THE FERNS OF NICARAGUA. 



125. 



Each of the three groups tends toward the same general 

 type. — namely a form with sori rounded, and indusium none 

 as in Phegoptcris and Polypodium. The first branch however 

 falls short of the development reached by the other two, the 

 DicksoniccB all retaining an indusium. 



The CythaecB and Vttiariazdo not form direct links in their 

 respective series, but are rather lateral branches, the Vti- 

 tariece from the Grammitidici?. and the Cxathece from the 

 Aspidiciv. 



The Cyathece are here recognized as a familv and not a 

 sub-order. The obliquity of the annulus is not constant, 1 the 

 development of a distinct caudex above ground does not 

 always take place, beside occurring frequently among the 

 Dichsoniecp. hence the elevated receptacle forms the onlv rea- 

 sonablv constant character by which the group can be dis- 

 tinguished from the Aspidiece. 



The following is a key to the Orders and Families of the 



Class FILICINE^E. 



Sub-class I. Leptosporangiatae. Sporangia developed from single epi- 

 dermal cells. Heterosporous, or when homosporous the spor- 

 angium with a more or less developed annulus. 



Sub-class II. Eusporangiatae. Sporangia developed from a group of 

 epidermal cells. Homosporous, the sporangium without ar 

 annulus. 



Sub-class I. Leptosporangiatae. 



A. Spores of two kinds Order I. Heterospor.e. 



Not represented in the collection. 



B. Spores all alike Order II. Homospor.e. 



a. Sporangia with a well-developed annulus. 



1. Annulus transverse. 

 i. Annulus equatorial. 



* Sporangia on filiform receptacles in marginal cup-like or two- 

 valved involucres; mostly laxly-cellular. 



Sub-order I. Hymenophyllace.c. 

 ** Sori dorsal, without involucre. 



Sub-order II. Gleichexiace.e. 

 ii. Annulus apical. . . • Sub-order III. Schizeace.e. 



1 Upon this point see also pp. 41 and 42 of " Histeria Filic urn"— John 

 Smith. 



