1958] 



THE BOTANY OF THE GUAYANA HIGHLAND PART III 



125 



stem dense and spreading as in C. hirtellum, but the leaves obtuse at the base 

 and with longer petioles as in C. pubifiorurn ; the peduncles are shorter as in 

 C. pubiflorum. Perhaps the two are not specifically distinct. 



Couma rigida Muell.-Arg. 



VENEZUELA: Maguire 4- Politi 28098, Cerro Sipapo (Paraque), Territorio Amazonas, 

 forest along lower Cafio Negro, summit of Cerro Sipapo, 1400 m alt., Jan 1, 1949, small tree. 

 Tate 732, summit of Mount Duida, dry slope of Savanna Hills, alt. 4400 ft, Aug 1928-April 

 1929, small tree, flowers red, juice milky. J. A. Steyermark 58243, summit of Serro Duida, 

 Savanna Hills, alt. 1025-1200 m, Sept 2, 1944 (MO.). BKITISH GUIANA: Maguire # Fan- 

 shawe 23232, Kaieteur Plateau, rare to occasional, bush island on savanna, May 5, 1944, tree 

 10-12 m high, 25 cm diam., copious white latex, fruit subglobose, green, not quite ripe, seeds 

 flat, oval, testa brown. 



The species were originally described from Bahia, Brazil. Its appearance in 

 the Guayana Highlands is most surprising. There is no record of its collection 

 in the vast intervening area. The collections from Mt. Duida {Tate 732, Steyer- 

 mark 58243) are erroneously cited as C. utilis in Bull. Torrey Club 58: 298, 453 

 (1931) and Fieldiana Bot. 28: 1043 (1957). The Ttate collection was discussed 

 by Monachino in Lloydia 6: 239 (1943). Steyermark 58256, cited together with 

 58243 under C. utilis in Fieldiana, is probably the same, but the specimen was 

 not seen by me. Steyermark 58243 contains flowers; the flowers are crowded in 

 the inflorescence and the stigma-lobes are about 1 mm long, as previously de- 

 scribed for the topotypical material. The calyx is puberulent. If the plants from 

 Bahia and Guayana are not precisely identical, they certainly are very closely 

 allied. 



Forsteronia gracilis (Benth.) Muell.-Arg., var. now ? 



VENEZUELA: Amazonas: B. Maguire 4' C. K. Maguire 35515, Serrania Yutaje, Rio 

 Manapiare, occasional at streamside near Base Camp, elev. 175 m, March 3, 1953, climbing 

 woody vine, flowers white, fragrant. BRAZIL: Estado do Amazonas: J. Murca Pires 60, 

 Maues, praia alagavel do rio, Nov 30, 1946, cipo, flores brancas, folhas triverticiladas. 



These specimens differ from typical F. gracilis in the anthers which are 

 glabrous or rarely with a few dorsal hairs, instead of barbellate, and in the 

 corolla-lobes which are much more densely pilose and with longer hairs within. 

 The leaves may be opposite or ternate. It has not been possible to see a specimen 

 of F. pilosa (Veil.) Muell.-Arg., known from Bahia, Rio de Janeiro, and Minas 

 Gerais; this species is related to F. gracilis and distinguished as having "corolla- 

 lobes conspicuously pilosulose within, anther tips glabrous or essentially so." 

 F. gracilis previously has been reported from British Guiana (many citations) 

 and only in the vicinity of Manaos in Brazil. Ducke 1890 from Esperanca, Ama- 

 zonas, Brazil, has typically barbellate anthers. Krukoff 8534 from Municip. Sao 

 Paulo de Olivenca, near Palmares, Brazil, and Buchtien 1896 from Mapiri, Bo- 

 livia, are probably also F. gracilis, but the flowers are too young for proper study. 



Forsteronia guyanensis Muell.-Arg. 



VENEZUELA: WurdacTc # Monachino 39733, Serrania Imataca, Territorio Delta Ama- 

 curo, occasional in forest 1-3 km north of Rio Guanamo, El Palmar-Raudal Trail, upper Rio 

 Toro drainage, elev. 270-470 m, Nov 23, 1955, vine climbing 20 m in trees, flowers white. 



This collection has a corymbose inflorescence, as described for F. brevifolia 

 Mgf. by Woodson in the key to the species of Forsteronia (Ann. Missouri Bot, 

 Gard. 22: 155. 1935). It seems, however, that F. brevifolia should be regarded 

 as merely a' synonym of F. guyanensis. The type was collected along the Tocan- 



