Ordinarily cold air falls to the ground; but not always, for under certain conditions cold, heavy air may actually rise, displacing warm, lighter air. But such conditions can be explained and there is no contradiction of the fundamental law that if acted on only by gravity, cold air, being denser, will settle to the ground and warm air, being lighter, will rise. And there must be a certain relation between the height of the level from which the cold air falls and the level to which the warm air rises. In other words, we have to apply the laws of falling bodies since a given mass of air, although invisible, is matter and as subject to gravity as a cannon ball. One of Galileo's most ingenious experiments consisted in swinging a pendulum and then by means of a nail driven in various positions intercepting the swing. He found that the bob always rose to the same level whatever circuit it was forced to take. But Galileo did not know what every schoolboy to-day knows, that air exerts pressure and is subject to physical processes like other matter, else he would certainly have given to the world a delicate air pendulum; and devised experiments on the movement of air that would have opened men's eyes to the fascinating flow and counter-flow of the air, even on a seemingly still night, one favorable for the formation of frost. The problem of the moving air mass, however, is more complicated than it looks. For with the air is mixed a quantity of water vapor. In a strict sense they are independent variables, and the view set forth in most text-books that air has a certain capacity for water vapor is misleading. We seldom meet with pure, dry air. A cubic meter of such a gas mixture would weigh 1,247 grams, at a temperature of 283 degrees A. (50 degrees F.). If chilled ten degrees, that is, to the freezing point of water, it would weigh 46 grams more. So that by cooling, air becomes denser and heavier. A cubic meter of a mixture of air and water vapor at saturation, at the first temperature above mentioned weighs only 1,242 grams, or five grams less, and if this were cooled ten degrees the mixture would weigh three grams less than the same volume of pure dry air. We see that in each case the mixture of air and water vapor weighs less than the air by itself. One would think that by adding water vapor which, while light, still has weight, the total weight would be the sum of both. It really is so, notwithstanding the above figures, and the explanation of the puzzle is that there was an increase in pressure with expansion, so that the volume of the air and saturated vapor was greater than one cubic meter. Since then a cubic meter of air and saturated vapor weighs less than a cubic meter of dry air at freezing temperature, speaking generally, we may expect moist air to rise and dry air to fall. Consequently, if in addition to falling temperature there is also a drying of the air, we shall have an accelerated settling or falling of cold