Leguminosce^ 
CHINA. 
181 
that the racemes of flowers are terminal as well as axillary. We ])lace little dependence on the slightly 
angled stems: indeed, if Dr. Wallich he correct in uniting the Hedysarum piirpureum and Desm. angulatum, 
we must almost suspect some error in the description given by De Candolle, as we have never observed the 
stems more angled than is usual in most species of the genus, where it is owing to the petioles being slightly 
decurrent. It is more than probable that several other species ought to be joined with it. 
1. Lespedeza Chinensis ; erecta, petiolis brevibus, foliolis oblongo-obovatis retusis mucro- 
natis subtus strigoso-pubescentibus, racemis brevissimis breviter pedunciilatis, legurainibus 
laevibus calyce aequalibiis. G. Don, in Mill. Diet. v. 2. p. 307. 
We cannot but consider this plant as distinct from Linneeus’ Hedysarumjunceum, {Lespedeza juncea,Vevs. 
and De Cand.) which has the leaves linear and cuneate; as it assuredly is from the L. jimcea of Dr. Wallich, 
Cat. n. 5743, of which we possess specimens from that munificent Naturalist. Another and Siberian species 
has been named L. Pallasii by Mr. G. Don, in his edition of Miller’s Dictionary, although he does not appear 
to be aware that it has been described and figured by Pallas, (Itin. App. 8. p. 394. t. 81. f. 4,) under the name 
of Trifolium? liedysaroides, and has actually been introduced into the genus Trifolium, in De Candolle’s Pro- 
dromus, v. 2. p. 204. n. 114. 
1. Abrus precatorius. Linn. 
1. Rhyiichosia ; caule lierbaceo volubili villoso tereti, foliolis subrotundo-ovatis 
supra molliter pubescentibus subtus velutino-tomentosis, racemis axillaribus folio brevioribus, 
leguminibus ovatis velutinis dispermis. — Low'. Cochin. 2. p. 562. De Cand. Prod.r. v. 
2. p. 385 ? 
We think there is no doubt of our plant being that of Loureiro; although De Candolle, who examined an 
original specimen, says that the racemes are longer than the leaves ; it is probable, however, that he only 
saw specimens in fruit, in which state the racemes might be elongated, although in the specimens we possess, 
with fully formed legumes, the racemes are still shorter than the leaf. 
1. Soja hispida. Mcench. 
1. Cajanus bicolor. De Cand. Prodr. v. 2. p. 406. Bheede, Hort. Mai. 6. t. 13. 
The specimen before us agrees exactly with what we possess from Dr. Wallich under the same name; 
but we cannot perceive any constant mark to allow of its separation from C.flavus, the number of seeds being 
very apt to vary. 
1. Pongamia Chinensis; foliolis 2-3-jugis ovato-oblongis acutis glabris, racemis axillaribus 
folio duplo triplove brevioribus, pedicellis verticillatis calycem subaequantibus ante anthesin 
recurvis. De Cand. Prodr. v. 2. p. 416. — Robinia mitis. Lour. Cochin. 2. p. 556. 
We believe this to be Loureiro’s Robinia mitis; and as he mentions that it is a shrub only three feet high, 
while the P. glabra is a lofty tree, we have been induced to give a character that may more effectually 
separate the two. 
1. Inga dulcis ; spinis stipularibus brevissimis rectis, foliis bigeminis, foliolis subdimiato- 
oblongis obtusis subretusis cum mucronulo, glandula in petioli dicliotomia et inter foliola, 
petiolo birto foliolis breviore, florum capitulis globosis racemosis, leguminibus tortis. — 
Willd. Sp. PI. V. 4. p. 1005. De Cand. Prodr. v. 2. p. 436. — Mimosa dulcis. Roxb. Cor. 
1. t. 99. 
2. Inga dimidiata; ramis angulatis, foliis bipinnatis pinnis 4-jugis, pinnulis inferioribus 4- 
5-superioribus sub 9-jugis, glandula ad basin petioli tetragoni communis et inter omnes pinnas 
