Etjobobenbron ^otietp ^oteg. 
rarely more and often much smaller, and they are oblanceolate, tapering gradually 
from apex to base. Frequently they are deflexed on the branch. The upper 
stratum of underleaf indumentum consists of shallow open bowl-shaped hairs 
with rounded somewhat bell-shaped base on short pluricellular stalks, the cells 
of the walls isodiametric and the margin of the bowl without a fringe of branches. 
It is these open bowls which give the foveolate surface remarked upon by Rehder 
and Wilson. They are deciduous more or less and leave exposed the whitish 
pellicle of under-stratum agglutinate rosette-hairs. In R. Rex we have the same 
kind of cup-hairs, and in R. Hodgsoni and R. regale the form is nearly 
approached, only in them the base of the bowl is somewhat funnel-shaped, not 
bell-shaped. The flower-truss is small for the series though there may be as 
many as 15 flowers. The coroUa is small, about 3 cm. long, and the spotting on 
the rose surface is conspicuous. The filaments of the 14 stamens are most 
minutely puberulous. The ovary is 7-chambered, the chambers wide with 
very narrow septa, and outside the ovary is densely coated with peculiar flesh}'^ 
stalked flock-like few-branched hairs (the branches curhng on themselves) which 
are most distinctive. . The curved capsule is narrow, about 3 cm. long, 6 mm. in 
diameter, sparsely clad with red-brown hairs. 
Forrest got seed of this in 1918, if not before, and we may expect therefore to 
have the plant in cultivation. The young shoots on the dried specimens resemble 
somewhat those of R. niveum. Hook, /., but that species has no close phyletic 
relation. 
Rhododendron decipiens, Lacaita. 
In May 1913 Lacaita collected in flower in Western Sikkim the plant which 
he names thus at an elevation of 10-11,000 feet, adding to his own specimens 
a fruiting one collected later in the year by Ribu for the purpose of his description 
in the Journal of the Linnean Society, of December, 1916. Ffe describes it as 
arborescent with leaves elUptic or obovate-elliptic, at base narrowed or subcordate, 
20-22 cm. long, 9-11 cm. broad, rugose above (but less so than in R. Falconeri) 
ferruginously tomentose below when young, the upper layer detersile leaving a 
pale as it were lepidote indumentum not altogether unhke that of R. Hodgsoni. 
Flower-truss 25-30-fl6wered. Corolla of the form and piuple-rose colour of 
R. Hodgsoni. Staminal filaments pilosulous. Ovary tomentose. Capsule 
4-5 cm. long, 10-13 mm. broad, intermediate in size to the capsules of 
R. Hodgsoni and R. Falconeri. “ On the spot,” he says, “ I assumed this, 
from its colour and general appearance, to be a variant of R. Hodgsoni, amongst 
which it grows promiscuously.” On his return he saw more of R. Falconeri 
in it, and concludes that whilst approaching R. Hodgsoni in foliage the shape of 
the capsule makes it impossible to assign it to a form of R. Hodgsoni, and that 
it would seem to be rather an undescribed variant of R. Falconeri wth purple- 
rose (not flesh-coloured) flowers and smaller capsule. It w^as growing too 
copiously in its home to be regarded as a hybrid of R. Falconeri and 
R. Hodgsoni. 
I have not examined thoroughly Lacaita’s specimens wLich are now in Kew 
Herbarium, and indeed their detailed investigation could be pursued usefully 
only if R. Falconeri and R. Hodgsoni in their several forms were passed in 
210 
