136 DRYOPTEROID FERNS [CH. 



Conclusion 



In the preceding pages the probable phyletic relations of the genera 

 included in the Aspidieae have been lightly sketched. It is apparent how 

 the great genus Dryopteris naturally takes a central position, which is con- 

 nected downwards by comparison with the Woodsieae to some probable 

 Cyatheoid source. It may be held as developing in various ways the lop- 

 sided type of a constantly superficial, mixed sorus. The Male Shield Fern 

 has been used as a central type within that great genus: provisionally that 

 seems at least a reasonable view, since notwithstanding their several lines 

 of specialisation the related genera and sub-genera readily fall into their 

 places as derivatives from it. That all of these Ferns referred to the Aspidieae 

 by Diels and Christensen are really akin (after certain obviously incompatible 

 genera have been excluded, such as Dipteris and NeocJidropteris) is witnessed 

 by the richness of their synonymy, and the great diversity of method shown 

 by systematists in the delimitation of the genera and sub-genera. 



No attempt has been made here to treat this extensive family in detail, 

 systematically. That would be an impossible task for anyone but a professed 

 systematist. What has been attempted has been to lay down those broad 

 lines which a natural classification must needs follow ; and indeed has followed, 

 though not always consistently in the past. The results have been checked 

 by reference to probable phyletic origin, and by biological probability. The 

 most important features of advance are held to have been (i) the advance 

 from a gradate to a mixed sorus, (ii) the adoption of a lop-sided or zygo- 

 morphic symmetry, (iii) the production of a peltate "indusium superiiml' 

 (iv) the partial or complete abortion of the indusium, (v) the spread of the 

 ex-indusiate sorus along the veins, and (vi) the overflow of the sporangia on 

 to the free surface. Most of these are progressions that may be seen illus- 

 trated in other phyla than this one: the fact that this is so adds confidence 

 to the present treatment. The production of the peltate indusium oi Aspidiuiii 

 and Polystichuvi, however, appears to stand by itself, though it results in a 

 structure strangely like that of Matonia, with which, however, the palaeonto- 

 logical history shows that it cannot have had any near phyletic relation. 

 Other characters, and notably the progression from an open venation to 

 occasional vein-fusions and finally to a close reticulum, follow with some 

 degree of constancy upon the widening of the leaf-area and the advancing 

 specialisation of the sori : but the sori themselves may be held as providing 

 the most distinctive features of advance. 



In accordance with the steps of specialisation which they show, the leading 

 genera of the Aspidieae may be grouped, not as a final or actually phyletic 

 arrangement, but so as to bring into prominence the changes which are 

 held to be the most important in leading towards that end. It will be found 

 that the progression thus roughly indicated runs parallel with similar pro- 



