns falls 

 short of demonstration. But the variety and cogency of the comparisons 

 may be held as indicating a reasonable probability of its truth. 



A general review of the Gymnogrammoid Ferns suggests, and at times 

 may indicate with some degree of clearness, their evolutionary place. They 

 are a rather varied group, but inherent truth may often be learned from 

 studying the general facies of a large body of organisms even if somewhat 

 loosely akin, as these Ferns appear to be. Their kinship is based partly on 

 positive characters of external form and internal anatomy: partly upon the 

 rather negative soral characters, and particularly upon the absence of a true 

 indusium, with its corollary, an indefinite soral construction. They may be 

 held to represent a plexus of phyletic lines, all traceable back with probability 

 to Ferns with marginal sporangia of larger size than theirs, such as the 

 present-day Osmundaceae and Schizaeaceae. A widespread transfer of those 

 sporangia from the margin to the surface of the blade appears to have 

 happened early: this is exemplified in Todea and Plagiogyria on the one 

 hand, and on the other it is suggested by Anemia and MoJiria. It is reflected 

 as an accomplished fact in Notholaena. In none of these is there any 

 specialised indusial protection, other than a recurved leaf-margin. We may 

 picture how, after that change from the margin to the surface had been 

 effected, the production of more numerous sporangia, diminishing in size as 

 their number increased, has spread first along the lines of venation from 

 the margin inwards: subsequently in some instances the areas between 

 the veins may also have produced sporangia, giving an Acrostichoid state, 

 as seen slightly in Adiantnm, and more clearly in Jamesonia, Trismeria, 

 and others. Occasionally a lateral linkage of sori resulted in intra-marginal 



