46 SISSOA. 
17. DALBERGIA COSOMANLELIANA Prain Journ. As. Soc. Beng, lxx. 2, 60 (1901). 
D. spinosa W. & A. Prodr. i 266 (1834), not of Roxb. . 
An erect glabrous shrub, the ultimate branches bifarious, horizontal, rigid, spinous, 
leaves fasciculate on small tubercles, *75—1°5 in. long; leaflets 7--9, rarely 11, elliptic 
от cuneate-oblong, obtuse or retuse, *25—-°35 in. long, "12--2 in. wide, glabrous on 
both surfaces, pale-green; rachis ‘S—1°2 іп. long, at first puberulous, but soon 
glabrous; petiolules very short, glabrous. Flowers minute, secund, in small recurved 
cymose panicles, fascicled on small tubercles, 27 in. long, “45 in. wide, peduncles 
puberulous, pedicels glabrous; bracteoles caducous, basal lanceolate, epicalycine ovate, 
subacute, embracing lower third of calyx-tube; calyz campanulate, teeth obtuse 
except lowest subacute, one-third as long as tube; corolla white, petals rather 
distinctly clawed; stamens 9, іп one sheath split along upper side or in two lateral 
Май ев or occasionally in two bundles with the obvexillary stamen also free; ovary 
glabrescent, distinctly stipitate, style short; ovules 2—4. Pod indehiscent, thinly 
coriaceous, ovate subacute, distinctly wide-reticulata throughout, base cuneate, rather 
long-stipitate, quite glabrous, l-seeded, 1°75 in. long, “б in. wide; scd large, roniform, 
compressed, * in. long, ‘3 in. wide, brown, hardly shining. | 
INDIA: Coromandelia; Shovagiri Hills, Wight! 
This plant, which appears to be rare, for it has been only twice or thrice collected, and in each 
case by Wight, must be looked on as ono of the doubtful species in this work. That it is not D. 
spinosa Roxb. an examination of the figure of that species will at once demonstrate. Wight and 
Arnott, following Wallich, were themselves uncertain as to the identity of the two, and the unqualified 
acceptance by subsequent writers of Wight’s D. spinosa as identical with Roxburgh’s, was not based on 
fuller knowledge. The affinity is with D. multiflora, and із во close that this may prove to be no 
more than an outlying form of that species, which, though it has no spines on the smaller branches, 
has a spinous main-stem. The size of the leaflets, much smaller in D. coromandeliana than in D. 
multiflora, is not a sufficiently differential feature, nor perhaps is (һө fact that the leaves are fascicled 
Jm the present plant, but not in D. multiflora. Tho pods differ in shape and sizə; but this again, 
having regard to the great variability within 2. multiflora, is not an absolute character. . The 
absence of all tomentum from the pods of D. coromandeliana must, however, be esteemed, at least 
for the present, as diagnostic, if for no other reason than because of its convenience. If, in spite of 
this difference, D. coromandeliana be reduced to D. multiflora, then D. phyllanthoides, which differs 
even less essentially than D. coromandelia does, must also be so reduced; while if p. multifiora, 
D. phyllanthoides and В. coromandeliana are treated as conspecific, it may be necessary to treat some 
of the other Malayan forms with which D. phyllanthoides has been so much confused as integral 
portions of the same protean plant. е 
Рглте 21. Dalbergia coromandeliana Prain,—1 ; Flowering branch, Wight 798 
(K. D. 821), n. s,; 2, bud x 4; 8, calyx, laid open X 4; 4, standard x 4 ; 6, wings 
X 4; 6, keel-petals X 4; 7, ovary x 4; 8, ovary, laid open x 4; 9, ovule x 10 ` 
10, fruiting branch, from Shevagiri Hills, Wight (K. D, 822), n. s.; 1l, seed m.s. 
18. DALBERGIA MELANOXYLON Guill, & Perr. Flor, Seneg, Tent. 227, t. 33 (1834); 
Benth. Journ. Linn. бос, iv. Suppl, 47, (1860); Dalz. & Gibs. Bomb, Flor. 
Suppl. 24 (1861); Prain Journ. As. Soc. Beng. lxvi, 2, 446 (1897); Ixx. 9 
59 (1901); Cooke, Flor. Pres. Bomb. i. 396 (1902). 
D, Stocks. Benth. Journ, Linn. Soc. iv. Suppl, 42 (1860); Bak. in Hook. f, 
Flor. Brit, Ind. ii. 934 (1876); Talbot Bomb. List 75 (1891); Woodr. 
Journ. Domb. Nat, Hist. Soc. xi. 426 (1897). 
