200 Evans: HEpATICAE OF PUERTO RICO 
gemmiparous branches, on which the leaves and underleaves differ 
materially from those described (FIGURE 4). On these modified 
branches the leaves are ligulate in form and are not falcate, the 
two margins being approximately straight and parallel; the apex 
of the lobe is rounded or truncate and the outer portion is much 
more coarsely dentate than on ordinary leaves. The cells of the 
lobe are of about the same size throughout (FIGURE 9g). The 
lobules of these modified leaves become more and more rudimen- 
tary on passing from the base of the branch toward the apex and 
are finally reduced to minute and indistinct basal folds. The un- 
derleaves on the gemmiparous branches are close together and 
more or less squarrose ; they are, moreover, strongly convex, and 
it is probable that the species derived its name from underleaves 
of this type. The margin in the broad apical region is frequently 
denticulate, and sometimes the two teeth nearest the middle are 
larger than the others, thus giving the appearance of a shortly 
bifid underleaf with a shallow apical sinus. The complicated 
synonymy of C. convexistipa is due largely to the occurrence of 
these peculiar branches, which sometimes completely mask the true 
characters of the species ; it has been ably elucidated by Stephani.* 
The perianth of C. convexistipa is also subject to considerable 
variation, although the form described may be considered typical. 
In some cases the postical keel is extended upward as a distinct 
auricle or horn, and in rarer instances an antical keel, also with a 
horn, makes its appearance. Usually these supplementary horns 
are shorter than the lateral, but in one observed case they were 
fully as long and gave rise to a four-horned perianth. The posi- 
tion of the horns, however, was different from what we find in Cera- 
toleyeunea, where two of the horns are developed from the upper 
angles of the postical keel. In other cases the floral leaves, which 
normally unite to form the perianth, fail to do so properly and 
leave an incomplete organ. Occasionally the abnormal union €x- 
presses itself in the development of irregular wings growing out 
from the surface of the perianth. 
It is not difficult to distinguish C, Chitonia from the species just 
described, although the two are rather closely related. C. convex 
stipa is a smaller plant, much more delicate in texture and usually 
of apaler color. It differs further in its less crowded leaves and 
* Hedwigia, 27: 284. 1888. 
rn 
amelie 
