New Species of Craibiodendron. 
BY 
W. W. SMITH. 
With Plates CVIII.-CX. 
S1ncE the publication of the ericaceous genus Craibiodendron 
in the Records of the Botanical Survey of India, vol. iv. p. 276, I 
have had the opportunity of examining additional Indo-Chinese 
material, chiefly in the Herbaria of Kew and Edinburgh. I 
found that Mr. S. T. Dunn, dealing with Chinese material in the 
Kew Herbarium collected by Dr. A. Henry, had marked certain 
sheets as representing a new genus of Evicaceae, an interesting 
confirmation of the view taken by Mr. Craib and myself in our 
examination of the Burmese and Siamese specimens. 
A survey of this new material and comparison with a wider 
range of related genera than was before possible have led me 
to consider the new genus as more closely related to Leucothoé 
than to Pieris or Lyonia, This is supported by the somewhat 
imbricate, almost free sepals, the muticous anthers and the 
winged seeds. I find the character of the large, unilaterally 
winged seeds to hold good in the three species of which fruits 
are available. 
Craibiodendron shanicum extends into Yunnan, W. China. 
In this province are two other closely allied species, C. Henryt 
and C. yvunnanense; a fourth species, the imperfectly known 
C. Forrestii, was obtained by Forrest in the upper Salwin valley. 
What is probably a fifth species is found in Assam, C. Mannit, 
also incompletely known. 
C. Henryi in its typical state is distinct enough from C. 
shanicum: in Burma, the latter, recorded from several localities, 
shows little variation, but in Yunnan there are forms (vid. Henry 
9505B, in Herb. Kew et Herb. Edin.) which suggest a transition. 
The Khasian species and the shrub collected by Forrest in the 
Salwin valley are very inadequately represented, but a short 
description has been given here for the sake of completeness. 
{Notes from R.B.G., Edin., No. XXIV., January 1912. ] 
E 
