Since the two columns of water A B and F C, by the supposition, will be suspended by some power acting within the tubes they are contained in, they cannot determine the water to move one way, or the other. But the column B F, having nothing to support it, must descend, and cause the water to run out at C. Then the pressure of the atmosphere driving the water upward through the orifice A, to supply the vacuity, which would otherwise be left in the upper part of the tube B C, this must necessarily produce a perpetual motion, since the water runs into the same vessel, out of which it rises. But the fallacy of this reasoning appears upon making the experiment. [Illustration] Exp. 1. For the water, instead of running out at the orifice C rises upwards towards F, and running all out of the leg B C, remains suspended in the other leg to the height A B. Exp. 2. The same thing succeeds upon taking the siphon out of the water, into which its lower orifice A had been immersed, the water then falling in drops out of the orifice A, and standing at last at the height A B. But in making these two experiments it is necessary that A G the difference of the legs exceed F C, otherwise the water will not run either way. Exp. 3. Upon inverting the siphon full of water, it continues without motion either way. The reason of all which will plainly appear, when we come to discover the principle, by which the water is suspended in capillary tubes. Mr. Hawksbee's observation is as follows: Fig. 2. Let A B F C be a capillary siphon, into which the water will rise above the level to the height C F, and let B A be the depth of the orifice of its longer leg below the surface of the water D E. Then the siphon being filled with water, if B A be not greater than C F, the water will not run out at A, but will remain suspended. This seems indeed very plausible at first sight. For since the column of water F C will be suspended by some power within the tube, why should not the column B A, being equal to, or less than the former, continue suspended by the same power. Exp. 4. In fact, if the orifice C be lifted up out of the water D E, the water in the tube will continue suspended, unless B A exceed F C. Exp. 5. But when C is never so little immersed in the water immediately the water in the tube runs out in drops at the orifice A, though the length A B be considerably less than the height C F. Mr. Hawksbee, in his book of Experiments, has advanced another observation, namely, that the shorter leg of a capillary siphon, as A B F C, must be immersed in the water to the depth F C, which is equal to the height of the column, that would be suspended in it, before the water will run out of the longer leg.