retch
our limbs. Much refreshed, we continued our voyage. I forgot to state
that at every island where we touched we engraved our names on the
trunks of trees, in the most conspicuous situation, and stated the
direction in which we were going. We had done this also on our own
island, as we called it, that should any vessel visit the spot she might
perhaps convey intelligence to Captain Frankland that we were alive, and
give him some clue as to where to look for us. Our friends understood
our object, and now added some sentences in their own language to the
same effect. The fine weather continued, and confident in the guidance
and protection of Him who had hitherto preserved us from so many and
great dangers, we launched forth again into the deep.

We passed several small islands; some had but a few stunted trees
growing on them; others again had scarcely soil sufficient to nourish a
few blades of long wiry grass; while others were barren rocks without
verdure of any description, their heads but lately risen from beneath
the waves. I believe that it was at one time supposed that these coral
formations rose from immense depths in the ocean, and that those
wonderful and persevering polypi worked upwards till they had formed
submarine mountains with their honey-combed structures; but it is now
ascertained that they cannot exist below at the utmost fifty feet of the
surface, and that they establish the foundation of their structures on
submarine mountains and table-lands, while they do not work above
low-water mark. How comes it then, it will be asked, that they form
islands which rise several feet above the sea? Although the polypi are
the cause of the island being formed, they do not actually form it.
They begin by building their nests on some foundation which instinct
points out to them. First they work upwards, so as to form a wall, the
perpendicular side of which is exposed to the point whence the strongest
winds blow and the heaviest sea comes rolling in. Then they continue to
work along the ground and upwards on the lee side of the wall, sheltered
by their original structure from the heavy seas. They also work at each
end of their wall in a curve with the convex side exposed to the sea.
Thus, at length, beneath the ocean a huge circular wall of considerable
breadth is formed. Storms now arise, and the waves, dashing against the
outer part of the walls, detach huge masses of the coral, six feet
square or mo