574 FILICALES 



These features indicate that Loxsoma is more nearly related anatomically 

 to the Dicksonieae and Dennstaedtiinae than to any other family of 

 Ferns ; but a reasonable analogy is also to be found with the more 

 advanced species of Gleichenia : the solenostelic structure seen in G. 

 pectinata as well as the origin of its foliar trace are points for comparison, 

 while structural affinities of a more remote nature are also indicated with 

 the Schizaeaceae and Hymenophyliaceae. 



It thus appears that Loxsoma is a generalised type, while its rare 

 and local occurrence countenances this view. In habit it shows similarity 

 to such genera as Dennstaedtia, Microlepia, and Dava/lia, a comparison 

 which finds support in the anatomy of the vascular system : not only do 

 the habit and anatomy support this, but also the form of indusium and 

 receptacle, and the basi petal succession and orientation of the sporangia. 

 An affinity with the Hymenophyliaceae is also unmistakable, but probably 

 not so close as has often been assumed : against it are the texture of the 

 leaf, the mode of dehiscence and the structure of the sporangium, 

 and the low output of the very large spores : in any case the affinity 

 is with the less specialised types (e.g. Hymenophylltim dilatatum] rather 

 than the more specialised (e.g. Trichomanes). The sporangium, and its 

 annulus and dehiscence point clearly towards the Gleicheniaceae and 

 Schizaeaceae ; and though the habit of the leaf is different from these 

 Ferns, the structure of the creeping rhizome shows a certain resemblance. 

 The similarity of position of the annulus, and constancy of orientation 

 are important, especially when taken with the very peculiar facts of induration. 

 For, as we have seen, the distal side of the annulus is indurated, while 

 the proximal can still be followed, though it is commonly thin-walled ; 

 but occasionally single cells, or groups of cells, of the proximal side are 

 also indurated : these cannot be functionally active, since they do not 

 form a connected series. In them I think we can only see a decadent 

 vestige of a completely indurated annulus, and conclude that Loxsoma was 

 derived from ancestors with a complete oblique annulus, probably with 

 a median dehiscence. Such ancestry might be found in the neighbourhood 

 of Gleichenia. In Gl. dichotoma we have a type in which the sporangium 

 and the sorus are similar in their main character. If we imagine these 

 sori to be marginal (as they are in Lygodinm), surrounded by a cup-like 

 annulus which is already suggested in some Gleichenias, with the annulus 

 modified as explained above to suit the more crowded sorus, and with a 

 smaller number of spores, balanced by a larger number of sporangia 

 produced in basipetal order, the sorus of Loxsoma would be before us. It 

 is not suggested that any living Gleichenia was a progenitor of Lo 

 but Loxsoma appears to be a link connecting the Gleichenia Scliiz 

 affinity with the type of Dennstaedtia and Microlepia. It should be 

 regarded as the sole representative of a distinct tribe : the attempt should 

 not be made to force it into any other tribe of living Ferns. 



