172 



MEMOIRS OF THE NEW YORK BOTANICAL GARDEN 



[vol. 10 



collection with "Miconia Duckei Hub. n. sp." on the label; the epithet was 

 penciled out. I believe this sheet to be the holotype of M. arirambae, not re- 

 annotated by Huber after he discovered Cogniaux' prior use of the epithet. The 

 species seems to be fairly common in Venezuelan Guayana, with specimens from 

 Bolivar (Aristeguieta 2219, Steyermark 59204, Wurdack 34373) and Amazonas 

 {Maguire & Maguire 35058, Cowan & Wurdack 31490-A), as well as from British 

 Guiana {Maguire & Fanshawe 32182) and the lower Trombetas and Rio Maro 

 in Brazil (Ducke 10957, Fires & Silva 4387). However, I can find no differences 

 in minute details from the isotype (NY) of M. albo-rufescens Naud. Unfortu- 

 nately the sheet of Claussen 154 has only young fruit and no flowers. Almost 

 certainly, a comparison of flowering material from Minas Gerais with the excel- 

 lent Guayana collections will establish this synonymy. 



Miconia wittii Ule. 



Unfortunately, the available isotype (MG) of this species shows only very 

 young buds. After comparison of foliar and inflorescence details, several recent 

 Brazilian, Colombian, and Venezuelan col led ions seem at least conspecific with 

 Ule's Manaos material. The leaf blades vary considerably, from narrowly oblong- 

 ovate to elliptical with shortly acuminate to bluntly acute apices, and from 3- to 

 indistinctly 5-nerved ; the recent specimens have generally smaller leaf blades 

 than the type. The anther connectives are clearly prolonged 0.7-0.9 mm basally 

 and with either a cordate dorso-ventral appendage or a blunt dorsal tooth ; I 

 cannot interpret if this accords with Pilger's (as M. micrantha Pilger) "antheris 

 . . . basi truncatis" statement. The general vegetative features of M. wittii 

 resemble those of the M. elaeagnoides Cogn.-Jf. dichrophylla Macbr. complex 

 discussed by Gleason (Bull. Torrey Club 58: 242. 1931), but the flowers are not 

 secund as in that southwest Amazon basin species. M. wittii is also suggestive of 

 the rarely collected M. elata (Sw.) DC. ; that West Indian sp'ecies differs at least 

 in the tetragonal young branches, undulate-denticulate leaf margins, and less 

 appressed lower leaf surface pubescence. Until topotypical material proves other- 

 wise, the following are referred to M. wittii: 



BRAZIL, Amazonas : Rio Icana, Sao Felipe, Froes 27884, 27892. COLOMBIA, 

 Vaupes: Soratama, Rio Apaporis, Schultes & Cabrera 12555. VENEZUELA, 

 Amazonas: Cano San Miguel, Rio Guainia, Wurdack & Adderley 43247; east 

 bank of Rio Casiquiare 300 m below mouth of Rio Pacimoni, Maguire & Wurdack 

 357,27>Savanna No. 1 at northwest base of Cerro Yapacana, Maguire & Wurdack 

 34496; Cano Cupaven, Rio Orinoco, Wurdack & Adderley 42807. 



Miconia perturbata Wurdack, sp. nov. 



Sect. Glossocentrum. M. tetragonae Cogn. affinis sed foliis breviter hebeti- 

 acuminatis hypanthiis amplioribus distincte lobulatis staminis minoris connectivo 

 basi ventraliter appendiculato. 



Arbor 5 m. Rami primum hebeti-quadrangulares demum tereti foliorum venis 

 primariis subtus petiolis infjorescentiis hypanthiisque primum sparse stellulato- 

 puberuli mox glabrati. Petioli 1-2.5 cm ; lamina 8-14 X 5-7 cm elliptica apice 

 breviter (5-7 mm) hebeti-acuminata basi late acuta vel obtusa tenuiter coriacea 

 integra eciliata 5-nervata jugo exteriori tenui nervulis laxe reticulatis reticulis 

 irregularibus ca. 2-3 mm latis, supra et subtus in superficie glabra. Panicula 

 8-12 cm longa lataque ramis primariis secundariisque plerumque 4 (-6) -verticil- 



