That didn’t fit. Anyone down here should have been looking for that wreck, not running across it by accident. I’d been expecting a working party sent out by the people under that cover.

Well, there’s more than one thing I haven’t been expecting about this business. Stop with the working hypotheses, brother, you haven’t enough data even for that yet. Just watch (I don’t even address myself by name).

So I watched. I watched her swim around the shattered bow, and into it and out again, and over it. Then I watched her unlimber an object which turned out to be a light, which had been clipped to her suit belt, and swim inside once more. That worried me a little; the camouflage for the tank had not been designed for that sort of inspection. Its clamps, its launching springs -

She came out again, with no more visible signs of excitement than before, and at that point something else dawned on me. It was a very minor point compared with what I had already seen — at least, it seemed minor when I first noticed it; as I thought, it became more and more a major puzzle.

Her scuba suit was, as I said, quite ordinary except for helmet and ballast. Its ordinariness included a small tank between the shoulders, whose upper end just touched the helmet and was presumably connected with it, though I could see no piping. All this was reasonable. The jarring note was that there were no bubbles.

Now, I’m familiar with rebreather systems, and I know about chemical supplies — mixtures of alkali metal peroxides and superoxides which react with water to give free oxygen and pick up carbon dioxide. I know them well enough to know that they must have, besides the chemical container and mixer systems, a sort of ‘lung” — a variable-volume, ambient-pressure gas bag or tank — with the supply chemicals between it and the user’s own lungs. The exhaled gas has to go somewhere until it’s ready to be inhaled again. That ‘lung” must have a volume great enough to take all the air a swimmer can exhale at one breath — in other words it must have as much volume, or nearly as much, as his own inflated

lungs. There was no such bag visible on this swimming outfit, and the back tank was not nearly large enough to have contained one. It seemed, therefore, that the unit did not involve a chemical oxygen supply; and unless some sort of microscopic pump was taking the gas as fast as she exhaled and squeezing it back into another part of that little tank at fantastically high pressure, there should be exhalant bubbles. I couldn’t see any reason for such a recovery system, but I couldn’t see any bubbles either. I had already been bothered about what gas mixture she could be breathing — at this pressure, half of one percent oxygen would have burned her lungs out, and there was nothing I knew of which could be used to dilute it. Even helium was soluble enough down here to make decompression a job of many hours.

It crossed my mind for a moment that people might be living permanently under this pressure, breathing a nearly pure helium atmosphere with a fraction of a percent oxygen in it; but if that were the story, I still couldn’t see why that girl’s suit didn’t give out bubbles. Granted there might be every economic reason to recover helium, there are engineering problems which I still don’t think could be completely solved.

No. All hypotheses inadequate. Keep on observing. Facts so far are only that she seems to be living and moving normally in a closed system at outside pressure, and that the pressure in question — skipping the old superstition about flattening a human body — is quite high enough to mess up any biophysical or biochemical processes involving gas dynamics.

There wasn’t much more to observe, though. The girl clipped the light back on her belt, took a last glance at the wreckage and began swimming away from it. She didn’t go back the way she had come, but continued on to my right, slanting away from the lighted region. In a few seconds she had disappeared, though I knew she couldn’t be very far away yet.

It seemed likely that she was off to get help in moving the wreck off the tent roof. How long before she would be back with it was anyone’s guess. There might be a tent entrance a few hundred yards away, or there might not be one for several miles. The former seemed a trifle more likely, but I wasn’t going to risk money on the question.

Just my future.

She might have noticed the gear that had held and launched my tank; she wouldn’t have had to be much of an actress to hide an expression of suspicion under the circumstances. If she had noticed and reported it, those who came with her were going to be very curious about the whole area. The outside of the tank was deliberately a little irregular in outline so that it wouldn’t be too obviously artificial, but it was not going to fool anyone who took a really good look at it. Maybe it would be better if I moved a little farther away. I wasn’t concerned with personal safety; I could always get away, but I wanted to see as much as possible before that became necessary.

So I told myself.

Moving would be a slow process; traveling ability was not really a design feature of the tank. There were two dozen of the legs, and I had enough stored power to retract them against their springs several thousand times (that had taken argument), but I had not been born a sea urchin. I had had a little practice rolling the thing around under water, but the purpose of the rig was to let me juggle into a better observing position, not to keep out of the way of searchers. If I were found, my only real recourse would be to drop ballast and start for the surface. That was a once-only operation, and I didn’t want to resort to it before I really had to. There was still some hope, I figured, of deciding what was going on down here in some detail.

Maybe it’s courage, or maybe just natural optimism.

I began working the legs, hoping that no instruments in the neighborhood were recording the D.C. pulses as I turned the retraction solenoids on and off. I had found during practice that I could climb a slope of five or six degrees if the bottom were hard enough to give the ‘feet” any resistance, but that near the limit of steepness the going tended to be tricky. If I overbalanced and started downhill again it took very fast work with just the right legs to stop the roll. The sphere had a respectable moment of inertia. Because of its outer irregularity, some positions were naturally more stable than others, and some were much less. Just now I was wishing that I had spent more time in practice, though I consoled myself with the thought that the boss wouldn’t have authorized the energy expenditure anyway.

I had worked my way between thirty and forty yards farther up the slope, with only one mistake that cost me any real distance, when the party I expected showed up.

It wasn’t a large one — four in all. One could have been, and probably was, the girl I had seen before; the other three seemed to be men, though it was hard to tell at this greater distance. One of the new ones was towing a piece of equipment about three feet long, cylindrical in shape, and a little more than a foot in diameter. It had a slight negative buoyancy, which was understandable — they’d make sure that nothing which got loose would find its way to the surface.

They swam over to the wreck, and two of them began pulling lengths of line from the cylinder. They attached these to convenient parts of Pugnose, while the third man pulled from the other end of the cylinder something that looked like a heavy bundle of netting with a collapsed balloon inside it. When the other lines were made fast he manipulated something on the cylinder, and the balloon began to inflate slowly. The wreckage didn’t have much submerged weight, and it wasn’t long before the balloon had it hoisted clear of the roof. Then all four of them got on the far side and began pushing it, swim fins fluttering violently.