IOWA PTERIDOPHYTA. 147 



terides was not adopted for any group because it is made 

 to include such divers forms as Lycopodium and Botrychium , 

 and hence does not represent a natural group. 



For the remaining classes Swartz's names Filices (used 

 by Ljnne in a broader sense) and Lycopodinece are adopted. 

 The group Filices as recognized by him* was assigned the 

 same limits as are here recognized. Later Willdenow 

 (1. c.) used the name in a more restricted sense, omitting 

 the eusporangiate groups now called OphioglossacccE and 

 Marattiacccp'\ , and also the Scliizcaccce and Osmundacccz . 



The name Lycopodincce% was the first name applied to the 

 correctly defined group, and it is retained in its original 

 form for the same reason, and because in that form it has 

 become a familiar name. 



The use of these names can scarcely cause confusion, 

 and it is warranted by the law of priority. 



Generic names, because entering directly into all binomial 

 combinations, present a much more serious problem. 

 Genera, at least among ferns, are not distinct entities, for 

 well-defined characters which would absolutely mark them 

 do not exist in most cases. That this is true has been amply 

 demonstrated by the wide divergence in the definitions of 

 genera offered by the many eminent pteridologists who have 

 sought to introduce order into the classification of this 

 fascinating group. Thus the genus Polypodium as defined 

 by Hooker was subdivided by John Smith into nearly 

 thirty genera, and the more restricted Polypodium of 

 American authors is now subdivided into five genera. 

 Swartz's Aspidium (with occasional combinations with 



*F1. Ind. Occ, vol. iii, pp. 1578-1758 (1806) and in Synopsis Filicum. 

 (1806). 



f If a separate class is created out of this family as has recently been 

 done by Bitter (Eng. and Prantl Nat. Pflanzenf., 195 L4ef., p. 422, 

 — 1900) it should bear the name Poropterides (Willdenow, /. c.) 



JSyn. Fil., p. 87 (1806). 



