THE FERN'S OF NICARAGUA. , 3^3 



The classification which is here presented contains no start- 

 ling innovations. Some changes are made in the relative position 

 of well-known groups with the view of securing a more con- 

 venient arrangement which will also give expression to exist- 

 ing relationship along the line of the characters upon which 

 the classification is chiefly based. 



A key to the larger groups is arranged, without, however, 

 any attempt at a full definition of these groups, as they are 

 sufficiently characterized by the standard authors. 



The development of the FiJiciiiCiV seems to have been 

 toward a suppression of the annulus. and toward a restriction 

 of the spore-bearing areas, resulting in diminished sori with 

 the indusia finally suppressed, or in limited specialized portions 

 of the frond. 



No linear arrangement of the groups will properly show the 

 relationship which exists between them. 



The suborders of the Subclass Leptosporangiatm^ for exam- 

 ple, form a branching series of which the HxmcnofihxUaccce 

 and GIcichcniacciF form the basal portion, or main stem, the 

 HymenophyllacecB clearly constituting the first node, the struc- 

 ture of the lamina of the frond (often but one cell thick), the 

 usually well-marked dorsi-ventral structure of the plant, the 

 occasional development of root-hairs (rhizoids) only, in place 

 of true roots, placing them near the Liverworts, and the pro- 

 duction of a protonema-like prothallium in some species of 

 Trichomona suggesting the Mosses. The Glcichcniacccc which 

 form the second node still recall the Liverworts by their often 

 marked dichotomy and by the frequent peculiar innovations of 

 the frond, but in other respects they are more nearly like the 

 higher groups of the sub-class. The species of both of these 

 sub-orders have a complete transverse, equatorial annulus. 



The development from this main stem is along two principal 

 branches, namely toward the suppression of the horizontal 

 annulus through the Schizeaccce, where it is apical but com- 

 plete, to the OsmundacecR where it is sub-apical but rudiment- 

 ary, on the one hand, and toward the vertical incomplete 

 annulus of the typical Polyfiodiaccce through the Family 



