THE RHODODENDRONS OE SIKKIM-HIMALAYA. 
271 
extends even to the Polar region?, diminish- 
ing in the size of the species and number as 
we recede from the Himalaya. In North 
America they appear again, though under a 
very different aspect from that they present 
on the sub-tropical mountains of Asia, * * * 
In north-west India the genus Rhododendron 
is first seen on the Kunawar hills, and, ad- 
vancing east, follows the sub-Himalayan range 
for its whole length, the species increasing in 
number as far as Sikkim and Bootan ; thence 
the genus is continued to the Mishmee hills, 
the eastern extremity of the range, crossing 
the Brahmaputra to that lofty range which 
divides the water-shed of the Irawaddi from 
that of the Brahmaputra. Though scarcely 
found throughout this long line of upwards of 
1,200 miles, below 4,000 feet, the Rhododen- 
drons still affect a warm damp climate, where 
the winters are mild. The English naturalist, 
who is only familiar with the comparatively 
small hardy American and European species, 
would scarcely expect this. *-* * R. arbo- 
reum, according to Captain Madden, inhabits 
various localities between 3,000 and 10,000 
feet : this is in Kamaoon, where, of course, the 
genus would descend lowest, and the range is 
incomparably greater than that of any other 
species, at least of those found in Sikkim. 
Dr. Griffith, after extensive wanderings in 
Bootan, gives the limits of the genus in that 
country as between 4,292 and 12,478 feet, 
which is a lower level by 3,000 feet than they 
are known to descend to in Sikkim. In the 
extreme east of Assam, where the Himalaya 
itself diverges, or sends lofty spurs to stem 
the Brahmaputra, on the Phien pass to Ava, 
Rhododrendrons ascend from 5,400 to 12,000 
feet, to the upper limit of arboreous vege- 
tation. 
" Westward again, as far indeed as the west- 
ern termination of the Himalaya, the species 
descend lower than in Bootan : an anomalous 
fact, for which, in our ignorance of the con- 
trasting features which may distinguish the 
Eastern from the Central Himalaya, I can 
only assign conjectural causes. Among these 
may be the proximity of the ocean to the 
Sikkim portion of the range, and the presence 
of heavy mountain masses, covered with win- 
ter, and even perpetual snow, to the south and 
east of the upper extremity of the Brahma- 
putra, whereas the genus is found nearly 2,000 
feet lower than in Sikkim. The descent of 
the snow line in Upper Assam to 14,000 or 
15,000 feet, is no doubt due to the same 
causes, and this is a most remarkable fact. 
Uniformity of temperature, excessive humid- 
ity, and a broken surface, produce the same 
effect here as in the high southern and ant- 
arctic latitudes — favouring the formation of 
snow and its permanence, and also extending 
the range of tropical forms upwards to a 
greater elevation, and the descent of temperate 
or arctic forms to a lower one. * * * 
" Only four species, R. Dalhousice, It. 
Campbellice, It. argenteum, and R. arboreum, 
grow near Darjeeling. The second and fourth 
form scattered bushes at 7,500 and 8,000 feet ; 
the It. argenteum is a small tree, at 8,000 or 
9,000 feet. 
" It was on the ascent of Tonglo, a moun- 
tain on the Nepalese frontier, that I beheld 
the Rhododendrons in all their magnificence 
and luxuriance. At 7,000 feet, where the 
woods were still dense and sub-tropical, min- 
gling with ferns, pothos, peppers, and figs, the 
ground was strewed with the large lily-like 
flowers of It. Dalhousice, dropping from the 
epiphytal plants, or the enormous oaks over- 
head, and mixed with the egg-like flowers of 
a new Magnoliaceous tree, which fall before 
expanding, and diffuse a powerful aromatic 
odour, more strong but far less sweet than that 
of the Rhododendron. So conspicuous were 
these two blossoms, that my rude guides called 
out, ' Here are lilies and eggs, sir, growing 
out of the ground !' No bad comparison. 
[Above this occurs R. arboreum']. Along the 
flat ridges, towards the top, the Yew appears 
with scattered trees of R. argenteum, suc- 
ceeded by R. Campbellice. At the very sum- 
mit, the majority of the wood consists of this 
last species, amongst which, and next in 
abundance, occurs the R. barbatum, with 
here and there, especially on the eastern slopes, 
R. Falconeri. 
" The habits of the species of Rhododen- 
dron differ considerably ; and confined as I was 
to one favourable spot by a deluge of rain, 
I had ample time to observe four of them. 
R. Campbellice, the only one in full flower 
early in May, is the most prevalent. Some 
were a mass of scarlet blossom, displaying a 
sylvan scene of the most gorgeous description. 
Many of their trunks spread from the centre 
thirty or forty feet every way, and together 
form a hemispherical mass often forty yards 
across, and from twenty to twenty -five feet in 
height ! The stems and branches of these 
aged trees, gnarled and rugged, the bark dark 
coloured, and clothed with spongy moss, often 
bend down and touch the ground : the foliage, 
moreover, is scanty, dark green, and far from 
graceful, so that, notwithstanding the gorgeous 
colouring of the blossoms, the trees when out 
of flower, like the Fuchsias of Cape Horn, are 
the gloomy denizens of a most gloomy region. 
R. Campbellice and R. barbatum I observed 
to fringe a little swampy tarn on the summit 
of the mountain — a peculiarly chilly-looking 
small lake, bordered with sphagnum, and half- 
choked with Carices and other sedges : the 
atmosphere was loaded with mist, and the 
