Having breakfasted, Gurney and I again betook ourselves aloft to the main topmast crosstrees, carrying the ship’s telescope with us, our object being to subject the reef to a thorough scrutiny, in the hope that, with the sun now high in the heavens, and the light as good as it was likely to be, we might be fortunate enough to discover a way of escape from our extraordinary prison.

As a matter of fact we did now get a much clearer and, on the whole, more satisfactory view of the reef than upon the previous occasion; but although we perceived a perfect network of channels—some so narrow as scarcely to permit the passage of a boat, while others were wide enough in places actually to allow a ship to work to windward in them—the inequalities in the surface of the reef were so great as to render it impossible for us to trace any of them for more than three or four miles at the utmost. There were four channels, wide enough to allow the passage of the ship, branching out from the basin in which we lay, one trending toward the north-east; another running off toward the north-west, and then, apparently, by a zigzag course ultimately leading to open water on the west side of the reef; a third running west out of the basin for a distance of about three miles, beyond which its farther course became untraceable; and a fourth, broad on our starboard bow, which looked the most promising of all. We counted seven pools, or lakelets, in addition and similar to our own—three to the northward, one to the eastward, two to the southward, and one to the south-westward of our own; but the one in which the Mercury floated seemed to be the largest of them all. The reef appeared to be composed wholly of rock, covered for the most part with weed, but with broad expanses of sand here and there, interspersed with mud banks; and its height above the ocean level seemed to vary from about a foot to ten or fifteen feet, with occasional isolated hummocks, rising perhaps as high in some cases as forty feet. With the aid of the telescope we were able to perceive that considerable quantities of fish had been stranded and left to perish by the sudden upheaval, and the appearance of them caused me a slight spasm of alarm on the score of our health, which was only partially dissipated by the fact, to which Gurney directed my attention, that already great flocks of sea birds had appeared and were busily devouring them.

We remained aloft for more than an hour, studying the reef and, so far as I was concerned, making copious notes and a rough sketch map of it in my notebook, and then descended to the deck, having come to the conclusion that the only thing to be done was to make a systematic exploration of the reef, and especially of the channels, by means of one of the boats.

For this purpose we selected the larger of the two quarter boats, a very handsomely modelled craft of twenty-six feet long by six feet beam, with a keel nearly eight inches deep in midships, and rigged as a fore-and-aft schooner. She had been the late captain’s fancy boat, used by him for sailing ashore from open roadsteads, and was fitted with air-chambers forward and aft and under each of the thwarts, thus being converted into a sort of unsinkable lifeboat. She was therefore in every respect eminently suitable for the duty upon which we proposed to employ her.

I was rather afraid that, upon learning our purpose, Grace Hartley might express a desire to accompany us; and this would be somewhat awkward, in view of the rough work which might possibly lie before us; but to my relief she expressed herself as perfectly content to remain aboard alone, upon being assured that no harm could possibly happen to the ship. We therefore bent the ensign on to the main signal halyards, showed her how to hoist it, and directed her to run it up to the main truck in the event of anything occurring to render our immediate return to the ship necessary. Then, having hastily stowed away a few biscuits and a piece of cold salt beef in the boat’s stern locker, and placed a small breaker of fresh water in her for ballast, we lowered her to the water, brought her to the gangway, rigged her, and got away about six bells in the forenoon watch, Grace waving her farewells to us from the poop.

We decided to begin our exploration by examining the channel which opened on the south side of the basin, as that was the widest and the most promising of the four. Accordingly, upon leaving the ship, we brought the boat close to the wind on the port tack—which just enabled her to point fair for the mouth of the channel—and at once proceeded to take soundings. But we had not been under way five minutes before we found that, under whole canvas, the boat travelled much too fast to enable us to sound with the frequency and accuracy that I considered necessary; we were consequently obliged to take in the foresail altogether, and sail the boat under the single-reefed mainsail and jib, at least during the outward journey.

Within ten minutes of leaving the ship we glided into the channel which it was our intention to explore, and found ourselves slipping along a waterway ranging from two hundred to a thousand feet in width, with an average depth of about five fathoms. The sides of the channel were very rough and irregular, its direction was also exceedingly erratic, varying from east-south-east to south by west. This irregularity of direction was the worst feature of the channel; for, with the prevailing direction of the wind at about due east, there were stretches of the channel looking so close into the wind’s eye that the ship could never be sailed through them. True, some of these stretches were so short that Gurney believed the Mercury could be carried through them by making a half-board; but this would be a somewhat hazardous experiment, unless the wind chanced to veer a point or two in our favour, while, even then, there were other stretches that could only be traversed by kedging. But, apart from this disadvantage, there was nothing to find fault with, the channel being everywhere wide enough to permit the passage of the ship, and the depth in it never less than twenty feet, with a fine sandy bottom.

We traversed this channel for a distance of about nine miles, during which the general trend of it might be said to be south-east, and then we arrived at a point where it not only widened out, but also abruptly took a south-south-west direction, to our great delight. For if the ship could by any means be coaxed as far as this, she could then proceed with a free wind. But, alas for our hopes, we had not traversed more than another mile and a half before we found ourselves in a cul-de-sac, the channel coming to an abrupt end.

This was a very severe disappointment to us, for after travelling so far, and meeting with so few difficulties, we were already beginning to congratulate ourselves upon having found a way of escape at the first attempt. However, there was no use in worrying about it, the only thing to be done was to retrace our steps and try one of the other channels.

It must not be supposed that the channel which we had thus traversed to its extremity was unbroken; on the contrary, there were several other channels branching out of it to right and left at various points in its length, two of which at least—one about three miles back, and another five miles back—had struck us as not altogether unpromising, and our idea was now to return and examine these. But before starting upon our return journey Gurney made a suggestion that was destined to exercise an extraordinary influence upon our future. As we lay hove-to in the cul-de-sac, discussing the question of what should next be done, our attention had been more than once attracted toward a large hummock of rock rising some thirty or forty feet above the general level of the reef, at no great distance from the margin of the channel; and Gurney’s proposal was that, before attempting anything else, we should land, make our way to the hummock, climb it, and ascertain whether any observations of value were to be made from its summit. The proposal had so much to commend it that it was agreed to forthwith. Laying the boat alongside the rock at a convenient spot, we all three landed, and set out to walk across the reef. The hummock in question was only some two hundred yards from the margin of the channel; we therefore soon reached it, despite the difficulty of the going, the surface of the reef proving to be exceedingly rough, and covered for the most part with weed so terribly slippery that it was positively dangerous to attempt to walk upon it. When at length we reached the summit of the rock we were not much better off; for although we could see from thence a great deal more of the reef than was to be observed from the boat, we made no discoveries that were likely to facilitate our escape.