CHINA. 
183 
pinnata, bijuga cum itnpari: foliola majuscula, petiolulata, oblongo-obovata, coriacea, nitida, glaberrima, 
minute reticulata, obtusa, emarginata, Stipulae obsoletce vel deciduce. Paniculae axillares, terminalesque. 
1. Layia emarginata. (Tab. XXXVIII.) 
Captain Beechey’s Collection contains only the foliage and ripe frnit of this plant ; but we have the good 
fortune to possess from Mr. Millett, specimens in flower, and are thus enabled to describe the plant as a 
genus hitherto unknown to authors. It is indeed probable that the Macrotropis of De Candolle, {Anagyris 
faetida, and A. inodora of Loureiro), may have affinity mth this plant ; but the rounded many-seeded fruit, 
and the “ folia multijugata” of the former, forbid the two to be united. 
Tab. XXXVIII. Layia emarginata. Fig. 1, Flower; Jig. 2, Calyx and Pistil Fig. 3, Legume; 
^g. 4, Seed; ^g. 5, Embryo : — natural size. 
1. Bauhinia variegata. Linn. 
In the specimen before us, the leaves are puberulous beneath, and not glabrous, as described by De 
Candolle; at the same time we think it must behisvar B. Chinensis. Roxburgh, in his Indian Flora, vol. 
2. p. 319, says, that the leaves are “ somewhat villous underneath,” so that they appear to vary considerably 
in that respect.* B. Candida seems a very closely allied species, if, indeed, it be really specifically distinct. 
De Candolle places the two in very different sections, but perhaps the only discrepancies are in the colour 
of the flowers, and the presence or absence of sterile filaments between the five fertile stemens. 
2. Bauhinia retusa. Roxh. Hort. Bengh. p. 31. (non Poir.) Flor. Ind. 2. p. 322. De 
Cand. Prodr. v. 2. p. 515. — B. emarginata. Roxb. MSS. et Icon, in Mus. E. I. C. n. 1237. 
Wall. Cat. of E. Ind. Plants, n. 5792. 
Only two leaves have been collected, but we trust there is no doubt as to the identity of the plant. 
From Roxburgh’s description, it seems to belong to De Candolle’s section, Phanera. 
3. Bauhinia corymbosa ; scandens, ramis teretibus cirrhiferis, foliis basi cordatis subtus 
in nervis petiolis ramulis. calycibusque rufo-pilosis, foliolis semiovalibus obtusis parallelis ad 
medium concretis 2-3-nervibus, corymbis terminalibus sessilibus, staminibus tribus fertilibus 
petala ovata stipitata margme crispa submquantibus, genuine stipitato, stipite tubo calycis 
adnato, leguminibus linearibus 6-12-spermis. — Roxb. Hort. Bengh. p. 31. FI. Ind. 2. p. 
329. De Cand. Prodr. v. 2. p. 515. Leg. Mem. p. 487. t. 70. — B. scandens. Burnt. FI. 
Ind. p. 94. (non Linn.) 
One of the most elegant and delicate of the genus. Roxburgh says of it, “ Stem scarcely any thing 
that deserves the name, but many long slender branches and branchlets climb and spread in every direction, 
to an extent of many fathoms, running over high trees,” &c. De Candolle’s figure and description represents 
the fertile stamens shorter than the petals : in our specimens, they are sometimes longer, but usually about 
the same length. 
4. Bauhinia scandens. Linn. {excl. syn. Rheede.) — B, Lingua. De Cand. Prodr. v. 2. p. 
516. Rumph. Herb. Anib. h.t. 1. 
Of this, only a fragment exists in the Collection, and we have referred it to the B. scandens, chiefly on 
account of the rusty coloured pubescence on the under side of the leaves, notwithstanding that each seg- 
ment has only three nerves. Much confusion prevails about this species, and we shall therefore make no 
apology for transcribing the following description from Rumphius: “ Tenera folia complicata sunt instar libri, 
interne penitus viridia et glabra externe quodvis segmentum per sex virides costas longitudinales distinguitur, 
atque ruffum, et ad tactum instar serici molle est, sine notabili tamen lanugine, et quodammodo splendens : 
seniora folia sese aperiunt seu explicant, suntque superne viridia, inferne gilva: eadem ruffa lanugo in 
petiolis et ramulis supremis observatur, quoque folia sint vetustiora, eo magis inferne glauca sunt.” De 
