104 ON SOUTH-AMERICAN APOCYNACEZ. 
tube, having a pilose ring in its mouth; segments lanceolate, with dextrorse convolution, 
reddish, almost glabrous, obsoletely pubescent on the median nerve; stamens exserted. 
Gardner's specimen agrees with the characters given by De Candolle and Múller. 
Casasetto's plant probably belongs elsewhere, as it appears to be a twining species. 
27. THYRSANTHUS RUFUS, nob.: Forsteronia rufa, Müll. l. c. p. 100, tab. 31. In Brasilia: v. v. et sicc. 
in herb. meo (n. 4039) ex montibus Organensibus ; v. s. in herb. Mus. Brit. ex eodem loco (Gardner 
530), in herb. meo (n. 2411) sine loco (Ackermann). 
I collected this species in company with Gardner in January 1838. It is tolerably 
well figured by Müller (Z. c.). The branches and branchlets are ferruginously tomen- 
tous, with axils 1-13 in. apart; the leaves are oblong-ovate, obtuse and rounded at the 
base, constricted near the summit into a short subacute acumen, chartaceous, entire, 
green and opake above, with many divaricately ascending nerves arcuately conjoined, 
puberulous on the midrib, paler beneath, pubescent, with yellow prominulent nerves, 
34-32 in. long, 13-13 in. broad, on pubescent petioles 3-4 lines long; the panicles, 
nearly erect, are opposite in the axils, or geminate at the ends of the branches, sub- 
pyramidal, 53 in. long, branching from near the base at intervals of about 4 in., closer 
gradually upwards, lower branches 1 in. long, bare and bracteolated at the base, again 
and again divided, all furnished with narrow acute bracteoles 4 lines long; flowers in 
each branch numerous, subcongested, on pedicels 2 lines long; sepals acute, externally 
pubescent, 13 line long, with few inner bifid scales; corolla 44 lines long, its tube eylin- 
drical, 13 line long; segments acutely oblong, reddish, membranaceous, slightly puberu- 
lous inside, with dextrorse convolution, 3 lines long; stamens exserted; filaments short, 
membranaceous ; anthers 2 lines long, linear, membranaceous at the apex, with 2 obtuse 
basal prongs; disk 5-lobed, as long as the 2 pilose ovaries; style very short, with a 
thickened clavuncle as usual. The fruit is unknown. 
Müller ascribes to this species 2 follicles more than a foot long, strangulated at remote 
intervals into 5 or 6 seminiferous lobes containing several seeds with a long apical 
coma; but these belong to his Forsteronia montana, a plant also from the Organ Moun- 
tains, which he confounded with it, and which has a very similar habit and inflorescence. 
This latter plant is fully described in a subsequent page, under that name. 
28. TuvnsawTHUS sEssILIS, nob.: Echites sessilis, Vell. Fl. Flum. p. 111; Icon. n tab. 85; A. DC. 
L. c. p. 476: Tabernemontana sessilis, Vell. l. c. p. 106: Malouetia sessilis, Müll. 1. c. p. 96. In 
Brasilia : v. s. in herb. Mus. Brit. sine loco (Claussen). 
The two plants recorded by Velloz under the same specific name are evidently iden- 
tical; and Claussen's specimens agree well with them. The floral characters are not those 
of Malouetia, but they accord in all respects with Thyrsanthus. Velloz found one of the 
plants at Cairurü, near Paraty, in the province of Rio de Janeiro, the other near 
Itagoahy, in the same province. It is a scandent species, with dichotomous, flexuous, 
verruculously lenticellate, lactescent branches, the axils being 1-4 in. apart; the leaves 
are elliptic oblong, roundish at the base, and suddenly constrieted upon the petiole, 
shortly acuminate, patently recurved, thinly chartaceous, very glabrous, obscure green 
