Described as a "cipo climbing in the trees, sends runners along the ground and roots there." The specimen is leafless, but has a still attached cauliflorous panicle at early anthesis. It differs in some details from O. deminuta as described by me: flowers fasciculate by 2-3 (not all paired); petals 6, but the alternate ones very small; androecium 2-merous (!), the filaments free, 0.8 mm long, dorsoventrally compressed, slightly dilated upward, the erect, collateral anther-sacs 0.4 mm long, vertically dehiscent.
In typical O. deminuta the flower has only 3 petals and the androecium is reduced
to a single stamen. I had supposed that this solitary stamen represented a reduced
3-merous synandrium of the type common in sect. Somphoxylon, but it now seems to
be remnant of an androecium composed of free stamens. The appearance of six
rather than only three petals is not considered taxonomically important. The same
variation has been established in O. tripetala (Barneby, 1970, p. 109).
Diels examined and annotated Traill 11, but misidentified it as Somphoxylon
wullschlaegelii, readily separable by its densely papillate staminate panicle and much