I22 BALFOUR—NEW SPECIES OF RHODODENDRON. 
I must point out also that in Rh. manipurense the androecium 
has in numerical symmetry and length of stamens and anthers 
the character of Rh. Maddeni. Hooker gives the number of 
stamens in Rh. Maddeni as 17-20, and they are as long as the 
tube. I find the same number and length in Rh. manipurense. 
In this Rh. manipurense approaches Rh. Maddeni more nearly 
than it does Rh. crassum and its allied forms named above, which 
all have 10-12 stamens. 
I wrote to Sir George Watt about his Manipur rhododen- 
drons, and he has been so good as to send to me the manu- 
script of his descriptions of all the species he met with during 
his journeying in Manipur. The manuscript came to Europe 
with his specimens in the eighties of last century but was not 
published. I give here a transcription of what Sir George 
Watt says about these forms which with reluctance he placed 
under Rh. Maddeni :— 
‘“‘ This is perhaps the most abundant bush on Japvo, Ching 
Sow, and other hills above 7500 feet in altitude. 
“On the young shoots its elliptical lanceolate leaves are 
very much larger and broader than on the flowering branches, 
bright shining green above and green with brown scales, becoming 
almost ferruginous, below. Under the microscope these spots 
are seen to be round peltate scales depressed like minute cups ; 
in the bottom of each is placed a round dark brown shining 
glandular body surrounded by a thin striated pale-coloured 
membrane. Flower-buds round, acute, solid, large, dark red 
and coated with a sticky substance; bracts reniform broader 
than long sometimes suddenly and shortly apiculate and having 
along the centre a band of closely packed scales. 
“Flowers appear in June to August and fruit ripens in 
September. They were described to me as white, and as this 
together with all the other characters agrees admirably with 
Hooker’s figure xviii [in Sikkim Rhododendrons] and descrip- 
tion of Rh. Maddeni, I have most reluctantly had to renounce 
my first impression that this lovely plant was a new species. 
“ Fruit not ro-celled (as described by Hooker) but 12-celled, 
bursting into 4-6 portions of one or more carpels and dehiscing 
from the 12 placental plates which after a time separate up- 
wards from the central axis. : 
“Seeds very large for a rhododendron. Hooker in his 
Rhododendrons of the Sikkim Himalaya states that this species 
is confined to ‘ the inner ranges, very rare, in thickets of the 
Lachen and Lachoong rivers at Choongtam.’ It is therefore 
extremely interesting to find that ‘this truly superb plant’ is 
by no means so rare as it has hitherto been thought, and that, 
although it has taken thirty years to discover the fact, Rh. 
