Caracas, Julio 29, 1931 
, Sllsworth P. zillip. 
Division of Plants, 
Smithsonian Institution, 
Washington, D.C. 
v *C *»■' V * 
r ur. xvi 
sample, a specimen of a 
rom the State of Palcon, 
With this mail I send to your personal address as 
parasitic plant which was brought me 
where it grows on Tribulus cist cides 
it first, I thought it might oelung to the Balanophcracae 
the dissection showeu, if 
has a calyx consisting of 
the length of the cerola. 
tips. The ten stamens a 
their filaments are very 
rated. The ovary is disc 
the style is columnar and 
It seems consequently tha 
but I have not been able 
thankful to you if you wo 
your first opportunity. 
but 
I have seen well, that each flower 
ive or more lineal segments, about 
This is gamcpetalous ’with ten narrow 
e inserted near the middle of the tube 
short and the anther-celds are sepa- 
- shaped, apparently with five cells; 
short, and it ends in a 5-lobed stigm 
t the plant belongs to the Sympetalae, 
to place it further. I should be very 
uld do better and send me the name at 
I have been working at your list with which I cannot 
always agree. Liy number 12003 cannot be Tibouchina barbigera ; 
its description corresponds with that of T. karste a li with the 
exception of the petals which are much larger in Baer's speci¬ 
mens . 
Both nos. 12865 and 12971 seem to come nearer to 
ii iconia tuberculata . On the whole I think Gleason is not very 
accurate in his determinations. 
1T° 11712 seems to correspond to the description of 
JJectandra pichurim, but the leaves and inflorescences look dis¬ 
tinct from those of n° 12768 which you identified also as IT. pi¬ 
churim. 
M° 12024 cannot be Gccco loba caracas ana according 
to the description. Under that name I Have"874o‘, 7460, 7734, 
12111, 12112, 12328 and Baer's 343. I have been trying to key 
H 
