BALFOUR—NEW SPECIES OF RHODODENDRON. 305 
The foliage-buds are all small, more or less oblong, with two 
or three small thick rusty lepidote scale-leaves outside and 
spathulate thinner ones within. The scale-leaves fall off at once 
on opening of the bud. None of them persist. The ptyxis of the 
leaves in the bud is involute or they may be plane, and the 
vernation is correspondingly implicate or complanate. This 
character in the bud is a tribe distinction of all these small- 
leaved Rhododendrons, marking them off from the whole of the 
large-leaved Rhododendrons, which have revolute ptyxis and a 
curious valvate vernation of the leaves within a hollow chamber 
formed by the numerous scale-leaves of the bud. 
The flowers are produced at the end of shoots either as 
solitary terminal flowers with hardly any stalk, or in groups of 
2-7 flowers equally and short-stalked. Where there are several 
flowers at the end of a twig they belong to one truss; there are 
never lateral accessory trusses or flowers. This is important, 
separating the group from Rh. dauricum, Linn. and forms 
within its orbit. There may be considerable variation in the 
number of flowers in a truss in the same species, but apparently 
some species have constantly a solitary terminal flower, e.g., 
the whole of the species in sub-series B (see p. 309).. 
In all cases the flower-bud is more or less globose, the outer- 
most sterile bracts very small ovate and lepidote, the ones within 
more or less rounded and hooded, mucronate or apiculate or not, 
lepidote outside and ciliate or lanate at margin and top. Varia- 
tions there are, and investigation must determine whether any 
point of diagnostic value is to be found in these sterile bracts. 
They are always more or less persistent until flowering is well 
over. The innermost spathulate more or less hairy bracts are 
also wonderfully uniform, as are also the thread-like bracteoles. 
usually longer than the pedicels, which are commonly lepidote, 
often reddened. 
In the flowers themselves the calyx is always cut to near the 
base into five lobes which are equal or unequal. If the latter, 
the postero-lateral are the larger. There is much variation in 
the size of the lobes. Sometimes the smaller are mere points. 
The larger may be membranous plates 5 millimeters long. 
They are green or reddened, may be lepidote on the back and 
margin, and ciliate or lanate atthemargin. As Rehderand Wilson 
have pointed out, forms like these Lapponicum Rhododendrons 
with sterile bracts remaining during flowering are apt to show 
divergence within a species in the calyx. One may recognise 
its oneness for the group, and that certain species have generally 
say large calyx lobes, others have small ones, but we do not yet 
know the limits of specific character in the calyx. 
It is otherwise with the corolla. It shows a short funnel- 
