FLOKA OE THE ISLAND OF noXGKONG. 387 
The single specimen is only the upper part of the plant, and has entire leaves, but the lowest leaf «hows 
the sub jrato-piunatifid form peculiar to the eauline leaves of this speeies ; moreover Zollinger's short dia- 
gnoais, 1. c., agrees very -n-ell with it. 
294. Blumea Javamca, Zollinger, Flora, 1847, p. hZ\,~BJumea lacera, Dc Cand.-Var h 
Blumet jy^ Cand. Prodr. vol. v. p. 436. n. 19 5. Victoria Peak (Champion, teste cL Benth.; 
Hance!). 
This is a much more villous plant than the preceding, and has much narrower scales of the involucre 
as stated by Zollinger, 1. c. ; it seems to be a good species. 
295. Blumea hieracifolia, De Cand. Prodr. vol. v. p. 442. n. 55. Hongkong (Hance !). 
The scales of the involucre in our specimens are really villous, and not " glabriuscuh-e." as said to be 
by De Candolle. The achienia are appressed, pubescent, almost sericeous ; otherwise De Candolle's short 
diagnosis agrees well. B. hiemcifolia seems to be closely allied to B. elongata, De Cand. Prodr. vol v p 
445. n. 75, but that has glabrous ach^nia. B. Chinensis, Wlprs. Nov. Act. Acad. Cses. Leop. Carol vol 
XIX. suppl. 1. p. 294 (not De Cand.) is also said to be allied to B. Ueracifolia, De Cand., but I have not 
seen it. 
296. Blumea holosmncea, De Cand. Prodr. vol. v. p. 442. n. 56. Hongkong (Champion, teste 
cl. Benth. ; Hance I) . 
The few specimens are only the upper part of the plant. The teeth of the leaves are somewhat rigid 
and pungent, therefore I am rather doubtful whether they belong to B, holosericea, De Cand. ; otherwise 
the brief diagnosis agrees well. Vainly I tried to identify our plant with any species of Dc Candolle's sec- 
tion Osyodonta^. None of the descriptions agree with it. It is however almost impossible to determine 
with certainty the species of such a difficult and by no means satisfactorily defined genus as Blumea, De 
, Cand., without examining a great number of authentic specimens. 
297. Blumea crinita, Arn., Dc Cand. Prodr. vol. vii. p. 283. Hongkong (Hance! Seemann!). 
298. Blumea Chinensis, De Cand. Prodr. vol. v. p. 444. n. 70 (non vol. vii. p. 283. n. 96, et 
Hook, et Arn. Bot. Beech, pp. 195 ct 265, et non Wlprs. Nov. Act. Acad. C^. Leop. Carol, vol. xix. 
suppl. 1. p. 294). Tolerably common, trailing in ravines (Champion, teste cl. Benth.; Seemann!), 
299. Blumea alata, Dc Cand. Prodr. vol. v. p. 448. n. 90. Hongkong (Seemann I). 
The single specimen, in a very young state, precisely agrees with my specimens, collected by Schmidt 
in the Nilgherry fountains. It seems to be closely allied to Bhmea venionioi'des, De Cand. Prodr. vol. v. 
p. 447. n. 87, which I have not seen. 
300. Bl-umea WigJitiana, De Cand. Prodr. vol. v. p. 435. n. 14. Hongkong (Hance I). 
The single specimen is only the foot-long summit of a plant in a young state. The flower-heads are 
two lines long and broad, crowded on the top of a short peduncle, arranged iu short racemes, and gradually 
passing into a dense, obtuse, and leafless ear, au inch in length, compound of crowded, abnost sessile 
flower-heada. The whole plant is hairy, of a somewhat greyish hue, and herbaceous. The upper leaves 
are obovate, an inch long, and almost as broad, irregularly and minutely toothed, wifh somewhat callous 
tips. 
301. Pluchea Indica, Less, in Linnica, 1831, p. 150.— De Cand. Prodr. vol. v. p. 451. n. 7. 
Hongkong (Hance!). 
3 c 2 
