ABSOLUTE ZERO — SIMON 



255 



parts connected by a tube. Upon pulling out the piston now the gas 

 will cool everywhere because this cooling is a homogeneous pro- 

 cedure. Let us cool now only the lower part to the low initial 

 temperature, at where we start (for example, to the temperature of 

 liquid hydrogen if we want to liquefy helium), leave the upper part 

 at room temperature, and pull out the piston again. Then the gas 

 will cool down within the lower part the same as if the upper part 

 were at the low initial temperature too. The atoms in the lower 

 part do not know whether the upper part is hot or cold and the 

 atoms do not know either if there is a cylinder and piston outside. 



ABC 



FiGDRB 3. — Principle of the expansion metbod. 



We get the same effect if I simply let it out by a valve, as in figure 

 3, C. I want to emphasize that the cooling arises within the cyl- 

 inder, and not at the valve, as in the Linde process. The procedure 

 described above has nothing to do with the Linde process, but it has 

 more resemblance to the Cailletet method. You remember, perhaps, 

 that he let a gas expand in a glass capillary tube, and with some of 

 the so-called " permanent gases " he could then see a little dust of 

 liquid drops, indicating that for some fraction of a second the 

 temperature had fallen considerably. 



Now, starting at high temperatures, the cooling effects obtained in 

 this way are very small. The chief reason for this is that a container 

 for high pressures has, at room temperatures, a heat content that is 



