In _Mechanical Inventions of To-day_ I have dealt with the uses which can be made of heat as a motive power. Here we have in some sense a reversal of the process. In the heat-engine the expenditure of heat produces motion. In the refrigerating machine motion produces heat, on the face of it a strange way of producing cold. Yet it is by the production of heat in the first instance that we are ultimately able to obtain the cold. One way to make a thing cold is to place it in contact with ice. But that process suffers from severe limitations. In the first place, we may not be able to procure ice when we want it. And in the second place, we may want to produce a temperature much lower than that of ice. Now a machine can produce any degree of coldness, almost down to the "absolute zero," the point at which a body is absolutely devoid of any heat whatever, the condition in which its molecules are absolutely still. That point is 274° C. _below_ freezing-point. Freezing-point on that scale is "zero," and so this _absolute_ zero is _minus_ 274°. Or, to put it another way, freezing-point is 274° _absolute_ temperature. The absolute zero has never been reached, and there is reason to believe that it never can be quite reached, but by methods about to be described a temperature within a few degrees of it has been attained. And all of this can be done without any cooling agent colder than water at an ordinary temperature. There are several systems, but the one which illustrates the principle most simply is that in which carbonic acid gas is the "working fluid." This is a very compressible gas, and so is well fitted for the purpose. First of all a pump or compressor compresses it. That has the effect of heating it. Such we might expect from the fact that heat is molecular activity: when by compressing the gas we force the molecules closer together, they naturally hit each other and the sides of the containing vessel harder than they did before, and the increased activity is manifested as increased heat. So the first effect, as was remarked just now, is to produce, apparently, increased heat. But then the hot compressed gas, by being passed through a coil of pipe surrounded by cold water, can be robbed of that heat. According to the speed at which it traverses the coil it will be more or less cooled: by causing it to travel slowly it can be brought down almost to the temperature of the water. So we start with the gas at atmospheric pressure and at somewhere about atmospheric temperature too. This we convert into compressed gas at a high temperature. After cooling it we have compressed gas at a moderate temperature.