164 ADDITIONS TO THE FLOBA OF CEYLON. 



referred by Thwaites to G. carinata Wall., with which, however, it 

 does not completely agree, and Hooker suggests that it may be a 

 form of G. coronaria Ham. The calyx-limb is angled, and truncate 

 at the mouth ; the corolla-tube 3-4 in. long, and the limb 7-lobed ; 

 and the fruit distinctly 6 -ribbed, with three placentas. 



Glossogyne pinnatifidaDG. On rocky ground in several places 

 about Nilgala and Ekiriankumbura, in January, 1888, mostly past 

 flowering. An easily overlooked plant, which probably occurs else- 

 where in the dry region. It is found in many j)arts of Peninsular 

 India. There is no published figure of this little Composite. 



Holostemma Rheedii Wall. It is remarkable that this beautiful 

 and conspicuous Asclepiad should have remained undiscovered till 

 now, and that it has been overlooked shows how slightly the more 

 remote districts of Ceylon have been botanically examined. I found 

 it frequent, climbing over bushes in the neighbourhood of Nilgala, 

 Bibile, &c., in Jan. 1888, and have introduced it to the Botanic 

 Gardens at Peradeniya. The species seems to have a wide range in 

 Peninsular India, and is also found in Burma. 



Ceropefjia Decaisneana Wight. I collected this in Sept. 1888, in 

 the forest-covered mountains of Bangala district, at about 4000 ft. 

 It has by far the largest flowers of any species in Ceylon. Wight 

 found this near Sispara, Nilgiris, and it is figured in his Icones, 

 t. 1259. 



Ceropegia parviflora, n. sp. Whole plant glabrous and 

 pale yellowish-green; leaves ovate, the lower ones broadly so, 

 rounded at base, the lower ones subcordate, tapering and very acute 

 at apex, thin; flow^ers small, in few-flowered umbels; calyx-lobes 

 linear-setaceous, glabrous; corolla | in. long, all pale yellow; tube 

 moderately inflated in lower part, nearly straight above, and not 

 dilated at mouth ; lobes very short, about one -sixth the length of 

 tube, oblong-lanceolate, not produced into tails, united at the tips 

 to form a short blunt cone ; coronal lobes linear, ciliate at base ; 

 processes long, spathulate, glabrous ; follicles 5-6 in. long, linear. — 

 Climbing over bushes at Anuradhapura, N. Cent. Province, Feb. 1888. 

 This species is allied to C. intermedia Wight (which is almost 

 certainly C. biflora L.), but differs conspicuously in its very small 

 yellow flowers, with very short tailless lobes, and linear coronal 

 lobes, as well as in its broader leaves. The small flowers suggest 

 C. bidbosa, which is, however, abundantly distinct in other respects. 



Wrifjhtia fiavido -rosea Trim. I am now able to add to my 

 description of this (Journ. Bot. 1885, 238) the characters of the 

 fruit, obtained from the same trees in Sept. 1888, very nearly 

 mature : — Follicles connate into a single fruit, 8 or 9 in. long, 

 stout, near | in. diameter, cylindrical, sharply pointed, glabrous, 

 dark green (black when dry) ; seeds very numerous, with a very 

 long coma. The connate stout follicles completely separate this 

 species from W. tinctoria Br., and connect it with W. tomentosa 

 R. & S. As in that species, the follicles separate when ripe, and 

 remain attached to one another for a while by their points only. 



Exacum petiolare Griseb. This rather ill-defined species was 

 recorded for Ceylon by C. B. Clarke (Journ. Linn. Soc. xiv. 427) 



