and the discovery of Fiji, and was soon to rob him
of his ship, now dealt him the unkindest cut of all. On August 13th, he
sighted the island of Vanikoro, and ran along its shore, sometimes within
a mile of the reef. There was no conceivable reason why he should not
have made some attempt to communicate with the inhabitants whose smoke
signals attracted his attention. Had he done so, he would have been the
means of rescuing the survivors of La Perouse's expedition, and of
clearing away the mystery that covered their fate for so many years. For,
after Dillon's discoveries, there can be little doubt that they were on
the island at that very time, and it is not unlikely that the smoke was
actually a signal made by them to attract his attention. The Comte de la
Perouse, who had been despatched on a voyage of discovery by Louis XVI.
on the eve of the Revolution, handed his journals to Governor Phillip in
Botany Bay for transmission to Europe in 1788, and neither he, nor his
two frigates, nor any of their company were ever seen again. Their fate
produced so painful an impression in France that the National Assembly,
then in the throes of the Revolution, sent out a relief expedition under
"Citizen-admiral" d'Entrecasteaux, and issued a splendid edition of his
journals at the public expense. We now know from the native account
elicited by Dillon that during a hurricane on a very dark night both
frigates struck on the reef of Vanikoro, that the _Astrolabe_ foundered
with all hands in deep water, and the crew of the _Boussole_ got safe to
land. They stayed on the island until they had built a brig of native
timber, in which they sailed away to the westward to meet a second
shipwreck, perhaps on the Great Barrier reef. But two of them stayed
behind for many years, and of these one was certainly alive in 1825. Now,
Edwards saw Vanikoro just three years after the wreck, and even if the
brig had sailed, there were two castaways who could have cleared up the
mystery.

After a narrow escape from shipwreck on the Indispensable Reef, he made
the coast of New Guinea, supposing it to be one of the Louisiades. And
here has occurred one of those curious errors in geographical
nomenclature which are perpetuated by the most permanent of all
histories--the Admiralty charts. Edwards gives the positions of two
conspicuous headlands, which he named Cape Rodney and Cape Hood, and of a
mountain lying between them which he called Mount Clarence. All t