14 INTRODUCTION. 
remaining species, JD. mimosoiles does not afford evidence of isolation; it extends to 
Assam and Sikkim, has a related species (D. stenophylla) in Central China and another 
allied form (D. Millettii) in Eastern China, D. Dyeriana is even less distinctive, since it 
appears to be a mere overflow from Central China: D. Henryana is hardly more distinctive 
than 2. mimosoides because, though itself undistributed, a very close ally (D. Kingiana) 
extends to the adjacent Kachin country in Western Indo-China, and another ally 
(D. Benthami), not quite so intimate, but still near, occurs in Eastern China. D. 
yunnanensis has a close ally (D. Collettii) on the Shan Plateau in Central Indo-China; 
D. wemaoénsis belongs to a group of species most strongly represented in Indo-China : 
the most distinctive species of Western Indo-China is D. polyadelpha. Ву the most 
liberal computation, then, the endemic factor for South-Western China is only 1%, or 
23. Eastern China, as defined above, though with fewer species, shows а higher 
degree of isolation. Неге there are nine Dalbergias: D. Mailleitii, D. Hancei, D. 
Benthami, D. tonkinensis, D. sacerdotum, D. rimosa, D. hupeant, D. Dalansae, D. torta. Of 
these, the first five are erdemic, though, for reasons already made clear, the evidence 
of D. Milletiii as to isolation is not conclusive, while that of D. Benthami is not 
great. On the other hand, the evidence of D. Balansae is greater than would at 
first sight appear, since it is widespread in Eastern China, but only overflows, and 
then perhaps only as a planted species, into Kiangsi in Central China; and 
D. hupeana, which only oecurs, and again perhaps only as a planted tree, in 
Kwang-tung, gives nearly as strong negative evidence in favour of the isolation of 
Eastern China as the positive evidence of D. Balansae, D. Hance, D. tonkinensis, and 
D. sacerdotum, which are all distinctively Eastern Chinese species. The two that remain 
are D. rimosa, a species widespread in South-Western China, Western Indo-China, and 
the Eastern Himalaya, which extends to Tongking without having so far been met 
with in Central Indo-China; finally, D. torta, which occurs on the coasts of South- 
Eastern China and of Tongking, and has been reported from every coast throughout 
our Asiatic region except those of the Sunda islands and of Coromandel, Тһе endemic 
factor for Eastern China is therefore $, or 55:5. bouem [o 
Excluding, then, the Central Indo-Chinese species that is Chinese only because 
it occurs in the Taping Valley, and including, as we must, within Chinese limits 
the province of Tongking, we have an area with 92 вресіев, of which 13 are endemic, 
во that its endemic factor is 13, or just under 60. The connection between China 
as a whole on the one hand and Indo-China or the Himalaya on the other is, however, 
almost entirely confined to South-Western China. If this district be excluded and the 
conjoined districts of Central and Eastern China be considered together, we have then 
an area with 11 species of Dalbergia, whereof 8 are endemic. The ‘endemic factor for 
Central and Eastern China is thus 1%, or 727, indicating a degree of isolation higher 
than is to be met with in any similar area in South-Eastern Asia, 
The Philippines, physiographically much isolated, with ten reported Species of 
Dalbergia, have only three peculiar forms, D. polyphylla, D. Mimosella, and D, Cumingiana, 
One of the reinaining species is the ubiquitous littoral D, torta ; another reported sea-shore 
species—D, spinosa—occurs also on the coasts of Indo-China and Coromandel. ‘The others 
аге—(1) D, tamarindifoliz, a species present both in Malaya and in Central Indo-China : 
. . . 2 
(2) D. parviflora, widely spread іп Malaya; (3) D. ferruginea, а species confined to 
Papuasia and the Caroline Archipelago; (4) D. discolor, а species common {о the 
Philippines, Celebes, and Borneo; and (5) D. Minahassae, а species common to the 
Pnilippines and Celebes. Тһе Philippine endemie factor is thus -3 от 30, and is too 
i0? 
