^totiotientiron ^ocietp Jtoteg. 
6. Indumentum surface pale-buff not scintillating 
somewhat woolly its funnel-shaped delicate cup-hairs 
deliquescing into many thin interlacing branches 
are not easily separable. Corolla colour unknown. 
Leaves oblong-lanceolate or oblong-oval narrow¬ 
ing to apex . . . . . . . . . . GALACTINUM. 
Leaves oblanceolate or oblong-obovate rounded 
at apex . preptum. 
For the differentiation of species the character of underleaf indumentum 
has been largely used in the preceding pages, and the features of the cup-hairs 
are not only diagnostic of the series, but also within it. It is difficult to give in 
a few words a satisfactory description of the hair-forms ; sketches or photographs 
are required to make clear their individual appearances and differences. These 
have been prepared and will I hope appear in an early number of Notes from 
THE Royal Botanic Garden, Edinburgh, with a fuller account of the indu¬ 
mentum and its development. 
The story which I have written is manifestly incomplete, and readers may be 
disappointed, as I am, over omissions where information is particularly desirable. 
But the time is not yet for the writing of complete stories of the several series 
of Rhododendron—and this because the old requires revision and the new to be 
assimilated is so large in amount. We have to learn much about the Himalayan 
forms. For too long we have been content to rest in the belief that the last 
word has been said upon them. This is far from being the case. Hooker 
recognised on the spot many distinct forms which Clarke subsequently com¬ 
bined at the sacrifice of precise diagnosis. These have all to be worked over 
again in the light of newer knowledge. And then the arrival of new species 
from Western China, which does not seem likely to abate for some time, is con¬ 
stantly extending our horizon and giving us new clues to relationship. Here I 
only claim to have brought together in one assemblage a number of forms that 
seem to be naturally related to one another more closely than they are to other 
forms. That the Chinese species will be found to vary as much as we know the 
Himalayan ones do we may expect and definitions based as at present upon single 
dried specimens will be found doubtless to require modification ; but I have 
confidence in the correctness of my interpretations of characters as phyletic and 
have hope that some at least of what I have written may be helpful to Rhodo¬ 
dendron lovers. 
1. BAYLEY BALFOUR. 
222 
