Wait a moment: just let us lay this one aside, but not forget it, as we shall want to refer to it in a few moments, and try another experiment which has some bearing upon the subject. I have here a tube just one foot square and two feet long, and one foot from the bottom there is what we will suppose to be an air-tight piston that can be moved without friction. Now, suppose we heat that air 490° (for the sake of easy remembering, say 500°); this would just double its volume--it would then be two cubic feet in size instead of one. Now, suppose that, instead of letting this air expand, we should put a weight on it, so as to keep it in its place, how much do you think we should have to place on? Two thousand one hundred and sixty pounds, or about one ton. Now, what do we find these 2160 pounds to represent? It is the weight of a column of atmosphere with a base of one foot square, or fifteen pounds multiplied by 144 square inches--it is the weight that would rest upon the piston if all the air was taken out from under it. Therefore, if you add about 500° of heat to a cubic foot of air, it makes it two cubic feet of air; or, if you attempt to keep it from expanding, you must put a ton weight upon it. Mark one thing, however, if it takes ten ounces of coal to heat that air to 490°, which we do by piling our ton weight upon it, it will take fourteen ounces of coal if we allow it to expand to two feet. In the former case, where the air remains stationary, it had done no work. It was ready to go to work, but it had not commenced. But in the case of its expansion, it had done a great work. What was it? Why it had lifted that ton of atmospheric air one foot in height, and that work was what used up the difference between ten parts and fourteen parts of coal (I don't trouble you with fractions). You see, therefore, to make the air quit the earth and ascend into the upper regions, requires a positive power, the same as it does to drive some poor simple people away from the fire on a cold day. We often say that, by heating air, we give it power to ascend; instead of which heating it destroys its power to maintain its position. It weakens--enervates it--so that its neighbors easily drive it out and take its place. One cubic foot of air, diluted to two feet, would be driven about two miles and a half high before it found any body as weak as itself, for every 350 feet in height, in round numbers, the pressure diminishes by an amount equal to one degree, or forced under water thirty-four feet reduces it to one-half its bulk. Now, let us go back and finish our syphon, or flue experiment.