302 BALFOUR—NEW SPECIES OF RHODODENDRON. 
by a large number of the species, e.g.—kRh. fastigiatum, Rh. 
impeditum, Rh. scintillans. I note that many of this type 
have the discontiguous under-leaf scales. A most graceful 
growth form is that shown by such plants as Rh. diacritum, Rh. 
telmateium, Rh. thymifolium, which have very thin erect twigs 
bearing rather small leaves, and these have punctulate under- 
leaf indumentum. 
In relation to cultivation of Rhododendrons in our gardens 
the series has special interest, for we have, according to the 
collectors, species which grow in clefts of limestone cliffs—such 
are Rh. cuneatum, Rh. pyvcnocladum, and Rh. rupicolum. Diverse 
from these in nidus are Rh. hippophaeotdes (boggy peaty pasture), 
Rh. impeditum (open peaty pasture), Rh. scintillans (open marshy 
pasture), and RA. telmateium (open boggy situations), and then 
Rh. diacritum is said to come from humus-covered boulders. 
Here, then, are species whose native habitat is definitely recorded 
—limestone in one set, peat and marsh in the other—and as 
Rh. cuneatum and Rh. rupicolum of the lime plants and Rh. 
hippophaeoides, Rh. impeditum, and Rh. scintillans of the moist 
peat plants are in cultivation, comparative experiment upon their 
growth in relation to soil conditions is possible, and may throw 
some light upon the problem of “ Rhododendrons and lime.”’ 
In all the species the short leaf-petiole is erect and is nearly 
adpressed to the stem so that the lamina stands off from the 
petiole at a considerable angle. In Rh. cuneatum alone do the 
leaves reach any great size. There they may be 6 cm. long and 
neatly 2 cm. broad. In the high alpines, Rh. nivale, for instance, 
they are very small—may be only a couple of millimeters across. 
The tint of the actual leaf surface above is dark green, beneath 
it is paler because there the epidermis always produces papillae 
coated with small particles of wax,* and these may give a glaucous 
look to the surface. The real tint of the leaf is obscured because 
both surfaces are covered with peltate scales forming an indu- 
mentum. The general construction of these scales is alike in 
all. There is a short stalk of several cells, often in two vertical 
rows, and this stalk expands above into a many-celled umbo, 
from the margin of which there extends all round a fringe of 
empty cells so connected as to run together to the margin of 
the fringe and to end altogether there, so that the edge is smooth 
and not toothed. The fringe is radially as broad as the umbo. 
This umbo may contain a resin-like excretion, which may become 
tinted yellowish, amber-coloured, or red, and is then glistening, 
or it may remain unmarked by change in content, and then the 
surface of the umbo is not glistening or sometimes the point 
*As Mr. Tagg, Assistant in our Museum, who is studying indumenta of 
