I also conveyed on board some candles, flint and steel, matches, and provisions, and as the last thing took in the stern anchor, so that the boat lay with the tide, tailing up stream. The other hawser I conveyed also—by means of the canoe and with a boathook—under the forward tank, so that I held the end within the interior, and could cast it off at anytime. It was by the sun about eleven o'clock when I gave one glance around me, and, standing on the last round of the ladder, I drew the manhole cover over my head and commenced screwing it down on the inside, which having done I lighted several candles, although I had a fair light reflected from the water and the bottom of the cove, formed of white sea-sand, directly beneath me, and distant, I should judge by the state of the tide, some twenty to twenty-five feet. My goats had become so accustomed to the boat that they showed little surprise at the rather dim light, and stood ready to perform their part whenever I should put them to their customary task.

My heart beat rapidly, not with fear, but with excitement and expectation. Here I was, already shut out from the outer air, and in a little world of my own. I hesitated to complete my experiment, and before going further I turned to my provisions and took a good long drink of claret wine to strengthen my courage and steady my nerves. If I was in a scrape I could get out of it, but my poor goats! they, I was afraid, would have to pay for any error in judgment on my part. Having regained perfect composure, I made up my mind to make the first test of the practical value of my boat, and that was to see if I could descend to the bottom of the ocean, that lay beneath me. By moving around I felt convinced that my calculations about the centre of gravity had been correct, and I felt that the boat would not capsize. It was remarkably stiff and steady, and would, I felt confident, remain so when submerged. This bugbear was already off my mind, and gave me confidence to proceed. So, moving to the place on the starboard side where all my pipes and stopcocks were congregated, I commenced by opening the stopcock of the water-pipe, which, as I had foreseen, brought no perceptible change. Some little water rushed into the tanks, but only what was sufficient to compress the air to the extent of the weight of the superstructure of the boat. This experiment did not sink it one particle; its buoyancy remained exactly the same, for the same air remained in the tanks, although compressed, and was not able to escape on account of the position of the outlet of the pipe that had opened communication between it and the water, pointing, as it did, directly to the centre of the earth. After waiting a little, and seeing that this all worked well, I placed my hand upon the stopcock of a more important pipe, namely the tank air-pipe, which led from the tanks to the outer surface of the superstructure.

Now, or never! Upon turning this cock I should descend or my theory would be incorrect. The moment my hand opened this valve the air would be expelled by the pressure of the boat upon the water, conveyed to the air in the tanks by the water-pipe, which was already open; and, as it was expelled, so the buoyancy of the boat would be decreased, and I should descend. The fatal moment had come, and with a firm hand I opened the tank air-pipe, and plainly heard the escaping air, the incoming water, and felt the boat descending, and saw the sandy bottom apparently approaching me. I cut off the discharge of the tank air-pipe, and with a slight rebound the boat arose again a few inches towards the surface, simply regaining its true position in equilibrio, that it had for a moment passed, by the momentum of its descent. By little turns of this stopcock I discovered, as I expected, that I could move the boat in a descending direction even an inch at a time. The movement was a perfect fascination, but each delivery of air was bringing me nearer the bottom, and as yet I had tried no means of rising again to the surface.

When I had gotten to within about six feet of the former I thought it time to see if I could again rise towards the surface. I was well aware that, having used up this air, it was so much loss to me, but I was in hopes to be able to replace it; and even if I could not replace it to make the boat rise to near the surface without it. I could do this in one way, by casting overboard the anchors and weights lying upon the wooden shelves; but this, if done to any great extent, might cause the capsizing of the whole affair. No; I had a better way than this, and at it I went. In the first place I closed the water-pipe, and then, having opened the screw-valve in the connected tanks, I screwed upon it the pump and commenced discharging the water from them—that had run in to take the place of the discharged air—into the water of the ocean, which formed, in one sense, the interior flooring of my boat.

