( 5 ) 
Bruguiera gymnorrhiza were frequent, and of a larger size than those which grow 
close to the shores, hut apparently agreeing with those which are everywhere 
found further up the creeks. All these (except the last mentioned) are trees 
which never occur in the mangrove swamps, hut in a sandy soil just behind 
them, or else bordering the sea where the shores are very steep and not 
adapted for the formation of mangrove swamps. This latter circumstance, 
however, cannot be the case here, the shores at the places where I made the 
observations being level for some thousand paces inland. 
A more evident fact proving the sinking state of these islands* can 
nowhere be seen more clearly than in these straits. One can there trace in 
several places the stumps of the sunken trees in the sea up to the state where 
the trees are just dying by the influence of the sea water, and the subsequent 
change of the soil by the formation of the mangrove swamp. 
The sinking state of the islands is further shown by another fact recorded 
in the report of the Andaman Committee, dated 1st January 1858, in sec¬ 
tions 8 and 9. There it is stated that the sea had encroached some 40 or 50 
feet since the first settlement at Chatham Island in Port Cornwallis, so that 
the store-house that stood there has been destroyed by the sea since the aban¬ 
donment of the place in 1796. An examination of the spot and of the docu¬ 
ments relating to that settlement might give us some guide to determine the 
rapidity with which these islands are becoming submerged. 
Lieutenant Jameson, of Chatham Island, has informed me that a similar 
encroachment of the sea is taking place at that island in Port Blair. 
There are nowhere, as far as I have been able to explore the rocky shores, 
signs of a recent upheaval. The mighty waves hollow out the soft sandstone 
and serpentine ridges, until undermined they fall into the sea, and form by 
their debris a natural barrier against the actions of the waves. An example 
of this process is afforded by the Bird-nest Cape. The caves of Hirundo 
esculenta at the same Cape exhibit semi-circular openings, and are nearly at 
an equal level with the sea. Had any recent upheaval taken place, they would 
most probably be of an oblong or elliptical shape, or more likely raised far 
above the sea-level. 
Long and broad flat fields of sandstone—layers, almost at a level with the 
sea and perfectly free from corals, stretch out in several places (especially 
along the western coasts, but also on Boss Island itself). These are submerged 
during tide and separate the coral reefs from the rocky coast, which, in other 
places, is immediately fringed by those reefs. 
An examination and careful determination of the corals and nullipores 
forming those reefs would surely throw much light upon the whole matter. 
The depth at which certain corals live is known, and the question of upheaval 
or submersion might be elucidated more easily and more conclusively by a 
zoologist than by a botanist. That a few coral banks (with living corals on 
them) are uncovered by sea during ebb I have stated already above, but 
nowhere along the whole length of the coasts visited by me are coral reefs 
raised above the tidal level. 
I have since visited the Arracan coasts (Akyab and the lower hills of 
Kolodyne District) which all show a decided rising, and are of very similar 
formation (sandstone) as those on the Andamans. The phenomena, however, 
presented to me were quite different from those I have observed on the Anda¬ 
mans, not only as regards the configuration of the coasts, but also its flora. 
Andaman Straits, which is impassable at low water for boats, will perhaps 
afford another good proof of this subsidence. I had no opportunity of visiting 
these straits. 
* They may he also stationary, but, from the evidence adduced, it is more natural to arrive at the conclusions I did 
9 
