From the pen of another eye-witness[3] we have a description of a visit at low-tide to the barrier-reef of Levuka, the old Fiji capital. We learn that the reef itself consists mainly of dead coral--“rough, uninteresting, shapeless limestones, with a very small covering of seaweed.” But after about a quarter of a mile of difficult walking, the travellers drew near to the “roaring surf on the outside,” where “fingery lumps of beautiful live coral began to appear, of the palest lavender-blue colour.” By the time that they were nearly within touch of the spray, “the whole floor was one mass of living branches of coral.” [3] The Hon. Ralph Abercromby. It was not, however, until they ventured actually into the water, beyond the outer verge of the “great sea-wall,” that they could see for themselves the true nature of the reef. This was the description given. “You look down, and see a steep irregular wall, extending deeper into the ocean than the eye can follow, and broken into lovely grottoes and holes and canals, through which small resplendent fishes of the brightest blue or gold flit fitfully between the lumps of coral. The sides of the natural grottoes are entirely covered with endless forms of tender-coloured coral, but all beautiful, and all more or less of the fingery or branching species, known as madrepores. It is really impossible to draw or describe the sight.” Equally impossible was it to photograph these fairy grottoes, “seen through twenty feet of surging water,” though some snapshots were taken in moments when the vast billows withdrew themselves. One difficult question in connection with coral reefs has been--How can they rise, as they do, out of considerable depths, when coral-polyps cannot live below some twenty fathoms from the surface? Two chief explanations have been offered, each of which has been warmly taken up and defended. One meets the difficulty by means of the theory of a rising and sinking earth-crust. A mountain-top might have been first heaved up to within twenty fathoms of the surface. Then later, while the polyps were building upon it, the mountain might have slowly sunk. In that case the polyps below would die, as the water deepened; but the wall of dead coral would remain, reaching in time very far down. The other explanation also admits that coral-reefs have been generally built upon sub-ocean heights. But it supposes that height to have been partly raised by volcanic forces, and afterwards built up to the needed height by a growing collection of oozes and animal-remains, deposited through ages. Many proofs have been found in recent years of the probable truth of the latter theory. Still, it does not of necessity do away with the earlier explanation. Here, once more, the question may be asked--Why should not both be true? If some submarine peaks have been lifted to just the right level by gradual building up of sediment, others may have been so raised by volcanic action or crust-crinkling.