and softer until, if the rod is not too thick and the flame is sufficiently hot, a drop of molten glass will finally fall from the end of the rod.

If the temperature of the rod had been measured during this process, it would have been found to be continually rising. This behavior, so completely unlike that of crystalline substances, is characteristic of tar, wax, resin, glue, gutta-percha, alcohol, carbon, and in general of all amorphous substances. Such substances cannot be said to have any definite melting points at all, for they pass through all stages of viscosity both in melting and in solidifying. It is in virtue of this property that glass and other similar substances can be heated to softness and then molded or rolled into any desired shape.

199. Change of volume on solidifying. One has only to reflect that ice floats, or that bottles or crocks of water burst when they freeze, in order to know that water expands upon solidifying. In fact, 1 cubic foot of water becomes 1.09 cubic feet of ice, thus expanding more than one twelfth of its initial volume when it freezes. This may seem strange in view of the fact that the molecules are certainly more closely knit together in the solid than in the liquid state; but the strangeness disappears when we reflect that the molecules of water in freezing group themselves into crystals, and that this operation presumably leaves comparatively large free spaces between different crystals, so that, although groups of individual molecules are more closely joined than before, the total volume occupied by the whole assemblage of molecules is greater.

But the great majority of crystalline substances contract upon solidifying and expand upon liquefying. Water, antimony, bismuth, cast iron, and a few alloys containing antimony or bismuth are the chief exceptions. It is only from substances which expand, or which change in volume very little on solidifying, that sharp castings can be made; for it is clear that contracting substances cannot retain the shape of the mold. It is for this reason that gold and silver coins must be stamped

rather than cast. Any metal from which type is to be cast must be one which expands upon solidifying, for it need scarcely be said that perfectly sharp outlines are indispensable to good type. Ordinary type metal is an alloy of lead, antimony, and copper, which fulfills these requirements.

200. Effect of the expansion which water undergoes on freezing. If water were not unlike most substances in that it expands on freezing, many, if not all, of the forms of life which now exist on the earth would be impossible; for in winter the ice would sink in ponds and lakes as fast as it froze, and soon our rivers, lakes, and perhaps our oceans also would become solid ice.

The force exerted by the expansion of freezing water is very great. Steel bombs have been burst by filling them with water and exposing them on cold winter nights. One of the chief agents in the disintegration of rocks is the freezing and consequent expansion of water which has percolated into them.

201. Pressure lowers the melting point of substances which expand on solidifying. Since the outside pressure acting on the surface of a body tends to prevent its expansion, we should expect that any increase in the outside pressure would tend to prevent the solidification of substances which expand upon freezing. It ought therefore to require a lower temperature to freeze ice under a pressure of two atmospheres than under a pressure of one. Careful experiments have verified this conclusion and have shown that the melting point of ice is lowered .0075° C. for an increase of one atmosphere in the outside pressure. Although this lowering is so small a quantity, its existence may be shown as follows:

Let two pieces of ice be pressed firmly together beneath the surface of a vessel fuil of warm water. When taken out they will be found to be frozen together, in spite of the fact that they have been immersed in a medium much warmer than the freezing point of water. The explanation is as follows:

At the points of contact the pressure reduces the freezing point of the ice below 0° C., and hence it melts and gives rise to a thin film of water the temperature of which is slightly below 0° C. When this pressure is released, the film of water at once freezes, for its temperature is below the freezing point corresponding to ordinary atmospheric pressure. The same phenomenon may be even more strikingly illustrated by the following experiment:

Let two weights of from 5 to 10 kg. be hung by a wire over a block of ice as in Fig. 171. In half an hour or less the wire will be found to have cut completely through the block, leaving the ice, however, as solid as at first. The explanation is as follows: Just below the wire the ice melts because of the pressure; as the wire sinks through the layer of water thus formed, the pressure on the water is relieved and it immediately freezes again above the wire.

n

FIG. 171. Regelation

Geologists believe that the continuous flow of glaciers is partly due to the fact that the ice melts at points where the pressures become large, and freezes again when these pressures are relieved. This process of melting under pressure and freezing again as soon as the pressure is relieved is known as regelation.

Substances which expand on solidifying have their melting points lowered by pressure, and those which contract on solidifying have their melting points raised by pressure.

QUESTIONS AND PROBLEMS

1. What the meaning of the statement that the heat of fusion of mercury is 2.8?

2. Explain how the presence of ice keeps the interior of a refrigerator from becoming warm.

3. How many times as much heat is required to melt any piece of ice as to warm the resulting water 1° C.? 1° F.? How many B. T. U. are required to melt 1 lb. of ice? How many foot pounds of energy are required to do the work of melting 1 lb. of ice?

4. If the heat of fusion of ice were 40 instead of 80, how would this affect the quantity of ice that would have to be bought for the refrigerator during the summer?

5. Five pounds of ice melted in 1 hr. in an unopened refrigerator. How many B. T. U. came through the walls of the refrigerator in the hour?

6. Just what will occur if 1000 calories be applied to 20 g. of ice at 0° C.?

7. How many grams of ice must be put into 200 g. of water at 40° C. to lower the temperature 10° C.?

8. How many grams of ice must be put into 500 g. of water at 50° C'. to lower the temperature to 10° C.?

9. Why will snow pack into a snowball if the snow is melting, but not if it is much below 0° C.?

EVAPORATION AND THE PROPERTIES OF VAPORS

202. Evaporation and temperature. If it is true that increase in temperature means increase in the mean velocity of molecular motion, then the number of molecules which chance in a given time to acquire the velocity necessary to carry them into the space above the liquid ought to increase as the temperature increases; that is, evaporation ought to take place more rapidly at high temperatures than at low. Common observation teaches that this is true. Damp clothes become dry under a hot flatiron but not under a cold one; the sidewalk dries more readily in the sun than in the shade; we put wet objects near a hot stove or radiator when we wish them to dry quickly.

203. Evaporation of solids, - sublimation. That the molecules of a solid substance are found in a vaporous condition above the surface of the solid, as well as above that of a liquid, is proved by the often-observed fact that ice and snow evaporate even though they are kept constantly below the freezing point. Thus, wet clothes dry in winter after freezing. An even better proof is the fact that the odor of camphor can be detected many feet away from the camphor

crystals. The evaporation of solids may be rendered visible by the following striking experiment:

Let a few crystals of iodine be placed on a watch glass and heated gently with a Bunsen flame. The visible vapor of iodine will appear above the crystals, though none of the liquid is formed.

A great many substances at high temperatures pass from the solid to the gaseous condition without passing through the liquid state. The vaporization of a solid is called sublimation.

204. Saturated vapor. If a liquid is placed in an open vessel, there ought to be no limit to the number of molecules which can be lost by evaporation, for as fast as the molecules emerge from the liquid they are carried away by air currents. As a matter of fact, experience teaches that water left in an open dish does waste away until the dish is completely dry.

But suppose that the liquid is evaporating into a closed space, such as that shown in Fig. 172. Since the molecules which leave the liquid cannot escape from the space S, it is clear that as time goes on the number of molecules which have passed off · from the liquid into this space must continually increase; in other words, the density of the vapor in S must grow greater and greater. But there is an absolutely definite limit to the density which the vapor can attain; for as soon as it reaches a certain value, depending on the temperature and on the nature of the liquid, the number of molecules returning per second to the liquid surface will be exactly equal to the number escaping. The vapor is then said to be saturated.

S

FIG. 172. A saturated vapor

If the density of the vapor is lessened temporarily by increasing the size of the vessel S, more molecules will escape from the liquid per second than return to it, until the density of the vapor has regained its original value.