92 GYMNOGRAMMOID FERNS [CH. 



Comparison 



The designation "Gymnogrammoid Ferns" covers a large number of 

 genera and species, if it be taken as connoting those Ferns which show the 

 general features of the old genus Gymnograiiime: i.e. superficial sori, more or 

 less extended along the veins, but without any specialised indusium. We may 

 now add to this from the more modern point of view \h& proviso, that there 

 shall not be any comparative ground for believing that an indusium existed 

 in their ancestry. This definition would cover the bulk but not all of those 

 genera which were grouped by Diels under his designation of Gymno- 

 gramminae, Cheilanthinae, and Adiantinae, together with Cassebeera: while 

 some Ferns previously included under Acrostichum would have to be added 

 on grounds of comparative study. But it excludes such genera as Ceterach, 

 Plettrosorus, Hypolepis, and Aspleniopsis\ while Plagiogyria would be held 

 apart as a specially archaic type. 



With characters so diverse as some of the included Ferns show, and a 

 common character so comprehensive, it would be against reason to assume 

 at the outset any single phyletic source. We ought rather to be prepared to 

 find here grouped together the products of a plurality of evolutionary lines. 

 By comparison among themselves, and by reference to types more archaic 

 than they, or of earlier geological occurrence, it may be found possible 

 to segregate some at least of them, and to assign them to their probable 

 phyletic sources. The old conception of Gymnogramme as a genus stood 

 next in indefiniteness to those of Polypodiuni and Acrostic/mm. Like them 

 it probably did not represent a single natural group. One general feature 

 which holds in the Gymnogrammoid Ferns more than in most others is that 

 its characters are negative rather than positive. In the Gleichenioids and 

 Cyatheoids, as also in the Hymenophylloids and Dicksonioids, the sorus is 

 a clearly circumscribed entity, with a definite receptacle round which the 

 sporangia are grouped, and in many of them more or less constant protective 

 indusial growths are present. It is true that the identity of the sorus may 

 be blurred or even obliterated by soral fusion, with or without subsequent 

 fissions: or by the adoption of an Acrostichoid spread in those types which 

 are held as the most advanced. But in the Gymnogrammoid Ferns the sorus 

 is not a definite entity in the same sense as in other Ferns: it is without the 

 circumscribing influence of an indusium, and the sporangia appear from the 

 first to have been liable to spread indefinitely along the veins, or even into 

 the areas between them. This indefiniteness, while it makes the whole group 

 more comprehensive, removes from the sorus those distinctive features which 

 have aided its comparison elsewhere. The effect of this is to accentuate the 

 importance of the sporangia themselves, and to concentrate comparison upon 

 the details of their position, structure, and spore-output. 



