ON SOUTH-AMERICAN APOCYNACER. 93 
21. MarovETiA Panamensis, Müll. in Van Huerck. Pl. Nov. 185; Benth. & Hook. Gen. ii. 707: Stem- 
madenia Guatemalensis, Müll. in Linn. xxx. 410. In Panama et Guatemala: non vidi. 
This species has none of the characters of Stemmadenia, but in all respects agrees with 
Malouetia, as Müller himself indicated. Its branches are terete, blackish red, its branch- 
lets compressed; the leaves are oblong, very acute at both ends, submembranaceous, 
with undulate-repand margins, with 10-12 pairs of divaricate nerves, are near 8 in. long, 
22% in. broad, on petioles 44 lines long; inflorescence terminal, in two opposite cymes; 
peduncle extremely short, almost pulviniform, supporting numerous aggregated flowers 
on stout pedicels 5 lines long, which are as long as the flowers; sepals ovate, obtuse, 
ciliolated on the margins; tube of corolla cylindrical, constricted in the middle, longi- 
tudinally subangular and striate; segments longer than the tube, lanceolate-ovate, fur- 
furaceous above and at their base ; stamens inserted in the contraction of the tube; disk 
cylindrical, of 5 lobes connate at their base, free above, surrounding 2 ovaries, which are 
somewhat shorter, subpuberulous at their apex, and obtusely ovate. 
THYRSANTHUS. 
The three genera Thyrsanthus, Parsonsia and Forsteronia are so much alike in habit, 
kind of inflorescence, and floral structure, that their difference would hardly be recog- 
nized, were it not for the divergence observed in the structure of the fruit and seeds. 
The Parsonsia of R. Brown (excluding the American species) was shown to be a distinct 
genus by De Candolle; but Zhyrsanthus and Forsteronia have been entirely or in part 
confounded together by most botanists, and even the author of the former genus has 
renounced it in favour of Forsteronia (Gen. ii. p. 710). Notwithstanding this, Thyrsan- 
thus appears to me a distinct and tenable genus. It was established by Mr. Bentham in 
1841, upon one of Schomburgk’s Guiana plants, acknowledged in 1844 by De Candolle, 
who described six species, one of which was figured in Delessert's Icones. Miller, how- 
ever, in 1860, cancelled it, absorbing all its species in Forsteronia; and Bentham, in 
1876, forgetful of his former accurate observations, revoked his genus Thyrsanthus in 
favour of Forsteronia. The latter, proposed by Meyer in 1818, was established upon a 
short diagnosis applicable alike to both these genera; but as two of his species belong to 
Thyrsanthus, a genus then unknown, it follows that his third species — floribunda, 
Sw., Parsonsia floribunda, R. Br.) remains the type of Forsteronia. 
Thyrsanthus comes near Malouetia, differing in little else than the character of its 
inflorescence; and there is no great dissimilarity in the structure of its flower; but in 
that of its fruit and seeds there is a wide divergence, which enables us at once to dis- 
tinguish the one from the other. In the absence of this test, there remains the :esti- 
vation of the corolla, which shows a dextrorse convolution in one, and a sinistrorse 
direction in the other. 1 have observed also a peculiar character in the leaves of all the 
species of Thyrsanthus—the presence of a peculiar granular yellow gland at the junction 
of the midrib with the petiole, either single or divided, which does not appear in 
Forsteronia. 
