— 47 — 
which is quite concealed by them. This arrangement may in some measure 
make up for the lack of specially differentiated cortical stem-cells in this species. 
In S. cuspidatum on the other hand, which is of more aquatic tendencies, such 
an arrangement is unnecessary and the stem is quite bare, the pendent branches 
being inconspicuous and little developed. The stem-leaves of S. cuspidatum 
are distinguished as somewhat fibrillose in the apical part, and as a rule this is 
true, but it is also a fact that the fibrils may be lacking in stem-leaves of S. cus- 
pidatum, while on the other hand, in immature or otherwise abnormal specimens 
of 5 . recurvum, the stem-leaves may show somewhat of the fibrillose condition 
of the branch-leaves, a condition of isophylly or hemiisophylly that must always 
be borne in mind in considering the stem-leaves of any species of Sphagnum, 
for these are obviously a secondary development from the type of the branch- 
leaves and are accordingly in immature specimens not adequately differentiated, 
while in other abnormal ones they may approach the same type by reversion. 
In section the chlorophyll cells of the branch-leaves of S. cuspidatum are de- 
scribed as trapezoidal, having a narrower exposure on the ventral surface of 
the leaf, while those of S. recurvum are triangular, the apex of the triangle barely 
reaching the ventral surface. Normally this is true, but unquestionable speci- 
mens of S. recurvum, particularly the aquatic forms called S. fallax Klinggraff, 
may be found with the leaf-section of S. cuspidatum, just as we have seen before 
that aquatic specimens of S. palustre may show such abnormal chlorophyll cells. 
I have even seen such specimens of S. i ecurvum var. tenue. The character which 
remains, which to my mind is the least subject to confusion, is the stem-cortex. 
In S. cuspidatum the cortex consists of about two rows of good-sized, fairly thin- 
walled cells clearly set off from those within. It is possible that aquatic con- 
ditions of growth may have some effect upon the cells, but that they can im- 
mediately produce a difference such as that between the cortical stem-cells of 
S. recurvum and S. cuspidatum is incredible, and S. fallax is an aquatic form of 
S. recurvum, which does not show it. 8 Artificial species that have been inserted 
between them with such names as S. fallax, S. pseudor ecurvum, S. pseudo cuspi- 
datum, etc., are apparently always satisfactorily referable to either the one or the 
other. Of other characters than those already mentioned it may be said that 
the branch-leaves of 5 . cuspidatum are more elongated to almost linear, often 
without undulations, and that their empty cells are accordingly narrowed and 
their pores reduced to the mere end-pores on either surface, or nearly so. Of 
further synonyms 5 . virginianum Warnstorf, 1900 is retained in 1911, 9 though dis- 
tinguished there neither in key nor description, nor so far as one can see in the 
specimen, from small forms of S. cuspidatum. S. ruppinense Warnstorf, 1908, 
originally Prussian as the name suggests, was given an American locality in 191 1 , 10 
but here also I am unable to find a valid specific distinction. The same is true of 
S. Faxonii Warnstorf, 1908, from New England. From researches of Bartlett * 11 
8 Cf. the note in Bryologist, XX, 87. 1917 and the literature there cited. 
9 Pflanzenreich, 5 1 : 269. 
10 Op. cit., 232. 
11 Rhodora, X, 113 f. 1908. 
