[Illustration: Fig. 222.] 330. =Description of the Process of Freezing.=--In order to make the process of freezing clear to you I will describe it as it ordinarily occurs, that is, from the action of cold air upon the surface of water. The uppermost layer of the water imparts some of its heat to the air in contact with it. This air rises and colder air takes its place, which being warmed in its turn rises to make way for more of the cold air. You have therefore a constant current of warmed air upward from the water. In the mean time there is a current of a different character in the water--a downward one. As fast as the water at the surface parts with heat to the air it falls, other warm water taking its place, to cool in its turn and go down. This falling of the cooled water goes on regularly until a portion of water becomes cooled down to 39°, that is, 7° above the freezing point. This layer does not sink, but remains at the surface, for it is lighter than the warmer water below. This is because the law that heat expands matter is now reversed. Beyond this point of the thermometer the colder the water is the lighter it is. As the cooling now goes on from the air coming, as before, in successive layers to the water, the cooled water at the surface continually increases. At first it is a mere single layer of particles, but after a while it is quite a body of cold water lying on the warmer water below. At length some of it is cooled down to 32°, the freezing point, and a thin film of ice now forms. The state of things just at this stage of the process may be represented by a simple diagram, Fig. 222. Let the line _a_ represent the film of ice. The space between _a_ and _b_ is the portion of water cooled down below 39°. The space below _b_ is occupied by the water which is above this temperature. In the space between _a_ and _b_ the cooler the water is the nearer it is to the surface. That is, from the line _b_, where the water is exactly at 39°, as you go upward, the water lessens in temperature, it being successively 38°, 37°, 36°, etc., till, just in contact with the film of ice, _a_, it is at 32°. The ice goes on to thicken gradually by additions below. But it is to be remembered that ice is a good non-conductor, so that the very first layer of ice makes the cooling of the water proceed more slowly than before. And the thicker the ice becomes the slower is the cooling. This secures against too great a formation of ice.