Evans: HeEpaTicaAE OF PuERTO RICO 9 
be most at home in moist forests, especially on the lower slopes 
of mountains. 
The stems are prostrate and freely but irregularly branched. 
In most cases the branch-leaves are smaller than those borne on 
the stem, and occasionally microphyllous branches are produced. 
The latter, however, are less modified than in certain other genera 
of the Leyeuncae. The lobes of the leaves spread widely from 2} 
the stem and are broad and rounded at the apex (PLATE 2, FIG- 
URES I-3). The margin is entire throughout, unless hyaline cells 
are present at the apex. The lobule is strongly inflated, especially 
in the basal portion, and forms a distinct water-sac. The free 
margin spreads widely from the axis and terminates in a slender 
curved tooth which is usually from three to five cells long and 
two cells wide at the base (FIGURES 7, 8). This tooth and the 
shallow lunulate sinus commonly bound the circular opening into 
the sac. The hyaline papilla arises from one of the marginal cells 
of the sinus but is displaced to its inner boundary and cannot be 
seen without dissection. The keel is arched and sometimes very 
strongly so. The leaf-cells have thickened walls, but the thicken- 
ing is uniform or nearly so and there are no distinct trigones. Of 
course when ocelli are present these have thin walls, and the same 
is true of the hyaline cells found at the apices of the lobes in NV. 
catenulata (FIGURES 5, 6). The underleaves are orbicular and 
undivided. They are attached by a short and slightly arched 
basal line, and their margins are entire (FIGURES I-3). 
The female inflorescence is borne on a short or somewhat elon- 
gated branch and usually innovates on both sides, the innovations 
being nearly always simple and sterile (FIGURE 1). The bracts are 
larger than the branch-leaves which precede them but smaller 
than the leaves on robust stems. They are distinctly complicate 
and unequally bifid, and their lobes spread obliquely. The 
bracteole is free and differs but little from the other underleaves 
€xcept in size. In most respects the perianth is similar to that 
described for Stictolejeunea. The postical keel, however, is much 
more variable and is sometimes produced upward as a broad and 
rounded sac, which may equal in length the two lateral auricles 
(FIGURE 12). In extreme cases the beak of the perianth may even 
be displaced to the antical surface of this sac. When the sac is 
