"Well, since you fear that method, I am ready to convince you of the facts by any rational course you may select, and if you wish to stake your entire argument on the general statement that a stream of water can not rise above its head, I will accept the challenge; but I insist that you do not divulge the nature of the experiment until, as you are directed, you make public my story." "Of course a fluid can be pumped up," I sarcastically observed. "However, I promise the secrecy you ask." "I am speaking seriously," he said, "and I have accepted your challenge; your own eyes shall view the facts, your own hands prepare the conditions necessary. Procure a few pints of sand, and a few pounds of salt; to-morrow evening I will be ready to make the experiment." "Agreed; if you will induce a stream of water to run up hill, a fountain to rise above its head, I will believe any statement you may henceforth make." "Be ready, then," he replied, "and procure the materials named." So saying he picked up his hat and abruptly departed. These substances I purchased the next day, procuring the silver sand from Gordon's pharmacy, corner of Eighth and Western Row, and promptly at the specified time we met in my room. He came, provided with a cylindrical glass jar about eighteen inches high and two inches in diameter (such as I have since learned is called a hydrometer jar), and a long, slender drawn glass tube, the internal diameter of which was about one-sixteenth of an inch. "You have deceived me," I said; "I know well enough that capillary attraction will draw a liquid above its surface. You demonstrated that quite recently to my entire satisfaction." "True, and yet not true of this experiment," he said. "I propose to force water through and out of this tube; capillary attraction will not expel a liquid from a tube if its mouth be above the surface of the supply." He dipped the tip of a capillary tube into a tumbler of water; the water rose inside the tube about an inch above the surface of the water in the tumbler. "Capillary attraction can do no more," he said. "Break the tube one-eighth of an inch above the water (far below the present capillary surface), and it will not overflow. The exit of the tube must be lower than the surface of the liquid if circulation ensues." He broke off a fragment, and the result was as predicted. Then he poured water into the glass jar to the depth of about six inches, and selecting a piece of very thin muslin, about an inch square, turned it over the end of the glass tube, tied it in position, and dropped that end of the tube into the cylinder. "The muslin simply prevents the tube from filling with sand," he explained. Then he poured sand into the cylinder until it reached the surface of the water. (See Figure 23.) "Your apparatus is simple enough," I remarked, I am afraid with some sarcasm.