284 the botany of the kachin hills north-east of myitkyina. 
that extend only westward from it, there are four that occur both 
in the Hukung and Taping Valleys, as well as in the central portion 
of the Kachio country. 
It seems then safe to conclude that the catchment-area of 
the Upper Irrawaday admits of being dealt with as a natural phytogeo- 
graphical entity, and that the plan of considering its eastern portion 
as Chinese and its western as Burmese must be abandoned. The 
question now to be settled is the precise relationship of this conjoint 
Hukung-Kachin-Taping district. Situated as it is where at least 
three well-marked phytogeographical sub-sub-areas-— those of China, 
Indo-China and the Eastern Himalaya-^meet, it may not inconceiv- 
ably be referable to any one of these, and from its position may even 
throw light on the relationships these bear to each other. The 
writer has, on more than one occasion, had reason to express his 
belief, based on an examination of the distributional features in 
particular genera like Pedtcularis and Gomphostemma^ that the 
hitherto accepted sub-division of Western Indo-China into a northern 
half termed Ava and a southern half termed Pegu, is at variance 
alike with pbysiographical and phytogeographical facts, and is as 
inconvenient as it is incorrect. The facts reviewed by him else- 
where have led to a proposal to treat as a natural sub-sub-area the 
block of mountainous country that intervenes between the valleys 
of the Brahmaputra river to the north and west, and of the Irrawa- 
day river to the east. This block certainly includes the hill-ranges 
known as the Patkoi, Barel, Khasi, Chin-Lushai, Chittagong and 
Arracan, and ends towards the south in the Andaman Islands, and in 
this sense possesses a natural rank equivalent to that of the Eastern 
Himalaya from the Sarju Valley, 82°E. Lon,, to the Dihong Gorge, 
95°E. Lon. The Mishmi-Kachin block, between the gorges of the 
Dihong and the Salwen, 95°E. Lon. to gg^E. Lon., may conceivably 
be referable to either one or other of the sub-sub-areas mentioned; 
but it may equally conceivably be regarded either as central Indo- 
Chinese— the tract of mountainous country intervening between the 
valleys of the Irrawaday and the Mekong— or as south-west Chinese. 
The unexplored character of the block in question has hitherto 
rendered any opinion on the point more or less conjectural ; in 
order to test the extent to which our present Kachin collection lends 
itself to the elucidation of this point, the distributional features of its 
elements have been tabulated so as to show the extension of each 
species westward into the Assam-Arracan sub-sub-area, and into 
the sub-sub-area of the Eastern Himalaya, beyond these into India; 
