He arose and abruptly departed. It was near morning, and yet I sat in my room alone pondering the story of my unique guest until I slept to dream of caverns and seances until daylight, when I was awakened by their vividness. The fire was out, the room was cold, and, shivering in nervous exhaustion, I crept into bed to sleep and dream again of horrible things I can not describe, but which made me shudder in affright at their recollection. Late in the day I awoke. On the following evening my persevering teacher appeared punctually, and displayed a few glass tubes and some blotting or bibulous paper. "I will first show you that liquids may change their levels in opposition to the accepted laws of men, not contrary to nature's laws; however, let me lead to the experiments by a statement of facts, that, if you question, you can investigate at any time. If two vessels of water be connected by a channel from the bottom of each, the water surfaces will come to a common level." He selected a curved glass tube, and poured water into it. The water assumed the position shown in Figure 11. [Illustration: FIG. 11.--A A, water in tube seeks a level.] "You have not shown me anything new," I said; "my text-books taught me this." "True, I have but exhibited that which is the foundation of your philosophy regarding the surface of liquids. Let me proceed: "If we pour a solution of common salt into such a U tube, as I do now, you perceive that it also rises to the same level in both ends." "Of course it does." "Do not interrupt me. Into one arm of the tube containing the brine I now carefully pour pure water. You observe that the surfaces do not seek the same level." (Figure 12.) [Illustration: FIG. 12.--A, surface of water. B, surface of brine.] "Certainly not," I said; "the weight of the liquid in each arm is the same, however; the columns balance each other." "Exactly; and on this assumption you base your assertion that connected liquids of the same gravity must always seek a common level, but you see from this test that if two liquids of different gravities be connected from beneath, the surface of the lighter one will assume a higher level than the surface of the heavier." "Agreed; however tortuous the channel that connects them, such must be the case." "Is it not supposable," said he, "that there might be two pockets in the earth, one containing salt water, the other fresh water, which, if joined together, might be represented by such a figure as this, wherein the water surface would be raised above that of the brine?" And he drew upon the paper the accompanying diagram. (Figure 13.) "Yes," I admitted; "providing, of course, there was an equal pressure of air on the surface of each." [Illustration: FIG. 13.--B, surface of brine. W, surface of water. S, sand strata connecting them.]