In each case the basic idea and error was in supposing that by some possibility the descent of a liquid through a given distance could be made to deliver more energy than would be required to elevate the same quantity of liquid the same distance. As a matter of fact, the descent of a liquid, the same as any other weight, through a given distance represents exactly the amount of energy necessary to elevate the same weight of liquid through the same distance measured vertically. Some loss by friction of the liquid in the containing tubes is inevitable as well as from friction in the working parts of the mechanism. Therefore, as this loss continues, some outside energy must be supplied. If all friction could be eliminated (which is an impossibility) and if the liquid were started in motion, the motion would be constant, but no energy could be taken from it for running other machinery without reducing the motion. There have been many arguments on this subject. We select one which was elicited by the publication in "Mechanics' Magazine" of an account of the device of the author of the "Voice of Reason." This argument was published in "Mechanics' Magazine" in 1831, and is as follows: I am induced to make an attempt to demonstrate the utter impossibility, under any circumstances, of making a water-wheel that will supply itself instead of having any surplus power. The accompanying drawing represents part of an overshot wheel in section, the buckets only part filled, by which the whole of the water expended continues to act through a greater portion of the circumference than it otherwise would do. The area of the vertical section of the complement of water to each bucket is made 40 inches; and taking the breadth of the wheel at, say 28 2/3 inches, gives 40 lbs. as the weight of water in each bucket; therefore, as there are 12 buckets containing 40 lbs. each, No. 13 30 lbs., and No. 14 only 20 lbs., altogether making a total of 530 lbs. acting on the wheel at the same time;--to show clearly all the effect that can be expected from this, I have divided the horizontal radius into a scale of 40 equals parts (there being 40 lbs. in each bucket); and from the gravitating centre of the fluid contained in each is drawn a perpendicular to the scale, where the effective force, or weight in each bucket, may be read off as on the arm of a common steelyard. The weights will be found as follows, viz:-- No. Lbs. 1 21½ 2 26¼ 3 30½ 4 33¾ 5 36¾ 6 38¾ 7 39¾ 8 40 9 39½ 10 38 11 35¾ 12 32½ 13 21 14 12