376 
AMERICAN JOURNAL OF BOTANY 
[Vol. 9. 
tions to insure cross-pollination by birds, is so plausible and appears to 
afford a satisfactory explanation for so many closely coordinated phenomena 
that it is not surprising that it should have been accepted without question 
by Hermann Miiller (1873), Schimper (1898),^ and others; particularly 
in view of the fact that the Marcgraviaceae are stated to be protandrous. 
Descriptive and Taxonomic 
The two species of Marcgravia growing in the vicinity of the Kartabo 
laboratory are characterized, as are other representatives of the genus, 
by having two distinct types of branches: (i) sterile runners (plagiotropic), 
and (2) pendulous fertile shoots (orthotropic) which bear terminal inflo- 
rescences. The former are provided with numerous clasping roots and 
small distichous leaves, whereas the latter have large distichous leaves. 
The plants scramble over the trunks and lower branches of trees which 
line the banks of the larger water courses. They are not lofty climbers, 
rarely rising more than fifteen or twenty feet above the surface of the 
rivers. Indeed, at high water certain of the inflorescences and clusters 
of fruit are completely submerged. 
The two species may readily be distinguished, even at a considerable 
distance, by conspicuous differences in color and habit of growth. In 
one species, which was designated in the field as A, the fertile shoots are 
relatively infrequent, relatively long, and bear dark green leaves which 
are so oriented that their upper surfaces are clearly visible (PI. XXIII, 
figs. I, 3). In the other species, B, the flower-bearing shoots are more 
numerous, shorter, and have yellowish-green leaves that are nearly hori- 
zontal and are considerably folded dorsally (fig. 4). Upon closer inspection 
the broadly oblong-elliptical leaves of species A are found to be strongly 
petiolate and to be inserted upon a zigzaged axis (fig. i), whereas those 
of species B, which are smaller and narrower, are nearly sessile and are 
borne on a straight shoot. Furthermore, the yellowish or brownish-green 
inflorescences of the latter species are attached close to the last leaf, whereas 
in the former species the dark, greenish-purple inflorescences form the 
terminus of a long, straight, terete peduncle.^ As shown in text figures 
2-5, the flower buds, fruits, pedicels, and nectaries of species B are larger 
and stouter than homologous members of the inflorescences of species A. 
The vines of the former bear inflorescences and fruits in various stages 
of maturity, whereas the fertile shoots of species A, at any given time, 
are all in equivalent stages of differentiation. In other words, species B 
flowers continuously, but species A exhibits a marked floral periodicity. 
2 Schimper, in quoting the above passage from Belt, substitutes Marcgravia umbellata 
for Marcgravia nepenthoides. He also figures the inflorescence of the former instead of 
the latter species. 
3 During the earlier stages in the development of the flowering shoots of species A, 
the peduncle may be provided with numerous small glandular leaves. Most of these 
more or less rudimentary leaves drop off before the inflorescences attain any considerable 
size. 
