244 BULLETIN OF THE UNIVERSITY OP WISCONSIN. 



HYPNUM.i 



I. Leaves spreading, or complanate, not secund. (II on p. 247.) 



A, Leaves costate half ivay or more. (B on p. 245.) 



1. Leaves acute or acuminate. 

 Margins denticulate. 

 Deourrent and anrieled, suddenly acuminate H. chloropternm, 562. 



Neither decurrent nor auricled, gradually acuminate. 



Cells uniform throughout .... H. Columbiae, 542. 



Alar cells dilated, distinct . . . H. Colambico-palustre, 688. 

 Margins entire. 

 Leaves acute or short acuminate. 

 Alar cells few. 

 Costa J^ length or double and short . . H. palnstre, 398. 



Costa vanishing just below apex . . . H. polare, 589, 



Alar cells more numerous, reaching costa, 



large H. psendo-uiontanam, B9S. 



Leaves long acuminate. 

 Erect or spreading ..... H. polgamnm, * 379. 



Reflexed squarrulose. 



Decurrent H. decnrsivulnm.^' 54I. 



Not decurrent. 

 Plants in dense tufts . . H. nnicostatnm,* 54O. 



Plants in loose, intricate tufts . H. chrysopliyllura,"' 378.V' 



2. Leaves obtuse,* entire. 



* Cells enlarged at the basal angles.^ (* * on p. 245.) 



Costa sub-percurrent. 

 Monoicous, sparingly branched, alar cells gradually en- 

 larged H. cordifoliam, 402. 



I Including subgenera XVII to XXVI of L. & J. Man. 377-406. 



^ See Appendix, nos. 544, 545. 



^ The separation of these three species is very unsatisfactory. H. chrysophyllum is a 

 very variable species and from the inspection of barren material of H. decursivulum 

 and JI. unicostatum and comparison with the characters as given by Kindberg, it seems 

 probable that they can not be regarded as distinct species but only as forms of M. chryso- 

 phyllum. 



* Sometimes aplculate in JT. sarmentoium. 



^ H. torrentis C. M. & Kindb., Appendix, no. 586, falls here. 



