70 MB. G. BENTHAM ON ANON ACE JE. 



Trigyneia angustifolia. Glabra infloresccntia excepta, foliis elon- 

 gato-oblongis longe acuminatis basi angustatis vix coriaceis, baccis 

 oblongis. — Arbor gracilis, 30-pedalis. Folia 6-8 poll, longa, vix 2 poll, 

 lata, acumine ultrapollicari, venis a costa obliquis utrinque 10-12. 

 Flores albidi, ad nodos dcfoliatos fasciculati, quam in caeteris specie- 

 bus majorcs, petalis exterioribus 4-4 £ lin. longis. Carpella 4-6. Baccrc 

 juniores tomentosa?. 



Tn the gapo of the Rio Uaupes in Venezuela (near the borders of Brazil) 

 (Spruce, no. 256/). 



Trigyneia grandis Glabra infloresccntia excepta, foliis oblique ovali- 

 oblongis acuminatis infra medium sublatioribus coriaceis nitidis, baccis 

 oblongis.— Arbor 20-pcdalis. Folia 6-9 poll, longa, 2i-'3i poll, lata, 

 rigidiora quam in praecedentibus, venis primariis a costa divergentibus 

 utrinque 7-10. Flores ad axillas fasciculati, albidi, circa 3 lin. diametro, 

 pedicellis semipollicaribus. Baccse erassae, 8-9 lin. longae, stipite 3 lin. 

 longo. Semina orbiculata, depressa. 



On the banks of the Lower Casiquiare in Venezuela (Spruce, no. 

 3163). 



It is possible that the Unona lucida, DeC, a Peruvian plant only 

 known from a fruiting specimen in Jussieu's herbarium (figured 

 in Delessert's Icones, vol. v. t. 89), may be a fifth species of Tri- 

 gyneia. 



In the above-mentioned four true species of Trigyneia the sta- 

 mens are closely packed, with the flattened top concealing the 

 cells, as in the majority of Anonacece. The plant distributed from 

 Mr. Spruce's collection as Bocagea canescens has them less close, 

 with an ovate tip, and the cells rather more exposed, nearly as in 

 Bocagea ; but the petals are strictly valvate in each row, as in Tri- 

 gyneia, whilst in Bocagea multiflora, Mart., they are very much 

 imbricate. AVe have no specimens of the two original species of 

 Bocagea (B. alba and B. viridis, St.-Hil.) ; and it does not appear 

 from the published descriptions and figure what is their real aesti- 

 vation; therefore it remains doubtful to which group the name of 

 Bocagea should remain attached. Should St.-Hilaire's species cor- 

 respond with the B. multiflora in aestivation, as they do in other 

 characters, the B. canescens may be transferred to Trigyneia, but 

 as a distinct section, wdth the following character : — 



Trigyneia? canescens. Foliis oblongis breviter acuminatis coriaceis 

 subtus pallidis canescentibusve, pedicellis brevissimis solitariis, sta- 

 minibus circa 12 ultra loculos ovatis, baccis globosis. — Arbor parva, ra- 

 mis horizontalibus. Folia breviter petiolata, 2-4 poll, longa, 1-1^ poll, 

 lata, venis parum conspicuis. Pedicelli axillares, 2-3 lin. longi. Flores 



