Strictly speaking, there is no soil upon the island:  what may, however, be considered as such consists of the disintegrated calcareous rock, on the low part mixed up with sand.  This rock, acted upon by the weather, has a tendency to fall down in large masses, leaving cliffs, steep and rugged in some places and smooth in others; in colour it varies from white to red, and is usually of a light pink.  Behind one of the beaches, a few feet distant from high-water mark, I observed a bank twelve feet high of slightly agglutinated coral sand in parallel beds, mixed up with large depositions of weather-worn shells:  Tridacna, Hippopus, Strombus, etc., all of species now living on the reef.  At one end this deposit appears to have been tilted up, forming a slight ridge stretching across the low part of the island.  The shores in some places are fringed with coral conglomerate composed of shells and sand, fragments of coral, and rolled pieces of rock from above.  The reef surrounding the islands does not dry at low water, and in crossing it in the boat very little live coral was observed, except on the outer margin, outside of which the bottom is a tenacious mud, effervescing on the application of hydrochloric acid.

I collected a few plants, among which are a yellow-flowered Cleome, a purple Pongamia, Convolvulus multivalvis, Evolvulus villosus, Guettarda speciosa, etc.  The only birds seen were a white-headed eagle and an osprey, neither of which were molested although the latter frequently came within shot, and followed me as if from motives of curiosity.  Almost the only insects seen were small grasshoppers, rising in numbers at every step, and green ants which have nests in the bushes, and appear identical with those of the Louisiade and Australia.

No fresh water was found here.  Some recent traces of natives were met with—­including two fireplaces where turtle and fish had been cooked on a framework of sticks over a fire—­precisely similar to one of large size, formerly seen on the Duchateau Islands.  I saw many places where turtle eggs had been dug out of the sand behind the beach, where besides were numerous burrows of a maritime crab (Ocypode cursor) which also appeared to feed upon the eggs—­judging from the quantity of empty shells about the holes of those creatures.

Of the two remaining islands of the group, one, less than a quarter of a mile long, is covered with trees, probably a Bombax or Erythrina—­at this time destitute of leaves—­on the other is a high bare rock with three other small detached, needle-shaped ones lying off it.  The observations with the theodolite having been completed we obtained some soundings and returned to the ship.