380 
AMERICAN JOURNAL OF BOTANY 
[Vol. 9, 
(2) 
During several weeks spent in exploring the banks of the Cuyuni River, 
I was considerably puzzled by the fact that, although the vines of M. 
cuyuniensis bore flower buds and fruits in various stages of maturity, there 
appeared to be none that was actually in flower. A more intensive search 
led to the finding of numerous inflorescences from which the calyptriform 
corollas had recently fallen, but still no flowering pedicels were in evidence. 
Finally, an inflorescence was encountered, one evening, which had shed 
only three of its numerous protecting capsules. Early the next morning 
nine more were missing, which demonstrated conclusively that the plant 
had flowered at night. At ii P.M. the next night five of the remaining 
buds were found to be in ''flower," with their pink stamens still attached 
at the base of the pistil. It seemed to be highly improbable, therefore, 
that M. cuyuniejtsis is cross-pollinated by birds, Trochilidae, Coerebidae, 
or Tanagridae. 
(3) 
The question suggested itself, accordingly, are the flowers cross-pollinated 
by moths, bats, or other night-flying animals? In order to throw some 
light upon this point, several inflorescences were entirely enclosed in spe- 
cially prepared cloth bags, and several others were divested of their necta- 
riferous appendages. These inflorescences subsequently flowered and pro- 
duced fruits. But how could the flowers form fruits if the Marcgraviaceae 
are protandrous, as maintained by Delpino and others? Obviously there 
were three possibilities to be considered in this connection: (a) that the 
flowers were cross-pollinated by small insects which worked their way into 
the bags through small openings; (b) that the fruits were abnormal and 
did not contain viable seeds; and (c) that the flowers were self-fertile. 
The first supposition did not appear to be a particularly reasonable one, 
since certain of the inflorescences (both of the bagged and unbagged speci- 
mens) were without their hypothetical attracting organs, i.e., nectaries. 
Furthermore, when certain of the inflorescences were in flower, there were no 
other flowering vines within a radius of several miles. ^ 
A detailed study of numerous inflorescences at different stages of ma- 
turity revealed the fact that dehiscence occurs within the calyptriform 
corolla. The stamens are so arranged (PI. XXIV, fig. 7) that the stigma 
is coated with a thick layer of viscous pollen before the protecting capsule 
falls off. This layer of pollen persists, and may be found adhering to 
the stigmatic surface on old fruits. Stamens examined just after the 
corollas have dropped are found to be more or less completely devoid 
■* The specimens of M. cuyuniensis occurred at infrequent intervals along the widely 
separated banks of the Cuyuni River, so that there were only a comparatively limited 
number of individuals in the vicinity of Kartabo. All these plants were carefully located 
and visited every morning and evening. By observing the deciduous corollas it was 
possible to determine what vines had flowered during a particular period. 
