1963] 



BOTANY OF THE GUAYANA HIGHLANDS PART V 



135 



1. Symphonia globulifera L. f . Suppl. 49, 302. 1781. 



Moronobea coccinea Aubl. PI. Guiane 778, in part, and as to Figs, a-j, pi. SIS. 1775. 

 Symphonia globulifera L. f. var. macoubea Vesque in DC. Monogr. 8: 229. 1873. 

 S. globulifera L. f. var. major Diels, Notisbl. 15: 49. 1940. 

 S. microphylla Schultes, Bot. Mus. Leafl. 17: 20. 1955. 

 S. utilissima Schultes, Bot. Mus. Leafl. 17: 22. 1955. 



Type. Habitat in Surinamo. C. G. Dalberg (Linn.). 



Distribution. Widespread in tropical America, normally growing in rain- 

 forest or swampy areas ; Costa Rica and Honduras south to Peru and east to the 

 West Indies, Venezuela, and Guianas and Brazil ; tropical Africa. 



MELASTOMATACEAE 22 



With the exception of several problems of megamonographic scope, this paper 

 completes consideration of the undescribed taxa collected by recent Guayana 

 Highland expeditions. Starting with Dr. Gleason's first large-scale consideration 

 of Guayana Melastomataceae in the Duida report, about 200 novelties in the 

 family have been described for this area. It is regrettable that such a great 

 portion of the Lost Worlds region is still completely unexplored and that even 

 the flora of those general regions which have been visited is still so little known. 

 Certainly for new morphologic variants, no other land portion of the world is 

 today comparable. 



Many of the following conclusions concern problems set aside during several 

 years for more considered judgment ; my two preceding familial installments in 

 the Flora of the Guayana Highland have included the easier taxonomic decisions. 

 The studies would not have been possible without generous loans of critical 

 material from various European and American herbaria (F, G, IAN, K, L, M, 

 MG, NY, P, R, RB, S, U, US, VEN, W) ; their cooperation has been greatly 

 appreciated. The illustrations have been provided by C. C. Clare and M. L. Solt. 



Ernestia pullei Gleason. 



Two Venezuelan collections (Maguire 32697, Maguire, Wurdack, & Bunting 

 35855), from iron-ore mountains (Cerro Bolivar, Cerro Altamira) in Edo. 

 Bolivar, agree fairly well with the holotype (from Surinam) and with Guppy 

 333 (Acarai Mts., British Guiana-Brazil frontier), differing only in the finer leaf 

 serrulation and somewhat more granular lower leaf surface pubescence. In the 

 original description, Gleason described the ovary as 4-celled, mentioning, how- 

 ever, Pulle's 3-celled adnot on the type sheet. The type is in poor condition, but 

 all of the non-disintegrated fruits seem to be 3-celled ; both the Acarai Mts. and 

 Venezuelan collections have predominantly 3-celled ovaries, with only one of the 

 many available fruits on 35855 being 4-celled. I have compared a photograph 

 and fragment (leaf and capsule, NY) of the type of Brachypremna petiolata 

 Gleason with E. pullei and the match is excellent. In his discussion of the new 

 genus, Gleason compared Brachypremna with Ernestia, but erred in differentiat- 

 ing the genera on the bases of flower-mery and habit. Absolute synonymization 

 of Brachypremna pends my personal examination of the holotype; since both 

 species were published in the same paper, no nomenclatural adjustment will 

 be needed. 



22 J. J. Wurdack, Dept. of Botany, U. S. National Museum. 



