either been volcanic or coralline, and along the coast there are
fringing coral reefs very dangerous to navigation. At the same
time, the character of its natural history proves it to be a
rather ancient land, since it possesses a number of animals
peculiar to itself or common to the small islands around it, but
almost always distinct from those of New Guinea on the east, of
Ceram on the south, and of Celebes and the Sula islands on the
west.

The island of Morty, close to the north-eastern extremity of
Gilolo, was visited by my assistant Charles Allen, as well as by
Dr. Bernstein; and the collections obtained there present some
curious differences from those of the main island. About fifty-
six species of land-birds are known to inhabit this island, and
of these, a kingfisher (Tanysiptera Boris), a honey-sucker
(Tropidorhynchus fuscicapillus), and a large crow-like starling
(Lycocorax morotensis), are quite distinct from allied species
found in Gilolo. The island is coralline and sandy, and we must
therefore believe it to have been separated from Gilolo at a
somewhat remote epoch; while we learn from its natural history
that an arm of the sea twenty-five miles wide serves to limit the
range even of birds of considerable powers of flight.

CHAPTER XXIII.

TERNATE TO THE KAIOA ISLANDS AND BATCHIAN.

(OCTOBER 1858.)

ON returning to Ternate from Sahoe, I at once began making