OF BRAZILIAN RUBIACEJX. 931 
placentis bifidis pendulis ultra medium dissepimento adnatis. Ovula 
numerosissima, sursum imbricata. Capsula septicide dehiscens, 
coccis subintegris intus hiantibus. Semina elliptica,- compressa, 
apice et basi in alam lanceolatam integram sepe inflexam v. diffor- 
mem expansa. Embryo longitudine fere seminis, cotyledonibus ova- 
tis, radicula supera brevioribus. 
Species unica: C. Spruceana, Benth.—Arbor undique (ovario excepto) 
glabra. Ramuli teretes v. subcompressi, leves. Stipule caducis- 
simæ cicatricem relinquunt annularem, parce at longiuscule ciliatum. 
Folia ovali-elliptica, breviter et obtuse acuminata, basi angustata, 4— 
6 poll. longa, 2-3 poll. lata, membranacea, utrinque viridia, penni- 
nervia et eleganter transversim venulosa, axillis venarum subtus seepe 
barbellatis, petiolo 1—1-pollicari. Coryméus terminalis, fastigiatus, 
trichotomus, basi foliatus, ramis glabris compressis apice dense di- 
chotome cymiferis. Bractee ovate, membranacez, coloratze (albæ ?), 
fere petaloidex, 4—5 lin. longe, per paria connate, cymas generales 
partialesque necnon flores ultimos laterales ante anthesin involventes. 
Flores alares, sessiles, ebracteolati, nunc sessiles, preeter bracteas cir- 
cumvallantes ebracteolati, nunc breviter pedicellati et minute bibrac- 
teolati. Ovarium dense sericeo-villosum, 2 lin. longum. Corolla 
preter villos faucis glaberrima, tubo 14 lin. longo, lobis late oblongis 
tubo paullo longioribus. Capsula oblonga, a latere compressiuscula, 
4—5 lin. longa, villosa. 
Found in the islands at the junction of the rivers Tapajóz and Ama- 
zon, by Mr. Spruce, who states it to be a tree of 50 to 70 feet, with a 
brown polished trunk, the bark being deciduous, as in the Plane-tree. 
It is often branched from near the base, and known under the Portu- 
guese name of Pao Mulatto. 
Gompnosia, Weddell. 
(Trib. Cinchonee.—Aspidanthera, Benth.) 
In vol. iii. of ‘Hooker’s Journal of Botany’ I published, under the 
name of Aspidanthera, a genus which, in the absence of the fruit, I 
placed among Rondeletiee. This mistake probably prevented Mr. 
Weddell from recognizing it as identical with his genus Gomphosia, 
described in his splendid as well as excellent monograph of Cinchona. 
As his name is so much better adapted to the genus, as well as shorter 
than mine, I trust I may be allowed to avail myself of my own inaccu- 
