52 A FEW CRITICAL, LITTLE KNOWN, 
be depended on ; for, while the lower axillary ones are usually abbre- 
viated and 3-4-flowered, the upper, terminating the branches, have fre- 
quently as many as 12 flowers, and are not in the least umbelliform. 
In fact, many of Dr. Thomson's Western Tibetan specimens of C. 
elliptica, which are referable to the var. integrifolia, are quite as de- 
pauperate in regard to inflorescence as the most marked examples of 
C. pauciflora. The main difference seems to me to be the usually leaf- 
less leaf-opposed (or axillary ?) lower racemes ; but they are not always 
absolutely leafless, and this character may reasonably be attributed to 
their abbreviation. From the above considerations, I believe the plant 
in question must be regarded as a modification of C. elliptica, var. 
integrifolia, which has acquired a peculiar, often pendulous habit, from 
growing in shaded, humid, alpine localities, 
2. Camellia Hongkongensis, Seem. Of this plant an excellent plate 
has been published by Dr. Seemann (Linn. Trans. xxii. t. 60), but I 
infer from his paper, and from Mr. Bentham's description in the 
‘Flora Hongkongensis, that neither of these authors has seen the 
ripe fruit, which Colonel Champion vaguely described as glabrous. 1 
have recently had an opportunity of examining five or six fresh ripe 
capsules, and find them to be spherical, about 22 inches in circum- 
ference, cinnamon-coloured, and densely furfuraceo-scabrous on the 
surface; the seeds are a little larger than those of the Tea-plant. I 
quite agree with Mr. Bentham and Dr. Hooker in reducing Thea, 
even as amended by Seemann, to Camellia. 
3. Sterculia lanceolata, Cay. The seeds of this shrub are occasion- 
ally, though rarely, met with, still enclosed in the brilliant scarlet 
follicles, in the Hongkong markets. They are eaten, roasted or boiled, 
exactly in the manner of the common Chestnut. 
4. Trifolium flavescens, Tineo. This species, which was described 
by Pres! under the name of T. villosum, was afterwards correctly re- 
by him, and also by Savi, to the T, pallidum, W. and K. 
Gussone, however, is unwilling to admit their identity, and writes 
(Flor. Sic. Synops. vol. ii. pt. 1. p. 331), “ Differt a 7. pallido, W. et 
K., habitu magis diffuso, capitulis omnibus sessilibus, corollis semper 
. ochroleucis, leguminibus l-spermis, non 2-spermis, tubi ealycini fauce 
mon prominula.” Ihave made a very careful comparative examina- 
. tion of excellent specimens of T. pallidum, from Istria and the Banat, 
