299 
lying north-eastern peninsula had originally been detached from the 
main island and, being an islet of considerable width, that a causeway, 
ultimately becoming an embankment, has been thrown up by wave-action 
from each of the two adjacent bays. Soil washed down from the adjacent 
slopes during the rainy season has in the form of fine silt closed up the 
porous shingle banks at either end till these can now retain the fresh 
water within them and prevent the percolation of sea-water from with¬ 
out. To the east side of this lake there is a small flat meadow covered 
with KylUnga and Fimbistylis along with some Gyperus polystachyus but 
very little grass. Whether this meadow was originally a naturally bare 
patch or is only part of the clearing made in connection with the 
abandoned settlement on the adjacent hill it is difficult to say. If, how¬ 
ever, it was artificially cleared, it is unlike the rest of the clearing in 
this, that no woody jungle is reappearing in it now. At the time of our 
visit a number of snipe frequented the meadow. Close to the edge of 
the lake is a continuous belt of Hygrophila quadrivalvis; within this, and 
extending into the water, is a belt of Polygonum all round the margin of 
the lake ; inside the Polygonum float large matted patches of Panicuin 
Myurus. Here and there are patches Limnanthemum indicum; there 
is also a considerable quantity of Nymphcea rubra. The ordinary white 
Nymphcea Lotus, so common in similar spots in the Andamans, is not 
present, a circumstance which inclines one to think that this red water- 
lily may have possibly been introduced during the attempt to settle in 
the island. The water is quite potable and apparently wholesome; 
neither Chara nor Zanichellia is present, perhaps the water is rather 
deep for these. 
Very different in many respects is the lagoon on Little Coco 
which is simply a mangrove creek that has been banked off fram 
the sea by a small sand-dune having been thrown up across its mouth. 
It is not more than 1^-2 feet deep anywhere, with also a level but at 
the same time a softer botlom than the Grreat Coco lake, and this bot¬ 
tom is covered uniformly throughout by a meadow of Ghara mixed with 
Zanichellia. Here the water, though perhaps potable on an emergency, 
and though used by native craft that call in for it, is slightly 
brackish, and the lake is fringed throughout by Bruguiera, Lumnitzera, 
Geriops, Avicennia. etc., while clumps of similar mangrove trees occur 
throughout it. Its area is considerably greater than that of the Great 
Coco lake, for it is about a quarter of a mile long and a furlong across 
at the widest part; it was haunted at the time of our visit by teal. 
Here, curiously enough, Panicum Myurus does not occur, its place ^being 
taken by Paspalum scrohiculatum which floats in great patches at its 
south-western corner. There is no Limnanthemum and the Nymphcea 
109 
