COLLECTED IN HONG-KONG. 313 
sata sub-biloba instruetz, staminibus alterne, iisdem duplo fere lon- 
giores. Ovarium vertice breviter exserto inferum, biloculare. Ovuda 
in loculis solitaria, pendula, anatropa. Styli 2, distincti, longissime 
exserti, divaricati, olii facie externa striati, interna stigmatosi, 31- 
lin. longi. Stigmata valdè incrassata, complanata, irregulariter lobata, 
carunculoso-stigmatosa, diio. Capsula seminifera, obovata, lig- 
noso-coriacea, 6 lin. longa, bilocularis, apice bivalvis, valvis dorso tan- ` 
dem fissis, endocarpio corneo, in coccós duos bivalves soluto. Semina 
in loculis solitaria, pendula. 
This plant was sent to me by Capt. Champion as a new species 
of Hamamelis ; but an examination of its characters soon led to the 
conclusion that, though more nearly related to that genus than any 
other belonging to the tribe, it has sufficient peculiarities to distinguish 
it as the type of a new one. These may be briefly stated as follows :— 
Pentamerous flowers, no corolla, extrorse anthers, and that excessive 
vri ipe being the only individual belonging to the Order in 
which it e 
Bëra in ee course of my examination of this genus, been led to 
make some investigations into the structure of the genera which at 
present compose the natural Order Hamamelidacez, and the affinities 
which they bear to other Orders, I shall in this place offer the result 
of these inquiries ; the more particularly as I have to propose the union 
of Hamamelidacee, Altingiacee, Helwingiacee, and Bruniacee mto one 
group. 
Until the year 1818, when Robert Brown established the Order 
Hamamelidacee, the genus Hamamelis stood among the genera related 
to Berberidee, where it had been placed by Jussieu. At the same time 
Brown pointed out the affinity which exists between Hamamelis and 
the Cape genera Brunia, Stavia, &c., and suggested that they should 
form a distinct but nearly allied Order. At that period the characters 
which distinguish these two groups were more salient than they are at 
present ; the discovery of new genera in this, as in many other instances 
in the system, having so reduced them, that habit alone remains as a 
mark whereby to distinguish them. This must not, however, be taken 
into our calculations unless supported by more solid characters ; else 
the tiny heaths of Europe would not stand side by side with.the lofty 
2s 
