After all, why was this pumping necessary? If a way could be devised to open a valve, all the air would rush out of my compartment as easily as beer runs out of a bung-hole. In fact, it did rush out a little at a time, which is what made the handle go down of itself. But any such new valve would have to be automatically closed, as it would be manifestly impossible to enter and shut it. I kept on thinking, and finally began examining the partition between the compartments. There seemed to be several long screws that went quite through it. "Doctor, did you ever hear of those wise people who, after every freshet, shipped the surplus water down the river in boats? Well, it strikes me this air-pumping is just about as useless labour. Help me pull in the bulkhead and I will show you something." I went at once to the cylinder we used for discharging things from the projectile. With a pair of pliers I chipped off a small piece of the edge of the closing lid in two places, one near each end. This made two little irregular holes into the cylinder about eight inches apart. Then I pushed it half way out, so that one hole was outside and the other inside. Of course the air rushed through the inner hole into the cylinder, and thence through the outer hole to the exterior. "Shut that thing!" cried the doctor, when he saw what I had done. "Do you wish to suffocate us? That will let the air out perfectly, but how are you going to close it to admit the condensed air?" "People unskilled in these matters are so hasty!" I said rather sarcastically. "Wait until I have finished and you will see." I found he had a screw-driver, and I loosened one of the long screws and enlarged the half of its hole toward my compartment. Then I whittled a block of soft wood, so that it would slide smoothly into this half of the hole. Driving the screw home again, I just allowed its tip to enter the end of the block. Then I fastened a piece of stout twine to the cylinder and the other end to the block of wood, which was almost opposite it. Pushing the cylinder half way out, I made the twine taut, and hastening into the doctor's compartment, I thrust in the bulkhead. The air was rapidly escaping. Waiting long enough for all of it to have leaked out, I then unscrewed the long screw, which gradually drew in the block of wood and the twine, and thus pulled the cylinder into the projectile so that there was no connection with the exterior. Then the doctor let in the condensed air to a barometric pressure of twenty-six, and the whole operation was over in a few minutes. My compartment must have been almost a complete vacuum. When it was over, I cried rather triumphantly to the doctor,-- "There, you see, one doesn't need a steam pump to make the water run over Niagara! At this distance from the surface, nature abhors a gas and prefers a vacuum!" He was inclined to be rather sulky at first, but he really did not like pumping any better than I did.