tively small pot.” It will be seen by our figure that the charm 
of this plant is not in the flowers themselves, which have no 
beauty to recommend them, but in the large bracteas or floral 
leaves, which in our living plant are of a full and bright rose- 
colour, and, as the branches are literally loaded with them, the 
effect must be very striking. In respect of colour however, I 
suspect these bracteas are exceedingly variable. In Paxton’s 
plant they are deep purple: so green in one variety, that M. 
Choisy has constituted of it a species (B. virescens), mainly de- 
pending on that circumstance. In some of our specimens they 
are brick-red. Mr. Tweedie, in a note accompanying his speci- 
mens, says, “the trees seem all on fire with them.” That all 
M. Choisy’s zew species are trifling varieties I am satisfied, and 
indeed I have authority for them in my own Herbarium. His 
B. virescens is merely a variety with greenish bracteas (a pecu- 
liarity perhaps due to the plant being much shaded); and in- 
deed Martius’ specimen, n. 64 in my Herbarium, quoted by 
Choisy under B. spectadilis, is this variety, as well as Mrs. 
Graham’s specimen, rightly referred to in my Herbarium as B. 
virescens. Mr, Gardener’s n. 5139 (B. glabra, Chois.) is simply 
a nearly glabrous state of spectadilis. Blanchet’s n. 2573 (B. po- 
macea) is a specimen of spectabilis, with smallish leaves, and 
‘flowers and bracteas not quite developed. 
_ I feel doubts in regard to Bonpland’s B. Peruviana as a dis- 
tinct species. The differences figured and described are but 
trifling, and the author conceives them to be of more weight, 
“car on connait & peine quelques plantes qui croissent a la fois 
au Brésil et au Pérou.” My own specimens from three loca- 
lities, on the western side of South America, lead me to the 
conclusion that they are specifically identical with the Brazilian 
plant ; and I possess specimens from the late Colonel Hall, from 
Zarumille, near Quito (I believe), from Dr. Seemann, collected 
at Quinos, Peru (n. 940), and from Warszewicz, gathered on the, 
Magdalena. The spines in these specimens are straight, or 
slightly curved, or absent. 
Dzscr. The main trunk is quite arborescent; the branches 
numerous, long and subscandent, flexuose, more or less downy 
and spiny: spines varying much in size, and in straightness or 
curvature. Leaves petiolated, ovate, acuminate, entire, varying 
somewhat in shape, more or less acuminated, blunt, or even cor- 
date or acute at the base, varying too in pubescence ; sometimes 
glabrous. Peduncles axillary from near the apices of the very 
numerous branches, shorter or longer than the leaves, solitary, 
simple or forked, each branch bearing three large, coloured, 
membranaceous, reticulated, cordato-ovate, glabrous or downy 
bracteas, costate and veined like the leaves, in our specimen 
