Again let us imagine ourselves up in the place where gravitation has no effect. Suppose we lay a nail on the surface of a bowl of water. It stays there and does not sink. This does not seem at all surprising, of course, since the nail no longer has weight. But when we put a cork in the midst of the water, it stays there instead of floating to the surface. This seems peculiar, because the less a thing weighs the more easily it floats. So when the cork weighs nothing at all, it seems that it should float better than ever. Of course there is some difficulty in deciding whether it ought to float toward the part of the water nearest the floor or toward the part nearest the ceiling, since there is no up or down; but one would think that it ought somehow to get to the outside of the water and not stay exactly in the middle. If put on the outside, however, it stays there as well. A toy balloon, in the same way, will not go toward either the ceiling or the floor, but just stays where it is put, no matter how light a gas it is filled with. The explanation is as follows: For an object to float on the water or in the air, the water or air must be heavier than the object. It is the water or air being pulled under the object by gravity, that pushes it up. Therefore, if the air and water themselves weighed nothing, of course they would be no heavier than the balloon or the cork; the air or water would then not be pulled in under the balloon or cork by gravity, and so would not push them up, or aside. [Illustration: FIG. 11. The battleship is made of steel, yet it does not sink.] WHY IRON SHIPS FLOAT. When people first talked about building iron ships, others laughed at them. "Iron sinks," they said, "and your boats will go to the bottom of the sea." If the boats were solid iron this would be true, for iron is certainly much heavier than water. But if the iron is bent up at the edges,--as it is in a dish pan,--it has to push much more water aside before it goes under than it would if it were flattened out. The water displaced, or pushed aside, would have to take up as much room as was taken up by the pan _and all the empty space inside of it_, before the edge would go under. Naturally this amount of water would weigh a great deal more than the empty pan. But suppose you should fill the dish pan with water, or suppose it leaked full. Then you would have the weight of all the water in it added to the weight of the pan, and that would be heavy enough to push aside the water in which it was floating and let the pan sink. This is why a ship sometimes sinks when it springs a leak.