POLYNESIAN RESEARCHES. 
3 
that surrounds the shore—a most wise and benevolent 
provision for the ingress and egress of vessels, as well 
as a singular phenomenon in the natural history of these 
marine ramparts to the islands. Whether the current 
of fresh water, constantly flowing from the rivers to 
the ocean, prevents the tiny architects from building 
their concentric walls in one continued line, or whether 
in the fresh water itself there is any quality inimical 
to the growth or increase of coral, is not easy to deter¬ 
mine ; but it is a remarkable fact, that few openings occur 
in the reefs which surround the South Sea Islands, except¬ 
ing opposite those parts of the shore from which streams 
of fresh water flow into the sea. Reefs of varied, but 
generally circumscribed extent, are frequently observed 
within the large outer barrier, and near the shore, or 
mouth of the river; but they are formed in shallow 
places, and the coral is of a different and more slender 
kind, than that of which the larger reef, rising from the 
depths of the ocean, is usually composed. There is no 
coral in the lagoons of the large islands. 
The openings in the reefs around Sir Charles Sander’s 
Island, Maurua, and other low islands, are small and 
intricate, and sometimes altogether wanting, probably 
because the land, composing these islands, collects 
but a scanty portion of water; and, if any, only small 
and frequently interrupted streams flow into the sea. 
The openings in the reefs around the larger islands, not 
only afford direct access to the indentations in the 
coast, and the mouths of the valleys, which form the 
best harbours,—-but secure to shipping a supply of fresh 
water, in equal, if not greater abundance, than it could 
be procured in any other part of the island. The cir¬ 
cumstance, also, of the rivers near the harbours flowing 
