176 EvANs: TAXONOMIC STUDY OF DUMORTIERA 
in D. irrigua, where they seem to be somewhat of ararity. Ernst, 
however, demonstrated their frequency in Javan specimens of 
““D. trichocephala”’ (4, p. 207). He found them also in D. velutina 
but much more rarely and associated this difference with the 
prevailingly monoicous inflorescence of ‘‘D. trichocephala”’ and 
the prevailingly dioicous inflorescence of D. velutina. Whether 
or not bisexual receptacles are frequent in American forms of 
Dumortiera remains to be determined. 
The spores of Dumortiera vary in color from pale to dark 
brown. The tetrahedral form persists until maturity, although 
the ridges separating the faces are sometimes difficult to distin- 
guish. The surface-markings are remarkably uniform. Each 
face bears a series of minute and crowded papillae or short lamellae, 
which are usually irregularly distributed but which sometimes 
show a slight tendency to be arranged in short rows. No reti- 
culum is developed. With respect to size statements in the liter- 
ature are not in accordance. Stephani (25) gives a diameter of 
25 u for D. trichocephala and of 34 u for D. hirsuta; for D. velutina 
he gives no measurements. Ernst’s figures are considerably higher 
(4, p. 178); he gives a length of 45-60 uw and a width of 35-50 un, 
without distinguishing between D. trichocephala and D. velutina. 
Campbell (2, p. 329) states that the spores of D. trichocephala are 
about 20 uw long while those of D. velutina are about 29 u. The 
writer has examined spores of various forms from widely separated 
stations and finds that the long diameter measures 20-30 p, a 
considerable range being often present in a single capsule. Ap- 
parently little help can be obtained from the spores in the separa- 
tion of species. 
The preceding discussion brings out the untrustworthy nature 
and inconstancy of certain differential characters which have 
been employed in defining the species of Dumortiera. Those 
drawn from the size and method of branching of the thallus seem 
especially unreliable. Those drawn from the female receptacle 
and the spores are scarcely more satisfactory. On the basis of 
characters drawn from the structural features of the vegetative 
thallus, the two following species may be distinguished, and these 
are the only ones which the writer would recognize at the present 
time: 
