-85- 



Even the name recurvum has been discarded by some recent authors, though 

 hardly with equal justice. Beauvois' name is distinctive, the description em- 

 phasizing the distinctive feature, and his type-locality was South Carolina, 

 which excludes other species with recurved leaf-apices and the several valid 

 segregates, whose range does not in any case extend further south than New 

 Jersey. 



Some of its characters have been indicated in my last paper: the cortical 

 cells of its stem are not strongly differentiated from those within, the stem- 

 leaves are relatively small, the branch-leaves show when dry not only recurved 

 tips, but also several undulations further down (a condition of things not evident 

 when they are moist), their chlorophyll cells are exposed broadly on the dorsal 

 surface, triangular in section with the apex ordinarily just about reaching the 

 ventral surface, their pores are few, on the outer surface small end-pores usually 

 at both ends of the cell with sometimes an additional pore or two in the lateral 

 corners, on the inner surface a few rather large round pores in cell-corners, the 

 latter rarely exceeding four to the cell, sometimes almost or entirely suppressed. 

 The plants are nearly always bright green, but may show a brownish pigmenta- 

 tion; the leaves of the antheridial catkins are strongly pigmented brown. Though 

 dioicous the species fruits by no means rarely, fruiting tufts often showing male 

 plants intermingled with those bearing the capsules. 



Though the structure of this species is so relatively simple that one might 

 be tempted (wrongly) to consider it the ancestral type of the order, it is in its 

 formal variation the most difficult and critical species after 5. suhsecundum. 

 Apart from the forms which can with a degree of justification be separated as 

 varieties or species there remain several synonyms to be explained. S. pulch- 

 ricoma Carl Miiller, 1848 is now accredited by Warnstorf to North America.'* 

 It was based originally upon South American forms. I have not seen the type- 

 specimen of S. pulchricoma, but have seen Brazilian specimens so named by 

 Carl Miiller together with various others from South America. These I should 

 not feel entirely justified in separating specifically from 5. recurvum. Warnstorf 

 has included under 6'. pulchricoma not only specimens from our Gulf states, but 

 also some from as far north as Massachusetts which are entirely normal 5'. re- 

 curvum. The characters that he gives for S. pulchricoma, lacerate stem-leaves- 

 and chlorophyll cells of the branch-leaves not reaching their ventral surface, 

 do not apply to all the specimens named S. pulchricoma by Warnstorf himself, 

 and in fact so far as they indicate a tendency at all indicate one within the normal, 

 lines of variation of the usual European and North American 5. recurvum, except 

 that the leaf-section of our most southern form is noteworthy. The character 

 is however sornewhat variable. Upon a specimen from Alabama Warnstorf 

 based in 1908 a new S. riparioides, which he still (191 1) retains. The type- 

 material is in every way identical with other specimens included by Warnstorf 



' The relative lack of porosity of this and related species, in fact of the whole group Cuspidata^ 

 makes them of little value in the present important use for surgical dressings. 

 4 Pfianzenreich, 51: i88ff. 1911. 



