"The instant the bottle was released the discharged air was immediately attracted toward it, and gradually surrounded it entirely. It was like a little planet with an atmosphere of its own, as they fell back to the Earth together." "But I couldn't see it; I had to pump," I complained. "We must do it again." "We shall soon have our bottled things all emptied out on plates to dry up and spoil," he objected. So I emptied a biscuit tin this time, and delaying for no message, I put it in the discharging cylinder. Then I bent over the port-hole and gave the signal for the pumping. As I thrust out the tin I was astonished to see the lid pop off the first thing. The quick expansion of the air inside it did that. This air, as well as the air from the discharge pipe, seemed to flee from it instead of surrounding it, as the doctor had said. I continued watching so long that he finally said,-- "Hasn't it fallen out of sight yet?" "No; it is not falling away swiftly as the air does. It is following the projectile! It is not gathering any air about it as you said it would. It does not quite keep up with us; but considering our speed, it is doing remarkably well!" The doctor was not inclined to believe me until he had looked for himself. He watched and pondered for a minute or two. Then his surprise ceased, and he spoke in that assured way which always irritated me. "Quite natural, after all," he said. "That biscuit can is made of thin sheet-iron with a surface coating of tin. The iron has become magnetized by induction, and the Earth repels the can just as it repels us. It will follow us to the dead-line, and probably on to Mars, unless the sheet-iron loses its polarization. If we had cast out a thing of solid iron, it would rush ahead of us, instead of falling a little behind, as this does, for it would have no dead weight to carry. But we could not put such a thing out of the rear end, for no force would make it fall that way. If we put it out of the forward port-hole, it would beat us in the race toward Mars." I remarked to the doctor that the air-pump seemed to be incorrectly built, for its action was strangely difficult in the reverse manner that it should have been. The down strokes went by themselves with a quick snap, but the up strokes were as if against pressure, and the moment the handle was released it flew down again. He had not tested the pump at the surface, as it was of a well-known make, but it certainly seemed to work backwards. Moreover, the more nearly we had a compartment emptied of air, the more difficult the pumping should become, but here again the reverse seemed to be the case, for the longer we worked the easier the up strokes became.