254 ANNUAL REPORT SMITHSONIAN INSTITUTION, 19 35 



system in which the state of order can be changed by some external 

 means. Then in the way described above, one is able to transmit a 

 part of the original disorder in the system to the surroundings, and 

 to cool it down, making the same change of this variable in the 

 opposite direction after having isolated it from the surroundings. 



Gases are the prototype of a disordered system, the existence of 

 which is necessary for the procedure of generating low temperatures, 

 and it is relatively easy with them to change this disorder by chang- 

 ing the volume. So practically all procedures for generating low 

 temperatures were worked with gases until recently. Of course, one 

 can make use of them only down to the temperatures at which they 

 liquefy, or more precisely, as long as their vapor pressures have still 

 practicable values. So in practice the generation of low tempera- 

 tures and the liquefaction of gases have become practically identical 

 conceptions, and every step toward a lower temperature has been 

 marked by the feat of liquefying a gas with a lower boiling point 

 than was previously possible. 



I will not speak now of the development of the real procedures 

 performed with gases in order to liquefy them, which, for technical 

 reasons, have to be much more complicated than the example I gave 

 you. I need only remind you of the names of Faraday, Cailletet, 

 Pictet, Olzewski, Linde, Hampson, Dewar, Claude, and Kamerlingh 

 Onnes. You know that it is now relatively easy to get down to the 

 temperature of liquid air, as the liquefaction of air has become im- 

 portant for industrial technique. But I want to mention that to 

 cool by one calorie at even such a relatively high temperature as 

 that of liquid air is already 500 times as expensive as to heat by a 

 calorie above room temperature; for instance, by burning benzine. 

 But at lower temperatures the difficulties increase enormously, so that 

 the use of liquid hydrogen has been restricted to a few laboratories 

 and that of liquid helium to still fewer big specialized laboratories. 



In recent years the study of the properties of matter at very low 

 temperatures has become increasingly important; and so, of course, 

 one has sought for ways of simplifying the low-temperature tech- 

 nique. We have in the last few years developed a comparatively 

 simple method for liquefying helium, which I have shown you al- 

 ready, and now I should like to say a few words about it. The pro- 

 cedure of expanding a gas in a cylinder, in the way already de- 

 scribed, is very simple. But at the very low temperatures it is diffi- 

 cult to realize technically a cylinder and a piston. 



But one can overcome this difficulty in a very simple way, which 

 I will now explain. In figure 3, A, we have a gas enclosed in a 

 cylinder with a piston. Upon pulling out the piston the gas will 

 cool down. In figure 3, B, you see the cylinder divided into two 



