182 



MEMOIRS OF THE NEW YORK BOTANICAL GARDEN 



| VOL. 10 



C. manoacensis Kranzl. Vierteljahr. Naturf. Gesells. Zurich 76: 157. 1931. 

 C. novemnervia (DC.) Tr. var. depauperata Pittier. Bol. Soc. Venez. Ciene. Nat. 11: 23. 

 1947. 



In a comedy of Kew Rule nomenclature errors, Cogniaux displaced De- 

 Candolle's name to accommodate an earlier Sprengel epithet within Clidemia, 

 thus necessitating- his name C. candolleana for the species based on Martius' 

 Para collection. I have examined the holotype (M) of C. bullosa DC. ; in one bud, 

 there were 15 stamens, indicating alignment with Bentham's species. The species 

 is quite variable both in density and development of upper leaf surface bullae 

 and in hypanthial pubescence; some of the flowers in the Martius collection (2 

 sheets) show sparse glandular setae on the hypanthia. Schultes & Cabrera 19255 

 and 12581, from Amazonian Colombia, are good matches vegetatively and repro- 

 ductively respectively. At present I am inclined to slur over any infraspecific 

 distinctions in this widespread species. For the south Brazilian species [C. bullosa 

 (Spreng.) Cogn.], the next available name is C. biserrata DC. I have examined 

 isotypes for the taxa proposed by Pittier and Kranzlein ; both are pleiostemonous 

 and accord well with Guiana material of C. rariflora. All of the four isotypes 

 (NY, US) of C. manoacensis (Busby <& Squires 239, from the lower Orinoco) have 

 5-merous flowers with ca. 15 stamens; Kranzlein 's vegetative and pubescence 

 details are correct. 



Clidemia japurensis DC. 



Examination of the holotype (M) of Leandra heterobasis (DC.) Cogn. has 

 revealed no difference from the C. naevula element of C. japurensis ; the Martius 

 collection has definitely predominantly 5-merous non-secund flowers and the 

 shorter dense glandular pubescence of C. japurensis as defined by Gleason. Since 

 the types of neither C. japurensis nor C. naevida have been_ personally studied, 

 the necessary varietal combination will not be made at present ; also the correct 

 generic disposition of this widespread but anomalous species is still unsettled. 

 Incidentally, after study of the Martius holotype (M) of Leandra intermedia 

 (DC.) Cogn., an excellent match for that rare species is de la Cruz 2702 (US), 

 from British Guiana ; both the holotype and the recent collection are unfortu- 

 nately fruiting only, but both have predominantly 5-merous fruit with an apically 

 sparsely glandular-puberulous ovary. 



Clidemia sandwithii Wurdack, sp. nov. 



Ex descriptione et photicone C. repenti Tr. affinis sed pubescentia breviore 

 foliorum subtus nervis secundariis evidentibus elevatis. 



Caulis scandens radicans teres petiolique modice setulosus pilis 0.3-0.5 mm 

 longis minute barbellatis primum rubris demura brunneis p.p. primum glandu- 

 liferis glandulis mox deciduis. Folia opposita non vel vix disparilia; petioli 5-9 

 mm graciles ; lamina 2.5-4.5 X 2-3.3 cm late oblongo-ovata apice per 3-7 mm 

 subabrupte hebeti-acuminata basi rotundata vel truncata coriacea nitida integra, 

 supra glabra, subtus in costa basim versus primum sparse setulosa demum 

 glabrata alioqui glabra, modice ciliolata ciliis barbellatis ca. 0.5-0.8 mm longis, 

 3-5-nervata venis primariis supra impressis subtus elevatis nervis secundariis 

 supra obsoletis subtus utrinque 8-10 cum tertiariis laxe elevato-reticulatis. Flores 

 in foliorum axillis singuli 5-meri; pedicelli 4-5 mm longi sparse setulosi ca. 2 

 mm supra basim et ad hypanthii basim bibracteolati ; bracteolae lineari-setaceae 



