EvaANs: TAXILEJEUNEA PTEROGONIA 125 
as it becomes old, is sometimes (as in so many other Lejeuneae) 
raised on astalk representing an elongation of that portion of the 
female branch situated between the involucre and the perianth. 
In the lower part the perianth is terete but in the upper part five 
more or less distinct keels are present, the dorsal keel being some- 
times less developed than the others. These keels may be nothing 
more than angular or rounded projections, but in most cases one 
or more of them will show a more or less distinct wing. These 
wings, in turn, may consist of nothing more than one or two pro- 
jecting and rounded cells, but in their best development, although 
still remaining short and narrow they sometimes show from one 
to three sharp teeth, one or two cells long. Between these two 
extremes are many transitional conditions. In the upper part 
the perianth is truncate but the angles are either sharp or blunt 
according to the nature of the keels; sometimes a perianth will 
be sharp on one side and rounded on the other. The keel is 
slender and rather long, being composed (when well developed) 
_ of four superimposed circles of cells. When the perianths become 
old and especially after the capsules have been extruded, the 
upper part becomes stretched out and lacerated, although it 
never assumes the strikingly companulate form found in 7. 
sulphurea and its alliés. 
A comparison of Fics. 12 and 13 with Fic. 1 and also with 
PLATE I, Fics. 1 and 2, will show how strikingly the vegetative 
organs of 7. densiflora resemble those of T. pterogonia and T. 
jamaicensis. The leaves, to be sure, spread at a somewhat greater 
angle and the auricles of the lobes (not shown in the figures) 
and of the underleaves tend to be less pronounced than in T. 
pterogonia, agreeing better with JT. jamaicensis, but Fic. 17 
shows that sometimes at least the auricles of the underleaves are 
_well developed. There is moreover a tendency for the leaves 
and the divisions of the underleaves to be blunter, and a prolonged 
search is sometimes necessary before a leaf tipped with two 
superimposed cells can be demonstrated. The dentations on 
the leaves, when present, show a similar tendency to be blunt 
rather than sharp. The sinus of the underleaves/is similar in all 
ree species and of about the same length. In T. densiflora, 
however, rounded sinuses are somewhat more frequent and the . 
