926 REPORT OF THE HARVARD AFRICAN EXPEDITION 
to three feet long, and among and slightly exceeding these are the pedicels 
surmounted by white button-shaped flowering heads. 
A salt water creek penetrates the lower sand belt. At high tide the creek 
overflows its low banks and forms a shallow lake that follows the course of 
the inlet some distance inland. In this locality, as might be expected in the 
tropics, grows the mangrove, Rhizophora racemosa. While the species is abun- 
dant at the mouths of the St. Paul, Du, and other rivers where saline con- 
ditions are prevalent, this locality is of especial interest since it shows the 
mangrove tangle in the process of formation, — from low isolated bushes to 
trees fifteen or twenty feet in height. But it iseven more interesting to find grow- 
ing on the flats that are periodically submerged by brackish water, plants which 
generally belong in habitats where only fresh water is available. Here two 
species of Xyris, one with purplish the other with yellow flowers, and a species 
of Hriocaulon, thrive in varying degrees of abundance. In a pool, shallow 
at low tide a species of Nymphaea, presumably N. lotus also persists. Twining 
among the low grasses and sedge, is a plant that produces clusters of beau- 
tiful sky-blue flowers that resemble those of a Lobelia. When the plants were 
carefully dug up and separated from the grass and mud, they proved to pro- 
duce pockets on the under side of the simple leaves, thus quickly placing the 
species in the genus Utricularia. So far as the writer is aware, such a species 
has not been reported from Liberia. At the edge of the flats, several spiny 
bushes of Ormocarpum verrucosum make themselves evident, not because of 
the flowers which are white and inconspicuous, but because the bushes, ten to 
fifteen feet tall, are made difficult of access by the long twisting and ascending 
branches. 
Because of the unexpected forms encountered, such a region has its fascina- 
tion and it is with regret that, because of the paucity of time, it must be left 
for future exploration. The land again becomes higher, and almost flat, and 
the soil becomes drier and more sandy. Except for a few gardens of cassava, 
and a scattering of palms and other trees, an open landscape is afforded in which 
the silky plumes of the grass Imperata cylindrica, wave in the breezes. An 
occasional termite hill stands out against the background of palms or forest 
trees. Again Anthostemma senegalense, previously seen in the neighborhood 
of Monrovia, grows in isolation. This time, however, the trees although on 
dry sandy soil, reach a height of about thirty-five feet. After going by patches 
of bracken fern, Pleridium aquilinum, a few clusters of low bushes, and scat- 
tered trees of the ‘‘monkey plum,’’ Duport is reached. 
Duport, a village consisting of a house, a storage shed, and five or six mango 
trees, while not in itself interesting, proved to be situated in an interesting 
locality. The sand plain on which it is established, slopes in all directions but 
one, toward a densely-forested swamp, so that all degrees from dry to wet 
sandy soil are present. In the moist pockets of the upper drier level is the tall 
yellow-flowered species of Xyris, X. indica, apparently confined to this special 
habitat. In the lower wetter parts of the plain is another species of Xyris 
with purple flowers, that may, on identification prove to be the same as one 
