66 FLORA VITIENSIS. 
carnose ; radicula incurva v. rectiuscula.—Arbores frutices v. suffrutices ; foliis imparipinnatis esti- 
pellatis v. stipellis minutis adnatis; racemis v. paniculis terminalibus; floribus albis flavis v. rarius 
violaceis; bracteis minutis deciduis; bracteolis plerumque nullis.— Edwardsia, Salisb. in Linn. 
Trans. vol. ix. p. 299. t. 26. 
Sophora, as now circumscribed by Bentham, includes Edwardsia, and is represented in tropical Poly- 
nesia by S. chrysophylla, Seem. = Edwardsia chrysophylla, Salisb., from the Sandwich Islands (Menzies! 
Macrae !), S. tetraptera, Linn. = E. microphylla, Salisb., the villose small-leaved form, from Easter Island 
(Captain Cook! in Mus. Brit.), and S. tomentosa, Linn., a widely-diffused seaside shrub. 
l. S. tomentosa, Linn. Spec. 533; fruticosa v. arbuscula, incano-tomentosa; foliolis 11-17 
late ovatis v. orbiculatis, obtusissimis v. retusis subcarnosis, quandoque subsericeis, rarius glabris ; 
racemis terminalibus, laxis; calyce truncato subedentato; corolla pallide flava; legumine indehis- 
cente distincte moniliformi 5—10-spermo.— Benth. Fl. Austr. vol. ii. p. 274; Griseb. Fl. West Ind. 
p.208. S. occidentalis, Linn. Spec. 533. S. Navarensis, Jacq. Amer. 118. t. 173. f. 1. flos. Astra- 
galus lanuginosus, Desc. Fl. vol. vii. t. 375. Nomen vernac. Vitiense, “ Kau ni alewa ^ (Women's 
tree).—Common on the seabeach all over the group (Seemann! n. 130; Storck! n. 886). Also 
collected in Tahiti (Forster! Barclay !), Uvea or Wallis Island (Greffe! n. 38), Botanists’ Island, : 
off New Caledonia (Forster!), and Tongan Islands (U. S. Expl. Exped.). Diffused over the east 
coast of New Holland, the West Indies, South America, tropical Africa, and the East Indies. 
SusonDo II. OZSSALPINIEJE.— Flores plerumque 5-meri. Corolla irregularis v. subirregu- 
laris, petalis imbricatis, superiore interiore. Stamina 10 v. pauciora, omnia libera. 
XXIII. Ceesalpinia, Plum. Gen. t. 9; Endl. Gen. n. 6765. Calyx tubo urceolato, limbi 5- 
partiti laciniis subzqualibus v. antica majore concava. Petala 5, wqualia v. subinzqualia. Sta- 
mina 10, libera; antherz conformes. Ovarium sessile v. stipitatum, 2-ovulatum. Stylus subulatus; | 
stigma minutum. Legumen ovatum, oblongum v. lineari-faleatum, compressum, inerme v. echina- 
tum, coriaceum, 2—]-spermum, 2.valve. Semina ovoidea, globosa v. compressa, exalbuminosa.— 
Arbores v. frutices scandentes, utplurimum aculeatz; foliis 2-pinnatis stipulatis v. exstipulatis; flori- 
bus racemosis v. paniculatis, flavis.— Guilandina, Juss. Gen. 350.; Endl. Gen. n. 6763. 
Bentham finds it impossible to retain the genus Guilandina as distinct from Cesalpinia, and the two 
are therefore united. Besides the two species enumerated below, there is in tropieal Polynesia the widely- 
diffused C. Nuga, Ait., which M‘Gillivray collected in Eromanga, and Lesson in Vanikoro. 
l. C. Bonducella, Flem. Asiat. Research. vol. xi. p. 159; scandens, aculeata villosa v. pubes- 
cens; foliolis oblongis v. obovato-oblongis; stipulis foliaceis deciduis ; leguminibus echinatis; semi- 
nibus griseis.— Guilandina Bonducella, Linn. Spec. 545. G. Bonduc. var. minor, DC. Prodr. vol. ii. 
p. 480; Rumph. Amb. t. 49. f. 1; Lam. Ill. t. 336. G. gemina, Lour. Fl. Cochin. vol. i. p. 345. 
Nomen vernac. Vitiense, *Soni."— Common on the seaside (Seemann! n.132). Collected by 
Forster, probably in New Caledonia (not New Zealand, as (by a misprint?) is stated in his * Prodro- 
mus’). Also found in the Sandwich (Menzies! Macrae! Nuttall !), Marquesas (Barclay !), and 
Society Islands (Barclay! n. 3284). 
The characters which Bentham (Fl. Austr. vol. ii. p. 277) assigns to Guilandina Bonducella and: Bon- 
duc hold good as far as I have been able to judge from the specimens examined. There are at the British 
Museum fruit specimens of both the grey-seeded (G. Bonducella) and yellow-seeded species (G. Bonduc). 
There is also an authentic specimen of G. gemina, Lour., which I have referred to C. Bonducella. 1 do not 
find specimens of Loureiro’s Guilandia Bondue, which, if his description of the fruit can be trusted, would 
seem to be no Cesalpinia at all, having 6 -7 oblong-ovate seeds, whilst Cesalpinia has not more than 2. 
; 2. C. Bonduc, Benth. ; scandens, aculeata, pubescens v. subglabra; foliolis ovatis v. ovato-acu- 
minatis; stipulis nullis; leguminibus echinatis; seminibus flavis.—Guilandina Bonduc, Linn. Spec. 
915. G. Bonduc, var. major, DC. Prodr. vol. ii. p. 480; Rumph. Amb. t. 48. Nomen vernac. 
Vitiense, * Soni,"—Seabeach of Vanua Levu (Williams !), 
` 
