-60 - 



The species are widely distributed over the whole world. Phylogenetically 

 at the head of the section stand, as it seems to me, some half dozen species, sharply 

 distinct from each other, but showing at same time a certain kinship. There 

 must then have been a considerable loss of ancestral forms. All of the species 

 last referred to are boreal in their distribution. Some of them have given rise 

 to considerable lines of development, including species that have achieved a 

 much wider distribution and themselves thrown off secondary derivatives of 

 more restricted range, while other have remained isolated and unproductive. 

 Whether one regards the last as monotypic "groups" or includes them in other 

 groups with which they show the closest relationship is largely a matter of in- 

 dividual choice. It is then in no substantial disagreement with Schliephacke's 

 and Russow's arrangement of the European species of Sphagnum (Warnstorf's 

 arrangement of the non-European species is, as already noted, unnatural) that 

 I divide the American forms of Acisphagnum into three groups: Squarrosar 

 Cuspidata, Acutifolia. Where this involves any deviation from current usage 

 my reasons will be given in the discussion of the individual species. It may be 

 said that the form (in section) of the chlorophyll cells of the leaves can not be 

 used as an absolute diagnostic character of the group, as the original type of 

 cell with central lumen and approximately equal exposure on both surfaces of 

 the leaf may occur within any group and actually does occur in them all, and 

 there is often some little variation within a single species. The highly specialized 

 stem-leaves and perichaetial leaves seem to me to furnish the most readily per- 

 ceptible indications of group-relationship. To name the one character that 

 appears most conclusive, not as a practical means of separating the groups but in 

 its real indication of natural relationships: it is the disposition of those mem- 

 brane-gaps in the hyaline cells of stem-leaves and perichaetial leaves commonly 

 characterized as "resorption" of the membrane. The stem-leaves of the 

 species of Squarrosa show such "resorption" mostly on the outer surface (as in 

 Inophloea) , those of Cuspidata and Acutifolia mostly on the inner surface (as in 

 Malaco sphagnum). Certain of the phylogenetically old species show resorp- 

 tion of the membrane on both surfaces, which accounts for the characteristic 

 lacerate condition of the stem-leaves of these species. In the perichaetial leaves 

 such membrane-resorption can be present only in caee at least a portion of the 

 leaf shows the normal leaf-structure of alternating hyaline and chlorophyll cells 

 and then only in such hyaline cells as are non-fibrillose. When these conditions 

 are present, and they are in most species of Acisphagnum, one will find in Squar- 

 rosa membrane-gaps on the outer surface, in Cuspidata on the inner, in Acuti- 

 folia on neither. The relationship of the three groups could be illustrated ap- 

 proximately thus: 



Acisphagnum 



Squarrosa Cuspidata Acutifolia 

 The Group Squarrosa Schliephacke 