To make this pump work I of course opened, and left open, the pump-pipe, so that the air from the interior rushed in and filled the tank as fast as the pump discharged the water, and at each stroke of the pump, after the first few, the boat, as I had hoped, began to rise; the water, having been just so much ballast to carry it down, being discharged by the pump, was just so much thrown overboard in weight to allow it to rise. By persistent pumping I made my boat rise quite near the surface, but not to the buoyant position it at first maintained, for I had in my descent used up considerable of the air in the tanks, which I had as yet not replaced, or rather what I had used from them had been replaced from the air of the interior when I pumped out the water, which I could only do by allowing the connection between the tanks and the interior to be open, so as to make the pump work. In short I had lost just so much buoyancy as was equal to the escaped air; but still I had been able to make the boat descend and ascend.

These experiments took me over two hours, and I commenced to feel the need of new air, and to notice that my candles began to burn a little dimly. I was thus warned that my air was being used up and charged with carbonic gas, and that it was time for me to renew it. So I unscrewed the pump and closed the valve, opened the water-pipe, and placed my hand upon the tank air-pipe and prepared to descend. One effect I should have noticed of my loss of air, and that was that the water in the interior of the boat rose considerably, and a large portion of the tanks was now submerged. A few turns of the stopcock of the tank air-pipe carried me near to the bottom, where I desired to be, to try my last and most important experiment.

Arriving to within a few feet of the bottom I rigged my spray-wheel, and connected it with the drum of the treadmill and set the goats at work. And it was time, for my breathing had become oppressive, and the animals themselves seemed dull and frightened. I had waited almost too long. My candles also commenced to burn more dimly, and I prepared to take my plunge into the water and come up outside of the boat should my experiment now fail. But wonder of wonders! my spray-wheel made but a few revolutions, dashing large quantities of minute spray into the interior by its rapid motion, before my lungs were relieved, the candles renewed their brilliancy, and the goats recovered from the lassitude under which they had a moment before seemed to be laboring.

The problem was solved. I had made my own air. I could remain below the surface as long as I desired. Everything about me was rather damp and moist from the dashing of spray about the interior, and several of the candles, that I had not protected, were put out; but two, in the extremity of the boat, were preserved, and now that my problem was solved I did not again light the former, the two remaining ones being all-sufficient. And in fact I did not need them; my own lungs, I found, were sufficient as a guide to tell me in future when to renew the air. Still it was fascinating to see these two candles burning brilliantly that had but a moment before been so dim. The reflected light from the pure sandy bottom just below me was amply sufficient for all purposes.

I imagined, by the slight shadow that the boat cast on the bottom beneath, from the brilliant sun that I knew was shining overhead, and from counting up in my mind all I had done since leaving the surface, that the air had lasted me, as nearly as I could judge, two hours; and that seemed to be the extreme limit to which I could go and not renew it. I also knew by the quantity of tallow consumed in the candles that it must be nearly that amount of time. I also noticed that the spray-wheel had not only purified my air, but that whilst it was in operation the boat had slightly ascended, proving that I had gained a lighter gas for the nitrogen and carbonic acid precipitated.

One more thing remained to be tested, and I should feel that my labors were complete. In the first place I made the boat ascend as far as possible, by means of the pump and stopcocks, as before described, and then I went to work and rigged my propeller and set the goats at work. I got the boat as near the surface as possible before communicating motion to it, so as not to run against any obstacle if possible. But then the body of water in which I was submerged was so pure, and free from anything of that nature, that there was little danger after all. With a feeling of confidence that I had not had in all the other experiments, I cast off the hawser affixed to the anchor that held the boat, and started the goats. Mechanics did not trouble me, and it was with no surprise, but only gratification, that I saw by the bottom that the boat was moving forward, and that it readily obeyed the helm. I turned it completely around by the tiller, and made an excursion of fully half a mile, I should think, up Stillwater Cove, once in a while getting out of the channel, when by stopping the goats and reversing the propeller I was able to back into the channel again, and finally to turn around by a series of forward and backward motions till I again arrived at the place from which I had started, which I knew by the anchor lying in mid-channel. By observation of the bottom I should say that the boat was propelled at least three miles an hour, which was sufficient for all my purposes.