94 ' BALFouR—NEW SPECIES OF RHODODENDRON. 
some small. Leaves thick and leathery lanceolate or oblong 
commonly about 5 cm. long 1 cm. broad narrowed to an acute 
point, margin scale-notched, base cuneate or obtuse ; upper 
surface concave the sides of the lamina bent upwards from the 
puberulous midrib, dark green glossy sprinkled with reddish 
distant scales, intervals between the scales much wider than the 
scales ; under surface tawny green or even brown lepidote, with 
many scales partly contiguous partly discontiguous the intervals 
never so wide as the scales. Flowers in a terminal 3-4-flowered 
umbel. Corolla rose-lavender red spotted, lepidote and epilose 
outside. Stamens and style longer than corolla. Style finely 
puberulous at base. 
As in all the Rhododendrons of the Triflorum series, the 
foliage varies according to situation of the plant. Grown in a 
greenhouse instead of outside, Rh. Davidsonianum is a different- 
looking plant. 
Rh. lochmium is most floriferous, and flowers early in spring. 
The corollas of the past flowers hang long upon the inflorescence, 
and, associated with the long persistent thread-like styles, give 
a ragged look to the plant after the first flush of flower. A 
noteworthy feature of the inflorescence is the length of the 
bracteoles. They are much longer than the pedicels and as 
long as the inner bracts, and when expansion of the inflorescence 
bud begins they push out between the bracts, apparently acting 
as wedges for separating the bracts and thus facilitating the 
Opening of the flower-bud. These bracteoles (prophylla) of the 
flower-axis are more or less developed in Rhododendron, differ- 
ing in form and size in the several species, and here as in other 
genera where they are found the question of their use has doubt- 
less occurred to many observers. It is a matter to which little 
attention has been given by botanists. Their name prophyll 
is reminiscent of archetypal hypothesis of their existence as the 
equivalents at the base of each phyton composing the plant of 
the cotyledons appearing on the embryo, which cotyledons must 
therefore be assumed to be foliar, and this is open to dispute. 
Beyond the valid statement that in vegetative parts prophylls 
may be protective—filling up the gaps at the side of the leaf- 
relation to flower-expansion, which is so evident in Rh. lochmium, 
will perhaps be found to be one of wide occurrence, it is so 
in the genus Rhododendron, and deserves to be investigated 
in other plants. ; 
