BALFOUR—NEW SPECIES OF RHODODENDRON. 307 
ovary is always more or less hairy—frequently each filament 
has a villous tuft there which projects over the ovary into the 
tube of the corolla. But the hairs may be few in number or 
they may spread up the filament to near the top (RA. capitatum). 
In most of the series the stamens are about equal in length to 
the corolla—longer or shorter—but in the forms with elongated 
corolla tube circling round Rh. intricatum, Franch. they are not 
longer than the corolla tube, and there show most markedly 
the apparent obdiplostemony of so many Ericaceae. 
The ovary in all cases is lepidote, but not only so, it may be 
also pilose. This pilosity is a fluctuating feature. In all cases 
the top of the toral disk at the base of the ovary is more or less 
pilose. The hairs may extend upwards over the ovary amongst 
the scales to the top in a varying degree. Sometimes (Rh. 
achroanthum) they obscure the scales; at other times are 
hardly visible. From the ovary the hairs may extend to the 
style nearly through half its length. On the same plant 
in some a style pilose and one epilose may be found. Yet at 
times the character is‘useful. The length of the style in 
relation to the ovary is of importance. In the forms where 
the stamens are not exserted from the corolla tube the style is 
always shorter than the stamens—it may be shorter than its 
ovary. In forms with long stamens the style may be about 
equal to them—longer or shorter. Then there is the majority 
of forms with style consistently much longer than the stamens. 
The fruit, apart from variation in size, has the same type in 
all species of Lapponicum Rhododendron where it is known—an 
ovoid capsule dehiscing by five woody valves to the base. 
Here follows an attempt to arrange the species in grouplets 
for the purpose of facilitating recognition of them by easily 
observed characters. The arrangement in sub-series and the 
keys seem to work all right in relation to dried specimens, and 
so far as I have tested they are effective for the few species in 
cultivation. They seem to bring together more or less the nearly 
allied species. But they are tentative. 
SuB-SERIES A. 
Scales of under-leaf indumentum brown, uniform, truly 
discontiguous. The species 
Rh, fastigiatum, Franch., 
Rh. flavidum, Franch., 
Rh. impeditum, Balf. f. et W. W. Sm., 
Rh. oresbium, Balf. f. et Ward, 
Rh. scintillans, Balf. f. et W. W. Sm., 
Rh. verruculosum, Rehd. et Wils., 
Rh. complexum, Balf. f. et W. W. Sm., 
