64 D. Prain — Flora of Narcondam and Barren Island. [No. 2, 



as extending to, but not beyond, Barren Island. To Blanford is due tbe 

 merit of having upset tbe fanciful hypothesis of tbe furtber extension 

 of the line across tbe Arracan Yoniab, and of baving suggested its pro- 

 bably true northern continuation. More recently it has been proposed* 

 by Berghaus and others to sub-divide the extended Sunda line of Blan- 

 ford into a Sunda Range proper, ending at the northern limit of Sumatra, 

 and a Pegu Range, containing Barren Island, Narcondam, Popah and 

 Han-shuen-shan. But it is obvious that if any sub-division be necessary, 

 the one proposed by Berghaus is erroneous. A sudden deep gap in the 

 line, with the furtber character of activity to the south of it, and non- 

 activity to the north, is a much more natural cleavage than merely a 

 number of miles of intervening sea, the nature of whose bottom is un- 

 known or has been misunderstood. If therefore Berghaus be justified 

 in differentiating a Pegu Range, it is clear that Barren Island must be 

 excluded from it, and that we must return to Yon Buch's view, that 

 Barren Island is the most northerly member of the Sunda Range. The 

 Pegu Range of very old and long extinct volcanoes begins then at 

 Narcondam, and extends at least as far as south-western Yunnan. 



The biological interest of these islands is not so great as the pby- 

 siographical, because, whether the ridge bei^e postulated exists or not, 

 there is little doubt that these sub-aerial portions never have been con- 

 nected with any of the adjacent lands. If Flat Rock has ever been sub- 

 aerial, and in a fit condition to shelter air-breathing creatures and 

 support vegetation, it is so no longer ; how great soever may be the 

 antiquity of the outer cone of Barren Island, it is probable from its con- 

 figuration, that at one time it has been the scene of a catastrophe like 

 that which in 1883 devastated Krakatau and totally destroyed its 

 animal and vegetable life. The only one that, from its topography, has 

 evidently remained for many ages in its present condition is Narcondam. 

 Already tbe writer has laid before this society some notes on the Fauna 

 of the islands f ; it remains now to be seen whether the biological facts 

 indicated by their Flora are in agreement with the deductions that 

 should follow from their physiographical configuration. 



All the plants found in the two islands are enumerated in the list 

 that follows ; running numbers are added to the locality so as to show 

 at a glance bow many species occur in each. In tbe discussion that 

 succeeds the list the peculiarities of each island are dealt with before 

 their common characteristics are considered. 



* Stieler : Hand Atlas, sheet 8. 



t Prain : Proceedings Asiat. Soc, Bengal, 1892, p. 109. 



