BALFOUR—NEW SPECIES OF RHODODENDRON. IIg 
life after the seedling stage. It has a juvenile stage recognisable 
by a structural colour-character from an adult stage. This is 
important for it makes legitimate the putting forward in 
explanation of the adult-leaf condition of Rh. Forrestii that 
it is a persistent juvenile form. We do not know the seedling 
state of Rh. Forrestii but we cannot doubt that its adult 
character of leaf-colour will be ascertained to be a continuation 
of its seedling character. 
I have suggested elsewhere * that the anthocyanin colora- 
tion of the leaf under-surface in these seedling Rhododendrons 
probably plays the réle of heat-regulator. It is conceivable 
that conditions of the environment might call for the retention 
of this controlling colouring matter through the later phases of 
life of the plant, and that in what we are calling Rh. Forvestii 
with its reddened leaves we have only a local casual rendering 
ofa type which normally exhibits the foliage with green under 
surface of what we are calling Rh. repens, that is to say, that 
we have before us an unstable variation. But that is not the 
limit of conception in the case. The evidence does not negate 
the possibility that the leaf-colour originating in this way in 
a seedling character has become fixed as an adult character 
and is a constant feature in the plant under consideration. 
The retention and the final fixing of juvenile character is a 
well - recognised phenomenon of plant - life affecting structure 
of deeper moment than mutable possessions like colour-tints 
even to the degree of differentiating genera. In Convallaria 
(lily of the valley) for instance we have the permanently 
juvenile form of Polygonatum (Solomon’s seal). In support 
of the view that Rh. Forrestii and Rh. repens are specifically 
distinct and that this leaf-colour is a fixed diagnostic mark 
three bits of evidence may be brought forward :— ; 
The observations of Forrest—‘‘the man on the spot “— 
who has been impressed by their difference and has found no 
intermediate states such as would be likely to occur were the 
leaf-colour a fluctuating one. His many specimens confirm 
is In only one adult twig of his Ka-gwr-pw specimens have 
I seen a suggestion of tinting on the under-leaf surface. 
The fact that the two sets of specimens of Rh. E orrestit come 
from different localities with difference of elevation of 2000 ft. 
seems to favour the idea of constancy of the leaf- colour 
character. Scpcee 
In allied species of the Forrestil series which leave no room 
for questioning of specific identity we find parallel develop- 
ments. Rh. evastum, Ball. f. et Forrest (see p. 60), and RA. 
porphyrophyllum, Balf. f. et Forrest (see p. 108), supply the 
* In Trans. Bot. Soc. Edin., xxvii (1917), 221. 
