In addition to the main islet, there are two smaller ones—Nepean, a low islet of coral-sandstone, a few acres in extent, lying about half a mile south of Emily Bay; and Philip Island, between three and four miles south, a volcanic mass about 900 ft. high, a mile and a half long from east to west, and three-quarters of a mile wide from north to south.

Norfolk Island is surrounded by cliffs, rising in the north and west to 300 ft. or 400 ft., and elsewhere to the height of 200 ft. The only exception to this is the coast-line between Bloody Bridge and Bumbora. Here the coast is protected for a distance of about a mile by rocks of coral-sandstone. This erosion does not seem to have been much greater on one side than the other, but it has obviously been very long continued, and the island is now much smaller than it originally was. This is further indicated by numerous outlying rocks, particularly on the northern side. In most places at the foot of the cliffs there are narrow beaches, usually of shingle. At Anson Bay the heavy shingle on the beach is sometimes replaced by sand. Thus Backhouse, who visited the island in March, 1835, records that “This [Anson Bay] was formerly a landing-place, but the sand has been washed away, and large stones remain, too rough for boats to venture upon.”^* This spot was chosen for the shore end of the Pacific cable, partly on account of the sand, as it was feared that boulders would chafe the cable. Shortly after the landing of the cable. some ten years ago, the sand was replaced by boulders, which were still there at the time of my visit. Such changes as these indicate the violence of waves and currents along the coast. On the south coast, opposite to Philip Island, there has been least erosion. Here there has been a considerable growth of coral-sandstone. This growth may have been due in the first place to the protection of Philip Island, which probably at one time extended farther towards Norfolk Island. The coral itself now forms an additional protection, though this too is undergoing rapid erosion. It doubtless at one time extended continuously to Nepean Island, which is now often subjected to a fierce sea that must soon destroy it. In spite of the present rapid formation of the coral,^† it can scarcely be doubted that it is now being destroyed much more quickly than it is being formed, for reasons perhaps not at present apparent. The long-continued period during which erosion has been going on is further indicated by the high cliff on the south side of Philip Island. This rises to the full height of the island, 900 ft., thus probably indicating that half of the island has been destroyed by wave-erosion, as in the case of such subantarctic groups as the Aucklands and the Snares.

The interior of Norfolk Island consists of comparatively level uplands from 300 ft. to 500 ft. high, intersected by narrow stream-valleys, sometimes forming ravines often 50 ft. or more deep. The streams in most places fall over cliffs into the sea—e.g., at the Cascades, Keripae, and Duncombe Bay. It is clear that this conformation is due to the fact that the ocean is wearing away the island at a greater rate than the streams can cut down their beds. It is only on the south coast of the island that any of them are able to reach the sea except by a waterfall over a cliff. Thus at Bloody Bridge, Bumbora, Town, and Beef-steak the stream-valleys slope down to sea-level, and are not truncated as elsewhere. In this region, but more especially between the cemetery and the pier, the coast-line is protected by coral reefs, and coral-sandstone, the product of the older reefs, as already

mentioned. The coral-sandstone here extends from a quarter to half a mile inland, and rises to a height of 30 ft. to 50 ft., with the strata lying horizontal or dipping slightly, usually inland. According to Etheridge (p. 116), the structure of this deposit is practically the same as the coral-sand rocks found on Ascension and Lord Howe Islands:^* “Touching the deposits at Ascension and Norfolk Islands, their descriptions would almost embrace that at our island. Of the former, Darwin says,' It consists of small well-rounded particles of shells and corals, of white, yellowish, and pink colours, interspersed with a few volcanic particles.'^† Mr. Carne's description of the latter is identical, almost word for word.” The consolidation of this rock is doubtless largely due to the percolation of rain through it. This dissolves out carbonate of lime, which cements together the particles below as it evaporates. Of the Lord Howe deposit, Etheridge (loc. cit., p. 124) says, “When first I examined this deposit in situ I regarded it as of aqueous deposition; but after due consideration of all the facts for and against I have abandoned this view in favour of an aeolian origin.” Unfortunately, I was unable to give much time to the investigation of this interesting formation on Norfolk Island, but have no doubt that it has been formed in the same way as on Lord Howe. My cursory examination of it did not enable me to determine whether there have been two distinct periods of formation, as at Lord Howe Island (Etheridge, p. 118, loc. cit.).

The main body of Norfolk Island consists of streams of basaltic lava, usually lying approximately horizontal. They are very much weathered, so that it is difficult to get fresh specimens of rock. The whole surface of the island except the rocky ridge of Mount Pitt and the precipitous portions of its western slopes is covered to a depth of 150 ft. to 200 ft. with a somewhat compact though porous layer of decomposed volcanic rock, through which wells can be dug without much difficulty. These wells, together with rain-water, provide the water-supply of the inhabitants. The deepest well is near Bloody Bridge, and goes to a depth of 186 ft.

This stratum apparently resembles the laterites of tropical regions in some of its characters. Unlike these, however, it provides a very fertile soil, often brown, and contains a grit, probably consisting largely of iron segregated from the volcanic earths; but, like the laterites, this surface earth has none of the impermeability of a clay; and, indeed, there is probably no true clay on the island. What has led to this immense weathering and decomposition of the volcanic rocks it is not easy to determine; but it can scarcely be doubted that it shows the island is by no means of recent formation. It is, of course, impossible to determine the age of Norfolk Island without data of a more definite character, and these are at present wanting.

Strewn over the surface of the soil in most parts of the island are boulders of all sizes, evidently the remnants of the harder undecomposed cores of the rock-masses. I was not able in the time at my disposal to detect any old crater-vents or fissure-lines. Possibly the rocky ridge of Mount Pitt may represent the portion of an old crater-wall. This lies about a mile distant from the west coast, and roughly parallel with it, and runs through the northern half of the island. Between Mount Pitt and the sea the country is rather rough and considerably broken by cliffs.