"Well," remarked the elderly Morey, shivering a bit in the chill air of the room, "loaded dice have long been noted for their ability to make money, but I don't see how that explains that working model of an Arctic tornado. _Burr_ it's still too cold in here. I think he'll need considerable area for heat absorption from the sun, for that engine certainly does cool things down! What's the secret?" "The principle is easy enough, but I had considerable difficulty with the application. I think it is going to be rather important though--" "Rather important," broke in the inventor's father, with a rare display of excitement. "It will be considerably more than that. It's the biggest thing since the electric dynamo! It puts airplanes in the junk heap! It means a new era in power generation. Why, we'll never have to worry about power! It will make interplanetary travel not only possible, but commercially economical." Arcot junior grinned broadly. "Dad seems to think the machine has possibilities! Seriously, I believe it will antiquate all types of airplanes, prop or jet. It's a direct utilization of the energy that the sun is kindly supplying. For a good many years now men have been trying to find out how to control the energy of atoms for air travel, or to release the energy of the constitution of matter. "But why do it at all? The sun is doing it already, and on a scale so gargantuan that we could never hope nor desire to approach it. Three million tons of matter go into that colossal furnace every second of time, and out of that comes two and a half decillion ergs of energy. With a total of two and a half million billion billion billions of ergs to draw on, man will have nothing to worry about for a good many years to come! That represents a flood of power vaster than man could comprehend. Why try to release any more energy? We have more than we can use; we may as well tap that vast ocean of power. "There is one thing that prevents us getting it out, the law of probability. That's why Dad mentioned loaded dice, for dice, as you know, are the classical example of probability when they aren't loaded. Once they are loaded, the law still holds, but the conditions are now so changed that it will make the problem quite different." Arcot paused, frowning, then resumed half apologetically, "Excuse the lecture--but I don't know how else to get the thought across. You are familiar with the conditions in a liter of helium gas in a container--a tremendous number of molecules, each dashing along at several miles a second, and an equal number dashing in the opposite direction at an equal speed. They are so thickly packed in there, that none of them can go very far before it runs into another molecule and bounces off in a new direction. How good is the chance that all the molecules should happen to move in the same direction at the same time? One of the old physicists of Einstein's time, a man named Eddington, expressed it very well: