Flora of Narcondam and Barren Island. 
63 
in this particular direction by the action of currents. It would seem 
easier, however, to explain these soundings by supposing that Barren 
Island formed the northern termination of a ridge on which Flat Rock 
with Invisible Bank, is situated. Here, too, the matter is easily capable 
of demonstration: soundings on a line bearing from Barren Island to 
Flat Rock, with one or two transverse lines of soundings will show 
whether such a ridge exists. 
The hypothesis that in Harcondam we see a continuation of the 
Sunda line of volcanic activity is not invalidated by the depth of the 
soundings between it and Barren Island. We know that there is a much 
deeper gap than this between two members of the same chain: in the 
well-known rift between Bali and Lombok, though the islands mentioned 
are only 15 miles apart, the narrow strait between is 2,100 fathoms 
deep.* And as a matter of fact, though the ridge is here deeper, it is by 
no means absent, for a sounding on the line bearing from Narcondam 
on Barren Island gives only 1,010 fathoms, while soundings to the west 
of that line, and between the supposed ridge and the Andamans, give 
1,140, 1,159, and 1,130 fathoms. Though our knowledge of the bottom 
contour of the southern part of the Andaman Sea—the portion to the 
east of the Nicobars—is very defective, the little that we know bears 
out the hypothesis of an eastern as well as a westeim ridge. At a point 
50 miles east of Little Nicobar a sounding of 1,284 fathoms is recorded, 
while 30 miles further east the bottom is only 1,000 fathoms deep. Then 
north of Pulo Rondo, in Lon. 95° 10' E., the depth is 990 fathoms, while 
20 miles further east it is only 930 fathoms. These soundings of 930 and 
1,000 fathoms not improbably indicate the ridge on which Flat Rock, 
Barren Island, and Narcondam are situated. The 990 and 1,284 fathom 
soundings must indicate the trough between the ridges; for to the west 
of the latter lies the Nicobar Glroup, and to the west of the former, in 
Lon, 94° 20' E., we find a depth of 975 fathoms, doubtless indicative of 
the western or Indian Ocean slope of the Nicobar-Sumatra ridge, since 
25 miles further north, in Lon. 94° 26' E., we have a sounding of 760 
fathoms indicating the crest of that ridge. The soundings referred to 
are shown on the two maps that accompany this paper. 
There is, pex’haps, some connection between the depth of the rift 
separating Narcondam from Barren Island, and the fact that from Barren 
Island itself southwards the volcanoes either still are, or have till re¬ 
cently, been active, while those from Narcondam northwards have long- 
been extinct. This has a certain bearing on another controverted point. 
Von Buch, as has been already stated, recognised the Sunda volcanic lino 
* Wallace, Island Life, 423 (map). 
277 
