MH. 0. BENTHAM ON' ANOXAOK.F.. 7] 



albidi, toraentosi, 2 lin. diametro. Ovula in earpellis floridis circa 1. 

 Baccse globosas, 1-2-spermae, tomentosae, subsessilcs, 3-5 lin. diametro. 

 In damp woods near Tomo on the Guainia (or Upper Rio Negro, above 

 its junction with the Casiquiare), in Venezuela (Spruce, no. 3549). 



A new species, a congener of Bocagea multiflora, may be de- 

 scribed as follows : — 



Bocagea espintana, Spruce. Foliis ovalibus v. elliptico-oblongis 

 glabris nitidis, floribus subsessilibus glabris, ovulis solitariis erectis. — 

 " Arbuscula 12-18-pedalis, trunco brevi vix brachiali, ligno tenacissimo, 

 ramis adscendentibus subpinnatis." Folia subsessilia v. brevissime 

 petiolata, obtusa v. subacuminata, 25-4-pollicaria,eoriacea, reticulato- 

 venosa. Flores albescentes, solitarii v. gemini. Pedicelli brevissimi, 

 bracteis squamiformibus obteeti. Sepala brevissima, orbiculata, con- 

 cava, glaberrima, duplici serie imbricata. Stamina circa 14, ineurva, 

 ultra loculos dorsales parum producta. Carpella 4-6, glabra, sessilia, 

 stylo capitato parvo. 



In woods near Tarapoto in Eastern Peru, and along the Huallaga 

 (Spruce, no. 4920). The poles used by the Indians for impelling their 

 canoes up the rapids are made of the light tough trunks of this tree 

 (R. Spruce). 



The plant distributed as Bocagea leucodermis, Spruce, remark- 

 able for the white bark of its branches, resembles the B. espintana 

 in its glabrous carpels with a single erect ovule in each. It is 

 described as a tree of 40 feet, with long slender branches. The 

 leaves are oblong, with long acumen, four to five inches long, gla- 

 brous and coriaceous. The inflorescence is that of B. multiflora. 

 I have not seen either the petals or the stamens, and am therefore 

 unable to characterize it more definitely, or absolutely to fix its 

 genus. It was gathered on the river Pacimoni, an affluent of the 

 Casiquiare, and distributed under no. 3352. 



Should the original Bocagea alba and B. viridis prove to have 

 truly valvate petals, they must be associated with Trigyneia canes- 

 cens, as a group intermediate between Trigyneia and the group 

 formed of Bocagea multiflora, B. espintana, and probably B. leuco- 

 dermis, which would in that case receive a new name, and be placed 

 either near Oxandra among TJmriece on account of the a?stivation 

 of its petals, or among Miliusece on account of its stamens. 



The species of Anaxagorea gathered by Mr. Spruce near Barra 

 do Eio Negro in North Brasil, and distributed under the name of 

 A. brevipes, Spruce, no. 1722, does not appear, on further com- 

 parison, to be specifically distinct from the A. acuminata, A. de 

 St.-Hil., which was again gathered by Spruce on the Casiquiare, 



and distributed under no. 3291. 



o 2 



