The high ridge that surrounds the depressed interior of the island, proclaims, at once, that it is the summit of an old volcanic cone, and an examination of the rocks confirms the supposition, and shows that it belongs to the trachytic class of volcanoes. The shape of the crater is singular, its length being more than twice its breadth, and the northern part being divided into two, by a ridge running from the edge of the crater towards the centre of the island, and directed nearly to the boat harbour, or that point where the wall of the crater is lowest. This configuration is probably owing to there having been two craters, the southernmost of which was the last in activity, and filled up the northern one with ashes; and the two valleys have been subsequently scooped out by subærial denudation. The direction of the dividing ridge, and the termination of the north-western valley in a narrow gorge prove the correctness of this view.
Nearly the whole of the island is composed of trachytic tuffs and breccia generally either white, or of a pale yellow or violet color, and enclosing here and there, fragments of trachyte and obsidian. These
tuffs are arranged in the cliffs that formed the wall of the crater, more or less horizontally, although of course much confused in places, and are but little intermixed with lava streams. To find these latter, we must go to the adjacent coast of the Great Barrier, about three miles distant, where, at the south side of Wangapoua Bay, we see thick beds of trachyte, and trachy-dolerite, interstratified with tufa, dipping away, from Arid Island at an angle of 35°; and farther inland, on the top of the dividing ridge of the island, we find tufa and agglomerate, most probably derived from Arid Island, at an elevation of 1550 feet from the sea level.
Now, these facts lead to some interesting deductions, which bear on one of the great questions of the day, in Geology, viz.:—Are volcanoes connected with a central fluid interior of the earth, or are their lavas derived from comparatively shallow depths below the surface?
It is evident that Arid Island, in its present condition—only 700 feet above the sea—could not have distributed tufas at an elevation of 1550 feet, neither could lava streams flowing from it now, have, at the same level as the crater, and at a distance of three miles, a dip of 35°. It follows therefore that either some other crater, in the direction of Arid Island, and much higher than it, but which has now quite disappeared, was the origin of these lavas and tufas; or, that Arid Island was once at a considerably higher elevation, not only above the sea, but with respect to the main land of the Great Barrier. The depth of the intervening sea, 12 to 17 fathoms, makes the latter supposition much the most probable; and assuming that it was so, we see that Arid Island must have sunk down, at least, 2000 feet below the level of the Great Barrier, because the tufa, at a level of 1550 feet, is evidently a submarine formation; while the shape and preservation of the crater of Arid Island shows that it was formed under the air, so that the bottom of the crater must have been above the highest level of the tufa. Now, it appears most probable, that the sinking of Arid Island was produced either by the breaking of the roof of the cavity, from which the lava and ashes had been extracted, or by a slipping down in mass of that part between it and the Great Barrier. If, therefore, we suppose that Arid Island was the centre of the subsided portion, we have a district six miles in diameter, which has broken, or sunk through by its weight, into a cavity below; and it appears to be impossible that such should have been the case, unless the thickness of the crust broken, or sunk down, was considerably thinner than the diameter of the cavity. For, if not, it would have been strong enough to have resisted the fracture, and the friction along the sides would have been too great to allow it to slip; so that it seems impossible that the region from which Arid Island derived its lava was so deep as six miles below the surface, and therefore it would be unreasonable to, infer that it was connected with a fluid interior; for most astronomers and geologists are now pretty well agreed that the solid crust of the globe is at least a thousand miles thick.