332 FLORULA HONGKONGENSIS. 
pearance. They are all found growing together and flowering about 
the same time, in April and May, and might readily be considered as 
mere varieties of one species, were it not for considerable differences in 
the form of the lobes of the corolla, and in the corona or scales which 
crown its mouth, which have even induced Mr. Hance to propose one of 
them as a distinct genus. All three, as well as M. monogynus, are tall 
creepers, with ovate-oblong or oblong-lanceolate, acuminate, smooth 
and shining leaves, and terminal corymbs of numerous more or less 
sweet-scented white flowers, with the throat yellow. The corona con- 
sists of five scales, which are either free or more or less united into a 
five-lobed cup, each scale being entire or more or less bifid. Major 
Champion has also observed a difference in the shape of the fruit, as 
to one species at least, but this point requires further investigation. 
The following are the Hong-Kong species*. 
1. Melodinus fusiformis, Champ., sp. n. ; cymis terminalibus, corollæ 
A lobis oblongis falcatis edentulis, coronze squamis 5 subdistinctis brevi- 
bus latis retusis subbifidisve villosis, bacca fusiformi.—Folia quam 
in M. monogyno breviora, summa interdum subtus leviter puberula. 
Cyme pauciflore, puberule. Zobi calycini obtusiusculi. Corolla 
multo major quam in M. monogyno, leviter odorata. Tubus intus 
supra staminum insertionem pilosiusculus. Corone squamæ nunc 
fere distinctæ, nune breviter in annulum connate. Bacca (ex 
Champ.) magnitudine Mali majoris, fusiformis v. interdum sub- 
pyriformis. 
2. Melodinus Zetus, Champ., sp. n. ; cymis terminalibus, corollæ lobis 
= 395 obovato-faleatis hinc obtuse 1-2-dentatis, corone squamis longius- 
pe cule exsertis semi-bifidis basi leviter connatis (bacca globosa?).— 
Folia omnino M. monogyni. Flores majores, in cymas amplas laxe 
. dispositi. Corona tubum corollæ lineis 13 superans, intus villosula, 
squamis lobisve erectis angustis bifidis. 
* "The following diagnosis, taken from Silhet specimens, will distinguish the true 
_ M. monogynus, Roxb.: cymis terminalibus, corollæ lobis oblongis falcatis edentulis, 
i ra dr rh ee mo ed lanceolatis integris longe pilosis, bacca e 
A. e, it is true, says are two scales i f th 
. making ten in all, but this is evident weite rcp caen 
ly a mistake, for i s 
have seen there are but five as describe by e, for in all the Silhet specim 
x f Roxburgh and figured by Wight. There 
. are indeed occasionally in all these Melodini ose oF two Ml ineo tollo M the 
~ base of the lobes of the corolla above the corona, but they appear to be quite inde- 
pendent and very irregular in their presence, number, or size, and might suggest the 
. propriety of re-uniting the genus Bicorona, A. DC., with Melodinus. 
