276 BALFOUR—NEW SPECIES OF RHODODENDRON. 
cells—the stalk shorter in Rh. grande than in Rh. sino-grande— 
from the top of which branch off many broad thin-walled cells 
of varying length, and it is these cells which become interwoven 
and pressed together to form the smooth surface of the under 
side of the leaf. The stalk cells retain a yellowish content, the 
branch cells are empty and uncoloured, hence the shining 
silvery look of the leaf under surface. Rh. grande is one of 
the Himalayan Rhododendrons easily diagnosed by its leaves 
amongst the large-leaved species. The leaves are always 
widest at the middle. They seem not to become obovate. 
The like feature belongs to Rh. sino-grande—which has usually 
broader leaves—except in the case of very large leaves. 
Rh. sino-grande is apparently the Chinese representative in 
South-Western Yunnan of the Himalayan ‘species. But the 
two species are readily separated by the total absence of glands 
from the pedicels and ovary of Rh. sino-grande and _ their 
presence in Rh. grande. The glands in Rh. grande are typical 
clavate glands, reddish or yellowish on longer or shorter stalks, 
and their excretion is extremely sticky. The thin long style 
of Rh. grande projecting the globose stigma from the corolla 
mouth is a familiar feature of the. species. In Rh. sino-grande 
the style is stout, thin, and hardly longer than the corolla 
tube. 
I understand Mr. Forrest has obtained seed of this splendid 
species. Though its home is in the South-West of Yunnan, the 
elevation—10,000-11,000 ft.—at which it grows may give us 
hope that in Britain it may prove to be one of the hardy large- 
leaved species. 
The dried flower bears out Mr. Forrest’s description of it 
attached to his No. 11,875—-white, with basal crimson blotch. 
That he says of the flower of No. goz1 “ crimson or rose? ” 
does not invalidate this, for the specimens are in fruit and no 
fresh flowers were seen by him. 
In the Flora of British India (1882) Mr. Clarke treated Rh. 
argenteum, Hook. f. (1849) as a synonym of Rh. grande, Wight 
(1847), and he gave similar treatment to Rh. longifolium, Nutt. 
(1853). Since that date the name Rh. grande has come into 
general use for all the forms which previously passed under 
Hooker’s and Nuttall’s names. It is in this sense I speak here 
and without prejudice of Rh. grande, Wight. The suppression 
of names did not pass without protest. Mr. J. H. Mangles, 
keenest of observers and unrivalled of growers of Rhodo- 
dendrons, writing in 1884* of “ Rhododendron argenteum and 
Rh. longifolium,” says: “ The Sikkim argenteum and its 
Bhotan ally (yet very distinct) Rh. longifolium are now in 
=. 1 The Garden, xxv (1884), 177. 
